0% found this document useful (0 votes)
685 views233 pages

Samuel A. Chambers - The Lessons of Rancière-Oxford University Press (2012) PDF

Uploaded by

killozap
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
685 views233 pages

Samuel A. Chambers - The Lessons of Rancière-Oxford University Press (2012) PDF

Uploaded by

killozap
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 233

UNIVERSITY PRESS

Oxford University Press is a department of the University of


Oxford, It furthers the University's objective of excellence in
research. scholarship. and education by publishing worldwide

Oxford New York


Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto

With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam

Oxford is a registered trade mark of Oxford University Press


in the Uk and certain other countries

Published in the United States of America by


Oxford University Press
198 Madison Avenue. New York. New York 10016

© Oxford University Press 2013

AH rights reserved. No part of this publication may be reproduced. stored in a retrieval system.
or transmitted. in any form or by any means. without the prior permission in writing of
Oxford University Press. or as expressly permitted by law. by license. or under terms agreed
with the appropriate reproduction rights organization. Inquiries concerning
reproduction outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above.

You must not circulate this work in any other form


and you must impose this same condition on any acquirer,

Library of Congress Cataloging-in-Publication Data.


Chambers. Samuel Allen. 1972-
The les sons of Rancière / Samuel A. Chambers,
p. cm.
Inc1udes bibliographical references and index.
ISBN 978-0-19-992721-0 (hardcover : alk. paper) 1. Rancière. Jacques-Political
and social views. 2. Poli tic al science-Philosophy. 3. Liberalism. 4. Democracy. 1. Tille.
JA71.C4332012
320.01-dc23 2012010282

135798642

Printed in the United States of America


on acid-free paper
For Rebecca
Acknowledgments ix

Introduction 3

Politics 38
2 Police 65

3 Litera rity 88

4 Critique 12 3

Afte rwo rd 157


Notes 17 1
Works Cited 201

Index 21 3
acques Rancière proc1aims that it is possible to teach what one does not
know. Applying this radical pedagogical principle to the teachings of a book
that a book can teach something the author does not know. What this
book teaches, if anything at an, is surely something other than, something
much more than, the sum of my knowledge. There is thus a profound gap
between the intelligence of a book and the will of an author. This book, in-
c1uding its errors, is "mine," then, in the sense that 1 have chosen the final
arrangement of chapters and sections and sentences and words. However, the
intelligence of the book belongs to no one, and it is shared by many. Here
1 would like to name and note my gratitude to sorne of those many.
The privilege of getting to do political theory at Johns Hopkins University
sometimes feels like cheating. Primary thanks go to my immediate faculty
colleagues in political theory: Jane Bennett, Bill Connolly, and Jennifer Cul-
bert. 1 am also grateful for the support of an my outstanding colleagues in the
political science department, inc1uding recent postdoctoral visitors. Numerous
colleagues across campus have helped to support me or to sustain this project;
in particular 1 would like to thank Amanda Anderson, Veena Das, Frances
Ferguson, Aaron Goodfellow, Siba Grovogui, Paola Marrati, and Adam Shein-
gate. Of course, the vibrancy of a graduate institution depends utterly on the
graduate students themselves; in this context, 1 am delighted to acknowledge
the vital contributions to my thinking made by seminar students over the past
four years. Nathan Gies, Jairus Grove, Chas Phillips, and Drew Walker aIl
contributed directly to this project in various, significant ways. 1 also thank the
many undergraduate students who have eagerly grappled with Rancière's writ..
ings and along the way taught me so much about them.
Perhaps most importantly, 1 want to express my deepest gratitude to the
many readers that 1 have been so lucky to have as this project has developed
over the years. Since 1 have sorne, the following is a
list of those who were kind enough to read, comment, criticize, or otherwise
respond to various portions of the manuscript at sorne point in its develop-
ment: Paul Apostolidis, Ben Arditi, Jane Bennett, Paul Bowman, Rebecca
Terrell Carver, Bill Connolly, Jennifer Culbert, Jean-Philippe Deranty,
Lisa Disch, Kim Evans, Eric Fassin, Alan Finlayson, Nathan Gies, Jairus
Grove, Stephanie Hershinow, Bonnie Honig, Adam Kennard, Patch en Markell,
Todd May, Kirstie McClure, Aletta Norval, Joel OIson, Michael O'Rourke,
Davide Panagia, Chas Phillips, Andrew Schaap, Michael Shapiro, Jon Simons,
Richard Stamp, Jeremy Valentine, Drew Walker, Liz Wingrove, Karen Zivi,
and John Zumbrunnen.
There are a few individuals whose contributions to the book constitute debts
1 could never repay but which 1 am very pleased to acknowledge here. Ran-
cière calls learning an act of translation, and 1 learned a great deal from Anne
Kantel (Gelman), Nathan Gies (Greek), and Rebecca Brown and Lisa Disch
(French). It has been an absolute joy to work with Angela Chnapko at Oxford
University Press. Her sympathetic understanding of the project and her respon-
siveness and professionalism have helped me enormously in bringing this
book to completion. 1 thank Barbara Price for her brilliant and deft editorial
work. 1 am also grateful to Tabitha Panter, who provided early and inspiring
mock-ups of the cover design. The completion of the project might never have
occuned were it not for the diligent, insightful, and always intelligent research
assistance provided by Nathan Gies. Thanks go also to the Krieger School of
Arts and Sciences and to the Department ofPolitical Science at Johns Hopkins
for providing the funds for Nathan's RA position. Stephanie Hershinow gave
the book one of its last, and thus most important, reads: 1 could not have asked
for a more learned, conscientious, and careful scrutiny of the text. Finally, 1 am
enOlmously grateful to Bonnie Honig for an untimely and invaluable reading
of the manuscript.
1 owe a special debt to Ben Arditi, who planted the seeds for this book just
over a decade ago when he invited me to give a paper on Jacques Rancière, a
thinker whose works 1 had never read, at a conference in London. 1 am deeply
grateful to participants and audience members at that 2003 Goldsmiths Col-
lege conference who helped germinate those seeds: Paul Bowman, Terrell
Carver, Mick Dillon, Alan Finlayson, Jim Martin, and Alex Thomson. Parts of,
and pieces related to, this project were given in a number of settings. 1 thank
aIl of the participants and audience members and note special thanks to the
hosts or conference organizers: Department of Government, University of
Essex, November 2006 (thanks to Aletta Norval, David Howarth, and Jason
Glynos); Department of Poli tic al Science, Johns Hopkins University, Decem-
ber 2007; Western Poli tic al Science Association annual meeting, March 2009;
American Political Science Association annual meeting, September 2009; Po-
litical and Moral Thought Seminar, Johns Hopkins University, March 20II

x 1 Acknowledgments
to John Northwestern conference on JL'.<'LlH-,'·"'"

April20I l (thanks to Dilip Gaonkar and Scott Durham); Distinguished Global


Ewha October 20 II; Political Colloquium,
November 20I l to Lawrie Balfour and "" .. ~~hc,~
\",norl,n-.""nT of Political and Economic Studies, of Hel-
sinki, December 20II (thanks to Jemima Repo); Department of Political Sci-
ence, University of Minnesota, 20I2 (thanks to Bud Duvall).
An earlier version of Chapter One was published as "Jacques Rancière and
the Problem of Pure Politics," European Journal of Political Theory, 10.3
(20 II), 303-26. Portions of Chapter Two are developed from "The Politics of
the Police: From Neoliberalism to Anarchism, and Back to Democracy," in
Reading Rancière, edited by Paul Bowman and Richard Stamp (London: Con-
tinuum, 20rI), I8-43. Sorne of the original thinking for Chapter Three was
published in a very different form as "The Poli tics of Literarity," Theory &
Event, 8.3 (2005). Portions of the Afterword rework sorne ideas that originally
appeared in "A Queer Politics of the Democratic Miscount," borderlands, 8.2
(2009).
The most important sustenance for writing cornes not directly, from col-
laborators and colleagues, but indirectly, from the family, friends, and com-
munit y that make it possible to write in the first place. Hal Janney had a deep
understanding of what Rancière calls the equality of intelligences, of what it
meant to read a text on one's own; 1 wish he had had the chance to read this
book. Our dog Luke provided plenty of support over almost the entire period
in which this book was written; a Husky's life is too short. Joel OIson was a
true colleague and comrade, and one of the sharpest thinkers and most lucid
writers 1 have ever known. He cannot be replaced, but he will continue to
inspire.
I am grateful for crucial distractions provided by the degenerate poker club,
a group of individuals firmly committed to teaching what they do not know.
I once again, but never enough, thank my parents, Tim and Jackie Chambers,
for their patience with, and understanding of, the inscrutable peripatetic life of
the academic; no one has taught me more about learning. David and Stephanie
Hershinow lessened the culture shock of retuming to America and helped us to
make a home again on this side of the Atlantic. Alan and Keri Finlayson and
Julian Brown and Jane Elliott assured us that we still have one on the other
side. Thanks to Phil Chambers for needed mountain air, and to Laurie Frankel
for timely video chats.
As always, Rebecca Brown read more drafts, gave more feedback, dis-
cussed more nascent arguments, and contributed more intellectually to this
project than anyone else. As my collaborator in life, Rebecca is the collabora-
tor in everything 1 write, think, or teach. It is thus a truism to say that this book,
like any other 1 produce, would not have been possible without her. Yet this
book is unique, in that at its heart lies a set of claims about pedagogy, about the

Acknowledgments 1 xi
of teaching - to to to the world. Those 1-''-''-,,"U~V~jl'V
claims emerge out of a reading of Rancière, but the claims themselves are not
Rancière's and the reading is not simply mine. use my own words to say what
l want to say about teaching, leaming, and pedagogy, but the animating
assumptions share with Rebecca. And we share them in the sense not that
each of us possesses them (and therefore we "agree" with one another) but in
the sense that the possession itself is joint; these principles can only be held
onto by holding them together. The Rancièrean "assumption of equality" that
makes a partnership possible cannot be an assumption that one person makes
or grants to the other. The assumption is itself the condition of possibility for
the "we," and for the "1."

xii 1 Acknowledgments
1 1 n

HIS BOOK HAS A trick title. The Lessons of Rancièrel sounds undernand-
ing and painless enough, since in echoing the cornrnon cliché, "the les-
sons of history," the title rnakes it sound as though this book will tell its
readers what they need to know about Jacques Rancière in the sarne way as
historians rnight tell their readers what they need to know about a particular
tirne period or event in history. In other words, the title says to the reader: there
is this person called Rancière, who has likely written sorne books; but you do
not really need to read his books since you can read this one; and it will explain
to you "the lessons" of Rancière's writings. You rnight also expect then, know-
ing that this is a book that fits into the broadly defined field of conternporary
political theory, that it would be a book "on Rancière" as a political theorist, or
that it would systernatically construct his overall political theory. Such a book
would naturally be structured by a series of explications of Rancière's writ-
ings, places where 1 tell rny reader what Rancière is saying, what it rneans, and
ultirnately draw sorne conclusions (i.e., the lessons).
This is not such a book, because an approach like the one described above
will not work in the case of Rancière. In order to write a book deterrnined to
give a general overview of the political thought of Rancière, or to develop and
defend "Rancière's political theory," one would have to either ignore or refute
Rancière's own explicit clairns about the relationship between his writings, on
the one hand, and systernatic theories of politics, on the other. 2 In 2009,
Rancière wrote a response piece to a series of articles on his work published as
a special issue of the journal Parallax. He chose to reply not according to the
tirne-honored tradition of the response essay: narnely, by clarifying certain
points, explaining his central tenets and precepts, and refuting criticisms of
"his theory." Instead, Rancière gives the essay its own deceptive title: ''A Few
Rernarks on the Method of Jacques Rancière," by Jacques Rancière. And he
writes the entire piece in the third person. Thus, he opens the essay by writing
never intended to a
literature, cinema or anything else. He thinks that there is already a good deal
of them and he loves trees enough to avoid destroying them to add one more
to aIl those available on the market" (Rancière 2009b: l Ran-
cière confirms that he has no "theory of" anything, but certainly not of politics
2009b: l r6). The Lessons of Rancière therefore cannot be a teach-
er's guide to Rancière's political theory, since that sort of class planning would
necessarily need to disregard one of Rancière's chief lessons.
The title of this book says and does more than it might first appear to say or
do because one of the centrallessons of Rancière's work - indeed, perhaps his
most important teaching is a lesson about the "les son" (Rancière and Oliver
2008: 172). In choosing the title 1 have, 1 do not mean to suggest that my book
is a systematic account of Rancière's political theory, but rather to convey a
crucial dimension of Rancière's work: his claims about "lessons," his argu-
ments about teaching and pedagogy, prove fundamental to his project. There-
fore in this book 1 avoid systematizing Rancière's writings; 1 resist morphing
those diverse and concrete interventions into a static or comprehensive "theory
of politics"; and 1 both follow and question Rancière when he cuts against the
grain of contemporary political theory. Rather than merely explicate Rancière's
own texts, 1 mobilize them for my own polemical interventions into debates
over democratic politics. 1 thereby refuse to draw out the lessons of Rancière's
texts in the manner of a series of morals, precepts, or mIes. It is because of this
approach, and not at an despite it, that 1 am committed to thinking through the
"les sons" of Rancière in another sense. Above aIl Rancière wants to "teach"
his readers something absolutely crucial about teaching. In making this claim
1 cannot overstate the extent to which Rancière advocates an utterly radical
pedagogy, one that completely reconceives aIl the central elements of "school-
ing," including the concepts of teacher, student, intelligence, and knowledge.
Rancière thinks it possible to teach without knowing; he believes that the best
schoolmasters can operate not on the assumption of their expertise, but on the
equality of intelligence; and this means ultimately that Rancière contends that
we can "teach what we do not know." The best schoolmasters are ignorant
schoolmasters.
Below in my discussion of equality, and scattered throughout the chapters
of this book, 1 will come back to this radical pedagogy and to the resistance to
a certain fonn of epistemological and ontological mastery - a resistance that
forms the backbone of Rancière's wlitings and critical investigations. Here 1
want to make just one specific and central point about the very idea of "lessons."
The lessons of Rancière are not the so-caIled les sons of history for many rea-
sons, but chief among them must be that for Rancière even the lessons of his-
tory are not the "lessons of history," as we typically understand that phrase.
Between 1975 and 1985 Rancière played a central role in the journal Les ré-
voltes logiques, wliting numerous single- and co-authored articles, contributing

1 The Lessons of Rancière


to collective
on the archive of
Rancière 201 In these early writings Rancière vVllühHviHly

the idea of the worker as the or the


'-"T1'"PI''' resistant. Whether in the guise of orthodox .uA .... ~".UJL.H' AllttmsserlaI1l1Srn,

or the "new philosophers," the poor (the plebs, the proletariat) are kept in their
place to the benefit of those who would represent them. Rancière '-''V'''''HU~''''UU
contests, above aH else, this confinement of the poor to a
sphere, this creation of a pure category for the poor.
the 2011 English translation of his contributions to Les révoltes logiques, Ran-
cière describes the collective project as an effort to hold on to a of
emancipation, of revolution, of revoIt, without jettisoning or ignoring "the
issues, complexities and contradictions of two centuries of struggle" (Rancière
201 lb: 9) - without ignoring, in other words, the fact that the particular lives
and struggles of workers rarely match the categories of political philosophy,
sociology, or revolutionary the01'y. Among many other things, this effort to
maintain the possibility of emancipation and revoIt while keeping one's ears
open to the voices of definite workers (not philosophical categories) entails a
steadfast refusaI to find in history mere "lessons" for contemporary political
programs (Rancière 201 lb: 12).
Rancière's career-long refusaI of the category "the pure proletarian" always
remains tied to his understanding of history, on the one hand, and to his radical
conception of pedagogy, on the other. A quotation from the editorial collective
of Les révoltes logiques captures these connections: at most, history teaches us
to "recognize the moment of a choice, of the unforeseeable, to draw from his-
tory neither lessons, nor, exactly, explanations, but the principle of a vigilance
toward what there is that is singular in each call to order and in each confronta-
tion" (Rancière et al. 1977: 6; quoted in Ross 2009: 29; and Ross 2002: 128).
History shows us that a choice has been made, but history's real teaching lies
in demonstrating that such a choice can never be "explained" post hoc (and
sUl'ely never predicted beforehand). We learn from history a vigilance toward
the singular, a capacity to sense that "something is happening." As we shaH
see, a moment of happening might serve as another name for "politics" in
Rancière's unorthodox and polemical account. Moreover, as Kristin Ross puts
it in her own gloss on this passage, "To happen, events must be perceived and
acknowledged as such" (Ross 2009: 29). This, then, gives us sorne indication
of the lessons we might learn from Rancière: first, an emphasis on the vigi-
lance that history te aches (and concomitantly an attentiveness to history itself);
second, a capacity to perceive and acknowledge events when they happen.
1 translate this second lesson into a capacity for surprise an understanding of
politics and history that remains open to the very possibility (the constitutive
necessity, even) of surprises. For Rancière, poli tics is always a surprise (see
Arendt 1958; cf. Beltrân 2009: 601).3

INTRODUCTION 1
To stakes of this it would contrast it with better
known, tradition al approaches to polities. Take liberal political theory, on the
one hand, and social science, on the other. These two names point to distinct
disciplinary fields with their own epistemologieal and ontologieal commit-
ments and their own everyday practices of knowledge production and dissemi-
nation. Nevertheless, liberal theorists and social scientists have at least one
thing in common: they do not like surprises. 4 ln the first case, the Lockean
theory of limited government, which forms the core of liberalism's basic
tenets, 5 emerges as a response to the tumult of the seventeenth century in
England, and it aims, especially, to provide safety and security for the growing
merchant class. The hemistic deviee of a "state of nature" - in whieh aIl au-
tonomous individuals are posited as free, equal, and capable of acquüing prop-
through the natural right of exclusion (the legitimate capacity to say "this
is mine") makes it possible for Locke to argue that aIl valid governmental
authority rests implicitly upon the consent of the governed. Locke thereby cre-
ates a purpose or telos for political society: namely, the protection of individual
life, liberty, and property. Thus, a liberal theory of limited government pro-
vides a check against any power that would encroach upon individual rights or
property. This tale will prove familiar to most readers, but 1 tell it again briefly
here to emphasize the liberal commitment to order, to structure, to a certain
form of hierarchy, and ab ove aU to the idea of political stability. Much like
seventeenth-century English Parliamentarians and the rising bourgeoisie, crea-
tures in Locke's state of nature would above alllike to be secure. Locke wants
to ward off the surprise of agents of the king knocking on the door - which is
not to make the mistake of characterizing Locke as wholly opposed to king-
ship, but to indicate his fundamental opposition to what he calls "Absolute,
Arbitrary Power" (Locke 1988: 284).
ln the second case, to the extent that they approach social phenomena seek-
ing systematic explanations, twenty-first-century social scientists search for
their own form of order. The social scientific modus operandi therefore con-
sists in the attempt to measure social and political life, to quantify it, to
"operationalize" ideas, concepts, or phenomena so that they can become "vari-
ables." No practice captures the heart of the social scientific enterprise better
than that of coding. Coding is the process of translating one form of data into
discrete and often quantifiable categories. The activity of coding survey or in-
terview responses wherein a subject's answers to a range of questions are
placed into categories and usually assigned numbers - proves particularly sig-
nificant for understanding the relationship between social science and the poli-
tics of surprise. Coding creates, quite straightforwardly and plainly on the
surface, a grid of intelligibility. Answers that do not fit the code are either
made to conform somehow (for example, taking an apparent non sequitur re-
sponse to a preference question as a sign of negative preference) or rejected
entirely. In this way, coding allows social scientists to create an order to the

1 The Lessons of Rancière


wish and also to avoid
and the code for scoring the answers have been created, nothing new or sur-
can show up: have to tell the investigator what she or he
wishes to at least to the extent that aIl answers have to be marked on, say,
a 0-7 scale. There can be nothing outside of this grid: no surprises. 6
Given this common ground, what might liberal theorists and social scien-
tists make of the events in Egypt in January and February, and throughout the
Middle East and northern Africa in the first half of 20 II? Above aH else, these
events took the world by surprise. No political pundits were calling for a revo-
lution in Egypt, and surely no social scientific experts or theorists had pre-
dicted it. There is no social science theory that shows the causal relation
between self-immolation in one country and major social transformations in
another. 7 It sounds almost comical to think that there could be such a theory
(even if there were, it would sUl'ely look more like sham than science), but one
cannot underestimate the amount of hand-wringing that went on, mainly
among political talking heads but also within the ranks of social scientists, im-
mediately following the ouster of Hosni Mubarak. Social scientists had not
merely failed to predict social transformation in Egypt, but had gone so far as
to repeatedly caH Egypt one of the most "stable" regimes in aU of the Middle
East. And it is not as if the idea of predicting so-caIled social unrest falls out-
side the purview of social science - quite to the contrary: "In the last three
years, America's military and intelligence agencies have spent more than
$ l 25 million on computer models that are supposed to forecast poli tic al unrest"
(Shachtman 2011).8
Despite the unexpected if not shocking nature of the events in Egypt and
elsewhere in the first part of 20 l l, l want to caH attention to two elements that
are not surprising. First, it is no surprise at all that the Egyptian revolution
would prove to be so novel and unexpected. In other words, it is not a surprise
that Egypt was a surprise. l risk this seemingly trivializing formulation in order
to calI attention to a very important point. Rather than suggesting that there
was something in the particular nature of the Egyptian revolution that made it
unpredictable - that threw off the CIA social unrest models, or surprised the
best expectations of social scientists - we must instead insist on a very differ-
ent, general claim. Democratie revolutions are not something one can predict
in advance; a mere transition of power may be subject to the social scientific
calculus of putative prediction, but a real revolution is always unexpected. 9
Second, in a very important way, the projects of both liberalism and social
science must be grasped not simply as avoiding or eschewing surprise but as
committed to a worldview that eliminates surprise, one that renders real novelty
a structural impossibility. For social science, nothing is ever truly surprising,
since prediction the explicit or (much more commonly these days) implicit
goallying at the he art. of the social scientific model of explanation -- renders the
future uniform (Connolly 2010). A model of explanation that min-ors prediction

1NTRODUCTION 1 7
entails an of the world that renders newness,
impossible. If we could always predict, we would never be surprised. As Ran-
cière himself says, "Political science['s] ... axiom is that nothing is ever surpris-
(Ranci ère 2010a: 12). Thus, the proper social scientific response to a real
surprise will always be to seek a post hoc explanation that would have elimi-
nated such surprise if only we had developed the explanation earlier. In this way,
social science operates in this bizarre conditional perfect tense of the "would
have happened," and this means that while social scientists are surely sometimes
surprised, they always strive to eliminate the possibility of such surprise. Liber-
alism, in complementary fashion, provides a framework in which aIl problems,
issues, confticts can be resolved, sorted out, and contained. The interest-group
calculus of negotiation and compromise commits liberalism to a math without
remainder, and therefore to a vision of a world without surprises (Honig 1993).
What, then, would it mean to rethink poli tics and history as full of sur-
prises?lP How could we restore to our understanding of politics - inject into
our study of politics the very sense of wonder and shock that the world
shared while watching the events in Egypt at the beginning of 20 II? What
would it mean for the word politics to point not toward legislative decisions
and judicial decrees, but to moments of irruption like those in Egypt?

Democratie Politics

As a precondition of such an approach, one must grasp politics as untimely


(Chambers 2003; Brown 2005). That is, if history runs a smooth linear course,
if events in the future can be causally predicted based on events in the past,
then politics can never rupture or interrupt that flow of events. We must have a
sense of history as nonlinear (Le., a sense of historicity) in order to understand
politics as a moment of interruption, as a possible event with repercussions
that can never be anticipated (Derrida 1994). As Ross puts it, "The temporality
of politics is not progressive, nor dialectical ... it is not continuous and it's not
over" (Ross 2009: 29). Democratie politics is untimely; liberalism is not. My
daim and Ross's both invoke Rancière's now widely recognized thinking of
democratic politics - a thinking that animates an the chapters of this book. The
book itself stages a series of diverse engagements with Rancière's unique ap-
proach to politics, as something irreducible to an exercise of power and never
contained within a set of institutions. Politics, for Rancière, cannot be under-
stood as the making of laws, or the decisions of courts. Instead, and at its root,
politics enacts dissensus. This means that poli tics cornes but seldom, and only
by way of a never predictable, and always insurrectionary, moment. This is a
moment when a given order of domination and a given regime of hierarchy are
radically called into question by the emergence of a political subject, a dëmos.
Importantly, the dëmos that undoes a given order does not and cannot exist

1 The Lessons of Rancière


and of 1-''-'~.lU,",'''.
Thus, the appearance of a political subject is always untimely, in the sense that
the political is alter the moment of !JViHL,","'.

formulation "Politics makes of politics" captures cru-


cial dimensions of Rancière's thinking, but only if we read this formulation in
the context of untimeliness can it escape the logical paradox within which it
appears to be Politics produces the subject that would seem to precede
poli tics. Another way of grasping this thinking of politics is to emphasize that
a radical democratic subject of politics does not just appear on the stage; it
brings the stage into being. As Peter Hallward argues, "Politics is about build-
ing a stage"; politics is a mise en scène (Hallward 2009: I42; Chambers and
O'Rourke 2009: 1).
This is what the Egyptian people did in January and February 20 II: they
created a scene of politics where befme there had been only a regime of order.
They made visible what had been invisible, made audible what had been inau-
dible, by appropriating a space that had not been theirs. We might say that the
Egyptian people thereby produced the stage on which political events would
afterward appear to unfold. However, Rancière's perspective would lead us to
see that the political moment was the creation of the stage itself. It is for just
this reason that we can grasp the Egyptian revolution as more decidedly politi-
cal in Rancière's sense of politics, as more distinctly democratic than many
other historical cases of "revolution." In the Egyptian case the revolution
occulTed not through the shifting of tides in a longstanding battle between
classes, but as the political creation and emergence of a new conflict where
before there had appeared to be no conflict at aIl. Undoubtedly Egypt had
meaningful underlying political interests and demands that were numerous,
long-standing, and often at odds with one another. 11 However, the revolution
did not merely express those pre-existing interests (it was not sorne sort of
natural outgrowth, development, or resolution of given interests) but rather
constituted new ones. For Rancière, this is the very essence of democratic poli-
ticS. 12 Poli tics, he says, must not be understood as a battle between existent
classes (it is therefore not "interest group politics" as practiced and modeled in
the USA, nor is it the "consensus politics" of European social democracy).
Rather, politics is that more originary moment of dissensus that brings about
the very existence of classes in the tirst place. In one of his most-quoted lines,
Rancière expresses this point in a thoroughly untimely language: "Politics
does not happen just because the poor oppose the rich. It is the other way
around: politics ... causes the pOOl' to exist as an entity" (Rancière 1999: II,
emphasis added).13
This book develops and elabol'ates this thinking of politics, exploring it in
depth, connecting il to other key debates in political theory, and extending it to
new domains of contemporary politics. The book's most important contribu-
tion lies in rethinking the very idea of democratic politics in a fresh way in a

INTRODUCTION 1 9
way that is free from the stale debates over liberal and deliberative and
Ilot yoked entirely to questions of institutional systems or regime structure.
This book implicitly asks (as Rancière so frequently does) the question, "What
is democracy?" The question proves significant because it is posed in terms
that have nothing to do with voting schemes and party systems, but everything
to do with politics in a far more radical and essential sense (what Rancière
calls, simply enough, la politique). In other words, and above aU else, this is a
book on democratic politics, wherein l insist on always returning to Rancière's
radical understanding of democracy as a "miscount." The democratic miscount
is untimely; miscount names the fundamental paradox of democratic politics
as the taking part in politics of those who have no part in politics.
The chapters herein focus first and foremost on seeking a distinct, theo-
retically fecund, and politically productive conception of democracy and
democratic politics. Rancière's powerful, polemical, and unique conception
of poli tics as dissensus motivates and enables this project to just the extent
that Rancière's conception makes it possible to distinguish decisively
between a liberal politics, on the one hand, and a democratic politics, on the
other. By linking politics fundamentally with the democratic - there is no
non-democratic poli tics for Rancière; for something to be "poli tics" accord-
ing to his definition, it must somehow bring about a democratic moment -
Rancière lets us see the unique and very much distinct logics of liberalism
and democracy. Viewed through the Rancièrean lens, a liberal order is an
order of the police: a hierarchical ordering and allotment of roles (and rights)
that operates according to a logic of domination. As l discuss in detail in
Chapter Two, Rancière does not merely denigrate police orders, since police
orders are not simply bad: such orders do a lot of good things, and we can
improve them. Nor does he mean to suggest a Manichean distinction between
the police as Evil and politics as Good. Nonetheless, according to the po-
lemical framing that Rancière provides, the constitution and maintenance of
a liberal order cannot be given the name "politics" and the two phenomena
must not be confiated. Politics is that phenomenon that disrupts the police
order any police order, even a liberal one - and confronts it with the logic
of equality.
The Lessons of Rancière thereby seeks to disentangle democratic politics
from liberalism. Its goal is to wrench a thinking of democratic politics free
from its entrapment within the confines of liberalism. In formulating this claim
in severe terms, l should also emphasize that my point is not to dismiss aIl of
liberalism, to suggest we eliminate liberal institutions, or to deny the impor-
tance of those institutions to citizens. Below, and throughout th~ book, l draw
a contrast between liberal police orders, on the one hand, and democratic poli-
tics, on the other. However, just as there are better and worse police orders
and just as we cannot do away with or ever eliminate policing, so also are there
better and worse liberal orders - and one cannot do away with liberalism

1 The Lessons of Rancière


without it worse order. Îs
aIl of liberalism; it is to argue for a democratic logic that is other to liberalism. 14
In picturing as constrained, or confined
invoke a number of related phenomena. point to a stifiing of the very
idea of democratic politics the theories of Rawlsian politicalliberalism and
Habennasian deliberative theory.15 In their central preoccupation with the
refonn of liberai institutions, politicalliberalism and deliberative theory have
the effect (intended or not) of reinforcing the equation of democracy with lib-
eralism. Such a collapsing of the distinct differences between democratic
thought and liberai frameworks is only encouraged, second, when liberal po-
litical institutions (voting, parliaments, courts, civil rights) take on the mean-
ing of democracy. Indeed, the common phrase "liberal democracy" expresses
concisely just the sort of confiation that 1 believe a Rancièrean conception of
"democracy" struggles against. For Rancière, "liberal democracy" is a contra-
diction in tenns; it is another way of saying "police politics."16
By insisting on this point 1 also distinguish my project from that of Chantal
Mouffe. 1 am obviously not the tirst to in si st on the difference between liberal-
ism and democracy: Mouffe herself points out that the distinction is argued for
most vociferously in Carl Schmitt's infamous critique of liberalism (Mouffe
2000; Schmitt 1988),17 ln a famous text of her own, Mouffe describes the dif-
ferences between liberalism and democracy while also rejecting what she calls
the "rationalist" approach to liberal democracy of Habelmas and Rawls; in this
way, her project and mine are sUl'ely complementary. However, for Mouffe,
the "democratic paradox" points to a tension between liberalism and democ-
racy that must be worked through, lived politically, and maintained theoreti-
cally. That is, on Mouffe's account "liberal democracy" is a meaningful and
important entity - a paradoxical entity, but an entity worth defending nonethe-
less (Mouffe 2000: 5). She challenges other approaches to liberal democracy
for their failure to note the tensions within it; hence she refers disparagingly
to "the recent attempts by John Rawls and Jürgen Habermas to reconcile
democracy with liberalism" (Mouffe 2000: 8, emphasis added). Mouffe insists
that the tension between democracy and liberalism can never be overcome,
but this does not stop her from trying to defend liberal democracy in a non-
rationalist manner. Mouffe's argument begins with a focus on the paradox of
liberal democracy, but this does not get in the way of her efforts to uphold,
maintain, and seek to augment liberal democratic regimes (Mouffe 2000: 18,
36). To do so means to exp and democracy and also to "consolidate liberalism"
(Mouffe 2000: 58).
As 1 will make very clear in the tirst two chapters of this book, the idea of
defending a liberal democratic regime and consolidating liberalism could
not be further from the thinking of Rancièrean democratic politics. Mouffe
distinguishes between liberalism and democracy but only in order to consider
their constitutive tensions. My book, in contrast, drives a wedge between

1NTRODUCTION 1 11
and liberalism in order to show
ti""i"Y'lrlr"l'Clr"U however much work side
side and however much they are thought together, they are not the same
work together. In her so-called paradoxical conception of
,"',""'U, Mouffe holds on to the idea of a liberal-democratic "regime." But as
!J'V

show in the example below, this is just the notion that Rancière rejects. The
gap that Rancière's thinking of poli tics opens between democracy and liberal-
ism is therefore one that mns aIl the way across or an the way down; it is not
p;.. ..
U i "... 'y or otherwise crossed. For this reason, 1 see Rancière's thinking of
(discussed in the next section) as expressing a more radical form
of political paradox than the "constitutive tensions" on which Mouffe centers
her account 2000: 5).
This unlike Mouffe's, does not provide another critique of Rawls, of
HabermalS, or of liberal institutions. One can already find a very long and excel-
lent list of such critiques. Instead of adding another one to the pile, this book
gerlenltes its own (very much oblique, but no less fundamental) challenge to lib-
eral and deliberative approaches to political theory. It uses a Rancièrean thinking
of democratic politics as leverage against the tendency to collapse democracy into
the broader terms of liberalism. Nor do 1 see a need to elaborate a critique of
Mouffe (or Schmitt): while 1 acknowledge a set of resonances between her proj-
ect and mine, the differences also prove stark. As 1 will show, the force of Ran-
cière's thinking is never aimed at the development ofliberalism - hence the power
of his writings to help us grasp the unique logic of democracy - and ultimately his
account of democratic politics simply proves incompatible with Mouffe's.
But if my goal is not to advance liberal democracy, nor to theorize, validate,
or reform it as a regime, this raises the question of why or how the difference
between Iiberalism and democracy matters in the first place. The book itself
offers the fullest answer to that question, as it makes the broadest possible case
for a Rancièrean democratic politics. Here, let me offer a brief and simple, yet
concrete and illustrative, example. Although Americans are often said to exer-
cise "full democratic rights" at the age of eighteen, citizens are expressly prohibited
by the Constitution from holding national office until they are older (twenty-five for
the House of Representatives; thirty for the Senate; thirty-five for the presidency).
John Seery has recently published a book-length argument in favor of Iowering the
age requirements for holding national office in the USA (Seery 201 la). To launch
the book, he published a short version of the argument as an article on Salon. Even
in this brief essay, Seery makes a large number of powerful arguments against
the age requirements, showing that they are di scri mi natory, unfair, out of
touch with the times, not in line with other democracies, inflammatory of
generational tensions, not meritocratic, and particularly offensive given that
much of the burden of military service faIls on the shoulders of the same citi-
zens who are prevented from holding office (Seery 201 lb). From a Rancièrean
perspective, it is not clear that one even needs to pile up arguments in this
fashion. Democracy is not mIe by any particular class; it is the mIe of

1 The Lessons of Rancière


the of democratic 1J'-'>iU,","',
be no argument for age restrictions. the imposition of a police order, and
the of would lead one to maintain such age restrictions.
clear the distinction between liberal institution al requirements and
1l!U.l>J.1.15

the of democratic see that demands the


full assent of democrats.
But this is not what it received at the hands of Salon's ostensibly left-leaning
commenters. While sorne applauded hi m, the great majority responded nega-
tively, and sometimes vociferously so. Commenter after commenter added to
the critique with a whole host of counterarguments: kids are stupid, brains do
not mature until people are oider, the Founders wanted it this way, the better
solution is age restrictions (to keep out the senile as weIl as the immature), and
so on. One after another of these "liberal democrats" (as 1 would like to think
of them in this example) kept giving reasons why the age restrictions should
stay in place. 18 But in every single case their claims added up, from a Ran-
cièrean perspective, to nothing more than an argument for inequality, for mas-
tery, for experts, for distinguishing by the measure of age between those who
know and those who do not. As 1 will show, democracy means a commitment
to the equality of intelligence. And as long as we are committed, instead, to the
alternative principles of mastery (and their attendant presumptions of inequal-
ity), we are not democrats.
To be clear, there is nothing inherent to liberalism that would dictate support
for the CUITent US constitutional age restrictions on office-holding, and one
might even argue that sorne of the Salon commenters, in defending age restric-
tions, were in fact making non-liberal arguments. Thus, liberals might respond
to my use of this ex ample by arguing that their version of liberalism provides
strong nOlmative grounds - on the basis of good liberal tenets such as full
human dignity and fundamental political equality - for rejecting the current age
restrictions. There is nothing logically wrong about this response, but in general
it misses the point. It is true: if we work within the normative framework of
liberalism, we could make a series of arguments for or against office-holding
age restrictions. 19 But, thought from the perspective of a Rancièrean democratic
politics, the rejection of the age restrictions is not based on any normative
grounds at an. Rather, the claim is different: there are no democratic grounds
for such age restrictions. Therefore, 1 offer this example not to enter a debate
over liberal age requirements. Moreover, setting lower age requirements for
holding office (whatever the age chosen) does not, in itself, constitute an act of
democratic politics. 1 am not, therefore, offering an example of democratic poli-
tics, but using an everyday political question as an example of how we might
think through the distinction between liberalism and democracy.20
For me, the implications of this case are much more circumscribed: the
example indicates the significance of the distinction between the liberal order
and democratic politics. While there may weIl be good arguments for restricting

1NTRODUCTION 1 13
age, these are never democratic The of is distinct.
redefining most of what we normally take to be poli tics under the banner
of "la police," Rancière makes space for this renewed thinking of democratic
politics. 21 He offers a much sharper sense of what it might me an to say "de-
mocracy" and not simply mean, "the given political order," because when
Rancière says "democracy" he always means moments of democratic politics.
Democracy, he insists, is not a regime (Rancière 2001: thesis 4; Rancière
2006c: 71). Rancière's approach therefore always demands we separate liber-
alism from democracy, since we can only understand liberalism as la police,
never as la politique. Thus, in their own determinate ways all the chapters in
this book conceptualize democracy and "the democratic" in ways that cannot
simply be recuperated by majoritarian representative institutions, nor relegated
to the terms of a Rawlsian (or other) liberal political theory.
Given these aims and goals, given the nature of Rancière's asystematic,
non-ontological approach to questions of politics, and given what he himself
explicitly says about systematizing an account of politics or producing a politi-
cal philosophy, it cannot be overstated that this is not a book on Jacques Ran-
cière nor on his theory of poli tics - at least not in any traditional sense of
devoting a book to a particular author. The chapters that make up the book are
themselves expressions and examples of Rancière's exemplary think.ing of
politics; this is why they eschew a systematic exegesis of Rancière's corpus
and avoid a general articulation of his political theory. There can be no grand,
unified theory of politics for Rancière, not simply (as his playful third-person
account might be taken to imply) because he has not chosen to write one (to
save the trees that he loves), but more important1y, because politics is made up
of polemical interventions. On these points, 1 take Rancière very much at his
word, but this means that 1 often work orthogonaIly to Rancière. Put differ-
ent1y, in order to take him at his word, 1 sometimes go against it. Therefore this
book demonstrates a certain fidelity to Rancière's thinking in the very refusaI
to treat him as a theorist in the "grand tradition," and specifically by appropri-
ating his writings in novel and unexpected ways - ways that often cut against
the grain of Rancière's own arguments. Throughout the book, my reading of
Rancière's texts aligns with his own remarks on his method in insisting that he
has no theOl"y of politics. This explains why these chapters on democratic poli-
tics do not simply add up to a book on Rancière's political theory. More than
this, in my approach to and reading of Rancière, 1 resist at aIl costs the idea of
articulating or constituting something like "the political thought of Rancière."
In other words, it is notjust that 1 have chosen not to write "a book on Rancière"
but that 1 refuse the very idea of producing such a text.
Of course, this refusaI to systematize Rancière should not be confused with a
hesitancy to engage with his work, a reluctance to develop certain of its themes,
or an unwillingness to make my own (sometimes polemical) daims about its
meaning and importance. In addition to pressing a series of specific arguments

14 1 The Lessons of Rancière


mnJU2~n Ci)nCTete, focused en,!~a.Q~errlents and
and alongside my eff011 to make the general case for distinguishing democratic
poli tics from liberal political The Lessons of Rancière also advances and
defends a particular reading of Rancière. While do not work through Rancière's
texts methodically or chronologically, in the various chapters of the book treat a
wide swath of his writings. focus pmticularly on those works that deal with
questions of politics, philosophy, wliting, and equality, and although certainly
address Rancière's most famous wlitings, sometimes emphasize less well-
known writings, and works that have only recently been translated into English. 22
Fm1her, while the secondary literature on Rancière has only begun to appear very
recently, and would therefore be appropriately called nascent, it is doubtless also
rigorous and rich. A vibrant and important conversation about Rancière's wlitings
has sprung up across multiple disciplines; my work here joins that conversation,
and seeks to help shape it in the future. Put differentIy, sorne of Rancière's "les-
sons" have already been enumerated by wliters other than Rancière; The Lessons
of Rancière is therefore about these thinkers as weIl.
ln the remainder of this introduction 1 will highlight a few terms that prove
most important for, and salient to, my particular appropriation of Rancière's
writings. "Laying the groundwork" is a poor metaphor for the work 1 want to
do here: since the project of the book is not to build an edifice, 1 do not require
a foundation. Rather, the book itself follows a certain path through Rancière's
writings, and includes a number of engagements with specific political mo-
ments and other particular thinkers. For this reason the better metaphor would
be "prepaling for a journey." Therefore, and as usual, the chapter summaries at
the very end of this introduction provide the roadmap. The sections below, on
concepts and themes, serve as the necessities for the tlip to come, the things
one has to pack - like water, snacks, clothes, coffee, music, and an extra pair
of shoes. This work thus proves prepara tory for the more detailed development
of concepts and the closer reading of texts in the chapters that come, but per-
haps more important, it will give the fiavor of my approach to Rancière's work
and indicate something of the nature of my engagement.

Subjects

Parties do not exist prior to the declaration of a wrong.


(Rancière '999: 39)
A subject is an outsider, or more, an in-between.
(Rancière 1995C: 67)

Of aIl the provocative, elliptical, polemical lines that Rancière has written
about politics, none has so transfixed me, and none has had so great an impact

INTRODUCTION 1 15
on my sense of Rancière's for than the tirst one
quoted above. The quotation is exemplary of Rancière's style. Rancière him-
self refers to his approach as "polemical," and in Chapter Four 1 discuss the
meaning of polemic for Rancière and relate it to the concept of critique. But
here it is important to be dear that for Rancière polemie does not mean lengthy
ad hominem attack. 23 And stylistically, to write polemically for Rancière
means to write provocatively by way of succinct formulations. Many of his
most crucial conceptual daims can be captured in short assertions, such as the
eleven words that make up this tirst quotation. And this line contains a number
of absolutely central elements to Rancière's thinking. First, the statement
refers to "parties," that is, the democratic subject, the subject of politics - a
category on which 1 elaborate below. These parties make a "dedaration," a
point that shows the centrality of language to Rancière's thinking. What they
dedare is "a wrong," and on the very same page Rancière says that wrong
"belongs to the original structure of aIl politics" (Rancière I999: 39). To
"dedare a wrong" is the essential work of polities itself. As 1 have identified
the elements here, one would think they combine quite easily into a simple
formula: subjects act politically. In other words, the democratic subject, in
dedaring a wrong, brings about polities; polities occurs because of this sub-
ject's actions. This formula sounds simple, but Rancière's first quotation above
completely unbalances the equation.
Rancière combines the various elements (subjects, action, and polities)
quite differently, for he brings in a temporal dimension that throws the formula
utterly off its axis. For Rancière, there can be no democratie subject to bring
about the political action - to create the politieal moment - because the daim
of the quotation above is specifically that such subjects do not exist before
politics. The "dedaration of a wrong" whieh is but another name for politics
in Rancière's framework brings about the subject ofpolitics, that is, "parties."
But how, we might immediately ask, can there be a dedaration with no subject
to do the dedaring? In Chapter Three 1 offer a detailed response to this ques-
tion, and in a certain sense 1 center my overall reading of Rancière on this
point. Here 1 want only to note the paradoxical nature of Rancière's seemingly
simple daim. With this short assertion, as in so many other places in his works,
Rancière makes a daim that simply seems impossible. 1 want to emphasize,
though, that he does not just make an impossible daim; he makes a claim for,
or of, the impossible.
And as with aimost aIl of Rancière's arguments, both the statement's "impos-
sibility" and its significance take shape against a particular context - in this
case, the context of tradition al approaches to modern political thought, in
which politics always starts with a subject. ln liberal theory, we begin by going
back to a state of nature, and we do so in order to locate a particular account of
the subject, to describe the characteristics of that subject within the state of
nature, a11 so that we may later offer an explanation for the origins of poli tics

16 1 The Lessons of Rancière


based upon the actions of that Traditional
counts of Marxism work with much the same structure: it is the human labor-
îng animal 1988), or the who is the source-point and the
for politics. A more structural and less humanist account of Marxism would
work similarly to the extent that the political agency of Marxist social transfor-
mation rests on the epistemological primacy of the proletariat. If we turn to
less orthodox accounts of poli tics, or better, if we move from the modern
period to the contemporary, we certainly have a celebrated relaxing of the
centrality of the subject, along with a whole host of efforts to decenter or de-
construct that subject. Nonetheless, many even so-called poststructural ac-
counts of poli tics today still refuse to accept the possibility of the impossibility
that Rancière names when he caBs the subject an "in-between." Even if the
subject has been de-centered by discourse or power, or the conditions of post-
modernity, many still read Michel Foucault, for example, as centering political
action on the subject. With these readings of the later Foucault,24 we have the
critique that the subject gives itself, and in this formulation, while the subject
may resist rather than create de novo, poli tics still starts with the subject. One
might extend this logic to Hannah Arendt's understanding of politics as action
in concert; her account multiplies the subjects who act, but those parties cer-
tainly seem to exist prior to and as a condition of possibility for politics.
Hence whatever Rancière might intend by the daim, "Parties do not exist
prior to the dedaration of a wrong," it certainly contrasts and conflicts with
traditional accounts of politics. Somehow Rancière means to convey the sense
of politics not so much without a subject - since the subject of politics, the
demos, proves to be an ineradicable element of Rancière's account - but with-
out a subject that cornes first, without a subject before the poli tics. This daim
that the subject does not precede politics, but rather the reverse, that poli tics
brings subjects into being - this is also a daim that prizes apart democracy and
liberalism. It does so with great force, by reconceiving and reimagining de-
mocracy and the democratic (and the politics proper to democracy) as a force
that does not start with a given, liberal subject. Rancière's daims about parties
not existing prior to politics serves to unravel and invert any state of nature
story, whether it be that of Thomas Hobbes, John Locke, or the twentieth-
century variant, John Rawls.
We cannot derive politics from any essential features of the human subject;
instead we must grasp politics as that irruptive force that brings about the sub-
ject. Rancière's subject is not just "decentered" but fundamentally dislocated.
lndeed, with the daims he makes in the quotations that open this section we
probably do better to read Rancière not as offering a specifie conceptualization
of the subject, but as refusing a them'y of the subject from the start. Rancière
consistently rejects, not just in his account of the subject of poli tics but in aIl
of his conceptualizations, the idea of an ontopolitical depth to his polemical
engagements. Rancière rejects aIl ontology.

1NTRODUCTION 1 ']
[One] cannot make any deduction from a theory of being as being to
the understanding of politics.
(Rancière 200gb: 117)

The current trend has it that you cannot think politics unless you trace
back its principles to an ontological principle .... My assumption is that
such a requirement leads to the dissolution of politics.
(Rancière 2011C: 12)

have already indicated that Rancière's thinking of politics pivots on a pri-


mary break with orthodox Marxist approaches, with liberal accounts, and also
with attempts to bind politics to ethics and the ethical turn. 25 However, none of
this ne(;essarily makes Rancière's writings unique, since many pluralist, post-
foundational and radical-democratic approaches to politics also rest on this
series of rejections or denunciations. And Rancière's understanding of demo-
cratic politics resonates on numerous levels - sometimes in powerful ways -
with these latter theories of politics. Rancière himself tends to eschew any
links between his accounts and those of other contemporary theorists, and
many a reader of Rancière has encountered resistance (from Rancière or from
Rancièreans) when trying to draw connections between his work and others.
At various moments in the chapters of this book 1 will consider sorne of those
possible connections, while also assessing Rancière's own hesitancy toward
them. Rancière tends to present his approach as singular, as utterly unique.
1 am not always persuaded by this move on Rancière's pmi; it may be, 1 sug-
gest, more rhetorical than factuaL Nonetheless, with the quotations that head
this section, Rancière makes a claim that proves absolutely essential to his
thinking and which surely does mark his distance from many other contempo-
rary accounts of politics. In fact, this issue of on toI ogy proves to be a sticking
point for many readers of Rancière who otherwise prove sympathetic and weIl
disposed to his account of politics - those who find his writings alluring and
persuasive and are attracted to his creative thinking of poli tics.
That sticking point can probably best be analyzed by looking at an impor-
tant argument advanced in the late 1980s by William Connolly and then re-
worked and refined by Connolly and many others on numerous occasions
since then. 26 In 1987 Connolly argued that it was impossible to conceptualize
politics without calling on an underlying "social ontology." He defined a social
ontology as "a set of fundamental understandings about the relation of humans
to themselves, to others, and to the world" (Connolly 1987: 9). One cannot
theorize politics, cannot think politics rigorously, without that thinking linking
up with calling upon, just as it also works upon, in a back and forth move-
ment between two registers - a set of ontological commitments. Later, Connolly

18 1 The Lessons of Ra ncière


Ontopolitical interpretation does not so much name a type of interpretation as
it caUs attention to the ineliminable fact that aIl poli tic al interpretation
terpretation) has a dimension of the ontos, the ancient Greek word for
(Connolly 1995: 16). For Connolly, the question is not whether or not to in-
dude the ontological within work on the poli tic al. Put simply, that is not a
choice we face. The ontological is required; it is ineliminable. As Patchen
Markell puts it, "A certain kind of attention to ontological concerns is a neces-
sary component" of aIl investigations of concrete political contexts v ....... U",'-"H

2006: 30; see White 2000; cf. Chambers and Carver 2008). Thus, when it
comes to writings in political theory, the real difference is not between onto-
political and non-ontopolitical works, but between "those who suppress the
'onto' in political interpretations" and those who engage the "onto" in distinct
ways (ConnoIly 1995: 9). This daim produces a critical perspective: anyone
who theorizes politics while eschewing ontology is really only concealing the
ontological dimension that his or her work necessarily secretes. 27 Again, ontol-
ogy is ineluctable.
Rancière is not unaware of these SOlts of arguments, as he makes clear in the
second quotation ab ove, where he suggests that the focus (by Connolly and so
many others) on the "onto" leads not to a deeper or fuller account of politics but
to just the opposite: a dissolution of politics. Rancière worries that in the hands
of political ontologists, ontology takes on the role of politics and starts to do the
latter' s work. He gives a clear example in the context of Hardt and Negri' s writ-
ings. Rancière glosses their debilitating political ontology as follows: "The
Multitudes are the real content of the empire that will explode it. Communism
will win because it is the law of being: Being is Communism." Rather than an
analysis of historical instances of political enactments, we have instead a "meta-
physical destination" (Ranci ère 201 IC: 12). Where Connolly sees the necessity
of the ontopolitical to aIl interpretations of politics, Rancière sees a thorough-
going resistance to ontology as a requirement of thinking the specificity of
actual politics. What Connolly considers an honest and indispensable articula-
tion of the ontological commitments that underwrite any politics, Rancière un-
derstands as the dissolving of politics into the ontological mixture. 28
We might even say that, for Rancière, to think politics aright, to grasp poli-
tics in history (and distinguish it from police) we must make a de-ontologizing
move - by refusing to deduce poli tics from ontology so as to take account of
politics in history. Ieven (2009) helpfully caUs this an "abstention from ont01-
ogy," since the only way to take a non-ontological position on ontology is
simply to abstain. Rancière repeats this de-ontologizing move, this abstention,
both in his polemical formulations of poli tics, where his bare, succinct, po-
lemical style resists any grounding for his concept of politics, and also in his
now very numerous meta-theoretical remarks, where he marks his distance
from an ontology in the clearest language he can muster. As the quotations

1NTRODUCTION l '19
to that is stark. on this point
the refusaI to ontologize, the rejection of an ontology - Rancière's approach to
poli tics may mark him not just as distinct, but perhaps even as unique. As
""-''-'''.1
<.CH UJ. HArditi has recently suggested, all of the various competing (and very
much conflicting) strands of post-foundational contemporary political theory
share sorne sort of commitment to (what Connolly caBs) the ontopolitical di-
mension. Here Rancière proves to be the exception; instead of accounting for,
or at least acknowledging, the ontological dimension or assumption of his
work, Rancière continually resists, refutes, or rejects the ontological, as the
occasion requires (Arditi 20I I).
ln my reading of Rancière's position on ontology, 1 emphasize not the
uniqueness of this daim as Rancière makes it rhetorically, but the centra lity of
this position to Rancière's broader body of work. In other words, the resistance
to ontology cannot be full Y grasped if we understand it as merely a resistance
on Rancière's part to have his work categorized in particular ways. Rather, the
refusaI of ontology proves essential to man y of Rancière's other arguments
(arguments that on the surface might appear to have nothing to do with daims
about ontoIogy). That is, many of Rancière's specific daims about politics,
about equality, about language, require and depend upon his refusaI of ont01-
ogy. His account of the subject, as discussed above, provides one such ex am-
pIe. For Rancière, there can be no ontological account of the subject that
grounds or conditions the activity of politics. Rather, it is the activity of poli-
tics, and only this, that indicates to us how we should come to understand the
subject. Any effort to take Rancière's statements about subjectivation29 as phil-
osophical grounds upon which to build a particular political project will stum-
ble upon his insistence that there is no ground to be found - there is no there
there. This is not to suggest that Rancière's approach to politics is groundless,
but rather to show that grounds are always political, historical, or contextual -
never ontological. Insisting on this point makes it easer to see that when
Rancière asserts that subjectivation is always and can only ever be disidentifi-
cation, his dedaration only makes sense within the context of anti-ontology
(Rancière I999: 36). Subjectivation has no grounding in being. The political
subject can never be the bearer of a politics, since the subject only emerges by
way of, through, and as articulated above, after poli tics.
ln taking his radical position against ontology, Rancière also thwarts those
efforts to appropriate his work by ontologizing particular daims that he makes.
His disavowal of ontology sets a certain bar for his readers, because Rancière
implicitly suggests that you can "ontologize" his work only by way of a poor
reading of it. In this way he seeks, intentionally or not, to foredose the possi-
bility of ontological interpretations of his work. His very intransigence on this
point means that he forces many of his readers to part ways with him. Ran-
cière's refusaI of ontology thereby expresses another dimension of his radical
pedagogy, as he makes it impossible to "follow him" except by breaking with

1 The Lessons of Rancière


him. As stresses, the abstention from '-'uC'V-L'V'j"-,

that its own share of problems for Rancière


to expand or further Rancière's thinking of politics - to consider
different contexts or elaborate distinct dimensions of it - may a
of commitment to ontopolitical interpretation greater than Rancière himself
would allow. discuss this phenomenon in more detail in Two, when
look at various ways that readers have attempted to offer such of
Rancière.
At this point 1 only wish to call attention to Rancière's own on his
claim that poli tics cannot be deduced from ontology and that the insistence on
"doing ontology" may lead to the disappearance of politics. He makes these
arguments consistently and repeatedly. Moreover, in the face of many efforts
to insist on Connolly's point - by showing that Rancière has no choice but to
make ontological commitments in his writings - Rancière refuses to back
down. Here 1 would stress that the steadfast nature of Rancière's refusaI to
ontologize politics points to a different, but absolutely crucial dimension of
Rancière's work: his fundamental commitment to historicity. Politics for Ran-
cière is never more or other than a political moment, a political event. Politics
occurs in history, and must al ways be understood in its historical dimension.
The historicity of poli tics, we might say, is one other reason that politics cannot
be traced back to a theory of being. If, then, Rancière always says no to ontol-
ogy, this is bec au se he always says yes to a certain priority of history, and he
does so, in tum, because of the very happening of historical events. If we
commit, as Rancière says we must, to the vigilance needed to witness those
events, then our rejection of ontology requires a concomitant tum to history.

History

Politics is always emplotted in historical configurations.


(Rancière 200gd: 287)
Historical method does mean something; 1 am not saying it does not
mean anything. It means you have to be located in this place, because
this object is social history.
(Rancière and Dasgupta 2008: 72)

The first "history" that matters under this heading is that of the publishing and
translation of Rancière's writings - a history that follows a unique and signifi-
cant path. To give a sense of that history, 1 break it out into five separate de-
cades (1960s, 1970s, 1980s, 1990s, and 2000S) and describe it as something of
a curve function, one that follows this formula: the earlier the work was pub-
lished in French, the later it is translated into English. In other words, and

1NTRODUCTION 1
to sorne
n'VY,,,,,I',h"'Mr~ we can map the r.ll!',I,r'"h,n1"\c and transla-
tions over these five decades as follows:
Rancière's one essay from the 1960s, despite its central importance to
Rancière's oeuvre, has not been properly published in translation at all. 30
His first book, from 1974, where he makes his crucially important and
definitive break with Althusser, was only published in translation in late
2011. As noted ab ove, Rancière's numerous writings for Les révoltes
logiques are only now beginning to appear, with Volume 1 published in
2011, and Volume II forthcoming in 2012.
Rancière's works [rom the 1980s were translated, very unevenly, and
sometimes roughly, over the same decade as they first appeared in print
(while the oider writings remained untranslated during this period). La
nuit des prolétaires appeared in French in 1981 and in English in 1989
(with a new English translation just out in 2012); Le philosophe et ses
pauvres was published in French in 1983, and while an English transla-
tion was promised in the mid-1990s it did not appear until 2004; and fi-
nally Le maître ignorant appeared in French in 1987 and arrived in
English only four years later. Numerous other texts from this period, of
course, remain untranslated today.
ln the 1990s a much larger number of Rancière's numerous publications
were translated into English, aIl within five years of first appearing in
French. Other translations appeared later than this, but the main point is
that in this decade Rancière's works start to appear with frequency in
English, and the English translations that did come out in this decade
were versions of French texts published very recently. Thus, this decade
marks the first time that English translations become both common and
current.
& ln the 2000s the majority of Rancière's works were translated very quickly
into English, and a number of new texts were produced directly in English.

One can see clearly th en that as we move forward through the decades, we go
from no translations or long-delayed translations, to frequent and timely
translations. After the 1990s, an even lm'gel' number of Rancière's works
appear regularly in English soon aftel' their original publication in French.
Moreovel', as we continue to go fol'wal'd in time, there is a simultaneous
reaching backward to produce translations of Rancière's earliest writings. As
of my own writing early in 2012, we have reached far enough back in time
that Althusser's Lesson, Rancière's first book, has just now finally appeared.
1 am not in the business of prediction, but if the pattern holds, we can expect
a new translation of Rancière's contribution to Reading Capital to appear
within the decade.
Now, aIl of this matters on the one hand simply because it situates my read-
ing of Rancière in the context of his reception in the UK and North America,

22 1 The Lessons of Rancière


and it to add sorne contour details to the map
here. But this publishing history has partieular significance for understanding
Rancière's own relation to as a field, a topie, a theme. Rancière first
made a splash (and a big with his writings from the 1990s, the writings
r\l1'r'lh"pllu "on politics." These texts were surely more amen able to translation

into other politieal and intellectual contexts, since in these works Rancière
engages with the tradition of politieal philosophy and with contemporary poli-
tics. These writings contain what many now think of as Rancière's trademark
style of sparse, often abstract, and commonly polemical formulations, and as
such, the writings themselves are easy to quote, easy to use in provocative
ways, and easy to reappropriate for a variety of distinct contexts.
Rancière's writings from the 1970S and 1980s, in stark contrast, were in one
sense "historical" writings, and they were deeply embedded not only in their
specific historical context, but also in a partieular French context. That is to say,
Rancière's work from this time came directly out of the archive, and while it
spoke weIl beyond that archive, it often spoke to specific French political de-
bates. Thus, to grasp what Rancière has to say about history requires a bit of
historical contextualization on the part of his readers. To be dear: it should be
obvious that 1 do not intend to apply to Rancière's thought the sort ofhelmeneu-
tics often used in the history of political thought. Nonetheless, 1 contend that we
can gain a crucial perspective on Rancière's thinking of democratie politics ifwe
do more than is often done to situate him in his context. One of the limitations to
be found in so many accounts of Rancière's work - induding, at times, his own-
is that they seem to come from nowhere and to be rooted to nothing. Rancière is
a polemicist. He writes so as to provoke, and especially in his writing on democ-
racy and politics, Rancière produces radical formulations. He thus makes bold
and startling assertions that often appear as if out of the blue; their suddenness
enables these daims to strike readers with real force. However, an engagement
with Rancière's writings needs to do more than merely repeat those formula-
tions; in writing on Rancière, one cannot simply rest with Rancière's own po-
lemical force. 31 It is not my goal here to repeat what Rancière already says, but
rather to extend his arguments through my own engagements, and to stage my
engagements in such a way as to make Rancière's thought resonate more broadly
and more powerfully for the readers of this text.
It is in this sense that context matters in reading Rancière. For me, no sec-
ondary source on Rancière has brought his understanding of poli tics into focus
nearly so weIl as Kristin Ross's book May '68 and its Afterlives (2002). In
making this daim 1 do not mean to rank books on Rancière or to offer sorne
filter on the secondary literature (e.g., read this one, not those). 1 have no gen-
eral interest in such an exercise, but it surely has no place with Ross's book,
since her text is not even a book on Rancière and does not constitute part of
that literature. Ross's book operates on one level as a history of May '68, but
it does its most significant work on a second level, wherein it engages the

1NTRODUCTION 1 23
contest '68 in the I970S and
up to the turn of the century. Ross does not try merely, or even necessarily at
aIl, to "get 11ght" in the historical sense; instead she makes her own
cal and theoretical argument about ' 68 so as to contest the dominant nar-
rative that has taken root. And the sense of ' 68 that Ross wants to defend
relies upon, just as it mutually informs, a thoroughly Rancièrean account of
poli tics. The themes of ' 68 that Ross argues for almost aU prove to be
central themes of Rancière's account of politics: challenge to mastery, disrup-
tion of order, verification of equality based upon its presumption, rejection of
the expert, political subjectivation as dis identification, and above aH, a certain
happening, a certain event-ness of history (Ross 2002: 6, 15,46, sI).
More than anything else, Ross wants to contest the interpretation of May
'68 that has become dominant in France in recent decades: the daim that
"nothing happened in May," that the event was a nonevent (Ross 2002: 6,
67).32 She challenges this interpretation by invoking (sometimes explicitly, but
frequently tacitly) a Rancièrean conception of politics. As a historian, Ross
herself does not polemicize, and she allows her larger daims to unfold through
her historical narrative. However, her argument throughout the book can be
conveyed by a statement that she makes early on, an assertion that takes a more
polemical fOlm. Against those who say over and over again that nothing hap-
pened in May '68, that there was no event-ness, no emplotment of politics in
history, Ross contends: "In May, everything happened politically - provided,
of course, that we understand 'politics' as bearing little or no relation to what
was called at the time 'la politique des politiciens' (specialized, or electoral
politics)" (Ross 2002: 15, emphasis added). Ross uses Rancière's arguments
explicitly to show that the daim "nothing is happening" is a form of policing.
She quotes from Rancière's well-known reinterpretation of the Althusserian
scene of interpellation, wherein Rancière derives from the scene an utterly dif-
ferent meaning than Althusser had produced. Rancière states: "The police say
there is nothing to see, nothing happening, nothing to be done but to keep
moving, circulating; they say that the space of circulation is nothing but the
space of circulation" (Rancière 2001: par. 22; quoted in Ross 2002: 22, Ross's
translation). Ross uses Rancière's account of politics to resist a peculiar and
potent policing of May '68 that would render it a nonevent. She reopens the
dispute in order to pro duce a scene of politics. In so doing, Ross's historical
work provides one of the most powerful elaborations and illustrations of
Rancière's thinking of politics, while at the same time both illuminating and
contextualizing Rancière's thought.
Rancière's work, however, cannot be reduced to an expression of, or reflec-
tion on, a set of historical events, whether they be May '68 or any other. My
point in highlighting Ross's work and in emphasizing the importance of May '68
for Rancière's thought is much more delimited: 1 wish to indicate at times the
location from which Rancière writes, and to draw sorne tentative links between

24 1 The Lessons of Rancière


his historical '68. uses
, 68 slogan as its epigraph, not to make the chapter speak only to that event,
but to show the traces of a connection to it. In a broader sense mine is an effort
to take Rancière's daims that can only ever be histori-
and that historical method means nothing less nothing
have to be located in this " Rancière always writes from
somewhere, even if he writes in such a way as to mask his location. In using the
archive to produce polemic, Rancière once again takes a distinct approach to
common categories and concepts. With respect to the category of "history" it
seems obvious that Rancière is not doing history of political thought and that he
is also not "abstracting" from history in the same way that analytic thinkers and
liberal theorists often do. At the very same time, Rancière's understanding of the
"event" remains distinct from a Deleuzean approach, given Rancière's commit-
ment to the archive and given the rich material that he draws from it. 33
When we turn to history, what we ultimately see in it are moments (or poten-
tial moments, or arrested moments) of politics. These moments, however, across
radical di fferences , are always marked, according to Rancière, by a meeting of
two logics: the logic of equality and the logic of inequality. It is in history, as
Rancière indicates in the quotation below, that we can locate the Imot that ties
inequality up with equality. The vigilance about history that Rancière demands,
the attentiveness to history that his approach requires, and the commitment to
untimeliness and historicity that are its implications - this treatment of history
allows us to get to, to get at, what Rancière caUs equality. As Rancière will repeat-
edly daim, we must understand equality not as given - and surely not as ontologi-
cal in any sense - but as historical verification. This vigilant approach to history
also means that Rancière takes a distinctive approach to equality. As 1 will now
discuss, Rancière argues that we must assume equality as a methodological
guideline in order to help us see its verifications and its demonstrations.

uality

A verification of equality is an operation which grabs hold of the knot


that ties equality to inequality.
(Rancière 200gd: 280)

If equality is axiomatic, a given, it is clear that this axiom is entirely unde-


termined in its principle - that it is anterior to the constitution of a deter-
mined political field, since it makes the latter possible in the first place.
(Rancière et al. 2000: 6)

No term better crystalizes the differences between Rancière's account of poli-


tics and the liberal account than "equality." Equality is a key term to help

1NTRODUCTION 1 25
BMOS84732
"""JHHé'-'-"_'" Rancière's thinking from liberalism because both ap-
proaches to politics make equality central and essential. Indeed, for liberalism
'-'rl"'Cl"'''' is so important that even authors who defend non-liberal political vi-
sions if they articulate a liberal conception of equality, a role to play in
the liberal tradition. With this description l am obviously thinking of Hobbes,
who defends an authoritarian vision of politics in his writings, just as he sup-
pOlted monarchy in his life. But Hobbes plays a fundamental role in the liberal
tradition because he starts with equality. Undoubtedly Hobbes makes one of
the clearest and most direct arguments in favor of equality in the en tire history
ofpolitical thought, when he claims that in astate of nature aIl men and women
are equal because even the weakest among them still has within him- or herself
the power to kill the strongest (Hobbes I994: chapter XIII).
From Hobbes onward, equality plays a foundational but at the same time
quite curious role within liberalism. It is foundational in the most straightfor-
ward s~nse: equality, paired with liberty, is the starting point of liberalism.
Liberal subjects are, by definition, all free and equal. To consider the human
being in his or her fundamental capacities and characteristics means to con-
sider human beings as equal, and everything else follows from this starting
point. We see this most strikingly in the linear logic of Locke's argument:
Locke famously tells his readers, on just the second page of his Second Trea-
tise that "to understand political power aright, and derive it from its original,
we must consider what state aIl men are naturally in" and that is a state of
freedom and equality (Locke I988: 266). However, l caB the function of equal-
ity in liberalism curious because, especially after Hobbes's very cIever argu-
ment, there is almost no argument whatsoever given to support this assertion
of equality.34 Equality is the given, it is the starting point, it is a substantive
requirement of liberalism, but it is not, for aIl that, a principle that liberalism
spends much time defending or even articulating.
This point cornes through most starkly in the work of Rawls, who really
makes no argument for equality at aIl. A Theory of Justice, for example, simply
stipulates on its tirst page that "each person possesses an inviolability founded
on justice that even the welfare of society as a whole cannot override" (Rawls
197I: 3). Equality is caught up in this "inviolability": we are an equal to just
the extent that we are inviolable. Rawls makes it cIear- as he asserts it repeatedly
that humans must be characterized, in their very being, as free and equal, but
Rawls always moves directly and rather quickly from this assumption to the
question that matters to him. He frames the question this way on the second
page of Political Liberalism: "How is it possible for there to exist over time a
just and stable society of free and equal citizens?" (Rawls I993: 4).
For liberalism, equality is best understood as a ground in the fullest sense
of the word. That is, Iiberal political theory will start with equality, but it will
not start with an argument about equality. And this means that equality is a
ground Ilot only in the sense that it serves as the basis for the theory, but that it

26 1 The Lessons of Rancière


must somehow
start. Equality is the condition of possibility for liberalism, but liberal
cannot establish it. Because the needs equality in order to the
way to begin is creating conditions of without establishing
equality as such. This extra-theoretical sense of equality as to
explain the role of the state of nature and similar devices, aIl of which indicate
the extent to which equality is both an essential and necessary part of the
theory, and yet somehow outside of it. Equality derives from the state of nature:
we locate it as a characteristic of human beings in that state and
Locke), or we model that state with a philosophical device like the original
position (Rawls), a device premised upon equality. at the same time,
and in a different sense, equality remains confined to the state of nature and
excluded from politics. In liberal theory, equality is a philosophical groundfor
politics, but politics is not really about equality. For liberalism, equality is not
at stake in politics. Hence, when Rawls discusses equality he is never talking
about what goes on in politics, but only about what makes politicalliberalism
possible. Equality functions as a given ground for a theory of justice, or lim-
ited government, because it is embedded in the framework of liberalism as a
characteristic of the human being. This makes equality substantive (in a philo-
sophical or ontological sense), but also somehow utterly unreal (in the sense of
political strife, conflict the stuff of politics).35
This sketch of the liberal understanding of, and approach to, equality, malœs
possible a much starker illumination of Rancière's account both of equality
and of his overall project. Like liberalism, equality is lodged at the heart of
Rancière's work, and it is repeatedly invoked by his writings. But to the ques-
tions of how equality gets there and how it operates, Rancière gives utterly
non-liberal answers. He rejects almost aIl other imaginable approaches to
equality. Thus, equality is neither a foundation nor a ground; it is not a sub-
stance or a goal. Furthermore, politics does not occur because of equality, and
poli tics cannot and does not seek to achieve equality. While it cornes as no
surprise that Rancière dismisses the idea of ontologizing equality, we can see
here how thoroughgoing is his resistance to any and aIl tendencies in this
direction. We do not experience equality as autonomous individuals, and
equality is not a substantive achievement of political communities. Also unsur-
prisingly, but still importantly, we can never measure it.
From Rancière's perspective, we might say that the question "What is
equality?" is simply misplaced or misguided, because this question takes the
"what is" form (ti estin, in ancient Greek) and thereby already gestures toward
the ontological realm. Rancière's understanding of equality can be better
captured if we completely avoid formulating it in an abstract philosophical
language. For Rancière, equality is a polemical term, not an analytical con--
cept. 36 Here then are sorne formulas that help to triangulate the sense, the
structure, and the function of equality in Rancière's thinking: we assume

1NTRODUCTION 1 27
serves as an
,:>rlll'=' 11tU equality is demon-
strated. As he does in many other places, in the second quote, above, Rancière
describes as axiomatic. Equality is an axiom that we assume and that
can be verified in practice. This assumption, however, is not a substantive
in the liberal sense, but rather a presupposition that could never take
'VU.'l'-'-Ul':;;'

the form of a "given" in a logieal proof (cf. Wittgenstein 1972). And the veri-
fication or demonstration of equality cannot be substantively grasped; we can
bear witness to equality, but we can never hold it in our hand or point precisely
to iL Equality, for Rancière, just is not that sort of thing.
If equality is anything, it is a 10gicY This explains how equality can be both
assumed and verified without ever being made substantive: both moments
occur when the logic of equality is operative. And the logic of equality never
ODlôrates on its own, in isolation, or in pure form. Rather, the logic of equality
operates in tension with the logic of inequality. Politics occurs only when these
two lügics meet, when they come into conftict. This is what Rancière means
when he says, in the tirst quote above, that equality is veritied through a pro-
cess that grabs hold of the knot that ties equality to inequality. Equality and
inequality always remain knotted together. Equality does not reside in or create
its own space; there is no sphere in which pure equality reigns. Instead equal-
ity only ever contaminates the space of inequality - what Rancière caUs an
order of the police. The police order is an order of domination and of hierar-
chy, and it operates according to its own presumption of a logic of inequality.
Polities occurs when that poliee order is interrupted by the heterogeneous
logic of equality. This makes equality a "one-off act" and it explains why
equality can never be institutionalized: "Equality tums into the opposite once
it aspires to a place in the social" (Rancière 1999: 34).
Equality in Rancière's account al ways proves to be equality of intelligence,
an equality of anybody and everybody. Here is Rancière's own description of
the presupposition of equality: "the mad presupposition that anyone is as intel-
ligent as anyone else and that at least one more thing can always be done other
than what is being done" (Rancière 20Ioa: 2, emphasis added). The emphasis
here is significant: equality is not a rational or logieal given, it is a "mad pre-
supposition." We do not start with equality as a given; we take a leap to arrive
at it as axiomatic. In this description, we see a supremely democratie form of
equality; it is not a characteristic of certain individuals under certain condi-
tions; it is an assumption that applies to anyone at aIl. Rancière's polities of
equality is always a polities of the anybody at aIl, of the anyone and everyone,
of the anyone whatsoever -- phrases that he uses interchangeably and with
great frequency across a wide range of his writings. 38
In this light we can see that liberalism gives equality a certain foundational
of place, making equality an organizing principle and a substantive goal,
whereas Rancière tums to equality as only someone who is tirst and foremost
a democrat cano For Rancière equality is not an ordering principle but instead

The Lessons of Rancière


a mechanism of that caUs any ofUVJlHHIUL,lVH

question. The axiomatic manifestation of equality thwmis a11 police


even liberal orders that would make fundamental. From Rancière's
perspective, we can rewrite the history of liberalism, showing that the move
from Hobbes's argument for to the lack of such an in Rawls
is also the move designed to create a system of politics based on consent (and
legitimate power) that keeps people in their place. Liberal theOl'y upholds this
particular (liberaI) form of police order, and in this process it strives to elimi-
nate (by controlling and constraining) the real force of politics. As discuss
further in Chapter One, in this sense liberalism is what Rancière calls "parapol-
itics": an effort to eliminate la politique by cordoning it off into one small
corner of a larger philosophy of o l'der. 39
Rancière's commitment to equality works hand in hand with his commit-
ment to politics. Whereas liberalism excludes equality from politics, Rancière
says that equality only appears in political moments. He explains: "Equality is
what 1 have called a presupposition. It is not, let it be understood, a founding
ontological principle but a condition that only functions wh en it is put into
action" (Rancière 2006d: 52, emphasis added). With this claim Rancière
widens the gap between his understanding of democratic politics and the ap-
proach favored by liberalism, since for him equality can only make sense
within the context of politics - and this means, in turn, within the context of
history. Here again, Ross's vigorous arguments in favor of a particular inter-
pretation of May' 68 can help to illustrate Rancièrean principles. Ross contests
the post-'68 interpretations that projected onto May '68 a narrow set of claims
for libelty; she does so by suggesting that we best interpret the events of May
'68 in terms of equality. But we must understand equality in a particular way,
a particularly Rancièrean way: "1 mean equality not in any objective sense of
status, income, function, or the supposedly 'equal' dynamics of contracts or
reforms, nor as an explicit demand or a program, but rather as something that
emerges in the course of the struggle and is verified subjectively, declared and
experienced in the here and now as what is, and not what should be" (Ross
2002: 73-74). And what "emerges," Ross stresses, is the central ide a oflinking
"intellectual contestation with workers' struggles" - that is, with the day-to-
day and on-the-ground activities of May '68 in which workers and students
formed a new stage of politics. This is a verification of equality as "part and
parcel of action" (Ross 2002: 74).
This particular understanding of equality thereby links up aIl three of the
previous concepts 1 have discussed. We cannot take equality as an ontological
ground, and we must actively resist efforts to ontologize it. This means that
equality only makes sense in the historical happening of politics. Moreover,
equality is never the property of a subject; it is not, contra liberalism, a charac"'
teristic of the political subject. Just as the declaration of a wrong is not some-
thing enunciated by political subjects but rather something that brings such

1NTRODUCTION 1 29
in similar fashion the of can be as the
condition for emergence of a political subject. And with equality my argu-
ments in this introduction spiral up, around, and through the broader theme of
"Rancière's lessons," since Rancière's account of equality initially emerges
most perspicaciously in his writings on pedagogy.
Rancière's critique of mainstream and dominant pedagogies from the nine-
teenth century to the present always centers on those pedagogies' presupposi-
tion and verification of the principle and logic of inequality. The work of the
teacher, traditionally understood, is to explain, to explicate, to tell the students
the lesson, to show them the meaning of the text: "Explanation, or the ordinary
routine of pedagogic practice, was above aIl a display of inequality" (Rancière
20I rb: 40). The master is the master because he or she can explain the texts
that the students are otherwise presumed not to understand on their own; thus,
teaching in this traditional sense presumes, just as it simultaneously demon-
strates, the inequality of intelligence between student and teacher (Rancière
201 lb: 40; Rancière 1991: xvi). Rancière caBs this entire framework that tra-
ditional teaching establishes "the explicative order," an order built on the
ground of the student' s inabilities. Explanation lies at the heart of the explica-
tive order. But just as he williater redefine the normal activity of politics as
police practices, Rancière had much earlier redefined explanation as stultifica-
tion. The argument pivots on a basic but powerful observation: "To explain
something to someone is first of aB to show him that he cannot understand it
by himself" (Rancière 1991: 6). Before any teacher opens his or her mouth,
the explicative order establishes the teacher as the master, as the one who will
help those to understand who cannot understand on their own.
Rancière's radical pedagogy, which he freely bOITOWS from Joseph Jacotot,
seeks the overturning of the explicative order. This new pedagogy rests on
nothing more than a reversaI of the explicative order's primary assumption.
What if the student can perfectly weIl understand for himself? What if the
student can read the text without the explanations of a master? It is the prin-
ciple of equality of intelligence - nothing more and nothing less - that unrav-
els the explicative order and founds a new, radical pedagogy. When a student
picks up a book and reads it for herself (even, as in the case of Jacotot's teach-
ing experiments, a book written in a language other than her mother tongue),
she is using the method of equality. Thiscapacity for anyone to read the book
without having someone else tell him what it means - this is the power of
equality, and this is aIl there is to equality.
What, then, does it me an to teach without stultification? How can there be
any "lessons" at aIl under the sign of a Rancièrean pedagogy? And ultimately,
what does it mean to title this work The Lessons of Rancière? Answers to these
questions begin to emerge when we note that Rancière does not give up on the
teacher or on teaching. He even, somewhat surprisingly, maintains a space for
masters and mastery. Rancière insists on the principle of equality of intelligence;

30 1 The Lessons of Rancière


this means that operate the of a teacher's
gence and a student' s ignorance. We assume the student is intelligent, and that
presupposition is the basis for our put more polemically, we
assume the of the teacher. The title to Rancière's book on "' ....
"'VL'-"L,

The Ignorant Schoolmaster, is not meant as a critique of cunent teaching (that


title would have been "The Stultifying Schoolmaster") but as a possibility.
Jacotot shows the promise of teachers who teach despite the fact that they do
not know - or teachers who teach as if they do not know. Nevertheless, the
teacher has to teach, and mastery has a place in Rancière's radical pedagogy
because the teacher is the one who imposes a mastery of will over his or her
students, creating an environment, a context, and a structure in which they can
learn (Rancière 1991: 13).
Therefore we must approach Rancière's teachings not with the assumption
that he knows something we do not. We must read Rancière not for a glimpse
into his superior intelligence or for the provision of an explanation of texts and
things that we cannot understand ourselves. We must operate instead under the
principle of equality of intelligence. But this do es not mean there is nothing for
us to learn. The "les sons" of Rancière are the lessons of an ignorant school-
mas ter and therein lies their radical significance. This book provides a struc-
ture and a context in which readers can grasp Rancière's teachings. More than
this, the book stages my own specific interventions and develops my own par-
ticular daims. Thus, what 1 caU The Lessons of Rancière retains a cruciallink
to Rancière's writings, but also goes beyond them. In the next section 1 give an
initial sketch of that structure.

G Coordinates

The Global Positioning System has numerous advantages over tradition al


maps, but the foremost difference for my plll-poses here - which is to trace the
basic structure of the book so as to orient readers and provide them with a
distinct sense of the path the chapters will follow - and for my metaphor, is
that GPS maps go beyond the spatial dimension. GPS provides both location
and time information. 1 have stressed at the out set what 1 will repeat immedi-
ately in the opening chapter: Rancière's writings aIl contain a crucial temporal
dimension. To render his thought static or formaI, to rob it of its temporal di-
mension, is to skew his arguments, to sap them of their polemical and political
force. Therefore, as 1 do in this section, throughout the book 1 consistently
avoid drafting a bluepIint for Rancière' s thought~ 1 always resist rendeIing his
writing and thinking stable or stationary. Instead, 1 mobilize dynamic readings
of a range of topics and engage those topics by way of selective appropriations
and creative interpretations of Rancière's own texts. Each chapter of the book
therefore centers not on a lesson as an explication (which would stultify) but

1NTRODUCTION 1 31
on a lesson as an up to other - as a emancipa-
tion. While always inspired and animated in some way by Rancière's writings,
the specifie engagements with ideas, authors, arguments, and contemporary
political questions are mine, not Rancière's. Therefore the book's foundation
lies in a series of concrete engagements with the particular problematics in
which find myself. As 1 discussed above, for Rancière history means that
"you have to be located in this place." The chapters of this book remain atten-
tive to radical temporality in their reading of Rancière, but more than this, they
retain their own temporal markers. They do not speak to a comprehensive
"political theory" that 1 would attribute to another author; they speak to the
time and place in which they are located, though they almost always speak to
that historical moment through the writings of Rancière.
In Chapter One, 1 start where so many readers of Rancière have started,
with "Politics." Rancière has animated and incited numerous readers, espe-
cially political theorists, with his radical rethinking of politics and his atten-
dant redefinition of what we usually caU poli tics as "police." In this chapter
1 insist that whatever else we might wish to say about Rancière's understand-
ing of politics, we must always refuse the idea of co-opting his account under
a model that renders poli tics pure. Even some of the best readings of Ran-
cière's polemical political works have failed to take account of the consistency,
subtlety, and dedicated nature of his resistance to the idea of politics as pure.
Because he gives a unique definition of politics (in contrast to police), many
contemporary theorists have been tempted to read Rancière as something like
an Arendtian - in the sense of taking him to be carving out a unique space and
preserve for "the political" as distinct from "police," in the same way that
Arendt shelters politics from the social and the economic. This chapter thus
engages with a small portion of the vast literature on Arendt, looking specifi-
cally at a few important recent texts that read Arendt and Rancière together.
1 suggest that those thinkers who blend Rancière and Arendt usually do so by
folding Rancière's project into Arendt's, and in making this move they must
pass over or abandon some of Rancière's keenest insights about democratic
politics. Ultimately, 1 show that - despite the many ways in which their thought
is complementary - to read Rancière as an Arendtian is to miss his most fun-
damental of points about politics: that it can never be pure.
Rancière's insistence on an impure politics holds a number of important
implications, many of which 1 unravel over the course of the book. Pirst, the
notion of impure politics thwarts not only Arendtian readings of his ideas but
also dialectical renderings. In this chapter 1 engage with the most important
early interpreter of Rancière's political theory in English-Ianguage writings,
Jean-Philippe Deranty. 1 resist the tendency in Deranty's interpretation to as-
similate Rancière's conception of politics to a dialectical model. 1 insist,
with Rancière, that there is no ground for politics; in tum, 1 argue, against
Deranty, that poli tics and police do not meet on the terrain of a separate term

32. 1 The Lessons of Rancière


encounter the of
HIV'-j'-''''HlJ

and the politicallogic of equality occurs in the only location it can occur: the
terrain of the police order itself. This means that contrary to to sys-
tematize or purify Rancière's account of politics, democratic politics must
always be bound up with police orders. For just this reason, Rancière's think-
ing of democratic poli tics requires of his readers far greater attention to his
concept of "the police."
Thus, Chapter Two, "Police," cornes at Rancière's well-known writings on
politics from a different tack - in this case, from the opposite direction. Most
contemporary political theorists are so taken by Rancière's novel and radical
approach to politics that they immediately move on from his fundamental dis-
tinction between politics and police, so that they can investigate more c10sely
the theOl'y of politics that this distinction makes possible. But since Rancière -
as he himself says, and as 1 have stressed above has no theOl)' of politics, it
does not make much sense to try to build a theory by piling up Ranci ère 's
statements about politics. Rancière says what he says about politics not in an
effort to ontologize politics. Instead, he uses the police/poli tics distinction
both as leverage for his polemical interventions into both contemporary and
historical political debates, and as a way to reorient our understanding of the
history of poli tic al philosophy. This chapter, then, does sorne of the basic work
to lay out the terms that prove fundamental to Rancière's broader conception
of democratic politics. However, my reading of Rancière refuses to turn any of
those temlS into established or analytic concepts that would fonn the basis of
a new model of poli tic al theory. Instead, 1 focus in this chapter on what 1 call
"the poli tics of the police," a phrase that helps me to get at not only what Rancière
means by police, but also how he mobilizes the distinction in order to make his
own set of critical interventions into politics.
This second chapter thus details three different ways of reading the phrase
"poli tics of the police." First, the "poli tics of the police" makes it possible to
analyze and unravel the meaning of the phrase that Rancière uses to describe
today's putatively "idyllic state of politics[:] consensus democracy" (Ranci ère
1999: 95)· As Rockhill helpfully puts it, consensus is "a particular way of
positing rights as a community's archë" (Rockhill 2006: 83). The victory of
what 1 would describe as a liberal consensus (rights as the principle ofpolitics)
leads to what Rancière caUs, echoing Thomas Jefferson, a loss of needed "militant
vigilance" conceming democracy (Rancière 1999: 97). Amled with the police/
politics distinction, Rancière can demonstrate that consensus democracy - which
1 show works by the same logic as US liberalism, that is, the logic of interest-
group pluralism is not poli tic al at aIl, but rather a model of policing. Consen-
sus democracy is founded upon the very elimination of politics, on the quelling
of the conflict, dissensus, and rupture that are central to Rancière's understanding
of democratic politics. Instead, consensus democracy merely implements a
new "police order" - a hierarchical regime of domination that institutes and

1NTRODUCTION 1 33
instantiates a of of the is thus a cri tic al
theOl'y, first and foremost. 1 th en offer an extensive but circumscribed engage-
ment with Todd appropriation of Rancière's thought for his anarchist
political project 2008). holds an understanding of police that would
render the phrase "polities of the police" nonsensieal. show that in
his defense of anarchism as a pure poli tics (one that would eliminate police
once and for aIl) May not only misreads Rancière on police but ends up posit-
ing an approach to politics that proves less than fully democratie (in Rancière's
sense). "The polities of the police" helps us to grasp the inadequacies of a
utopian model of pure polities. Finally, then, 1 suggest an account of "politics
of the police" that always insists on their mutuai imbrication, that definitively
refuses the ide a of a pure polities, and that points us instead toward a Ran-
cièrean theOl-Y of radical democracy.
Chapter Three, "Literarity," returns to the beginning, as 1 open this chapter
at the same place as Rancière begins his most important text on polities and
philosophy. ln his I995 book, La mésentente, translated as Disagreement
(I999), Rancière begins with Aristotle's account, in his Politics, of man as a
poli tic al animal. Aristotle can traditionally be read as grounding a theory of
politics on anthropology: anthropos (man), for Aristotle, is that creature that
possesses logos (reasoned speech), unlike animaIs that only have phonë (mere
voice). The logos makes man a politieal animal by endowing him with the
powers of deliberation and judgment, the quintessential activities of politics.
Thus, Aristotle, the so-called first politieal scientist, grounds his theory of poli-
tics on a particular theory of language a tool to be possessed and wielded by
the human being, the only properly political creature. But Rancièrean polities
begins in disagreement with Aristotle by rereading these famous passages at
the opening of the Politics. Rancière asks how we can distinguish between an
animal with mere voice and a human being with logos. Is the phonë/logos
distinction given in nature (as most readers of Aristotle would have it), or is it
precisely polities itself that allows us to draw the line between phonë and
logos, between the political and the apolitical? If it is the latter, as Rancière's
approach to politics always contends, we are led to a completely different un-
derstanding of the relationship between language and the subject. The political
subject, as 1 have shown above, cannot precede language and use it as a tool;
rather, the politieal subject only cornes to be in and through language and poli-
tics. There is no pre-given subject who uses language to declare a politieal
wrong, because before the wrong, there is no subject for us to speak of. Demo-
cratie subjectivity cornes about through democratic politics, and not the re-
verse. Language can therefore no longer be thought as an object for human
use; it must be understood instead as that medium through which the human
politieal animal emerges and in whieh it crystalizes.
In making this argument, 1 do a great deal of clarifying work concerning not
only what Rancière argues about language, but also the language in which

34 1 The Lessons of Rancière


Rancière's have the how
through a series of mistakes in, or questionable choices of, translation, the
English translation of Disagreement opens with a very serious misrepresenta-
tion of Rancière, making him appear to say just the opposite of what he says
in the original French. Rancière's interpretation of the famous Aristotle pas-
sages serves as both the jumping off point for his book and one of the central
arguments of Rancière's en tire conception of politics. 1 try at the opening of
my chapter to reset the terms of interpretation by outlining and correcting the
problematic English translation. Second, 1 provide sorne context for Rancière's
discussion of the process of subject formation. This argument lies at the ClUX
of Rancière's understanding of democratic politics, but here again, certain in-
consistencies in translation have blurred and confused a set of tenns, the mean-
ings of which need to be worked through with more care.
This work on language helps me to make my own argument about the role
of language (and the relation to more formaI, ontological theories of language)
in Rancière's thinking of politics. The chapter tries to make sense of Ran-
cière' s radical conception of democratic politics by refusing to focus solely on
what he says about either poli tics or democracy. Such an approach wards
against the tendency to generalize (or ontologize) from a few key statements
that Rancière makes about politics. And while this chapter rereads Rancière
through the question of language, it also actively resists the idea of talking
about Rancière's "philosophy of language" or of implying that Rancière's
work can be understood on the philosophy of language model. Language is not
Rancière' s ontology any more than any other concept would be, but language
does reveal crucial aspects of Rancière's thought that do not otherwise come
easily into view. Therefore, rather th an conclude with a précis of Rancière's
contribution to philosophy of language, 1 close the chapter by tuming to an
engaged discussion of Paul Apostolidis's recent work (2005; 2010) on immi-
grant meatpacking workers in the Pacific NOlthwest - taking up the account of
language and democratic subjectivity that 1 have just read in Rancière and
mobilizing it to consider the potential moments of democratic politics that
emerge in the union activism of these workers. 1 put Rancière's conception of
democratic politics to work by thinking through Apostolidis's account of im-
migrant union workers as a polemical scene for the emergence of democratic
subjectivity.
In the fourth and final chapter, "Critique," 1 retum to the very beginning of
Rancière's career and re-ask a question that he himself has raised, dropped,
and raised again, but certainly never answered definitively: the question of
critical theory. Rancière's first publication has become infamous: he contrib-
uted the opening section to the collaborative work Reading Capital, produced
by Louis Althusser and four of his students at the École Normale Supérieure
in 1965. The book by that title is certainly famous, but 1 caU Rancière's contri-
bution "infamous" for at least two reasons: tirst, the ltalian and later English

1NTRODUCTION 1 35
translations of the book rlrr',,,,,·,orl the contributions bS1EaOlet, and
Macherey and rearranged the order of presentation. Through this process and
the history of the reception of the text that followed, Reading Capital became
a text by a single author, Althusser. almost immediately after contrib-
uting to the volume, and in the wake of the events of '68, Rancière broke
decisively from Althusser. In rejecting quite completely the Althusserian
approach, Rancière implicitly disavowed his own contribution to Reading
Capital.
However, as l show in this chapter, Rancière never abandoned the subject
that he first broached in that lengthy (almost book-length) essay: the concept
of critique. This chapter considers three distinct contributions to a theory of
critique or critical theory that Rancière has made over the course of his career:
(I) the early essay in Reading Capital; (2) the arguments made in the later
chapters of Disagreement; and (3) a more recent essay on critical thinking
(Randère 2007c). Working sometimes with and sometimes against Rancière,
but always remaining in dialogue with him, l advance an argument against the
very notion, so well-established and commonly accepted in contemporary
theory today, of critical theory as operating by way of a logic of inversion. This
is the logic that underlies common understandings of critical theory as a prac-
tice of demystification; it rests on the classical opposition between appearance
and essence, suggesting that undemeath the false appearance lies a true es-
sence. The task of the critical theory becomes one of inversion to just the
extent that inversion can reveal the underlying truth the very truth that is
masked by outward dissimulation.
In a preliminary effort to move critical theory onto new and sustainable
coordinates, largue that we must reconsider this entire logic. More th an this,
l reject the logic of inversion (and the metaphysics of truth/appearance that
sustains it), suggesting that a new critical dispositif must function on the basis
of a different logic. Rancière points the way toward developing such a critical
dispositif when he tantalizingly hints at the ide a that the tradition of cri tic al
thinking has had it wrong an along, "that there is no hidden secret of the
machine" (Rancière 2007c: I5). l move beyond the steps Rancière has taken,
as I give the sketch of a renewed sense of critical theory, a distinct critical
dispositif that would eschew the logic of inversion. This would be a Rancièrean
critical theory (even it is not Rancière's) to the extent that it takes equality as
axiomatic and seeks its verification.
The book closes with an Afterword, which serves as a performative illustra-
tion of Rancière's concept of the "one too many." On the one hand, the After-
word threatens to be one chapter too many in the sense that it deviates mu ch
more significantly from Rancière's own arguments. On the other hand, Ran-
cière himself caUs the dëmos the "one too many," and therefore the Afterword
expresses an important dimension of democracy (Rancière I999: I88). Like
democracy, the Afterword involves a miscount. In both name and content, the

1 The Lessons of Rancière


Afterword is not a conclusion. it includes my
ment, as make the case for the complementarity between a
standing of queer and Rancière's thinking of fJVi.HA""'",

Rancièrean thinking of democratic in order to articulate


certain sense of queer politics, queer theory, and queer activism. This means
go directly against Rancière, who has himself had nothing good to say about
queer politics or queer them'y. Despite the fact that Rancière's few remarks
about queer theory have aIl been negative, nevertheless assert that Rancière's
approach to democracy, on the one hand, and queer theory's to sexual
identity, on the other, can mutually inform one another. Rancière's non-liberai
account of democracy is a thoroughly queer one, regardless of what Rancière
has had to say about "queer theOl'Y" as a field.
1 argue that Rancière's conception of the democratic "miscount" can be
understood as a queering of democracy because Rancière's refusaI to reduce le
compte des incomptés (the count of the unaccounted, those who are of no ac-
cou nt) to the marginalized or excluded, produces a queer politics. To develop
this argument 1 first show that, cmcially, Rancière refuses the notion that his
politics could be reconciled with or co-opted by a liberai theory of victimiza-
tion. At the same time, the very ide a of queer theory (or queer relational, ines-
senti al identity) depends upon thinking a category of "the queer" or the
"unintelligible" that is not simply excluded ("excluded" in such a way that the
political answer to the problem of exclusion would be nothing other than in-
clusion). Rather, that which is rendered marginal with respect to the norm is
essential to the very operation of the norm. It is both already induded, on the
one hand (since the margins of the norm are a necessary part of its very opera-
tion), and not-includable, on the other (since the margins cannot be brought to
the center without the very collapse of the nOlm). 1 argue that a politics of the
unintelligible - or better, of unintelligibility is a queer democratic politics.
This last claim within the book is not the definitive daim of the book but
merely "one more claim," and it is for this reason, among others, that 1 make it
in the Afterword and not in the last numbered chapter (or in a conclusion). Put
simply, the book has no telos. The path that the chapters follow does not lead
up, as if we would reach the tmth of Rancière's work, but rather through. For
this reason the final sections of the book take up concerns that are more mine
than Rancière's. "Through" therefore indicates not only a horizontal path
across Rancière's writings, but also an anival to a certain extent "on the other
side" of Rancière - at a place to which he may or may not have any interest in
travelling (Rancière 2003b). This is fitting. The non- or un-Rancièrean endings
to the book express best, in their own way, exactly what the book seeks to
demonstrate: the lessons of Rancière.

INTRODUCTION 1
Democracy is in the streets.

ANC1ÈRE FIRST CAUGHT THE attention and sparked the imagination of


political theorists with his writings from the I990s: these texts specifi-
cally conceptualized politics; they engaged with the historical tradition
of poli tic al philosophy; and above aIl, they offered a radical new definition of
politics. La mésentente appeared in 1995, a full three decades after Rancière's
first publication, yet it was the first work of Rancière's - when translated into
English four years later as Disagreement - that was widely read within the
field of political theory in the UK and USA (Rancière I999).1 For political
theorists, the appeal of Disagreement surely lay then, as it does now, in Ran-
cière's decision to rewrite the meaning of politics. In doing so, Rancière suc-
ceeded in provoking thought and debate: readers were by turns inspired or
offended by this new conceptualization of politics. Rancière's rethinking of
politics appeals to those theorists who seek an understanding of politics that
avoids the abstraction of Rawlsian proceduralism and Habermasian delibera-
tive democracy, while refusing to repeat the philosophical platitudes of certain
forms of poststructuralism. The power and potential of Rancière' s approach to
political philosophy lie in his ability and willingness to treat politics afresh -
to come at the question of politics from an as-yet-unexplored angle. Rancière
thereby gives political theorists new material, new resources for thinking poli-
tics, simply because he redefines not just what politics means, but what it does.
1 put the point this way because for Rancière the question of politics is not a
question of definition, and it is never, as 1 discussed in the Introduction, a
matter of ontology. Politics always involves interruptions, interventions, or
effects. Poli tics is not; politics disrupts.
This approach to poli tics has significant and sometimes worrying impli-
cations. First, it means that much of what political theorists study, and a
great deal of what they argue, will turn out - according to Rancière's new
definition - to be something other than politics. Second, as Rancière himself
corlcHlCle:s. it means that the ancient of is
highest and noblest pursuit that both SOCl'ates and Strauss made it out to be
(Strauss 1959). Quite to the Rancière dedares directly: am not a
political philosopher" 2003a). And he has reasons make
such a statement, because within the terms of his arguments, "political phi-
losophy" has a very particular meaning, Rancière rejects the idea of taking
political philosophy as a branch or "natural division" of the broader field of
philosophy (Ranci ère 1999). Rancière goes so far as to contend that the telos
of the political philosophy project is the very elimination of politics. 2 This
daim holds, according to Rancière, across the canon. From Plato to Aristo-
tle, from Marx to Arendt, political philosophers have sought to supplant the
an-archic disorder of politics with a hierarchical order of the philosopher
(Rancière 1974b, 2007c).3 As readers of Rancière have duly noted, aH ofthis
l'aises the question of how to understand the relationship between what Ran-
cière caUs poli tics and what we have been caUing politics for a very long
time, but which Rancière renames la police (Thomson 2003; Valentine
2005). Given the broad-sweeping nature of Rancière's redefinition of the
everyday activities of politics under the category of "police," it is tempting
to read Rancière as either a sort of Arendtian or a sort of Hegelian.
In the first case, in his effort to get at the specificity of politics, Rancière
could appear to be circumscribing a specifie sphere for politics while relegat-
ing aIl other phenomena to the non- or apolitical. Thus, as Arendt gives us the
categories of labor/work/action, Rancière would offer us politics/police. Like
Arendt, Rancière would be interpreted as seeking a purer conception of poli-
tics - a reading that dovetails perfectly with Zizek's critique of Rancière as a
"partisan of 'pure politics' " (Zizek 2006: 75). Thus, as Arendt protects against
the encroachment of the social onto the realm of action, Rancière would pro-
test against the expansion of police orders in such a way as to crowd out poli-
tics. And, in fact, many writers working on Arendt or on Arendtian approaches
to politics have folded Rancière and his thinking into their projects - some-
times lumping Arendt and Rancière together (Beltrân 2009; Halpern 2011),
sometimes marking certain differences (Schaap 20II; Honig 2001). One of
my primary goals here at the st art of this book is to darify what 1 insist are
fundamental and ineradicable differences between Rancière's thinking of poli-
tics, on the one hand, and certain sorts of Arendtian approaches to politics, on
the other. With this daim, 1 am neither making nor implying a critique of
Arendt. 1 am concerned not with the salience of Arendtian theories of politics,
or with the fecundity of Arendt's own texts, but with the common presumption
that Rancière's thinking of politics can be folded into a general Arendtian
framework. Both thinkers see politics as novel, creative, surprising, and unex-
pected. Both use a kind of analytical definitional work as a critical instrument;
in redefining things they challenge everyday conceptions of things. However,
the temptation to read Rancière as an Arendtian - which for me means the

POLITICS 1 39
inclination to read his of of a pure, pro-
tected politieal sphere - can lead readers astray in their approach to Rancière.
1t inclines readers to see Rancière's politics as pure, despite his very direct
claim that politics can never be pure 201 lC: 3).
Rancière' s conception of politics strikes most readers as so free- floating, so
unmoored from aIl other conceptions, that many naturally seek to ground his
thinking of poli tics somewhere. An Arendtian reading of Rancière solves this
problem, but at too high a cost. Another way to moor Rancière's conception of
politics is by fixing it within a broader conceptual vocabulary. Here the strat-
egy is more German idealist than Arendtian, as if calls for the introduction of
a third term to balance and to ground the polities/police opposition. Rather
than an Arendtian pure sphere of polities, we would thus have, on this reading
of Rancière's works, either a Kantian version of Rancière (wherein the third
term serves fonnally as the condition of possibility for polities/police) or a
HegeIian version (in whieh politics and police stage a confrontation, the syn-
thesis of which determines history). A third term serves to stabilize and fix
Rancière's conception of politics by giving politics and police a space of
mediation.
In this chapter 1 will first show briefly that sorne of Rancière's best inter-
preters (and sometimes with encouragement from Rancière's own texts) have
followed one of these tempting readings and either taken Rancière's work to
support a pure theory of politics or supplemented his account with a third term
that would somehow mediate the relation between poli tics and police. How-
ever, 1 will then go on to argue that Rancière's potential contribution to politi-
cal theory lies not in Arendtian or idealist veins of thought. Rancière rejects,
perhaps above aIl else, the very idea of a pure polities. Politics is precisely that
which could never be pure (and for this very reason, politics is never some-
thing that can be measured or predicted with "precision"). Rancière's critique
of the entire tradition of political philosophy (including Arendt) centers on his
resistance to the pu rit y or purification of politics. 1 will frequently bring my
reading of Rancière into sharper relief by contrasting it with those approaches
to his work that would (explicitly or tacitly) fold his arguments into a pure
thinking of politics.
But 1 wish to state clearly from the beginning that these hermeneutic con-
flicts rai se important political stakes. The emphasis on impurity matters a great
deal because, for Rancière, politics is an act of impurity, a process that resists
purification. That is, and as 1 will explain in the text below, poli tics makes a
supplement possible in the face of a social order that says it has no supple-
ment. Polities makes visible that whieh a social order wishes to render invisi-
ble, and it does so in such a way that it does not just "add" to what is already
given. Instead, the logic of politics undermines the purity of the given. To think
politics as "impure" in this sense means, on the one hand, to reject any model
of unalloyed politics (whether it be anarchism or Hegelianism or anything

40 The Lessons of Rancière


to insist
can be fully known and incorporated into the social order. To use a Rancièrean
H.uLM~'4M~ that will take up and in much more detail in '--'H'''IJL'-'~
to an "excess of words" that both makes politics
nr"'UAl1,t" its closure.

work out the for these conclusions reading Rancière against


not only his critics but also sorne of his apparent Throughout,
insist on a subtle and precarious conceptualization of the relationship, in Ran-
cière's between la police and la politique. This sort of understanding
requires a serious contextualist engagement with the translation of Rancière's
works into English (and curiously, also, as will explain, the translation of his
work from English back into French), yet the delicate issues of translation
have an important and substantive impact on how we understand Rancière's
thinking of politics. 1 argue that la politique and la police do not name sepa-
rate, sealed spheres that are mutually exclusive. At the same time, 1 insist that
the relation between them cannot be mediated, grounded, or sublated by a
third telm. The terms of Rancière's political writings are multiple and multi-
plied. They can never be reduced to two (same/other) or even to three (thesis/
antithesis/synthesis), since their impurity al ways resists such a reduction.
lndeed, it is in those dimensions of his thought that consistently thwart efforts,
even by his own readers, to render politics pure that political theOlists may find
the best resources for thinking politics anew today.

The Use of Politics

Rancière's most direct writings on politics and political philosophy written


in French in the early 1990s, but, as noted in the Introduction, translated much
more swiftly into English than his earlier writings rapidly grabbed the atten-
tion of many political theorists for at least two reasons. First, Rancière poses
to himself and to his readers a question that is clearly fundamental for most
political theorists. In the preface to Disagreement he frames that question as
follows: "What can be thought of specijically as politics?" (Rancière 1999:
xiii, emphasis added). Perhaps more importantly for political theorists, and
this is the second and most powerful way that Rancière gains our interest,
Rancière argues that a response to this question "force[s]" a distinction upon
us: we must distinguish politics "from what normally goes by the name of
politics and for which 1 propose to reserve the term policing" (Ranci ère I999:
xiii). When Rancière refers to "what normally goes by the name ofpolitics" he
means the actions of assemblies and parliaments; the decisions of courts; the
work of politicians; and aU the efforts of bureaucrats. Rancière not only re-
names an of this under a nonpolitical heading, but also gives it a name
that surely sounds pejorative on first reading: he calls it policing. Much early

POLITICS 1 4'
cOlmnlerltm'y on Rancière's thinking with to has been so
up in his distinctive definition of poli tics as to lead to relative neglect of his
COflcelDt of la police. While in Chapter Two take up a much more detailed and
involved reading of Rancière's concept of the police, it is essential to make a
few points explicit here.
Rancière clarifies his conception of police by showing that while it is re-
lated to the idea of uniformed officers riding in patrol cars and walking the
street, it must neveliheless be analytically distinguished from "the truncheon
blows of the forces of law and order" (Rancière 1999: 28). Rancière uses
"police," "policing," and "police order" to name any order of hierarchy. And
thus he invokes this broader concept of "policing" to indicate both policy-
making - as the term in English, though not in French, already connotes - as
weIl as a wide array of economic and cultural arrangements. In order to stress
the broad nature of his concept of la police, Rancière (uncharacteristically)
emphasizes the link between his use of police and Foucault's work. Foucault
argues: first, to the extent that any police order determines hierarchical rela-
tionships between human beings, "the police includes everything"~ second, to
the extent that it sets up a relationship between "men and things," the police
order also constitutes a material order (Foucault 2002a).
These links make it clear that Rancière calls on the concept of la police to
connote the vertical organization of society:4 the dividing up and distribution
of the various parts that make up the social whole. A police order is not just an
abstract order of powers (of laws or principles), it is "an order of bodies that
defines the allocation of ways of doing, ways of being, and ways of saying, and
sees that those bodies are assigned by name to a particular place and task~ it is
an order of the visible and the sayable" (Rancière 1999: 29, emphasis added).
In this key passage Rancière elaborates on the concept of the police in terms
that will become central to his later writings, since "an order of the visible and
the sayable" glosses Rancière's vital notion of le partage du sensible - usually
translated as "the distribution/partition of the sensible" but also connoting both
sharing and division (Panagia 2010).5 Policing is a way of dividing up and
linking up, of making visible and making invisible, the various parts of the
social order.
In "Ten Theses on Poli tics" Rancière suggests that "there are two ways of
counting the parts of the community." The first way of counting he calls police,
and he describes it as follows: it "only counts empirical parts actual groups
defined by differences in birth, by different functions, locations, and interests
that constitute the social body" (Rancière 2001: par. 19). At this point, the
English translation of "Ten Theses" moves on to Rancière's next sentence (and
the next way of counting). However, in the original French, as Maria Muhle
has very helpfully pointed out, Rancière continues: "à l'exclusion de tout sup-
plément" (Ranci ère 1998a: 176~ cf. Muhle 200T 4; see also Rancière 2010b:
36). The police, then, is not just a way of counting the actual groups that make

42 1 The Lessons of Rancière


up social is a manner of that excludes the !-,V,"".1lJJ.H'-]
any supplement to that order. There can be nothing more to count; the police
must count aIl. concur with when she go es on to that the
phrase left out of the translation is actually the "central of Ran-
cière's argument, because "what politics does, is to make this pos-
sible" (Muhle 200]: 4). asserting that the police's way of counting must
preclude from the start any remainder, Rancière shows again that "policing"
points to a particular type of partition of the sensory realm. This passage under
discussion immediately precedes Rancière's seventh thesis, which "the
police is a partition of the sensible whose principle is the absence of a void and
of a supplement" (Rancière 2001: par. 19). The police order distributes bodies
without remainder and without exclusion (à l'exclusion de tout supplément);
there is nothing it does not account for, nothing left over or external to its pro-
cess of counting.
Rancière's "second way of counting" the parts in any order is his definition
of politics: having taken most of the everyday occunences that ordinary lan-
guage refers to as "politics" and recategorized them under the heading of the
"police," Rancière may then identify politics as the very disruption of the
police order. The police order is hierarchic and implicitly built upon the as·-
sumption of inequality (an inequality based on the very differences that legiti-
mate the domination within the social order). The logic of politics is based on
the presupposition of equality, understood in the particular way that 1 discussed
in the Introduction. In this sense, politics serves to challenge, disrupt, and
consistently intenupt the smooth fiowing of the police order. Put simply, as
Rancière does in his seventh thesis, "Politics is specifically opposed to the
police" (Rancière 200I: par. I9). Politics is therefore dissensus: the disruption
of the given order of domination (the police order) by a political subject that
only emerges after the moment of politics, a subject that cornes to exist only
through the act of politics. 1 will return below to a much fuller exploration of
what Rancière might me an by la politique, how it relates to la police, and how
one might understand it in the terms of political theory. But first 1 want to inter-
rupt that discussion with the question that Rancière's initial definition imme-
diately caUs for: having defined politics uniquely, what does one th en do with
that definition?
For heuristic purposes, let me lay out a set of simplified answers to this
question ofhow one proceeds politically and theoretically when confronted by
Rancière's polemical definition of politics. 1 think we can see three general
approaches:

1. Use the definition as a critical tool; leverage it for one's own argumenta-
tive purposes.
2. Ontologize the definition; make it function as an ontological ground and
build a comprehensive theory atop it.

POLITICS 1 43
1",,...,,.,.,.,,,,,,.,,1-0 the definition into an extant make it fit
within a better-known framework. There could potentially be an almost
limitless number of varieties of this strategy, but here mention only
three that prove salient in readings of Rancière:

a) Anarchist theory
b) Arendtian poli tics
c) Politics of recognition
The first overall option, to use the definition of politics as a rhetorical, argu-
mentative weapon, is one that Rancière's own texts sometimes invite, and one
that sorne of his very best readers have surely found enticing. Vnder this head-
ing one starts with Rancière's succinct daims about politics so as to put his
novel definition of politics to work, thereby using the definition the way Fou-
cault said knowledge was used, "for cutting" (Foucault 1984: 88). 1b do this is
to take Rancière's new definition of politics and make it function as a weapon
of critique. Vndoubtedly Rancière does this himself when in the latter half of
Disagreement he shows that neoliberal consensus models of politics amount to
nothing more, though surely nothing less, than "orders of the police." This ap-
proach would make Rancière's definition of politics into a tool of politico-
theoretical argument, but it remains undear how it could serve more than a
negative function. It would invite aIl those questions posed to Foucault's work:
"What are your normative grounds?" "What positive foundation for actual
politics do es this definition provide?,,6
The second option, to ontologize the definition of politics, is an aIluring
choice, particularly if one finds the "normative grounds" complaints plausible
and meaningful. To exercise this option means that rather than putting the defi-
nition to work, one instead works it up into a full-blown alternative theory of
politics. This choice provides a direct response to requests for grounds, since
it turns Rancière's daims about politics into ontological foundations. How-
ever, the limitations of this approach are obvious from my discussion of ontol-
ogy in the Introduction. From the very start this move would have to violate
what we might caB one of Rancière's primary "mIes" - the one that prohibits
recourse to ontology. As 1 have already shown, Rancière's tactic will always be
to de-ontologize, but the effort to build a full political theory on the grounds of
Rancière's definition of politics would surely be an ontologizing move. It
would found the new theory on the principles of Rancière's conception of
politics. Rancière himself does not simply eschew this move; he consistently
rejects it when he sees it in other thinkers, and he repeatedly resists it wh en he
identifies it in appropriations of his own thought. None of this prevents one
from using Rancière's conception of politics in this way. However, such a
strategy requires that one not merely part ways with Rancière's own thought,
but double back and go against it. Perhaps this choice will be taken up by
future commentators on Rancière, but 1 do not follow it here.

1 The Lessons of Rancière


third and most choice to do with radie al
nition of polities is to incorporate it into an already extant poli tic al theory or a
broader framework. This has often to be a nrp'TPt''t"Pr1
among philosophers and politieal theorists who have taken up sedous en~~agement
with Rancière's writings on The incorporation strategy is exercised
in Tbdd impOltant book on The Political Thought of Jacques
Rancière, the first of what willlikely be many books in English that begin to con-
stitute the secondary literature on Rancière 2008; see also Davis 2010; Tanke
2011). title might suggest that he pursues the second (ontologizing) option,
that he systematieally builds up Rancière's "political thought" from the grounds of
Rancière's thinking of politics, equality, and so on. However, May seeks not neces-
sarily to develop Rancière's own unique political theory, but rather to appropdate
Rancière's arguments -- pmtieularly arguments about equality and democracy - as
a pdme resource for elaborating the extant theory of anarchism (May 1994, 2008,
2009, 201oa). 1 contend that in his effort to appropriate Rancière's thinking and
place it in the service of anarchist politics May winds up defending a vision of pure
(anarchist) politics. 1 take up this argument in Chapter Two, through a more de-
tailed engagement with May's important book, using his work as a contrast for my
own approach to Rancière's conceptions ofpolitics and police.
In this chapter, 1 set May's arguments aside and look instead at two other
important variations on the "incorporation" strategy. My overall concern is that
the incorporation of Rancière's writings into a broader body of thought tends to
reduce his conception of politics to a vision of what he and 1 caU "pure politics."
The first incorporation strategy emerges specifically with the work of those theo-
rists who read Rancière alongside of, or within the tenns of, Arendt's politieal
theory. The second appears in efforts to cladfy Rancière's two key telms (poli-
tics/police) by adding a third term that can stabilize the first two. In both cases,
the reading of Rancière renders his account of politics too statie, too de-histori-
cized, and, in a general (if grammatieally questionable) sense, "too pure." My
alternative approach to Rancière's thinking of poli tics begins to emerge here in
a number of distinct but related ways: in my insistence, with Rancière, that poli-
tics cannot be pure; in my attempt to show that Rancière's thinking of politics is
always a thinking of politics in history; in my contention that Rancière's under-
standing of the relation between la politique and la police can neither be grounded
in any ontology nor mediated by any third terrn; and in my argument (in the final
section) that democratic politics always and only goes on within police orders.

re Politics

In asserting that Rancière's work has frequently been incorporated into an Ar-
endtian framework, and in trying - in response to this trend to prize apart the
arguments of Arendt and Rancière, 1 am making a particular and circumscribed

POLITICS 1 45
Mm-kell's work
V<:>!'r>h,:>n me to the terms of that claim.
writes, "If there is anything like a standard reading of The Human Condition,
it is safe to say that it is built around the theme of separation. The point of the
book, we aIl know, is to div ide things that have been blurred together inappro-
priately" (Mm-kell 20II: 20; see also Holman 20II). Markell caUs this stan-
dard reading a "territorial" interpretation of Arendt's most famous book,
because it insists on "sorting" human activity into its proper categories (Markell
20 l 1: 16). 1 would add that in the effort to overcome the improper blurring of
categories and activities, this territorial interpretation of Arendt points toward
a pure conception of politics (either in Arendt herself or in the conception of
politics that the territorial interpreter tries to build). It is just this territorial ac-
count, and the conception of pure politics to which it is committed, that 1 wish
to reject. More to the point, 1 want to resist the tendency to fold Rancière's
writings on politics into this "standard" account of Arendt. In so doing, 1 am
not suggesting that the tenitorial reading of Arendt is the only one on offer.
After aIl, Markell himself clarifies the contours of this standard reading so that
he can provide an alternative (Markell 20 Il: 35), and 1 see a number of impor-
tant connections between the Rancièrean account of "impure politics" that 1
develop here and the alternative interpretation of The Human Condition that
Mm-kell advances. 7 My point is more subtle, then, since as Markell himself
shows, the territorial reading of Arendt does in fact prove to be the standard
one. For just this reason, the tendency to read Rancière alongside, or even
tacked on to, Arendt, makes it tempting to interpret Rancière's work territori-
ally. That is, the incorporation of Rancière into an Arendtian framework entices
one to read Rancière as a theorist of pure poli tics, despite his explicit rejection
of that very notion.
Perhaps the best ex ample of folding Rancière's thinking of politics into an
Arendtian structure is Cristina Beltrân's 2009 article in Political Theory. First
and foremost, this essay offers an interpretation of the widespread protests and
demonstrations by immigrants across the USA in 2006. Cutting across the
praise and condemnation of these events by commentators on the left and
right, respectively, Beltrân argues for the specifically political nature of the
protests (Beltrân 2009: 597). She uses Arendt's political theory to make this
case, as it lets her show that the demonstrators (a group that included legal and
illegal immigrants and their numerous allies) brought f01th the power of initia-
tion - a novel, creative political force. The protests, then, were "inaugural"
acts in the Arendtian sense of bringing forth something new and unexpected,
and the demonstrators were not "merely" protestors; in Beltrân's account they
emerge clearly and with sorne force as poli tic al actors.
As 1 noted at the beginning of this chapter, this Arendtian account of poli tics as
novelty (as surprise) and as the creation of something new the production of a
political stage where before there had apparently been none - resonates in various
ways with Rancière's thinking of politics. In at least this minimal sense, then,

1 The Lessons of Rancière


there are affinities between Arendtian account of the 2006
protests and Rancière's approach to politics. But Beltrân takes a further and sub-
step when she caUs on Rancière as suppOli for the account she
provides. In a crucial move contained within one sentence near the middle of
Beltrân's article she makes the shift fromArendt's account of "new beginnings" in
The Human Condition to Rancière's thinking of dissensus in Disagreement
trân 2009: 604). Beltrân suggests that the Arendtian framework of political action
in concert can include within it the Rancièrean conception of the miscount.
Beltrân herself then Im'gely drops Rancière and returns to more detailed
readings of other texts from Arendt only mentioning Rancière once more, in
the brief conclusion to the article. Beltrân is therefore not offering a reading of
Rancière, and she is not making a claim for the compatibility of his thought
with that of Arendt. And for just this reason, have no intention here of pre-
suming to offer a critique of Beltrân's interpretation of Rancière, as it would
be unfair of me to add up the few sentences in ber article where she mentions
his work as if they amounted to a full reading. Nonetheless, there is something
symptomatic and significant about Beltrân's treatment of Rancière: she quite
comfortably folds Rancière's far less well-known work on politics into the
tenns of Arendt's political theory. It is worth emphasizing here that within the
disciplinary subfield of political theory, more books have been written over
the past twenty-five years on Arendt's political theory than perhaps any
twentieth-century thinker besides Rawls. In other words, despite not presum-
ing to make any large claims about Rancière's writings, by pulling them in so
easily to the frame of Arendt's thought, Beltrân's essay has the effect of bol-
stering the idea that Rancière is some sort of Arendtian. l aim here not to criti-
cize Beltrân's essay, but to read it as suggestive of a larger trend (see also
Halpern 2011).
Beltrân's gesture toward Rancière may have taken its inspiration from an
earlier reading of Rancière by another prominent interpreter of Arendt, Bonnie
Honig. Indeed, Beltrân points toward Rancière's arguments at the specific mo-
ments that she wants to show that the 2006 demonstrators were not simply
demanding recognition, not asking to be granted rights, but were engaged in
an act of democratic taking (Beltrân 2009: 598). No one has done more to
develop the idea of "rights as takings" than Honig, and she does so by drawing
specifically from Rancière. Unlike Beltrân, however, Honig offers her own
interpretation of Rancière teasing out and developing the idea of "takings"
that Rancière's work surely gestures toward, but never explicitly articulates -
and she does not attempt to fit Rancière's arguments about politics within a
larger theoretical structure (Arendtian or otherwise). Although Honig uses
Rancière's work only sparingly within the terms of her much broader argu-
ment about demoCl'acy and foreignness, her work on Rancière proves exem-
plary because she allows it to remain distinct from other strands of contemporary
political theory. Honig's goal is not to develop the theory of theorists, but to

POLITICS 1 47
make a very and - while also rich and informed quite
concrete argument about democratic politics. In this way she invokes Rancière
and a whole host of other thinkers to help her achieve that end, but she remains
very attentive to the differences among those theorists. Honig's argument
therefore provides an illustrative contrast with Beltrân. Rather than closing
down the possibilities of Rancière's thought by folding it into the frame of
another thinker, Honig's concise reading of Rancière opens up Rancière's
thought for her readers. 8
Perhaps the most complete and subtle treatment of the relation between
Arendt and Rancière comes From Andrew Schaap (2011), who makes an argu-
ment concerning Rancière's own reading of Arendt's claim about "the right to
have rights" in her Origins of Totalitarianism (Arendt 2004: 376; Rancière
2004d; see also Ingram 2008; cf. Zivi 2012).1 will not summarize or rehearse
Schaap's work here in any detail, for the simple reason that 1 myself parallel
his line of argument in what I write below. Schaap shows that Rancière at
times misinterprets Arendt in order to polemicize his own point (Schaap 2011:
33), while also indicating nicely that the question of the purity of poli tics is
just what is at stake in Rancière's rejection of Arendt (32, 43). By approaching
the relation between Arendt and Rancière through Rancière's own reading of
Arendt's work, Schaap's article achieves a number of important ends for my
own pm-poses here. First, he obviously refuses the move to collapse Rancière's
conception of politics into Arendt's. Second, he shows that the purity of poli-
tics constitutes the central issue of contention between Rancière and Arendt. I
con tend that Rancière is often misread as thinking of poli tics as pure, of con-
ceiving of politics as somehow utterly unique or special, and as privileging a
so-called political sphere over other spheres of order and action. In turn, I see
approaches to Rancière such as Beltrân's as encouraging this line of thinking.
Schaap's work therefore proves crucial because he shows that the further into
Rancière's text we go, the more clearly we see his rejection of an Arendtian
conception of the pure political sphere. In turn, I want to develop this notion of
an impure politics beyond Rancière's polemical rejections of Arendt.
These various accounts of Arendt and Rancière help to shed sorne light on
the general relationship between their work that has begun to fOlm within the
field of political theory, and it reveals something of the tendency to try to think
their thought "together." But wh ether such thinking takes the problematic form
of fitting Rancière into the Arendtian framework, or more clearly marking
their differences, it nonetheless attests to a general tendency to try to under-
stand Rancière by seeing him as at least complementary to Arendt. While 1
have emphasized the difference between Beltrân's approach, on the one hand,
and that of Honig and Schaap, on the other, even in the case of these last two
authors, it might be tempting for sorne readers to collapse the differences be-
tween Rancière and Arendt by way of the fact that Honig and Schaap also
draw from both of them for their distinct accounts of politics. That is, although

1 The Lessons of Rancière


have out, and wish to the distinctions npr"'Pl=-"
proaches to Arendt and in aIl these cases we have a broadly Arend-
tian to politics that includes insights from Rancière this adds to the
general sense that whatever Rancière is up to in his account of politics, it ap-
pears to be Im'gely compatible with Arendt's theory. And bec au se the standard
account of Arendt is the territorial one, the decision to read Rancière alongside
Arendt may take readers in the wrong direction, that is, toward pure politics.
This explains my decision to make a strong argument for de-linking Ran-
cière from Arendt, and for doing so on the grounds of his own insistence that
"there is no 'pure' politics" (Ranci ère 201 IC: 3). The connections are not an
arbitrary choice of my own, as Rancière makes this crucial and explicit claim
about impure poli tics in the same talk in which he explicitly distances his work
from a traditional reading of Arendt: "1 wrote the 'Ten Theses on Politics'
primarily as a critique of the Arendtian idea of a specific poli tic al sphere and a
political way of life" (Rancière 201 le: 3), What, then, does it mean to think
politics as impure? First, l want to suggest that if politics has no ground, it
cannot be self-grounding either. Rancière explains that "poli tics has no 'proper'
place nor does it possess any 'natural' subjects" (Ranci ère 2001: par. 25). We
must say, then, not just that politics is not pure, but more, that poli tics is that
which renders impure. Rancière forrnulates the point in many different ways.
Politics is dissensus (Rancière 2001: par. 24; Rancière 20IIC). Politics is a
splitting into two (Rancière 2006c: 61). Politics is a rupture of the logic of the
archë (Rancière 2001: par. 8). Politics is subjectivation in the forrn of disiden-
tification (Rancière 1999: 37; Rancière 2007d: 559-60; Rancière 1998b: 29;
Rancière 1995c). And with regard to the relation between politics and police,
Rancière makes a very consistent argument: "The opposition between poli-
tics and police go es along with the statement that poli tics has no 'proper'
object, that aIl its objects are blended with the objects of police" (Rancière
20IIC: 5). Politics cannot be uncoupled from police; it only appears in this
"blended" form. But because politics is not simply impure itself but that
which renders impure, this "blended" forrn must not be confused with hy-
bridity or the mere amalgamation of different parts. ln blending politics with
the police, Rancière refuses to merge the two; he gives us a blending that is
always also an othering.
lt seems simple to conclude with the thesis that poli tics can never be pure,
a point that many commentators on Rancière have rightly made (Muhle 2007;
Panagia 2010; Rockhill 2006). But the argument proves more complicated
than such a conclusion would suggest, since impurity is, by definition, never
simple. And the impurity of poli tics produces a paradox for Rancière' s thought.
On the one hand, poli tics must not be pure. On the other, politics as that which
disrupts the police order must somehow remain "other" to that order; this is
why the blending is never a merging. For the disruptive force of politics to be
preserved, it must somehow remain extemal to the police order that it would

POLITICS 1
Yet as pure would the necessary
of the heterogeneous" that enacts politics (Rancière 1999: 32). politics
must be other to police, but not purely other. The to responding to this
paradox is to refuse to overcome it. Instead, Rancière's account of politics
must be understood as thinking the paradox, as capturing its fiavor and mobi-
lizing its force, rather than attempting to erase or resolve it. To defend this sort
of paradoxical argument means starting with a rejection of the idea of pure
politics a rejection that Rancière makes repeatedly and one that 1 have de-
fended in this section - but it is impossible to rest here. We must also grasp the
relation between politics and police. And while we need to understand the re-
lation as precisely as possible, we must al ways also insist that the relation
itself can never be specified with mathematical precision. On the topic of this
relation, so central to Rancière's thought, no English-language commentator
on Rancière has shed more light than Jean-Philippe Deranty - and Deranty
offers the clearest example of a third incorporationist strategy in his reading of
Rancière.

The Three~Term Model

Deranty has written a number of articles (in English) that provide comprehen-
sive overviews of the political thought of Rancière. Arguing in particular that
Rancière's work can be best understood within the context of the politics of
recognition, Deranty compares Rancière's arguments directly to the work of
Axel Honneth (Deranty 2oo3a). For the most part (and in the context of my
particular work here) 1 am not overly concerned with Deranty's broader argu-
ments; however, 1 do note that Deranty's reading is doubtless inflected by his
own desire to link Rancière with the tradition of the politics of recognition.
Surely Hegel is the most prominent theorist of recognition and the most sig-
nificant figure in the tradition of dialectical thought (and, anecdotally, Hegel is
the other author upon whom Deranty has focused most of his work). More-
over, although Deranty would never reduce Rancière' s thought to the dialectic,
he still sees Rancière's conception of politics within the terrns of the dialectic.
He refers, for example, to "the dialectic between equality and inequality" that
is punctuated by Rancièrean politics (Deranty 2oo3a: 153). The context of
Deranty's reading thus suggests an "incorporation" of Rancière's thought into
the already established politics of recognition, linked to a long history of
Hegelian and dialectical thinking.
My own reading of Deranty does not try to generalize from this broader
context, but rather focuses specifically on a particular set of powerful and im-
portant claims that Deranty makes about how we should understand the rela-
tion between politics and police. 1 am somewhat surprised that Deranty's
argument on this front has not had a bigger impact on the English-Ianguage

50 1 The Lessons of Rancière


literature on that proposes a radical re-evaluation
of Rancière's account of politics - just the issue that animates most political
theorists wh en they tum to Rancière. mentions at the end of his ac-
count (and in a footnote), a set of facts that think should be foregrounded. As
explains, Rancière was invited to participate in a seminar run Jean-
Luc Nancy and Philippe Lacoue-Labarthe in February 1982. This was the
seminar at which Nancy and Lacoue-Labarthe argued for the distinction be-
tween la politique and le politique (Deranty 2003b: fn. 27; Lacoue-Labarthe
and Nancy 1997).9 As 1 will show, this distinction will prove central to Der-
anty's reading of Rancière.
However, before digging into Deranty's specific interpretation, it seems
prudent to take one step back. There is a rich and varied tradition in contempo-
rary political theory of insisting on a difference between "politics" or "policy"
(la politique), on the one hand, and something like "the political" (le poli-
tique), on the other. In his 2007 book, Post-Foundational Political Thought:
Political Difference in Nancy, Lefort, Badiou and Laclau, Oliver Marchart pro-
vides perhaps the definitive history of what he caUs "political difference" from
Ricoeur, through Arendt, Schmitt, and Mouffe, aIl the way to Nancy and Lacoue-
Labarthe - with side-trips into Wolin, Sartori, and others (Marchart 2007).
Significantly, Marchait has now published a revised and significantly extended
version of this book, in German (Marchart 2010). The more recent Gelman
text covers more authors, and as 1 will discuss below, it also includes a direct,
if quite short, engagement with Rancière. For now, 1 will confine my remarks
to the more widely read and widely accessible work of Marchart's, the 2007
English text. In this book Marchart assigns himself the task of tracing the dis-
tinction between politics and the political through aIl of twentieth-century po-
litical theory. He sees his work as something of a comprehensive account and
provides lengthy commentary on most of the "big names" in Continental
thought. His references to Rancière, therefore, prove important precisely for
their sparseness. Early on, Rancière's name appears in a list of many who rec-
ognize a split between politics and the political, but there is no direct discus-
sion of any of Rancière's work (Marchart 2007: 7). ln fact, Rancière receives
no substantive attention until very late in the book. The few references that do
appear prove thin, but nevertheless very significant.
So how does the general understanding of "political difference," which
Marchart practically defines in his opening page as the difference between la
politique and le politique, manifest itself in Rancière? On first mention, Mar-
chart suggests something striking, but which goes unremarked as such in his
own text. He writes: "What Rancière calls la politique ... others would caU the
political" and this would be distinguished, for Rancière, from politics as police
(Marchart 200T 120). But "the political" is the English translation of le poli-
tique, and thus, according to Marchart, Rancière uses la politique to refer to
"the political" at exactly the moment when others would use le politique. In

POLITICS 1 51
terms of Marchart's own broader work (a book on
this moye makes sense, since it allows Marchart to show that while not ",r",1-rH..,"'"
with or insisting upon the distinction between politics and the (and as
would note, while not explicitly or formally using the term le politique at aIl)
Rancière, in a way, stumbles upon what Marchart caUs "political difference."
For other thinkers this is the difference between le politique and la politique,
whereas for Rancière it is the difference between la politique and la police.
And l should note that Marchart himself does not dwell on this point - Rancière
is certainly not one of the central thinkers of "political difference" -- and there-
fore Marchart may weIl be more alert to the slippery and impure thinking of
politics in Rancière that l try to track and assert here.
Neyertheless, in terms of my own argument, Marchart's key quote on Ran-
cière (from aboye) to the effect that la politique becomes "the political" for
Rancière, does seem odd, especially considering that the English translators of
Rancière's key texts on politics (the works from the early 1990s) haye seen fit
to translate, and do so rather consistently, la politique as "politics" and not as
"the political." That is, for those reading Rancière's work in English transla-
tion, "the poli tic al" makes no appearance at aIl. Thus, l would suggest that it
might make more sense to leave Rancière out of the "political difference"
model entirely, rather than trying to make him fit into it in inverted fashion.
The fact that Marchart quickly moves on from Rancière may be taken as evi-
dence that he too sees Rancière as not quite fitting the model.
In any case, things grow stranger when one turns from Marchart's brief
comments on Rancière and political difference to Marchart's reading ofZizek's
account of Rancière. In this context, Marchart tells us that Zizek finds his own
version of political difference in Rancière, which Marchart summarizes as
"the difference between 'la politique/police' and 'le politique'" (Marchmi
2007: 145). This formulation, especially in light of the earlier one, requires
sorne sorting out. Here again "the political" is distinguished from police. But
whereas in the earlier version we had the formula la politique =the political,
we now have a return to the more standard le politique =the poli tic al. At the
same time, this second formulation requires making la politique a synonym
for the police. Zizek's claim lO here strikes me as quite simply untenable since
Rancière's writings on politics consistently refer, in French, to a difference
between la politique and la police. Zizek asserts that la politique and la police
are the same for Rancière, when in fact the difference between them is the
fundamental and driving force in aU of Rancière 's writing on politics. If there
is to be any obvious "political difference" in Rancière it surely is to be found
here .- in the difference between la politique and la police that Rancière so
strongly insists upon. Again, one wonders wh ether the project of finding a
politics/the political distinction in Rancière is merely futile.
Perhaps, then, the distinction between politics and the political has no obvi-
ous place in Rancière's work. Indeed, at first glance it would seem that no

52 1 The lessons of Rancière


commentators on, no t1'ans1ators Rancière have
tinction in his own work in French between la politique and le politique. Il is
more than notable, that not only locates this difference
argument, but also makes it central to his of Rancière.
structures his articulation of Rancière' s concepts by both
tics and police in a rather typical fashion. But in defining "politics" as that
which breaks with the order of la police, refers to la poli-
tique. Thus, the opposing tenI1S, according to Deranty, are la police and la
politique. This brings Deranty back to the dilemma of pure how to
understand the interaction between these two terms that are diametrically op-
posed? Deranty's answer is striking and original:

This tension between la police and la politique creates a necessary place where
they can and must be mediated. Rancière caBs this third term le politique. It
is the place where the underlying equality operating within social inequality
is verified pragmatically in struggles and demands of equality. In this place is
therefore also verified the wrongness and "wrungness" of a social order that is
otherwise presented as naturally ordered. It identifies victims of the tort and
those who perpetrate the tort. In simple words, le politique is always a demand
for justice. Le politique is in essence polemic (Deranty 2003b: par. 6, italics
original, bold added).

In a subtle yet stark departure from the standard reading, Deranty says there are
not two terms in Rancière's conceptual frame; there are three. Le politique is
the third tenn. 12 It identifies and points to that place in which the logic of domi-
nation contained by la police meets the presumption of equality mobilized
by la politique. As Deranty explains in his later note, this reading contends
that Rancière has taken on board, in a serious way, the Lacoue-Labarthe- and
Nancy-inspired distinction between le politique and la politique. However,
and utterly unsurprisingly, Rancière has mobilized the distinction for distinct,
if not opposite, ends. Whereas le politique for Nancy suggests something like
the very essence of "the political," which has been eroded or 10st within moder-
nity, Rancière, according to Deranty, uses the third term, le politique, precisely
so as to en able an anti-essentialist understanding of poli tics (Deranty 2003b:
footnote 27).
By introducing what he caUs the third tenn in Rancière's thinking of poli-
tics, Deranty offers a powerful and persuasive argument for how to interpret
the relation between la police and la politique. But we still have to ask whether
the reading is supported by the text; that is, is this difference between la poli-
tique and le politique actuaUy present in Rancière's writings? If the difference
does exist in French, it has by no means been preserved by Rancière's English
translators: neither Julie Rose, who translated La mésentente, nor Rachel
Bowlby and Davide Panagia, who translated "Dix thèses sur la politique" refer
to such a distinction; they make no effort whatsoever to calI attention to a

POLITICS 1
difference between le and la lJolltl~rue. The French texts themselves
UVI,H-HdV",-,

include very sparse references to le politique and make no mention of any ex-
plicit or meaningful distinction to be drawn between la politique and le poli~
tique. For example, aIl of the theses in the "10 Theses" use the French la
politique and never le politique (Rancière 1998a, 1995a). AIl of this evidence
leads to one central question: where, if anywhere, can we actually locate the
distinction in Rancière between la politique and le politique?
ln the case of Zizek we can safely say that the distinction is imposed upon
Rancière's writings from outside in an attempt to make his thought fit into a set
ofpremade categories (cf. Parker 200]: 71). But the answer in Deranty's case
cannot be that simple, because Deranty does not start with "political differ-
ence" and then fit Rancière to this mold; instead, his argument for the "three
terms" emerges directly from his own reading of Rancière. But if a strong
distinction between la politique and le politique cannot be found in either La
mésentente or "Dix thèses sur la politique," where did Deranty find it?
In repeatedly asking this question of myself 1 had begun to wonder, against
my better judgment, if he made it up. But, of course, he did not. The answer is
that Deranty very likely found the distinction (although he does not tell his
readers this) in Rancière's works that were originally written in English. To
reiterate for clarity's sake, the distinction that Deranty draws between la poli-
tique and le politique originates not in Rancière's better known writings on
poli tics but from a few lectures that Rancière presented in English. 13 ln I 99 I
Rancière gave the paper "Politics, Identification, Subjectivization" at a confer-
ence in the United States. The conference, as Rancière tells readers in the
preface to the second French edition of Aux bords du politique, was devoted to
"the American debate over the question of identity," and circled around issues
of nationalism and racism (Rancière I998a: I3, emphasis added).14 The con-
ference organizers posed specific questions to the paper-givers, and Rancière
chose to structure his talk directly in response to one of those. In just his second
paragraph, after clearing the ground concerning his having to give the lecture
in English, a language that is not his own, Rancière writes, "1 quote from the
third point of the list of issues we were asked to address: 'What is the politi-
cal?' " (Rancière 1995c: 63).15 Thus, the very idea of thinking about "the po-
li tic al" cornes to Rancière from outside, from what was at the time a very
American-centric debate over multiculturalism, and it is voiced in a foreign
language, English. Rancière then goes on to give his answer to the question,
"What is the political?" This answer corresponds perfectly weIl with Deran-
ty's commentary: "The political is the encounter between two heterogeneous
processes." The first process Rancière calls "policy"; the second is "equality"
(Rancière 1995c: 63). As Deranty will echo more th an a decade later, Rancière
says, "We have three terms," but at this point those terms are "policy, emanci-
pation, and the political" (Rancière I995c: 64; cf. Deranty 2003b: par. 6). Ran-
cière then suggests we name the process of emancipation "politics." Finally, if

1 The Lessons of Rancière


into as la 16 wind up
terms that Deranty argues for: la police, la politique, and le politique. La poli-
tique is that logic of equality that encounters the order of domination consti-
tuted la police; le politique is the or space of such an encounter.
the logic here
''-'''''-'LLL'-'H-'UU, looks less than straightforward, and thus
remain skeptical that the "three terms" approach of both Deranty and the
lecturer-in-English, captures adequately the thinking of politics that
Rancière's original French texts provide. Before complicating matters with the
help of just those texts, l should bring the story of "poli tic al difference in Ran-
cière" to some sort of close. Rancière does mention the difference be-
tween politics and the political in other places, but the references are not only
limited in number but also confined to texts originally produced in English. In
another lecture also given in English, in 2003, Rancière refers directly to the
previous talk: "In an earlier text, 1 proposed to give the name of 'the political'
to the field of encounter - and 'confusion' - between the process of poli tics
and the process of police" (Rancière 201 lC: 5). Yet this lecture does no work
with the politics/political distinction and certainly does not maintain a three-
term model; rather, it invokes this earlier work in order to make the point,
which l have discussed above, that politics "has no 'proper' object" (Rancière
20IIC: 5). Finally, in a 2005 talk given after the death of Derrida, Rancière
also mentions "the political," but, as l will discuss below, this text actually
pushes even further away from the three-term model that Deranty proposes
and that Rancière's 1991 text would appear to support.
Given the fact that Rancière himself never develops the idea of "the politi-
cal" as a third term that would ground the encounter between poli tics and
police, l would hypothesize that the idea of three terms in Rancière gained
traction not because of these three short lectures given in English, but because
of Rancière' s decision (or that of his edit ors ) to have the first lecture translated
into French and included in the second French edition of Aux bords du poli-
tique. The 1991 lecture was translated into French so as to make the distinction
between la politique and le politique perfectly clear. Moreover, in that edition
the editor chose to italicize the le and la that precede politique in order to em-
phasize the "political difference." More than this, the new structure for this
second edition of the very popular book (the first edition included Rancière's
earliest writings centered most directly on politics, the ones on which La
rnésentente directly built) placed the 1991 lecture at the center of a new "part
one." Part one was titled "Du politique à la politique": from the political to
politics. Finally, Rancière wrote a new preface for the second edition in which
he explicitly discusses, for the first (and to my knowledge) only time in French,
the difference between la politique and le politique. If aIl that were not enough
to make it seem as if this distinction had always been central to Rancière's
thinking concerning politics, the blurb on the back of the book excerpts the new
preface at just the place where it specifies the distinction between le politique

POLITICS 1 55
la politique. The blurb reads: le itself as an of
philosophical thought, it is without doubt that this neutral adjective conve-
niently signifies a variation with the substance of la politique, in its
sense of a fight of the parties over power and the exercise of that power. To
speak of le politique and not la politique indicates the principles of law,
power, and community and not the activities of government" (Ranci ère
199 8a : 20).17
Given the importance of the second edition of Aux bords du politique to
Rancière's corpus, it seems likely that a reader picking up Rancière's work in
French over the past decade (a reader such as Deranty) would easily assume
that the difference between la politique and le politique had an important role
to play in Rancière's overall thinking of politics. For example, in the updated
and expanded German version of his book on poli tic al difference, Marchart
(2010) clarifies his account of Rancière specifically by adding a discussion of
this i 998 text. First, Marchart makes explicit the point that 1 suggested above,
that the fundamental "difference" for Rancière must certainly be that between
la politique and la police. He th en goes on to add that "the category of the
political, le politique, does not completely vanish" (Marchart 2010: 180, my
translation). 18 As evidence, Marchart cites the 1998 French translation of Ran-
cière's 1991 talk (originally in English), specifically where Rancière names
the "three terms" and suggests that "the political" is the terrain 19 for the en-
counter between the other two terms (Rancière 1995c [1991]: 64; Rancière
1998a: 84; Marchart 2010: 180). Unlike Deranty's texts from 2003, Marchart's
recent book thereby points directly to the "source" of Rancière's comments on
"the political."20 However, Marchart.'s German text is not widely available
and will surely not reach as many readers as Deranty's two earlier English-
language articles have done. For this reason sorne degree of mystery remains
concerning the three terms, and Deranty's reading still possesses a certain
gravitational force.
1 wish to resist that force by offering my own interpretation of "the politi-
cal" in Rancière's work. Before turning to that reading in the next section, 1
should emphasize here that despite the two very brief references to "the poli ti-
cal" in talks Rancière gave in English in the 2000S, it remains the case that the
vast majority of Rancière's writings on politics maintain no such distinction at
an. Most significantly, Rancière's central works on politics from the 1990S
were aIl produced after the 1991 lecture that had suggested three terms (in
English), yet Rancière did not bother to fold that terminology into La Mésen-
tente - clearly the central text of Rancière's devoted to politics and engaged
with the tradition of poli tic al philosophy. 1 therefore contend that there is
something very problematic about making the three terms of politics funda-
mental to one's interpretation of Rancière. To do so would be to take the French
translation of one short lecture coupled with an eight-page preface to a second
edition of a collection of essays, and use those texts as sorne sort of central

1 The Lessons of Rancière


corpus on poli tics. Without any further reason to think, or evidence to support
the that le politique mediates la politique and la police in Rancière's
main texts, it seems a mistake to structure an about Rancière's con-
,,'-'~JU'-"H of politics around this (not to mention that there seems to be
no evidence that Rancière himself wishes to reinterpret his past works through
this lens; since the preface, there have been no more writings in
so far as know, that maintain a difference between la politique and le poli-
tique). goal here is not to provide sorne sort of definitive refutation of
Deranty. lndeed, Deranty is not necessarily "wrong" in his approach to Ran-
cière, but 1 wony that his presentation of the three terrns in Rancière can have
a somewhat distorting effect, particularly for those readers who do not have
easy access to Rancière's writings in French.
1 must stress, however, that my resistance to the idea that le politique pro-
vides a ground or space of encounter between the conflict between la police
and la politique does not rest only on a contextualist argument concerning the
production and presentation ofthese texts. While 1 contend that the contextual-
ist work provides reasons to be wary, 1 also argue that the insertion of the third
term into Rancière's writings fails to account for the subtlety and power of his
thinking of politics. It actually blunts the incisiveness of his conception of la
politique to conceive of il within the three-term model. To insist that Ran-
cière's thinking of politics requires three terms is to limit its polemical force;
it is to make Rancière far more Hegelian than he is. Moreover, to make le poli-
tique into the space where the "fight of politics" is played out, is to push Ran-
cière far too close to the territorial Arendtian model. Both incorporationist
strategies of reading Rancière must ignore or erase crucial dimensions of his
thought. Rancière always resists the idea of a "sphere of action," and the pro-
motion of le politique to a mediating, grounding term runs the risk of trans-
forming Rancière's conception of politics into a specific sort of act that must
occur in its "proper" space (Rancière 2001: par. 4; cf. Rancière 201 IC). Even in
its formallogical structure, the three-telm approach proves overly symmetrical
and balanced for a thinker who consistently insists on a lack of balance -- who
thinks in and through paradox, not symmetry. The three-term model creates a
set of proper spheres for Rancière's concepts, wh en those concepts are always
designed to thwart the ide a of proper spheres.

The Dou bling of Politics: Democratie


Pol itics with in the Pol

To prevent closing down Rancière's thinking of politics by assimilating it to


such a model, 1 will build on Deranty's arguments but at the same twist or
"wring" them in a particular way. As Deranty stresses, the "wrongness" that

POLITICS 1 57
""',,,,,,",,'<, asserts in the face of a order is also a a
or torsion of the police order and its logic of inequality. In other words, 1 am
trying to apply to Deranty's own reading the anti-ontological torsion that Der-
anty so helpfully identifies in Rancière's wode Thus contend that there are
not really three distinct tellliS in Rancière's argument. Ifthere were three terms,
all three of them could be pure: a realm of domination (police), a realm of dis-
sensus (politics), and a ground upon which they meet (the political). But this
would be to center an essential conception of politics, le politique, as an onto-
logical foundation. This, as Deranty stresses and as 1have been arguing through-
out this chapter, could not be further from Rancière's project. As Rancière and
both consistently emphasize, Rancière avoids aIl ontology (Rancière 20IIC;
Rancière 2009b; Ieven 2009).21 Hence my argument: we do not have three
terms (police, politics, the political) but merely a doubling of one of the two
terms. In his recent essay on method, Rancière refers to "a doubling up of the
notion of politics" (Rancière 2009b: 121). And it seems more than anecdotal
to point out that this essay, specifically devoted to method and dealing at length
with the question of political theory, makes no mention whatsoever of a differ-
ence between politics and the poli tic al (nor one between la politique and le
politique).
PoUtics is doubled, always and already. It is "doubled" in that it is never
singular and never pure - "always and already" because the doubling is not a
secondai)' process that happens to a pre-given poli tics, but an essential feature
of la politique in the first place. Politics, like the logos 1 will discuss in Chapter
Three, is subject to an original taint - split from itself, split into two from the
beginning (Rancière 1999: 16,61). Thus, politics cannot be pure in Rancière's
thought, nor can there be a clear "poli tic al difference" in his work (la poli-
tique/le politique) because in his writings, politics doubles itself; that is, we
cannot distinguish le politique from la politique, given that neither is singu-
lar. 22 Let me unpack this dense formulation of the argument and explore sorne
of its implications.
First, 1 make the case for a "doubled politics" as a solution of sorts to the di-
lemma of how to translate - which is nothing other than the question of how to
read - Rancière: if la politique is never simply itself (never pure, never one),
then we can understand how it is both the disruptive other to the police order and
somehow simultaneously a part of the police order. La politique is always ulti-
mately opposed to and transformative of la police, but since the former is never
simply itself, it cannot be taken to be wholly external to and outside of the latter.
ln grasping for a third term, le politique, Deranty actually goes sorne distance
toward bringing this dimension of doubling into play in Rancière's work, but
1 insist on the crucial importance of the doubling of la politique, rather th an the
preservation of its purity through the introduction of a third teilli.
This approach to reading Rancière works not only for those texts that do
not mention le politique but also for those that do. 1 have been insisting that

58 1 The Lessons of Rancière


COll1Cf~ptJlon of rneans that
third term that would constitute a pure space of encounter for politics and
police. it is also clear that in a few of his texts Rancière does in-
troduce the term political and it serves a function distinct from that of the word
politics. How then to account for these brief appearances of the political, with-
out resorting to approach? Let me respond offeIing a brief read-
ing of the most recent text of Rancière's to mention "the political." In 2005
Rancière gave a talk, in commemoration of DerIida, at Birkbeck College in
London. The talk, given in English, was titled "Does Democracy Mean Some-
thing?" It covers a number of common themes in Rancière's thinking of poli-
tics and democracy, and, more to the point, it makes use of the word political
in a subtly distinct way.
In this essay Rancière reiterates a reading of Plato's Laws that he gives in a
wide variety of contexts (Rancière 2006c 39-40; cf. Rancière 1999: 64). In
Book III of the Laws Plato crea tes "a list of an the necessary qualifications of
mling" (Rancière 2010b: 50). The first six qualifications an indicate a specifie
archë for mling; that is, they demonstrate a principle by which one group (the
strong) shaH mIe over another (the weak) in a variety of forms. So far there is
nothing surprising in Plato' s account, sin ce, as Rancière puts it, "Government
seemingly requires an account of its arkhë," a justification of "why sorne take
the position ofmlers and others" mled (Rancière 2010b: SI). But Plato's read-
ers are in for a surprise when they get to the seventh item on his list, where they
discover that the final qualification to mIe turns out to be a principle that is no
principle at aIl: "'the drawing of lots,' or democracy" (Rancière 2010b: 51).23
Adding this item to the list produces something of a paradox: democracy is not
an archë, not a principle of mIe, but merely mling itself, kratos, an almost
random "prevailing." Any good Platonist would like to insist on the necessity
of an archë, and thereby remove this seventh qualification from the list. But
Rancière shows that Plato keeps it there for a reason, and not merely because
of the empirical fact of democracy' s existence. Plato must add this seventh
item to the list because the absence of archë within the democratic title to mIe
serves a purpose: it "rebounds on the 'good' qualifications" (Ranci ère 2010b:
5 1). What Rancière means by this is that aIl the other pIinci pies for mling can
surely use the archë they provide in order to justify their mIe, but those prin-
ciples in themselves cannot explain political ruie.
Let me elaborate on this claim, and try to explain Rancière's provocative
assertion that the other titles to mie are not political. Basing mling on age
gives us gerontocracy; basing it on knowledge gives us epistemocracy. None
of this, however, gives us politics: "Missing from the list of fonlls of govern-
ment, however, is the political foml. If the idea of political government means
anything, it must imply an extra something" (Rancière 2010b: SI-52). What
makes mling political for Rancière is that it partakes of this supplement, this
title to mIe that is no title, this principle of rule that is no principle - this kratos

POLITICS 1 59
Îs archë. Rancière says, refers a power of the dëmos
that is not authorized by any archë at aIl. Furthermore, Rancière here in-
sists that the political can only emerge out of democracy. All politics is demo-
cratic politics because "the political" appears only when la politique confronts
la police. Rancière expresses the point this way:

Power must become political. For that to happen the logic of the police has to
be thwarted by the logic of politics. Polities means the supplementation of an
qualifications by the power of the unqualified. The ultimate ground on which
rulers govem is that there is no good reason as to why sorne men should rule
others. Ultimately the practice of ruling rests on its own absence of reason
(Ranci ère 2010b: 53).

Thus we might say that in one sense Deranty was right aIl along: le politique
provides the ground for politics. However, this "ultimate ground" tums out to
be absolutely no ground at aIl, and le politique can therefore in no way be
thought as a prior space that mediates the confrontation between politics and
police. Quite the contrary: the political only cornes about because of the irrup-
tion ofpolitics within a police order. Notice here that Rancière repeatedly uses
the word political - again, in English - as an adjective. He does not refer to
"the political" as a fundamental, ontological category that makes poli tics pos-
sible. Instead, he shows how the dissensus of politics brings about "political
govemment" or "political power."
And in this way Rancière's own use of the adjective political points directly
toward the impurity of politics. It is not politics but rather policing that would
seek to predetermine the field of le politique: "The logic of the police consists
in delimiting the sphere of the political" and it does so "in the name of the
purity of the political." Against this notion of purity Rancière caUs on a demo-
cratic poli tics of impurity: "Democratie logic ... consists in blurring and dis-
placing the borders of the political" (Rancière 2010b: 54). This account of
what 1 would prefer to caB politicalness (to distinguish it from the category of
"the political"), pulls together a number of strands of my argument, since we
can see here that the very idea of having "the political" name a demarcated
space - this idea belongs to or at least partakes of police logic. After an, de-
marcating spaces is what policing does. Democratie logic, on the other hand,
unsettles those demarcations, and it often does so by finding or making "politi-
cal" that which did not seem political beforehand: it makes non-qualifications
into qualifications and rend ers the apolitical political. In short, it politicizes.
This aIl means that politicalness is not a third terrn that Rancière would
conceptualize or define. Put differently, "the political" is not a part of his con-
ceptual vocabulary or critical apparatus in the way that la police and la poli-
tique clearly are.lnstead, politicalness would be better described as a dimension
of the struggle of dissensus; it is an aspect of certain distributions of the sen-
sible, and it is a result of conflict, not a theoretical condition for politics. To say

1 The Lessons of Rancière


this is criticisms of the in le
tique as a prior, delimited space for the happening of poli tics is not merely that
this gets the between poli tics and I-"_Jc'H~''-''''UHV,J'''
backwards fails to grasp "politicalness" as itself an of contesta-
tion). More than this, in le politique as a mediating space,
actually attributes to Rancière a way of using "the political" that Rancière
himself criticizes within the tradition of political philosophy.24
This reading returns me to my broader daim for a "doubled politics," since
we can now see this doubling through the idea of le politique as that which
emerges out of the confrontation between la politique and la police (rather
than providing a plior stage for that confrontation). In this context, want to
insist that the argument for a doubled politics cannot be confined to the level
of hermeneutic debates or semantics (even if it arises there, even if that is the
location from which we can begin to make the argument in the first place).
The notion of a doubled politics not only solves interpretive problems but
also casts helpfullight on Rancière's conception of politics on, as I said in
the opening of this chapter, not just what it means but what it does. Rancière
does not, à la Nancy and others, wish to separate something like "the politi-
cal" from politics; nor does he seek to establish "the political" as more origi-
nary, more fundamental to politics (cf. Marchart 2010: 178). Perhaps
encouraged by Rancière's early use, in English, of the phrase "field of en-
counter," Deranty is too quick to assume that le politique is the "necessary
place" for politics and police to me et. Such a rendering of Rancière' s thinking
does not convey forcefully enough what Deranty refers to in the very next
paragraph as the "anti-ontological" shape of Rancière's argument (Deranty
2003b: par. 7). As I have shown, even where Rancière introduces the term
political he by no means proposes a radical difference or dichotomy between
le politique and la politique. 25 More to the point, the potential ambiguity of
"politics" in Rancière's English-translation texts helps to convey something
of the impurity of politics that Rancière himself insists upon in English (Ran-
cière 201 IC: 2; cf. Rancière 2009c). If there can be said to be such a thing as
"political difference" in Rancière, it is surely this doubling of politics itself,
rendering it always already impure.
Deranty's attentiveness to the difference between le politique and la poli-
tique shows that the mediation between la police and la politique is not one
that would be transcended or sublated by a third term. Deranty himself holds
to this line, saying that the "mediation" provided by le politique "must not be
thought as a synthesis, since the logic of the tort [the logic of the 'wrong' that
makes for dissensus] is decidedly nondialectical" (Deranty 2003b: 144). By
shifting Deranty's observation of the difference between la politique and le
politique from the introduction of a third tenn to the doubling of one of the two
main ternIs, my reading insists on the paradoxicallogic of Rancière's workjust
as it maintains the impurity of rus account of politics. It also supports Deranty's

POLITICS 1 61
'-'Bq,nJlU".'" on the non-dialectical nature of Rancière's thought. from
itself (containing the traces of both la politique and le politique within it), poli-
tics could never be pure. am arguing both with and against Deranty here:
agree that to render Rancière's telillS in dialectical fashion is to misconstrue
them badly, but 1 think we best avoid this false construal by avoiding alto-
gether the hypostatization of three terms in Rancière. This gets at a more gen-
eral point: it always proves very hard to avoid dialectics simply by daiming a
position as non-dialectical. Any opposition to dialectics always remains sub-
ject to being captured by dialecticallogic, of turning that "opposition" into the
negative moment on the way to dialectical synthesis. Thus, my goal here with
respect to Deranty's interpretation of Rancière centers on making a more con-
certed effort to avoid a dialectical rendering of Rancière's thought. 26
The three-tenn model tends, as 1 have shown, to purify politics. In addition,
that approach also seems to ignore quite blatantly a crucial point that Rancière
himself frequently makes: politics and police meet within the police order
itself. Politics goes on in the only place where it can go on: within the social
formation where it occurs, that is, within the space of the police order. And
politics must be doubled because of this very fact about its spatio-temporal
location because politics is that which opposes the terms of the police order
but does so within its terms. Only an impure form of politics could do such a
thing. Zizek tries to fit Rancière's thought into the category of "pure politics"
(Zizek 2006: 75) because he wants to lump Rancière together with other "post-
Althusserians" such as Balibar and Badiou - while distinguishing Zizek hi m-
self from the lot of them. But as I have shown, Rancière does not fit into this
"pure poli tics" category, and in trying to force him there, Zizek badly mischar-
acterizes Rancière's thought. Rancière could not be more direct: "There is no
'pure' politics" (Rancière 201IC: 3). This daim does more than just refute
Zizek's characterization, showing why Rancière is no mere "post-Althusserian";
the daim demonstrates that in trying to grasp the meaning and importance of
Rancière's conception of politics as that which iITupts into any given police
order, we must see the interconnected nature of poli tics and police.
In a crucial passage that responds to critics who would (mis)read him as
proposing the purity of politics, Rancière writes (and 1 comment in brackets),
"Politics does not stem from a place outside of the police .... There is no place
outside of the police. [And hence there need not be a third place where politics
and police meet aIl 'meeting' is conftict within the police order itself.] But
there are confticting ways of doing things with the 'places' that it allocates: of
relocating, reshaping or redoubling them" (Rancière 20IIC: 6, emphasis
added). I italicize Rancière's use of "redoubling" because the word itself is a
doubling of double, connecting directly to my broader argument about how to
understand his conception of politics. To remain both impure and non-dialectical
(politics' impurities cannot merely be waiting to be removed through a process
of sublation), politics must always be "redoubled" in this way.27

62 1 The Lessons of Rancière


This "cr,ri1rH'I la as doubled while
of poli tics in democratic politics - brings to light both the impurity of poli tics
and the inherent opposition between and police. Thus, and as have
shown textually if le politique - or politicalness - plays a role in
theory, it cannot be a grounding role, that of providing a spaee for
politics; here, as elsewhere, we find no ontology. Raneière refutes directly both
the ide a of politics having a proper space and the notion of the politieal provid-
ing a space for politics to occur: "The exceptionality of politics has no specifie
place. Politics 'takes place' in the space of the police [again, no third space is
needed], by rephrasing and restaging social issues, police problems and so on"
(Rancière 201 IC: 8) The eneounter between la politique and la police is Hever
definitive, never final, and never produces a new "stage" of history. It is always
a renegotiation of the very police order in which we live (Thomson 2003: 6).
In the end, democratic politics has no other choice but to negotiate in this
way; it cannot do otherwise. Those readers of Rancière who see in (or project
into) his works a radical alternative to every form of politics as we know and
have known it will find this conclusion utterly unsatisfying. 1 suspect that
many readers of Rancière would like him to provide what a certain form of
orthodox Marxism once offered: a utopianism stripped of the label "utopian,"
a vision for a poli tics wholly other to what lies before us in our own political
conjuncture - in a word, hope (May 20IOb). 1 share the sense ofhope's impor-
tance to politics (a point 1 take up in Chapter Four), but in certain readings of
Rancière 1 wony about the emergence of a particular kind of hope: a pure
hope, a hope untainted by the unruliness of democracy. And yet, nothing could
be further from Rancière's own vision of democracy: for him, unruliness is
Just what democracy offers. But that is not a11 it offers, for in the unruliness of
democracy we can locate the verification of equality through the "excess of
words"28 (Rancière 1991, 1994), and the only genuine hope there is: not the
hope that poli tics will save us, but that democratic poli tics will change what is,
will alter what is given.
For this reason, what we might call- even in the face of his own reluctance
to give it such a name "Rancière's political theory," must be a particular kind
of political theory. Not despite, but due to the nature of its radical commit-
ments to equality and the people, this is not a full blown "theory of politics" or
the political. As 1 discussed in the Introduction, Rancière has argued strongly
that he has no theory of politics, nor any intentions to proffer or develop one
(Ranci ère 2009b: 114).29 Claims such as these make it obvious that Rancière's
thinking of politics resists the trajectory of any pure, formaI account of what
politics should be or become. But this is not just because Rancière has chosen
to do something else other than "produce a theory of politics." It is because
democratic politics is never a pure poli tics. 30 To insist that politics is not pure
is surely to reject the idea of a formaI poli tic al theory that would lay normative
grounds or predict historical processes. As Rancière puts it, "1 was not willing

POLITICS 1
say how we must think and act" 20IIC: But it is no means
to reject "political theory" in a more broadly conceived form, for a commit-
ment to the impurity of polities is a commitment to another task, a reraising of
the question, "How are we to reinvent politics?" (Rancière I995c: 70).
A Rancièrean reinvention of polities must remain committed to the impu-
rity of polities; it must consistently resist the temptation to shelter polities in
its own proper sphere. Above aIl, a reinvention of polities, for Rancière, must
never lose sight of, and must never fail to understand, the sites of domination
and inequality, for these are the very locations of those rare moments of poli-
ties. Polities happens on the telTain of the police. Therefore, a reinvention of
politics must begin not with the ground of equality or the conditions that char-
acterize the free and autonomous individual. Instead, politics begins with the
hierarchy, inequality, and structural domination of aIl social orders. A new
thinking of polities can only start with la police.

1 The Lessons of Rancière


The barricade blocks the street but opens the way.

HE PREVIOUS CHAPTER CONCLUDED with the polemical daim that dem-


ocratic poli tics can neither supplant nor erase the police order: politics
can never be pure. Instead, politics, as an act of impurity, must al ways
be both tied to and engaged in conflict with the police order. Politics can do
nothing else than this: renegotiate and reconfigure the police order. The fact
that politics always concerns a dissensus with an inegalitarian and hierarchical
social order means that la police proves far more important to Rancière's
broader theory than sorne of his readers have perhaps appreciated. For just this
reason this chapter offers a doser analysis of the concept of police in Ran-
cière's thought, and it argues for the absolute centrality of the police not only
to Rancière's thinking of democratic politics but also to his broader body of
work. The alternative to reading Rancière's account of politics as producing a
purer model of the political, or as carving out a protected sphere for politics, is
to interpret his daims about poli tics always in the context of his account of the
police. This means, at the same time, taking seriously Rancière's own asser-
tions about the import and consequence of police orders; it means emphasiz-
ing that "police order" is not merely - and sometimes is not at all - a term of
disparagement. Here l demonstrate a polities of the police, in the sense that
policing matters to politics - in the sense that police is a part of politics (and
not just politics' other).
To see why police matters in this way, one needs a fuller sense of Rancière's
broader understanding of what his fOlmer teacher, Althusser, called the social
formation, and what Marx simply called "society." Jean-Philippe Deranty
makes an important comment in this context, when he writes, "Rancière has
borrowed the metaphor of society as a gravitational order, a kind of Aristote-
lian nature, where objects always end up falling to their proper places" (Der-
anty 2003b: par. 30). The idea of society as marked by a nature or ruled by a
law like gravity willlikely strike most readers as rather antidemocratic. In fact,
that has a law-like order appears to of either an aris-
tocratic or elitist political ontology, an ontology that surely proves incompati-
ble with Rancière's radical thinking of equality as outlined in the Introduction.
Of course, have already suggested that Rancière's method is not to co un ter a
poli tic al ontology he opposes with one that he SUppOltS, but rather to resist aIl
ontology. And his eschewal of ontology offers a clue as to how to read the
quo te from Deranty above, for Deranty's observation is astute: Rancière do es
describe society as a social order, as a hierarchical, structured whole.
However, Rancière does not take up Aristotelian ontology; rather, Rancière
borrows a rnetaphor. 1 want to insist, in my reading of Rancière, on a particular
way of understanding this metaphor, for although Rancière does, at times, men-
tion such a metaphor, he always identifies it as a metaphor. Those committed to
the hierarchy of the police order may wish to literalize this metaphor, to naturalize
the inequality of a given police order. But, in a supplement to Rancière's argu-
ments about the "presupposition of equality," we can argue that police orders are
built upon an assumption of inequality. Those "stultifiers" whose existence de-
pends on that presumption will try to "verify" this inequality by repeatedly dem-
onstrating their own superior intelligences. Their repeated presuppositions of
inequality mean that the social order is always in fact marked by domination.
Yet this claim does not commit one to the idea that such domination is natu-
raI, that the inequality we see in a social formation is underwritten by a prin-
ciple, an archë, that legitimates the domination. Rancière will always insist
that in the face of the inequality that structures a police order, the verification
of the equality of intelligences will always expose the wrong of that order.
Thus, there is nothing natural about inequality; it is nothing like gravity, de-
spite the metaphor. As Rancière himself says, "Convention alone can reign in
the social order" (Rancière 1991: 78). Thus, to caB the social order a "quasi-
natural order," as Deranty does, requires us to put enormous weight and re-
peated emphasis on the quasi- part: the social order passes itself off as natural;
it has recourse to the metaphor of gravity. But natural it is not. Therefore we
can say the following: While the social order is always already being natural-
ized, this work of naturalization, of passing off as natural that which can only
be conventional- such is the work of the police, of policing. Rancière's vision
of politics can only make sense within the terms of those naturalized, hierar-
chical social orders that he caBs police orders.
Police orders are nothing more nor less than the very social orders in which
we aIl live. We encounter them every day. This marks a crucial distance be-
tween police and politics, since politics, for Rancière, is not at aIl a common
or routine happening. As Rancière (in)famously puts it in a passage that 1 will
retum to, "Politics doesn't always happen - it actually happens very little or
rarely" (1999: 17). That politics happens so infrequently demands a response
to the question of how we are to understand the workings of the police. It leads
us to ask how "police orders" function, what role they play in Rancière's

66 1 The Lessons of Rancière


and what is at in the of for.E.'<.U.U'-'ivl

of politics and for his broader "lessons.


This chapter three distinct, but at limes overlapping, answers to
this question. In the first section 1 show that Rancière initially introduces his
concept of "the police" in a way that allows him to redefine neoliberal consen-
sus models (interest-group liberalism) as nothing more than "orders of the
police." This gives Rancière the space to articulate his nove} of
politics, posed in stark and consistent opposition to police. ln this way, police
serves both as leverage for critique and as a foil for politics. despite its
apparent centrality to his entire politico-theoretical framework, in just these
works where he introduces the concept of police, Rancière seems content to
leave it somewhat under-theorized. More than this, Rancière appears untrou-
bled by the fact that most of what we typically take for poli tics has been rede-
fined by this minimally developed concept of police. Of course, one could take
this to mean that poli tics is special because it is, as Rancière says, rare. How-
ever, in this chapter 1 want to ask: if politics happens but rarely, and if the
world we live in can only ever be a world of police orders, do we not need to
think more carefully and critically about the nature, extent, structure (and
structural weaknesses) of those orders?
A second answer is to push "police" from foil for poli tics to Manichean
"other" to politics. We see one foml that this type of response can take in the
work of Todd May, whose book on Rancière develops an account of anarchism
as the only true democratic politics, the only politics committed fully to the
Rancièrean verification of equality in the face of social orders (i.e., police
orders) of hierarchy and domination. l To make this case, May must give a
particular account of police orders because his anarchist poli tics commits him
to the complete obliteration of police. Along the way, 1 suggest that May makes
a significant contribution to a theOl·y of the police - not least because police
remains such an underdeveloped concept in Rancière's political theOl-y. In
May's account, the concept of "police" takes on more prominence as it serves
the role of an enemy to be defeated by poli tics.
In the final section, then, 1 offer a third response, as 1 pull together this
chapter's various strands in order to think through, augment, and reorient Ran-
cière's notion of "the police." 1 base my case partly on a critique of May's
anarchist project, as 1 argue that May's commit ment to anarchism requires him
to depart from Rancière's line of argument at exactly those places where Ran-
cière offers his most subtle insights into democratic politics. May seeks to
supplant police with politics. Therefore, for him, politics must destroy police.
Put differently, May embraces the element of impropriety that proves central
to Rancière's thinking of politics, but he fails to retain any faith in Rancière's
concomitant commitment to an impure politics - to a rejection of any and aIl
philosophical projects that would render politics pure. In tying poli tics to the
elimination of police, May allows a pure poli tics in through the back door,

POLICE 1
without there would be to contaminate I-.H.H"""''''>J.
This chapter therefore caUs for a shi ft from anarchism to a rearticula-
tion of Rancière's allegiance to democracy. 1mportantly, Rancière's "democ-
racy," as he frequently reminds his readers, is not a regime. As he polemically
explains, "We do not live in democracies" (Rancière 2006c: 73). Therefore, a
theory of democracy inspired by Rancière - which may or may not remain a
"Rancièrean" theory of democracy - depends not on dismissing or rejecting aIl
police orders, but on investigating and grappling with them. It demands further
development of Rancière's provocative but ellipticai comments concerning the
"neutraIity" of la police and about the superiority of sorne police orders to
others.

ing Politics: la police


Rancière's perhaps path-breaking, or perhaps merely curious, definition of poli-
tics is now weIl known in contemporary English-speaking poiiticai theory, and
because of this, Rancière's concept of "the police" or "police orders" also has
sorne currency within the field. The two go together, of course, because Rancière
redefines most of what we typically take to be poli tics, and relocates it under the
broad heading of la police. This key move appears early on in Rancière's most
weIl-known text in English, Disagreement. There he writes, "Politics is gener-
aIly seen as the set of procedures whereby the aggregation and consent of col-
lectivities is achieved ... [It denotes] the organization of powers, the distribution
of places and roIes, and the systems of legitimizing this distribution." He then
goes on to state quite flatly that he would Iike to apply a whoIly different name
to such a system: "1 propose to caU it the police" (Rancière 1999: 28). If the re-
versaI enacted here does not come through loudly enough, Rancière had already
announced in the preface that he will "propose ... the term policing" for "what
normally goes by the name of politics" (Rancière 1999: xiii).
This unique redefinition of almost everything we usually caU politics, this
renaming of a broad swath of phenomena under the category of "police" obvi-
ously opens up the space for a new way of thinking politics. lt cornes as no
surprise, then, that most accounts of Rancière's political theory immediately
move on to his definition of politics: alogie antagonistic to aIl policing, a logic
that disrupts and rearranges that order by countering the police order's logic of
domination with the political assumption of equality (Rancière 1999: 29). But
in making this move so quickly, these commentaries on Rancière account for
his concept of police as little more than a foil for the more important argument
about politics (e.g., Panagia 2006; Dillon 2003a). It seems worth noting, then,
that in his own presentation in Disagreement, Rancière spends a great deal
more time sorting through the meaning of police before moving on to his con-
ception of poli tics.

68 1 The Lessons of Rancière


wager is as follows: that doser attention to these i-'m)è)i::li"C;è)
texts will produce a subtle but significant reorientation of our
conception of The formaI starkness of Rancière's daims
about politics may his readers into up the '-HH,';;;".""'.'"

thinking, particularly wh en it cornes to a definition of poli tics from sorne


angles, looks like nothing they have ever seen before. Perhaps Rancière en-
courages his readers to see his concept of politics as " but this ap-
proach also runs the risk of rendering his thought less salient in sense
of the political world we inhabit. To put it bluntly, if aIl we take from Rancière
are rare and beautiful political moments (which are too easily boiled down to
historical revolutionary moments), th en how do we orient or action
within the realm of police orders? The question proves particularly salient
given that aIl of our actions occur within this realm. In other words, if we take
Rancière's concepts of police and politics seriously, do we not also have to
admit that we live in police orders, not in a space of politics? As 1 showed in
greater detail in Chapter One, politics is not really a space in Rancière, as it is
for Arendt; politics is a disruptive force that emerges within a field foreign to
itself. We cannot live in, nor even aspire to live in, Rancière's "political," for
he offers no sphere to shelter us. Our realm is that of la police, and it therefore
seems prudent for us to take seriously Rancière's understanding of police
orders.
Critical attention to Rancière's "police orders" has often been avoided by
reducing the idea of "the police" to little more th an a creative renaming exer-
cise. There can be no doubt that the equation "Rancière's police =our politics"
makes space for the new term, "Rancière's politics." And it is hard to blame
Rancière's readers if they quickly move forward in his text to try to figure out
what this new term might be or do. Moreover, as I explore below, as weIl as in
other chapters in this book, Rancière's rhetorical move to capture our everyday
politics with his new term, "the police," serves as a powerful critical tool for
him. Nonetheless, to take police as nothing more than a substitution for "regu-
lar" politics would be a mistake. lt would prevent us from seeing what else
changes when, for example, Rancière reinterprets an activity like polling
within the rubric of police orders. And by swiftly moving on from police to
politics in Rancière's lexicon, we fail to see the important connections that the
concept of la police draws between Rancière and other thinkers.
In other words, taken in context, Rancière's approach to the police may not
turn out to be so strange or curious as it has appeared to some of his North
American readers. We should note then, the ex tant body of work that already
understands la police as something far broader, something more historically
and politically significant than officers on the streets. On this point, Rancière
uncharacteristically signaIs his own conti nuit y with this history and with other
thinkers. In other words, while in general Rancière appears to studiously avoid
citing other thinkers (especially contemporary French thinkers), and while he

POLICE 1
resists his associated with more
famous) French theorists, when it comes to his concept of the police, he notes
its connections to one of the most famous French thinkers of aIl, Michel Fou-
cault. Foucault's work shows that "the police" may include an vertical rela-
tionships between human beings, while also bringing in material relations -links
between humans and the world (Foucault 2002a).
Rancière has this context in mind when, immediately after introducing the
term la police in Disagreement, he admits that it "poses a few problems." It is
here that Rancière first insists that we dissociate his thinking of "police orders"
from the actions on the ground of either beat cops or feds. But Rancière stresses
that the distinction should be drawn not as a matter of definitional fiat, since a
"nalTow definition [of police] may be deemed contingent" (Rancière 1999:
And we know this because of the work of Foucault, who se lectures on
seventeenth- and eighteenth-century "reason of state" showed that "the petty
police is just a particular form of a more general order that alTanges that tan-
gible reality in which bodies are distributed in community" (Rancière 1999:
28; see Foucault 2002a; cf. May 2008: 41 and Muhle 20072). The term polic-
ing can therefore be used to designate a broad set of phenomena concemed
with structuring and ordering a social formation. The police arranges reality,
in the sense that it distributes people and things into locations and roles.
The connection to Foucault and the argument made conceming him both
prove crucial at this juncture. As l showed briefly in Chapter One, Rancière's
concept of a "police order" cannot be dissociated from his larger concept of le
partage du sensible. The multivalent phrase contains at the same time both the
sense of dividing up the world, of ordering it, of structuring it, on the one han d,
and the sense of connection, of linkages, and sharing, on the other. Rancière
describes le partage du sensible as that which "define[s] the modes of percep-
tion" that make any order both visible and say able. Any configuration of the
world that we might encounter depends on a prior distribution/partition/shar-
ing of the sensory realm. 3 The distribution is a distribution of "parts" and of
places, of spaces for being in that world. In a crucial formulation, Rancière
writes, "Le partage du sensible is the cutting-up of the world and of 'world'"
(Rancière 2001: par. 20). l call this a crucial way of putting it, because in one
succinct statement Rancière has combined two absolutely essential elements
of his thinking of the partage. First, the idea of a "cutting-up" points to the
multiple senses of partition (of the world) as simultaneously separating and
connecting. To put up a partition means to wall off one area from another, but
erecting partitions also creates a new space where there was not one before.
The partition excludes, and at the same time it makes a new form of participa-
tion possible. Secondly, the "cutting-up" in this statement caBs attention to
sense and the sensible (the "world"); this is a cutting up that determines what
can be seen and what can be heard. Le partage du sensible determines a certain
sort. of intelligibility, of what "is" because it is made legible by the partage. 4

70 1 The Lessons of Rancière


The as l indicated in is of
partage du sensible. Ta see "the police" in this way, means, on a generallevel,
to recognize that "police order" designates phenomena not only more general
th an the but also ilTeducible ta simple domination or inequality.
account of Rancière's understanding of la police needs ta start with the
crucial distinction between police orders, on the one hand, and the force or
violence wielded by police officers, on the other. Tanke is therefore surely
right when he begins his overview of the police by saying that "'the police,' in
Rancière's sense, does not refer ta the truncheon-wielding caps who crack the
skulls of striking workers or unruly students" (Tanke 20 II : 45). lndeed,
Tanke's paraphrase is almost a direct quote of the language of Rancière that
1 quoted directly in Chapter One: "Police normaBy evokes ... the truncheon
blows of the forces of law" (Rancière 1999: 28). In short, we must not reduce
la police ta the actions of those individuals that we caB "the police." Tanke's
glass continues, however, in a way that 1 would contest. He caBs the police
"the means by which a society enforces its distribution of the sensible" (Tanke
20 II: 45, emphasis added). My broader argument here should already serve ta
indicate why 1 would reject this particular presentation by Tanke. "The police"
is not a name for sorne power that cornes after a particular partage has been
put into place, thereby serving ta maintain it. Rather, "the police" is a distribu-
tion of the sensible (a particular and distinct type of distribution, as 1 discuss
below). ln Rancière's untimely account, there can be no c1ear-cut difference
between the distribution and its enforcement. Rancière argues that the pres-
ence and activity of more police officers indicates not the strength but the
weakness of a police arder (Rancière 1999: 28). Ta make sorne bodies visible
and others not is precisely ta maintain (ta enforce) a set of distinctions be-
tween them. ln turn, ta challenge the police is not ta undermine their enforce-
ment of a given distribution but ta disrupt it sa radically as ta create a new
distribution. Davide Panagia eloquently explains this subtle point in Rancière's
thinking: "Sustaining and encouraging the circulation of things, Rancière's
police officer guarantees the continuation of the organic correspondence that
constitutes our regimes of perception and the pmiitions of the sensible that
make circulation possible" (Panagia 2009: 121, emphasis added; see Watkins
2010).
Policing is therefore not just the action of caps, and it is not merely a set of
powers; it is a distribution of bodies (Ranci ère 1999: 28). The May '68 slogan
that serves as the epigraph ta this chapter captures this sense that policing (and
the battle with policing) depends on bath the halting and opening of flows of
circulation. It is injust this sense that the police is a partition/distribution/sharing
of the sensible, le partage du sensible. Moreover, the police is not just a particu-
lar type of partage: it is unique. will elaborate on the central importance of the
unique partage that is the police order, by looking again at Rancière's assertion
that there are two ways of (ac)counting (for) the parts of the community: "We

POLICE 1 ]1
will call the first policer: counts actual groups defined
by differences in birth, by different functions, locations, and interests that con-
stitute the social body, without any supplement" 2001: par. The
second, politics, therefore tries to make possible, to bring about, the SUlJPl(::;mem
that police will always seek to exclude. the police order is notjust any sort
of partage du sensible, since what we might cal1 a "police partage" is one that
attempts to account for and to contain aIl. A police order is a partage du sensible
that admits of no remainder, that excludes the possibility of any supplement. The
dissensus produced by politics is indissociable from its introduction of a supple-
ment where there is supposed to be none.
But this does not, according to Rancière, make the police order totalizing in
the sense of determining a repressive state order - hence the importance of the
other intellectual context that Rancière provides, when he insists that the con-
cept of police order must not be confused with Althusser's concept of "state
apparatus" (Rancière 1999: 29; see Althusser 1971).5 State apparatus, says
Rancière, cannot be disconnected from a conception of state standing in op-
position to society, a notion that depends, from Rancière's perspective, on con-
fusing poli tics with police. More than this, to take police as a repressive force
is to miss the crucial disclaimers that Rancière offers concerning his concept
of "police order." In other words, while as readers of Rancière we tend to cel-
ebrate politics in its very opposition to police, it would be too easy to simply
dismiss or denigrate aIl police as repression or violence. Rancière warns his
readers to avoid such a faulty interpretation, yet l wonder if we have paid
proper heed to these alerts. Perhaps they are worth enumerating:

1. Police is a neutral and non-pejorative term (1999: 29).


2. Police can be reduced neither to repression Bor even to "control over the
living" (2001: par. 19).
3. Police is not a leveling mechanism; not aIl police orders are the same
(1999: 30).
4. "There is a worse and a better police" (1999: 30-31).
5. Police orders may make more or less space for the emergence of demo-
cratic politics (2006c: 72).6

This list opens up an enormous area of inquiry for explaining and developing
Rancière's understanding of police, its role in his politico-theoretic work, and
its salience for a broader thinking of contemporary politics. l will come back
to these dimensions later, for now l simply want to bring my logic here to sorne
closure by pointing out the limitations to an approach that would take la police
in Rancière as nothing more than a counterweight to politics. If we refuse to
reduce police to a mere tirst postulate, a given necessary to Rancière's thinking
of politics, we are left with a different set of questions. Most important among
them is this: what work does the concept do for us? In the context of Ran-
cière' s writings from the mid- 1990s, largue that police serves to specify and

72 1 The Lessons of Rancière


to
language sense: a polemic meant to challenge the form of the political regime,
the actions of its or the of its processes. This suggests one
possible of the of the police." There is a politics
of the in that Rancière' s of the serves ends
within the political circumstances in which he publishes. Put in an-
other way, we might say that Rancière's use of "the police" has significance in
the way that it speaks to a particular political context. This daim can be made
to resonate more widely when we consider that Rancière introduces the term
police somewhat late in his career and he drops it relatively quickly.7 ln other
words, does it matter that Rancière proposes the term police at a particular
historico-politicaljuncture? 1 contend that the timing proves more than coinci-
dental, as it appears concomitantly with Rancière's critique of what he often
calls consensus politics, or "post-democracy." As Todd May helpfully articu-
lates, Rancière's critique of consensus democracy - which Rancière caUs a
"conjunction of contradictory terms" (Rancière 1999: 95) implies a radical
challenge to neoliberalism (May 2008: 146).
As we know, "interest groups" and the interactions among and between
them are the basic building blocks of today's liberal and neoliberal politics.
But for Rancière, "conflicts of interest," the very core of what we caH "interest
group politics," are exemplary of the police order because they are exdusively
matters of policing. What Rancière caUs politics has nothing to do with the
coordination of interests: "The political dispute is distinct from aIl conflicts of
interest between constituted parties of the population" (Rancière 1999: 100).
A neoliberal interest-group politics forms and founds a particular police order.
However, to say only this would be to miss the force of Rancière's critique: he
is not content to throw names at this contemporary political regime (by calling
it "police"). Neoliberalism, or consensus democracy, articulates a particular
arrangement between any given police order and the potentially disruptive
force of democratic politics. ln other words, it is not just that neoliberalism is
not politics, but that neoliberalism seeks the end ofpolitics. As leverage for his
critique, Rancière refers to this contemporary form of consensus democracy as
"post-democracy." He defines it as follows:

Postdemocracy is the govemment practice and conceptuallegitimation of de-


mocracy after the dëmos, a democracy that has eliminated the appearance,
miscount, and dispute of the people and is thereby reducible to the sole inter-
play of state mechanisms and combinations of social energies and interests.
(Rancière 1999: I02).

Rancière's understanding of democracy, as l will discuss in greater detail


below, always remains bound up with a fundamental disruptive conflict, a dis-
sensus. "Consensus democracy," in contrast, is committed to the degradation,
co-optation and possible elimination of this element of dissensus; it seeks to

POLICE 1 73
transform a11 confiict into a fOIm of Consensus is thus
the end of democracy - "in a word, the disappearance of politics" (Rancière
1999: I02; see also 2008: I46).
The concept of "the police" provides the crucialleverage for the critique of
consensus democracy. Rancière's ability to take the neoliberal marketing of
"consensus" and show how it boils down to the curious and feeble fonn of
"post-democracy" depends upon the critical lens provided by la police. Ran-
cière's polemic here amounts to much more than merely decreeing that consen-
sus democracy contains no politics, that it is only police. That is, of course, true.
But on Rancière's terms it would a1so be true of almost aIl institutionalized
political regimes. The key to the critique lies in showing that consensus democ-
racy commits itself to the elimination of politics. It is a police order devoted to
its own pure and perpetuaI preservation, a police order that strives for its own
perfection as a police order. Thus, consensus democracy is post-democracy in
the same way that the Platonic kallipolis would be post -democracy. It does not
merely exclude politics from policing; it puts an end to politics. Doubtless, it
does this self·consciously in calling for just that: "the end of politics." This
means that despite operating on very much distinct terrains, post-democracy
functions in a similar manner to political philosophy. For Rancière, the latter is
a philosophical ordering project designed to replace politics with police. The
former is a putatively "political" project airning for the same goal. Both use the
name "politics" as a banner under which to seek the elimination of politics. 8
Clearly, then, Rancière's concept of police works in the service of his own
political interventions. In this context 1 want also to emphasize the different
levels on which Rancière's most explicitly "political" texts operate (i.e.,
Shores, Disagreement, and Hatred). Disagreement, for ex ample, can easily be
read as an abstract and detached philosophical work. After aIl, it opens both its
preface and its introduction with quotes from Aristotle; it operates on a dense
and philosophically obtuse level of logical reasoning; it seems to cite only
historical examples of politics (and not contemporary instances), and to do so
in the service of very broad and general points. And most commenta tors read
Disagreement the way they rnight read Arendt's The Human Condition or any
other work in political theory: as a project of philosophy or, at most, political
ontology. This approach can be encouraged by assigning only the tirst sixtY
pages of the text to students. (1 plead guilty.) Perhaps this is why most explora-
tions of Rancière's political theory center themselves on the tirst half of Dis-
agreement, supplemented by sorne of the theses from "Ten Theses."
The worry is that such an approach tums Rancière into a political philoso-
pher, when he himself has mounted a damning critique of the project of politi-
cal philosophy. How do you theorize politics while avoiding the trap of political
philosophy? Perhaps you link your conception of politics to your assessment
of and engagement with the contemporary political situation. Understanding
Rancière's thinking of politics requires working with his political interventions

74 1 The Lessons of Rancière


as weIl.
peppered throughout the very texts in which he articulates his concepts of poli-
tics and police. For five of Disagreement which contains
the critique of post -democracy that have just been discussing .-- troubles those
readings that would tum Rancière into a thinker of concepts. In that '-'H'ÀIJLl~'
Rancière engages not with Plato or with Marx but with the contemporary dis-
course of "consensus democracy." The essays in Shores and Hatred also fit ill
with any attempt to make Rancière into a philosopher: these are direct poli tic al
engagements; many of the essays that make up those two books were
ously published in popular periodicals. And Rancière has himself argued that
we might read Disagreement backwards, seeing it first as a political interven-
tion (Rancière 2oo3a: par. 4; see Thomson 200 3: 9).
Therefore, to get at the politics of the police, we need not only to take the
concept of the police more seriously and to read it much more broadly, but also
to draw the connections between Rancière's thinking of the police and his
political interventions. In this section, 1 have connected the dots between Ran-
cière's own use of the police and his critique of consensus democracy. 1 now
tum to a distinct understanding of the impOltance of the police to politics, by
offering an analysis and assessment of a politics of anarchism drawn from
Rancière' s work.

Defending Anarchism and the Retreat to Pure Politics

Perhaps 1 should remove aIl possibility of confusion at the outset of this sec-
tion: it is not 1 who will be "defending anarchism," but May who does so in his
book, The Political Thought of Jacques Rancière (2008). Indeed, as 1 hinted at
the beginning of this chapter, while May's book offers a vibrant, engaged, and
always thought-provoking set of arguments in and around Rancière's writings,
it seems a very oddly titled work. Simply put, it does not set out to articulate
"the poli tic al thought" of Rancière; rather, it seeks to mobilize a particular
reading of Rancière's work in support of a spirited defense of anarchism (see
May 1994) hence the title of my section. 1 say this not as a critique of May
but rather to set the stage for an engagement with his work. Thus, my daim
that May's book does not provide an overall account of the "political thought
of Jacques Rancière" is not itself a polemical daim. A reviewer of May's book
as sympathetic and knowledgeable about Rancière as Miguel Vatter has said
the same thing, right at the start of his review: "In reality, the book is less an
in-depth interpretation of Rancière than it is a plaidoyer for the daim that
democratic politics belongs within the tradition of anarchism, as opposed to
those of liberalism or Marxism" (Vatter 2008: par. r).
1 tum to May not because his book was the first secondary sourcebook on
Rancière (and remained the only one for a number of years), but because his

POLICE 1 75
aOt,ro,a.ch to Rancière proves invaluable for the stakes of
Rancière's concept of la police. appropriation of Rancière's thought for
anarchist purposes requires a very determined and distinct interpretation of
and this means that his argument helps me to work through what re-
ferred to above as the "poli tics of the police." Let me state the argument suc-
cinctly before unpacking it through my reading of May. "elevates" politics
to a pure form of action, while reducing police to an anti-political and implic-
itly repressive order of domination and injustice. This leads, 1 argue, to an
unproductive conception of "the police" in the service of a limited theory of
politics.
The steps to reach this conclusion prove subtle, because in so many ways
appears to be an exemplary reader of Rancière. Most praiseworthy is
refusaI to make Rancière into a philosopher; May sees clearly, and fre-
quently reminds his readers of, the political stakes of the Rancièrean project.
Bm on my reading, Rancière's politics are not the same as May's, and May is
thereby often forced to appropriate creatively -- or sometimes simply to misread
- Rancière in order to get to the anarchist conclusions that May had quite
clearly decided on from the outset. 9 This divergence likely begins with May's
account of "the police," which, as mentioned above, proves notable because
May gives the police so much attention. May begins by emphasizing Ran-
cière's own point, that the idea of a broadly understood "police order" can be
tied back to Foucault's lectures from the 1970s, where Foucault traces the
genealogical origins of the tenn in seventeenth- and eighteenth-century Euro-
pean thought. May shows the extent to which Rancière's use of the term over-
laps with Foucault's, while noting that Rancière develops the term quite
differently. But when it cornes to that development, May makes a very signifi-
cant interpretive choice. At just this juncture, he writes, "Policing, as Rancière
defines it, is deeply embedded in Western political philosophy" (May 2008:
42). May contends that Rancière locates the first occurrence of such policing
in Plato, and May explicates the concept of the police primarily through a
summary of Rancière's interpretation and critique of the Platonic philosophy
of order (May 2008: 42-43).
But this is a curious exegesis, because when Rancière himself introduces
the term police, he defines it in a context outside the project of political phi-
losophy. As 1 have already argued above, for Rancière, police names an order
of intelligible bodies, a distribution and counting of the parts of society. Police
is "a symbolic constitution of the social" (Rancière 2001: par. 19). Now, it is
very much true that in both Disagreement and "Ten Theses" Rancière goes on,
later, to make a crucial argument about Platonic political philosophy and po-
Iicing. But the claim is not that Plato is an example of policing. The claim is
that Plato's political philosophy substitutes police for politics, that the struc-
ture of Plato's philosophical project operates in such a way that it identifies
poli tics with police. And this identification of the two amounts to the elimination

1 The Lessons of Rancière


..-"v~~c,'V,-" as understands it. of 1nAI,h .....
nj

losophy, particularly in the Platonic project that Rancière names (as


rightly "archipolitics," the is substituted for politics. But this
substitution occurs as a of - as it were the philosophical
nowhere does Rancière say that it is essential to the constitution of a
order as such. And this fact renders own definition of the police ex-
tremely problematic. Immediately after summarizing the critique of Platonic
archipolitics, cornes to his most succinct statements on the police. He
writes, "In the end, the goal of policing is precisely that of eliminating poli--
tics" (May 2008: 43, emphasis added). Thus, this claim makes policing, by
definition, a mechanism for the destruction of politics. Unsurprisingly, then,
May proceeds to interpret Rancière's conception of politics as follows: the
goal of politics is to eliminate the police. Politics, according to May, not only
disrupts the police order (as Rancière clearly contends) but also says "no" to
that order in its entirety - something Rancière never asserts (May 2008: 49).
Such a reading fits perfectly, of course, with an anarchist project in which true
freedom and equality come only from the people and after the elimination of
govemment.
But that is getting ahead of the story. Let me step back then and try to assess
May's claim that the police order seeks the elimination of politics. This read-
ing of policing is tied directly to May's understanding of archipolitics; he re-
peatedly refers to the latter as a "form of policing" (e.g., May 2008: 43, 45).
That is, archipolitics is one configuration of the police in its effort to eliminate
politics; metapolitics is another, and so on. lO But to put the relation between
archipolitics and police in this way is to miss the entire bru nt of Rancière's
critique of the Platonic project. If the police order always and somehow natu-
rally sought the elimination of politics, there would be nothing especially
problematic or even interesting about Plato. It is the fact that political philoso-
phy seeks to replace politics with police that makes the Platonic (and Marxist,
and Hobbesian) project so dangerous. The critique of political philosophy
proves necessary, within Rancière's framework, because of the need to chal-
lenge, question, and alTest this substitution. Rancière's entire approach to "po-
litical philosophy," his effort to think politics outside its terms, therefore
depends on conceptualizing police as an order distinct from politics, but not as
an order with an inherent drive to supplant politics.
By failing to take account of the difference between an empiIically given
police order and the mobilization of the police within the Platonic project,
May tums politics and police into versions of matter and anti-matter: they can
never actually meet except in sorne final, universe-altering confrontation, but
they stand ultimately opposed to one another. 11 Hence May's claim that "the
goal of policing" simpliciter is the destruction of politics. Put directly, how-
ever, Rancière says no such thing about the police. In fact, he explicitly rejects
the notion of police or politics as pure forms in this way. After laying out the

POLICE 1 77
basic terms of as an to and rh ".,."r-rn.ra
Rancière reminds his readers that poli tics remains inescapably twined with
police. He writes, "We should not forget either that if politics implements a
logic entirely heterogeneous to that of the police, it is al ways bound up with
the latter" (Rancière 1999: 3 1; cf. Rancière 200 1: par. 21) Y
At this point one might accuse me of being either unfair to (in the nar-
rowness of my hermeneutic criticisms) or at least somewhat pedantic. 1 would
respond by arguing that the stakes of the se differences in conceptions of the
police turn out to be qui te high. They begin to emerge in May's reading when
he develops his anarchist account of democratic politics. May's anarchist
framework provides him with a structure in which to interpret sorne of the
most complex, subtle, and/or vexing elements of Rancière's thought. Primary
among these may be the fact that in taking almost everything we thought was
politics and calling it police, Rancière provokes his readers to ask the question
of when or where his poli tics happens. As 1 noted at the beginning of this chap-
ter, Rancière offers a direct response when he openly admits that politics
occurs "rarely" (Rancière 1999: 17). But for most readers, especially political
theorists, such an answer seems necessarily unsatisfying unless and until we
provide one of two possible supplements to it: (1) we can supplement the
response with an account of how to bring about such political moments; or
(2) we can add to this answer a further elaboration of why and to what extent
we should concern ourselves with phenomena that do not always add up to
political moments.
The first option seems the obvious choice for May. Indeed, one of the ways
in which Rancière's writings clearly do resonate with the project of anarchism
is on this point. Whereas so many readers of Rancière balk at the notion of poli-
tics happening so little, May positively likes the fact that democratic politics is
made rare in Rancière's account. Why? Because the rarity of politics fits with
the revolutionary and self-sustaining project of anarchism. May wishes to
define democratic poli tics as a process that enhances the lives of those who
engage in it. Democratic politics should be attractive to potential political actors
for just this reason, and the fact that there are few historical examples of demo-
cratic poli tics, the fact that democratic politics is rare - aIl this only makes it
more attractive. Why be an anarchist if anyone can do it or if it has aIl been done
before? May therefore wants to work within a space of political thought that
calls for an active self-creation of equality - and thereby perhaps implies or
even plans for the possibility of a revolution to come. 13 That Rancière defines
politics so as to make actual occun-ences of it scm"ce poses no problems for
May's political thought; rather, this dimension turns out to be an asset.
May's reading draws out a conception of democratic politics as a pure poli-
tics of the people thereby rendering democracy compatible with anarchism.
Anarchist thought, of course, depends upon maintaining a crucial distinction
between government and the people. 14 For example, May approvingly cites

1 The Lessons of Rancière


communist anarchism: name a of
conduct under which society is conceived without government" (Kropotkin
1995: 233; cited in 2008: like this shed light on read-
ing of Rancière's politics/police distinction. Police takes the place usually re-
served for government within traditional anarchist thought, and poli tics
the role of anarchist action. This leads to the essential twist that 1 have outlined
above: police must not just be disrupted or it must be eliminated.
Perhaps the most telling line in cornes early on, in his first and primary
elucidation of Rancière's concepts of politics and police - that prior to his
exploration of the history of anarchism as providing the "roots" for Rancièrean
poli tics. May turns from police to poli tics (just as Rancière does in Disagree-
ment), arguing that "politics ... is not a matter of how distributions occur." This
is straightforward enough: politics is not police; it is not about ordering and
distributing, not about the counting of those parts that already have a part.
May, however, completes the logic in a striking move: "Distributions are what
governments doL b]ut they are not what people do" (May 2008: 47, emphasis
added; cf. May 2009: 113). This final claim jibes perfectly weIl with the proj-
ect of anarchism, but it does not fit at aIl into the broader frame of Rancière's
project. In Rancière's terms we would have to say that, of course, distributions
are things that people do. Rancière goes further, insisting that "the police can
procure aIl sorts of goods" (Ranci ère 1999: 31). This, Rancière reminds us,
does not mean that we should confuse police with politics, but it ought to give
us pause when May claims that police stands opposed to people. Why does
May make this seemingly unwarranted shift?
The answer, unsurprisingly, likely lies in May's commitment to anarchism.
Anarchism, as May articulates it, requires a fundamental separation of spheres:
anarchist politics must commit itself to the elimination of aIl injustice, the
destruction of aIl orders of hierarchy. And this means, in addition, that anar-
chist politics must attempt to bring about a substantive equality. May force-
fully defends a conception of anarchism as maintaining a radical commitment
to equality. But again, none of this meshes very weIl with Rancière's under-
standing of politics or police. First, Rancière has himself actively resisted the
idea of reducing his thought to anarchism and has explained "the important
difference of perspective" between his thought and that of anarchist theory
(Rancière et al 2008: 175). May would no doubt point to the places in Ran-
cière's thought where he notes his explicit affinities with the anarchist tradi-
tion. 1 would certainly not deny these affinities, as there can be no doubt that
the vely idea of democracy as an·-archic proves central to Rancière's account
of politics. 1 would nevertheless insist on the distinction that Miguel Vatter
helpfully draws. Vatter distinguishes between, first, a certain sense of anarchy
that circulates in contemporary French political thought, in which "anarchy
means the absence of a metaphysical foundation for the distinction between
who rules and who is ruled ([hence] the necessary failure of aIl hegemonic

POLICE 1 79
pnnclplE~S of of "classical anarchism from
Godwin to Kropotkin" (Vatter 2008: par. 1). This first idea of anarchism plays
a role of undoubted importance in Rancière's but problems emerge, as
1 have shown, when tries to interpret Rancière within the terms of this
second tradition of anarchist thought.
Moving past the terminological and historical questions of anarchism,
would insist on a series of more substantive points. For example, when Ran-
cière argues, as 1 quoted above, that politics will always be "bound up" with
police, he continues as follows: "The reason for this is simple: politics has no
objects or issues of its own" (Rancière 1999: 31). These are not isolated
remarks; in later writings Rancière expands and develops this notion in an
attempt to clarify the stakes of his project. In a passage 1 quoted in the previ-
ous chapter, Rancière claims that "the opposition between politics and police
goes along with the statement that politics has no 'proper' object, that aIl its
objects are blended with the objects of police" (Rancière 20IIC: 5). For Ran-
cière, there is no politics without police.
In contras t, May seeks a politics not only de-linked from police, but also
t'ully self-referential and committed to the substantive ground of equality.
Equality, for Rancière, is nothing like a substantive ground. Deranty puts it
succinctly: "Equality is not an essence, a value or a goal" (Deranty 2003b: par.
I). Equality proves to be an assumption that can be verified, but it grounds
nothing at aIl in Rancière's thought. For Rancière, politics therefore proves to
be the demonstration of the assumption of equality; it occurs if and only if
there is an encounter between the logic of equality and the logic of domina-
tion. It i8 just this assumption of equality that makes possible the clash be-
tween heterogeneous logics; the verification of equality results from such
conilict. This explains why equality can be understood in Rancière neither as
a substantive good nor as an ideal telos. But this means, contra May, that poli-
tics does not occur because of equality, nor does politics achieve equality. AIl
politics does is to challenge, to thwart, to disrupt or dislocate, and perhaps
finally to change the police order. 15
As 1 have already discussed in the previous chapter, and as 1 will elaborate
in my final section, below, 1 think this is enough for politics to do. However, at
times, it is not enough for May. Because he insists on reading Rancière with
and against the grain of distributive theories of justice, May repeatedly asks
Rancière for normative grounds (e.g., May 2008: 119). And while he remains
very sensitive to Rancière's own understanding of equality as not providing
such grounds, May still frequently implies that perhaps equality serves this
function in (a reconstructed) Rancièrean thought, after an (cf. May 2008: 118).
Moreover, May frequently hints at the notion that equality might, someday, be
achieved as a substantive end (May 2008: 75; May 2009: 117).
Given that close readers of Rancière will tend to reject the foundationalist
approach of Rawlsian or Habermasian normative political philosophy, the

80 1 The Lessons of Rancière


continued demand for will strike
book as odd. Rancière' s conception of politics as dissensus, his understanding
of subjectivation as disidentification, and his rejection of political ontology all
seem to point away from the "nonIlative grounds" approach to political
And when Rancière argues famously in Disagreement that "parties do not
exist prior to the declaration of wrong," he would appear to take the ground out
from undemeath the feet of foundationalists a key point that discuss in
much more detail in the next chapter (Rancière 1999: 39). Here 1 simply want
to insist that the need for normative grounds is not intrinsic to Rancière's argu-
ments. Instead, this need is a product of May's own logic. May's project re-
quires what he caBs "nonnative force," because of the unique way in which
May comes to understand democratic poli tics itself. While May insists that
democratic politics will always be connected to the world in some way, he
defends a conception of politics that seems extremely inward-looking and
self-referential. That is, democracy for May seems to be primarily about de m-
ocratic actors. While politics might (or might not) change the world, its mean-
ing, according to May, cornes not from the world but from its agents. Thus,
although poli tics has effects, May insists that we should not "confuse having
social effects with the existence of politics." He continues with a striking
fonnulation:

A polities rnay or rnay not effect change. It is not in the consequence but in the
acting out of a presupposition of equality that politics occurs .... A dernocratie
polities is defined by the actions and the understandings ofthose who struggle,
not by the effects upon or actions taken by those the police order supports.
(May 2008: 72, ernphasis added; cf. May 2009: 1 I3)

Despite the fact that Rancière's own examples of politics aIl seem to involve
new partitions of the sensory realrn, the radical disruption and reordering of
the police order - "the essence of poli tics is to disturb" the police order, says
Rancière (Rancière 2001: par. 21) - May insists here that the ultimate defini-
tion of democratic politics is found not just in the actions but in the under-
standings of democratic subjects. 1 have called this a "self-referential"
definition of politics, not because it is circular, but because it refers politics
back to agents rather than to political effects.
It is just this dimension of May's argument that suggests a need for aug-
mentation in the form of nonnative grounds. In other words, if poli tics only
exists when agents struggle, there needs to be sorne leverage, sorne motivation,
sorne way to mobilize that struggle. Within the ternIS of May's logic, the sub-
stantive comrnitment to equality provides that nonnative edge. However, this
redefinition of democratic politics as emerging out of the self-understanding
of democratic actors has the curious but significant result of further denigrat-
ing the police. If poli tics refers only to itself, police is only important as a foil
for understanding politics. And May himself stresses this point when he insists

POLICE 1 81
orders. Here echoes his lines
from above, during his discussion of anarchism. He argues that anarchism
does not strive for a change in government, a new form of government or a
different set of people in power, but the overcoming of power. first quotes
Colin Ward: "Anarchism ... doesn't want different people on top, it wants us to
clamber out from underneath" and th en goes on to insert the following: "(Com-
pare this statement to Rancière's position that a democratic politics seeks to
undermine police orders, not change or modify them)" 2008: 96; quot-
ing Ward I982: 22).
This seemingly innocent parenthetical requires serious comment. First,
reference to "Rancière's position" is actually an internaI cross-reference
within May's own text, because it is May himself, not Rancière, who argues on
page 72 (cf. 43) ofhis book that politics does not seek to change police orders.
But neither there nor at the moment of this parenthetical statement does May
citc Rancière at aIl. And as 1 have shown, in Rancière's hands politics has no
autochthonous goal that it seeks of its own volition. Politics stands opposed to
police, but always in relation to police. And this opposition always manifests
itself in the form of transformed police orders, not undone police orders. But
second, what 1 say here is surely no surprise to May, or any other reader of
Rancière. Just three pages later, May writes, "Democratic politics ... does not
lead to a final state of justice but perhaps only to better conditions in a police
order" (May 2008: 99). What accounts for the difference, and apparent contra-
diction, between these two quotations from May? The tirst appears as a paren-
thetical commentary on a summary of anarchist thought; the second emerges
during an attempt to sort out Rancière's conception ofpolitics relative to anar-
chist goals. In other words, May's attempt to make Rancière play the role of
resource and support for anarchism leads May to stretchings and distortions of
Rancière's thought that prove readily apparent to most readers of Rancière-
including, at other points in his text, May himself.
Ultimately, however, my disagreement with May centers not on working
toward the "right" reading of Rancière. My primary concern is how to under-
stand the relation between politics and police - how to grasp that relation so as
to give an account of democratic politics. May articulates that relation as fol-
lows: "Politics arises as an alternative to the police. It must arise in a police
order, since otherwise there would be no need for it (because there would be
no inequality). It engages with the police, but not primarily as a struggle
against it. It engages primarily as an expression of equality" (May 20I2: I).
For May, any disruption of the police order, any production of a new distribu-
tion of the sensible, will only come about as side-effects of the internaI cre-
ation of equality that is politics. For me, this entire account severely depreciates
the importance of police orders and along the way undermines the power and
importance of democratic politics. May's commitment to a pure concept of
politics, one that is self-sustaining and never "parasitic upon the police" (May

82 1 The Lessons of Rancière


2012: leads - as argue in my below-
away from democracy. Democracy, as the paradoxical condition of !-,v.uU'"'''',
requires not a dismissal of police as the "other" to politics, but rather a re-
newed critical engagement.
In an effort ta draw sorne conclusions from my critical intelTogation of
interpretation of Rancière, 1 would suggest, then, that in the hands of
henneneutics, the opposition between police and politics undergoes
something of a Manichean transformation. Politics becomes a pure force, ut-
terly and radically distinct from and in opposition ta any and all police orders.
Zizek caUs this "ultrapolitics" (2006: 71, 75). As an obvious but significant
corollary, the concept of police is denigrated in May' s hands: one utterly loses
Rancière's sense of the police as "neutral" or "non-pejorative," and instead one
cornes to see police as the evil other to politics. The police arder is quite simply
that which must be destroyed. While it remains inevitable, it must take the
shape of the big Other precisely so as to motivate and mobilize the utopian
anarchist vision of the future. May' s reading resists any sense of a meeting
point between the logic of poli tics and the order of the police: politics, May
argues, sUl'ely stands opposed to police, but it becomes unclear how politics
would ever encounter the police. 16 That is, like any Manichean view, May's
rendering of the politics/police dichotomy precludes an active engagement
between the two realms; the only form that battle can take is the ultimate
batde, in which an anarchist utopia will replace aIl police orders once and for
aIl. 1 should stress that May's logic points toward this conclusion despite
May's own clear understanding that, in Rancière's thought, politics and police
must meet. May argues that the democratic dissensus creates two worlds.
However,

if the worlds were entirely distinct, if they had no point of contact, every po-
litical struggle would be a fight to the death. Every democratic political strug-
gle would reduce itself to a struggle between two competing visions, only one
of which could prevail. There could be no democratic politics that wasn't en-
tirely revolutionary. (May 2008: 112)

But this is just what happens within the tenns of May's interpretation of poli-
tics and police. Police bec ornes pure domination and politics becomes purely
revolutionary. Further, 1 would suggest that May articulates this problem so
accurately because it plagues his anarchist reinterpretation of Rancière. May
recognizes that democratic politics does not work this way, yet May thinks that
the Manichean separation of worlds is overcome by a sort of dialectical media-
tion of a third tenn: "There is at least one cornmon normative element in any
nominally democratic society that binds those who struggle and those against
whom they struggle. This common element [is] a commitment to equality"
(May 2008: 112). May's misreading ofpolitics and police (as radically separa te)
must therefore be supplemented by a rnisreading of equality (as substantive

POLICE 1 83
as have shown Rancière's own
does not need a third tenu, since in Rancière' s understanding, politics and
never form separate worlds; are always and already "bound up"
with one another. Rancière says it "There is no politics"
cière 201 IC: 3).
And it is in this binding, in this unavoidable meeting of the logic of poli-
tics with the logic of the police, that we may locate a viable and salient
thinking of "the politics of the police." In the final section, then, make the
case for this rendering of police, and 1 counter May's anarchist vision of
poli tics with a reassessment of a democratic politics that retains both impu-
rit y and impropriety.

In this chapter 1 have at times advanced my analysis of Rancière's concept of


"police" by trying to think through the meaning of the phrase "the politics of
the police" that is, by trying to unravel the implications that la police has
upon and for Rancière's broader sense of politics. Some readers of Rancière
might fairly contend that, given the fundamental opposition between politics
and police in Rancière's thought, there can simply be no politics of the police.
They might ask, "Is poli tics of the police an impossibility?" As 1 indicated in
the Introduction, Rancière is a thinker of paradoxes, and his is sure1y a para-
doxical account of politics. 1 have therefore used "politics of the police" as
something of a watchword, designed to keep the discussion focused not on
resolving paradoxes, but on working them out in a way that may only deepen
them. Put differently, if a poli tics of the police is impossible, the question for
readers of Rancière becomes, how do we think that impossibility?
My reading of the Rancièrean concept of the police starts with the basic fact
that politics is nothing other than that which stands opposed to and interrupts
any order of the police. This also means, simply but importantly, that police
itself can never be understood as politics; whatever we might say about a
police action (and there can be many reasons to praise actions within police
orders), we cannot caU it politics. 1 have then pushed that reading a step further
in my effort to demonstrate that the li ne between poli tics and police cannot be
understood as impenetrable. The divide must be porous; it must allow a cer-
tain type of movement over and back. Politics can never be purely other to
police, for the simple reason that politics itself cannot be pure. In Chapter One
1 argued against a three-term model for understanding Rancière's writings on
politics, and in this chapter 1 have offered a reading of "the police" consistent
with those earlier daims. It is only by insisting on a certain sort of con-
tamination of politics by police (and vice versa) that one can avoid the
need for a third term. If the line between politics and police is uncrossable, we

1 The Lessons of Rancière


up with a fixed account, account can
through the introduction of a third term. This is
tologize equality: he needs a normative
initiate the battle of politics the
interpretation of Rancière pushes the politics/police OPlJOSltlC)ll
extreme. In making polities out as a force designed to eliminate police,
required to defend, perhaps des pite himself, a vision of pure IJVÂHÂveJ.

here has not been to dismiss or even solely to refute but rather to use an
engagement with his quasi-dialectical rendering of the police/politics distinc-
tion so as to call for a reassessment and rethinking of Rancière's own catego-
ries. In other words, the fact that ends up, as showed to
such non-Rancièrean conclusions while working from Rancièrean premises
should lead us to a reinvestigation of those premises. 1 have tried to carry out
such an investigation here by taking seriously Rancière's own understandings
of la police, and by pushing his analysis into a deeper exploration of the mean-
ing and political stakes of his concept of the police. The politics of the police
is not a third tenn, but rather a way to think about the motion already present
between the tirst two terms - a way to consider what is at stake in Rancière's
conception of an impure politics, one necessarily, and always already, bound
up with police orders.
In closing 1 wish to delineate one other crucial dimension of la police and
its politics: namely, "the politics of the police" in the most banal sense. Jump-
ing off and fOl'ward from my discussion from May, 1 insist on a commitment
to and concern with the polities of the police in the quite elemental sense of
changing, transforming, and improving our police orders. As Alex Thomson
has very nicely put it, "There is doubtless much to do in tellliS of developing
better rather than worse forrns of police" (Thomson 2003: 1 1). When it cornes
to police, we require a democratic vigilance, not a utopian dismissal. Rancière
provokes his readers with his succinct assertion that "we do not live in democ-
racies"; instead, and as the only alternative, we live in police orders (Rancière
2006c: 73). May reads these claims as a utopian caU to fashion, someday, a
pure democracy. He reads them as not merely a critique but a denigration of
the world we do live in (a police order) in favor of the ideal of a true democ-
racy. Obviously, 1 read Rancière very differently. On my reading, we do not
live in democracies, and we never will. We never will, not because we will
never achieve what we ought to achieve, not because of failures on our part,
but because that is not what democracy is about. l ?
Democracy is not utopia. For this very reason, "struggle" means something
different to Rancière than it does to traditional Marxism - or even, 1 suspect,
to May. As my engaged reading of May makes clear, he places a certain sense
of "struggle" at the very heart of his conception of politics. More significantly
for my reading, May insists that the importance of struggle lies in its meaning
for those who do it. Rancière says almost the exact opposite when he treats the

POLICE 1
of a workers' strike. strike Îs the qumti;:sSeil1tlal eXélm1Dle of
'-'fHUUIIJL'V

gle" in the sense that uses it. But Rancière says that it is not a given that
a strike and therefore a certain sort of "struggle" in the traditional sense of
workers mobilizing against capital - is political. Rancière "A strike is
not political wh en it caUs for refOlTIlS rather than a better deal or when it attacks
the relationships of authority rather than the inadequacy of wages. It is poli ti-
cal when it reconfigures the relationships that determine the workplace in its
relation to the corn munit y" (Rancière 1999: 32). If a strike does nothing more
than demand better wages, then it simply does not produce a moment of poli-
tics, and in this case it can be clearly and legitimately understood as a negotiat-
ing move within a police order an attempt to improve the terms of that police
order without changing its structure in any significant way. On the other hand,
a strike may become political if it caUs into question the very organization of
the workplace. If the strike marks a moment of dissensus, a disruption of the
given partition/distribution of the sensible, then we can understand the strike
as a moment of politics. But this means that, in the sense used by traditional
accounts of workers' politics, the presence (or lack thereof) of so-called strug-
gle does not determine the possibility of poli tics. Struggle in this limited yet
traditional sense may be a central element of a police order. 18
Rancière has a different sense of struggle. He uses the word as another of
his synonyms for dissensus. Thus, when he writes, "To understand what de-
mocracy means is to hear the struggle that is at stake in the word," he points to
a more radical, more constitutive struggle than a mere dispute and negotiation
over resources (Rancière 2006c: 93, emphasis added). To hear "struggle" in
democracy means, as Rancière continues, to hear not only "anger and scorn"
but also, and more importantly, "the slippages and reversaIs of meaning that
[the word democracy] authorizes" (Ranci ère 2006c: 93). The struggle of de-
mocracy is a struggle over terrns, over places, over roles - over "propriety"
itself. For these reasons, the democratic meaning of struggle in Rancière' s
sense, which 1 am distinguishing here from more traditional Marxist accounts
of "struggle" as any collective action of workers, must not be understood as a
struggle that contains its own telos. What matters, as Rancière shows so lu-
cidly in his example of the strike, is not just that individuals come together, but
what is at stake, and what is put into question, in their coming together. This is
not a struggle merely for the sake of those who engage in it. It is the struggle
that is at stake in democracy.
Understood this way, we also see that democracy does not create equality,
and it does not eliminate government. Democracy is, instead, "the paradoxical
condition of politics" (Rancière 2006c: 94). And as 1 have shown in detail
above, a paradoxical politics is an impure politics. Democracy is both: "Demo-
cracy really means ... the impurity of politics" (Rancière 2006c: 62). This ex-
plains why democratic poli tics necessarily produces, and will continue to
create, "hatred of democracy" (Rancière 2006c: 94). Rancière fights against

86 1 The Lessons of Rancière


this hatred with almost every word he but says
not live in democracies, he should not be read as calling for us to "one
live in democracies. To put it succinctly, democracies are not ta be lived in.
Rancière "There speaking, no such thing as democratie gov-
ernment" (Rancière 2oo6c: 52). Once again, might take this as reason
enough to reject aU government, but 1 take it as reason to cultivate a demo-
cratie politics more, not less, attendant to the possibility of transforming the
police order.
Democratie poli tics is that which emerges in such a way that it disrupts our
given poliee order. Furthermore 1 would contend, without complaining, that
we will always live in police orders. Democratie poli tics proves to be a trans-
fomlative force, one that requires cultivation, care, and direction. Poli tics is
therefore absolutely vital. At the same time, however, we cannot forget that
"nothing is politieal in itself," and while "anything can become political," such
becoming political only occurs when the logic of equality is made to meet the
order of the police (Rancière I999: 32). ln the Rancièrean spirit of provoca-
tion, let me wrap up with something of an enthymeme. If we alllive in police
orders, and if "polities acts on the police," then the polities of the police (la
politique de la police) must be central not only to political theory but also to
our polities of ordinary life (Rancière I999: 33).

POLICE 1
Write everywhere.

y TWO PREVIOUS CHAPTERS have revolved around concepts doubtless


central to Rancière's writings, weIl known to Rancière's readers, and
extensively discussed in the growing secondary literature on his
corpus. ln this chapter 1 shift my attention to a term, lite ra rity, that has re-
ceived far less attention either from Rancière himself or from those who engage
his work. Mine, however, is not an effort to "fill a gap" for the sake of adding
to the commentary. 1 focus here on Rancière's idea of "literarity," and on the
related conceptions of the relationship between language and the human
animal that this concept crystalizes in Rancière's arguments, for a number of
distinct reasons. Despite the fact that Rancière himself has not discussed the
term at length, or made it central to his self-understanding of his project (in
contrast to concepts such as equality, politics, and le partage du sensible that
he has used in this way) 1 contend that literarity can be read back into and
through Rancière's wider body of writings in such a way as to shed important
light on his larger project. To put the point succinctly, the idea of an "excess of
words" provides a crucial counterbalance to Rancière's indispensable notion of
police orders, and thereby proves central to Rancière's philosophical project. 1
Above aU else, Rancière's thought always opposes itselfto philosophies of order,
and literarity must be grasped as the force that always already haunts such phi-
losophies. Literarity is that which philosophers always seek to, and always fail
to, contain. For this reason alone, a deeper analysis of literarity not only comple-
ments the discussions of poli tics and police from the preceding chapters but also
significantly extends and at times fundamentally alters our understanding of
what Rancière means by police/poli tics, democracy, and equality.
Moreover, in turning to literarity, the Lessons of Rancière takes a turn of its
own, as 1 begin here the pro cess of adding a certain torsion to my reading of
Rancière. Deranty's use of torsion (the act of twisting) as a metaphor for think-
ing about the "wrong" in Rancière's account of Disagreement can be extended
this and zo09; cf.
forthcoming). In solid mechanics, torsion describes not just a twisting, but a
in an the of across that object.
Torsion thus results in what is called "shear stress," which arises because tor-
sional force is to the object. wish to apply a certain
torsional force to Rancière's a force that will twist and extend it but
without leading to any shearing (in physics, an act of deformation that is irre-
versible). The torsional force takes two forms: namely, context and questions.
The concept of literarity makes it possible to contextualize Rancière's work in
a manner that proves both illuminating for Rancière's overall thought and pro-
ductive for political thinking - and does so in spite of Rancière's consistent
resistance to either the general project of contextualizing or the specifie efforts
to circumscribe his work (Ross zo09: z8; cf. Davis Z010). And 1 pose specific
questions to Rancière's work, ones that he does not explicitly wish to answer.
Or, thought the other way around, 1 appropriate his work in order to build my
own responses to such questions. My focus on literarity allows me to place
Rancière in dialogue with a series of larger considerations and concerns that 1
find crucial for how we think poli tics and theory today, but these are questions
in which Rancière himself has shown little interest.
This chapter locates Rancière's arguments about literarity in a number of
broader historical and theoretical contexts. First, 1 frame Rancière's famous
reading of Aristotle's more famous lines on "man as a political animal" in
terms of wider debates about language, and particularly about the proposed or
presumed relation between the human being, on the one hand, and the "fac-
ulty" of language, on the other. Rancière makes a series of radical, polemical
daims about the well-known passages in Aristotle's Politics, and he then
quickly shifts to a series of broader daims about his (Rancière's) "politics." In
this chapter, 1 pause before that shift and ask sorne broader, philosophical
questions about language -, questions that Rancière himself avoids, when pos-
sible, and resists, when necessary (ZOIIC: 16). Thus, 1 attempt to position or
locate, but surely not to p, Rancière within this very large set of philosophical
debates initiated by Aristotle. Second, 1 contextualize Rancière's work by
looking comparatively at other twentieth-century French theorists. 1 illuminate
Rancière's concept of "subjectivation" in relation to a wider and deeper series
of discussions and debates over the use of the related terms assujettissement
and subjectivation in Foucault's work. Here 1 focus mainly on the work of
translation in an effort to darify terms. Third, 1 also contextualize Rancière by
reading his daims about speech and language in the context of Jean-François
Lyotard's concept of le différend. 1 use Lyotard's explicit theory of language to
create productive comparisons and contrasts with Rancière's elliptical and po-
lemical remarks on language, discourse, and literarity.
This contextual work helps me to pose general questions about the relation-
ship between language and politics, and specifically to use Rancière's arguments

LlTERARITY 1
to ask what it would me an to THTn,,,,,,,, non-,mt:tlrC>Dc)centr'ic account of lan-
guage, and what the implications of such an account would be for politics. For
of course, a non-anthropocentIic conception of the language/poli tics
relation cannot be based on or fixed in an ontology. As have already noted in
previous chapters, Rancière's eschewal of ontology always places a certain
bm'den on his readers, as it makes it hard to get outside of his own polemics
without violating his work in a certain way. Here try this trick - the trick of
saying more and other than Rancière would say, but without rejecting or un-
dermining his own claims - by reinterpreting Rancière's arguments in the con-
text of debates over language and by posing to them the question of
anthropocentrism. 2 How can we think politics in a way that is non-anthropo-
centric and also not rooted in ontology? Rancière does not ask this question, of
course, but it can nonetheless be constructed from a reading of his work, and
his writings contain a provocative and rich answer - an answer that takes form
in the concept of literarity.
Rancière writes, "Humans are political animaIs because they are literary
animaIs" (Rancière and Panagia 2000: lIS). This chapter unpacks, unravels,
and twists this provocative and polemical claim. The link between the "politi-
cal animal" and the "literary animal" can only be drawn by literarity; to dem-
onstrate this point is to reveal the politics of literarity at the heart of Rancière's
project. lt goes without saying that the poli tics of literaIity is a democratic
politics and a politics of dissensus. Moreover, it is also a poli tics attentive to
the constitutive political importance of language a discursive politics. How-
ever, this is certainly not a discursive politics in the sense that deliberative
democrats use the word discursive. In clear contrast to a "discourse ethics"
approach that would seek to redeem liberal institutional practices "norma-
tively," the politics of literarity sharpens the contrast between liberal politics
and Rancière's democratic politics. Just as no philosophy of order can contain
"the excess of words," so in its own way does literarity undo liberalism. Liber-
alism cannot manage or process the force of literarity, despite its best efforts to
do so. Literarity opens a split between democracy and the liberal structures
that would narrow and constrain it. For aIl of these reasons literarity proves
centrally important to my broader reading of Rancière and to the arguments of
this book. It provides another layer to the arguments for politics and police that
l have already proposed in the first two chapters, and it sets the stage for my
account of critique in the final chapter.
This chapter traverses much terrain; hence a roadmap is in order. In the first
section, l start where Rancière and so many before him have started: with the
famous opening lines of Aristotle's Politics. l show that Rancière's interpreta-
tion of Aristotle has been fundamentally mis-presented in the English transla-
tion of Disagreement, making it more challenging than necessary for Rancière's
readers to grasp this, his central and opening argument in his most important
text on politics. Having clarified terms, l th en cast Rancière's polemical

1 The Lessons of Rancière


a deconstructive - one
istotelian logos as contaminated at its core. Only an impure politics can arise
from this impure of language. In the second section turn to a dis-
cussion of Rancière's understanding of the political and linguistic process
which political subjects are formed - subjectivation - and it is here that read
Rancière in the context of Foucault's significant work in this area. In the third
section 1 make the case for a non-anthropocentric understanding of language,
and 1 darify the meaning of anthropocentrism by distinguishing it from an-
thropomorphism, anthropology, and humanism. In the fourth section, 1 turn
away from Rancière, and toward a reading of Lyotard, using the latter's work
to make an argument for a non-anthropocentric understanding of the relation-
ship between the subject and language. In the fifth section 1 focus directly on
the central theme of this chapter, 1iterarity. 1 use my understanding of anthro-
pocentrism and Lyotard's challenge to anthropocentrism as leverage both for
bringing into the open Rancière's crucial arguments about literarity and for
demonstrating their centrality to his understanding of democratic politics.
1 show that poli tics requires, caUs upon, and paradoxically also brings about,
an excess of words. The sixth and final section offers a specific example of
what 1 call a "discursive polemical scene" in order to illustrate my broader
argument about literarity and democratic politics. Literarity is not ancillary to
politics~ the "excess of words" is not a secondary concern of Rancière's, or a
matter merely for questions relating to film and literature (cf. Rancière 2006b).
Rather, there is no politics without literarity.

Phonë, the Sign of the Logos, and Anthropos


Rancière opens Disagreement by telling his readers that he will "begin at the
beginning." And so the text quotes at length the "celebrated sentences" from
Book lof Aristotle's Politics, the place where Aristotle locates his own theory
of politics squarely - although even for Aristotle, still somewhat implicitly --
upon a theory of language. It is hard to underestimate the importance of Aris-
totle's words for Rancière's book: he not only begins his own work by quoting
Aristotle at length (an uncommon practice for Rancière) but also devotes the
first few pages of the book to a very detailed reading of those Hnes. And in a
sense the entire first chapter is an extended interpretation of the Aristotle pas-
sage: Rancière moves out from the level of Aristotle's sentences in the quote,
to the wider argument in Aristotle's Politics, and then on to the broader context
of Athenian democracy. This chapter gives my own particular and unique pre-
sentation of Rancière's reading of Aristotle, as 1 make the case for the central-
ity ofliterarity to Rancière's project.
ln order to get to that argument, let me begin with my own polemical daim
about these opening passages of the first chapter of Ranci ère ,s first breakthrough

L1TERARITY 1 91
on poli tics and his work most often read (or misread) as a book of political
claim: the first few lines of this book do not make sense. am not
suggesting that the 1ines are hard to follow, that their argumentative logic is
their meaning obscure, or their implications obtuse. am saying that a
careful reading of these lines shows them to be nonsensical: not just difficult,
but fully self-contradictory.
The problem lies not in anything that Rancière himself wrote, but in the
overall production of the English translation. In particular, let me spotlight the
coincidence of two different editorial choices made by the translator of Dis-
agreement, Julie Rose. Given the weight placed on Aristotle's words and the
significance of them for Rancière's argument, the words that actually appear,
in on page one of Disagreement prove to be something of a disap-
pointment. Rose says at the beginning of the notes that aIl translations in the
book are her own, but in the case of the Aristotle, that is simply not the case.
Rather than translating Rancière's French or Aristotle's Greek, Rose instead
substitutes a completely unattributed and uncited, but nonetheless verbatim
copy of T. A. Sinclair's 1962 Penguin Books translation (Aristotle 1981). This
is Rose's first choice. At best, it constitutes sloppy scholarship, but that is not
my criticism. 1 am concerned instead with the fact that Sinclair' s is a far from
literaI translation: it is a somewhat dated, somewhat wordy translation that
seems overwritten, especially for the purposes of Rancière's argument. The
Sinclair translation that opens the English version of Disagreement diverges
significantly, sometimes dramatically, from what appears in Rancière's text in
French.
These differences can make it more difficult to grasp Rancière's arguments,
but more importantly, they also lead to a fundamental misrepresentation of
Rancière's reading of Aristotle. This proves to be the case because Rose's
choice of the Sinclair translation has problematic implications for the render-
ing of Rancière's own opening lines. As 1 will discuss in greater detail below,
Aristotle specifies the distinction between voice and speech and the distinct
functions thereof. In Sinclair's translation, voice expresses, whereas speech
indicates. 3 Rose then makes a second choice when, in the lines that follow the
Sinclair version of Aristotle, Rancière uses the words indiquer and manifester.
Below 1 will show that these words, for Rancière, are clearly direct references
to the Aristotle lines that Rancière has rendered himself in French. At this
point 1 focus on only what the reader of the translated English text sees. Rose
has decided to translate indiquer and manifester as "indicate" and "express,"
respectively. Thus, in the English version of the text, one first reads of Aristo-
tle's distinction between voice, which expresses, and speech, which indicates,
and then one reads Rancière's reference to "speech, which expresses, while
the voice simply indicates" (Rancière 1999: 2). This is the specific claim that
makes no sense.

1 The Lessons of Rancière


And it does not make sense, me be
ment we have before us, Rancière appears to get Aristotle not just wrong, but
absolutely backwards. WhereAristotle Sinclair's English) says that the voice
expresses and speech indicates, Rancière says (in Rose's English) that the
voice indicates and expresses. Rancière is from con-
text, not presenting this as a refutation of Aristotle but as an apparent direct
reading of him -- a reading, repeat, that can only look completely incoherent
and unintelligent. Of course, the apparent contradiction is just that: apparent.
The contradiction in the English text proves to be utterly and completely a func-
tion of the two choices made by Rose. ignoring Rancière's own translation
of Aristotle, Rose fails to see, as 1 will explain below, that Rancière has iinked
the phëmë/logos distinction to a difference between indiquer and manifester.
And she also fails to notice that the English translation that she borrows marks
a distinction between expressing and indicating. Finally, by choosing to trans-
late manifester as "express" she completely reverses Rancière's meaning. To-
gether, these two separate decisions by Rose help to pro duce a text that, in
English, makes it appear as if Rancière is incapable of comprehending, on a
very basic level, the very lines of Aristotle that he has just quoted at length.
To see why Rancière's arguments are not nonsensical, and to get at his
actual reading of Aristotle, we need, then, to retum to the original texts. It
seems clear that in the original French, Rancière, unlike Rose, does offer his
own translation of Aristotle: Rancière cites no translator for his quote from
Aristotle and his rendering of the lines from Aristotle hews very closely to
Aristotle's Greek. 4
First, then, a direct translation of Aristotle' s Greek:

Of aIl the animaIs, only humans have speech [logos]. On the one hand, the voice
[phonë] falls to the other animaIs and is a sign [esti sëmeion] of the painful and
the pleasant. (Their nature has developed to this point: they perceive pain and
pleasure and can indicate these [literaIly: show by a sig n, sëmainein] to each
other.) On the other hand, speech [logos] makes visible [dëloun] the useful and
the harmful, so as [to make clear] the just and the unjust. This, in relation to the
other animaIs, is peculiar to humanity. The human animal alone has a perception
of good and bad, just and unjust, and so on. And the communion of these makes
a household and a city. (Aristotle 195T 1253a9-18, translation mine)5

In Aristotle we can see one sharp distinction, with other subtle differences
mapped onto it in tum. Fundamentally Aristotle means to distinguish between,
on the one hand, mere voice, phonë, which "faIls to" or "accrues" to aIl ani-
maIs, and on the other, "reasoned speech," logos, which only man possesses.
In addition, the relationship between types of animaIs and types of language
works differently in the two different cases. That is, phonë is given (passively)
to aIl animaIs, whereas logos is possessed (more activeIy) only by the human
animal. And finally, phonë and logos function differentIy. Phonë "is a sign"

LlTERARITY 1 93
sëmeion) or "indicates way of a (sëmainein), whereas logos can
make clear or make visible (dëloun). This means that phonë can only indicate
pleasure and pain, whereas logos can show or demonstrate what is just or
unjust. Unsurprisingly, these differences will prove central ta Rancière's read-
ing, and it will be no more of a surprise ta find Rancière, in his translation of
Aristotle, establishing his own distinctions in French, sa as ta set up that
reading.
Here is what Rancière's French translation looks like when rendered in
English:

Alone of aIl the animaIs, man possesses speech [parole]. Without a doubt the
voice [voix] is the method for indicating [indiquer] pain and pleasure. It is also
given to other animaIs. Their nature proceeds only up to a point: they possess
the feeling of pain and pleasure and they can indicate [indiquer] that among
themselves. But speech [parole] is there for demonstrating [manifester] the
useful and harmful and, as a result, the just and the unjust. Ifs this that is
man's own, compared to the other animaIs: man alone possesses the feeling of
good and bad, just and unjust. And it's the community of these things that
makes the family and the city. (Rancière I995a: I9, translation mine)

Thus, the logos/phonë distinction emerges here as the parole/voix distinction.


Where phonë in Aristotle's text can only show by a sign, can only indicate
(sëmainein), so voix, in Rancière's text can only indicate (indiquer). As logos
for Aristotle makes visible (dëloun), parole for Rancière demonstrates or man-
ifests (manifester). To repeat, the above passage serves as the opening lines of
Rancière's first chapter. Therefore Rancière's translation of the Aristotle estab-
lishes the terms for his argument. 6 The translation sets Rancière up for the
claim that lies at the heart of his reading of Aristotle, a reading out of which
Rancière weaves his own arguments about democratic politics. Rancière
writes,

The supremely political destiny of man is attested by un indice: the posses-


sion of the logos, that is, of speech, which manifeste, while the voice simply
indique. (I999: 2)

Let me note right from the start that by quoting the central ternIS of this claim
with Rancière's French left intact, and by presenting it in the context of Ran-
cière's own translation of Aristotle, we can clearly see that Rancière offers a
perfectly coherent, perfectly lucid interpretation of Aristotle. lndeed, in one
sense Rancière is simply repeating in this short quote what his careful transla-
tion of Aristotle already demonstrated: that the voice can only indicate while
speech can demonstrate. Yet he also adds one new dimension, which will prove
to be a very significant wrinkle: in addition to the paired terms that have been at
play from the beginning (animais/man, voice/speech, indicates/demonstrates)
Rancière suggests a third term. Above we see notjust the pair indique/manifeste,

94 1 The Lessons of Rancière


but also the extra term, indice.
in nature, between man and animal, between logos and phonë, between mani-
festing and indicating this difference must be "attested a sign [un indice]."
In this "attesting," in this reading of the difference between phonë and logos,
we will be able to locate the Rancièrean torsion applied to Aristotle. the
passage then fmther comment and elaboration, on a number of dis-
tinct levels.
Rose translates indiquer as "indicates" - a logical and obvious choice that
has the advantage of circling back to Aristotle (presuming of course that one is
working with Rancière's rendering of Aristotle, or using a more literaI transla-
tion th an Sinclair). While "indicates" proves unproblematic, in choosing to
translate indice as "sign," Rose loses the connection between indiquer and
indice that could be preserved with the word indication. Another choice for
indice would be "clue," and while this might be a less common, more idiom-
atic rendering of the word, it would connote exactly the sense of "reading the
sign" - that is, "looking for clues" - that, as 1 argue below, Rancière works
through in other writings that set their sights on Book 1 of Aristotle's PoUtics.
Cherche des indice, "searching for clues," also caUs to mind the police, a term
never far removed from Rancière's accounts of politics.
Rose then translates manifeste as "expresses." As 1 have shown, this causes
enormous problems given her use of the Sinclair translation of Aristotle: it
Ieads to the false appearance that Rancière has completely misread Aristotle.
However, even putting that issue aside, the choice of "express" for manifeste
proves questionable. "Manifest," "display," "show," and "demonstrate" aIl
make better options than "express" as aIl of these other terrns offer two signifi-
cant advantages. On the one hand, they an contai n, as "express" does not, the
sense of "make visible" that exists in Aristotle's Greek, dëloun. On the other
hand, these terms aIl connote the idea of appearing, of visibility, contained in
Rancière's sense of le sensible, a connection that Rancière himself draws just
a few lines later when he says that the difference between these two kinds of
animaIs is also the "difference between two ways of having a part of the sen-
sible [différence de deux manières d'avoir part au sensible]" (Rancière 1995a:
20, my translation),7 On the strict level of Rancière's first reading of Aristotle,
logos, unlike ph8në, has a role to play in le partage du sensible.
As 1 will show below, Rancière will exp and this account greatly (giving
ph8në its own role), but even here, as Rancière carefully maps out Aristotle's
distinctions, he already begins to go beyond the Aristotelian account. For Ran-
cière, the logos is not merely a possession of the political animal. The sign of
the logos makes politics possible, but for reasons quite distinct from Aristot-
le's. That is, Aristotle's account of politics cannot be dissociated from his fun-
damental distinction between a politicallife (the good life, eu zën) and mere
life (zën). This difference rests upon the possession of the logos (Aristotle
1996: 12S2b30; cf. Rancière 2001, par. 2).

LlTERARITY 1 95
On my own account of Aristotle - and Rancière's aside
briefty - the logos functions as a tool of sorts, one that the human animal can
wield so as to distinguish between good and bad. For citizens to fulfill their
functions within Aristotle's understanding of politics, must actively
engage in the life of the polis through judging and deliberating; to do so they
need the use of language (cf. Arendt 1958: 3). In this account, then, the very
nature of the political cornes to rest upon a certain conception of language,
because the political nature of Aristotle's citizens would prove unrealizable
without language. If one were to start with Aristotle's conception of man the
political animal using logos - and build out from it his the01)' of language, one
would run the risk of both instrumentalizing language (as the possession of
man) and presupposing an autonomous subject. That is, if language bec ornes
a tool to be wielded by human subjects, those subjects must precede the lan-
guage that they can possess and use for their own ends. This is a rather anthro-
pocentric theory that takes the subject as ontologically given and that tums
language into an object that can be controlled by people (Chambers 2003). It
is also, quite clearly, an approach that either proceeds by way of, or at least
invokes along the way, ontology.
Rancière therefore takes a distinct approach. Although he begins with Aris-
totle, Rancière, we might say, begins in disagreement. ln eschewing aIl ontol-
ogy, Rancière also rejects the idea of grounding politics in any theory of
language, and he most certainly denies Aristotle's conception. ln the "Ten
Theses on Politics" (2001), Rancière offers a deconstructive reading of Aristo-
tle, one that makes his "disagreement" with Aristotle stark. Here again, Ran-
cière says that political animaIs are marked by a sign (signe): "their possession
of the logos" (2001: par. 23).8 And Aristotle claims that this sign makes evi-
dent the distinction between the political animal and aIl others. Rancière, how-
ever, wants to provoke Aristotle, to polemicize. Thus, he asks a series of
questions of AIistotle that, l suggest, boil down to the following: How do we
read that sign? Is this signe that putatively marks the difference between
human political animaIs and aIl others so clearly legible as Aristotle seems to
think? Or, is the legibility of that sign precisely what politics would render
problematic? In response to Aristotle, in disagreement with him, Rancière
writes,

The only practical difficulty is in knowing which sign is required to recognize


the sign [which indice is required to recognize the power of indiquer]; that is,
how one can be sure that the human animal mouthing a noise in front of you
is actually voicing an utterance rather than merely expressing a state of being?
If there is someone you do not wish to recognize as a political being, you
begin by not seeing them as the bearers of politicalness, by not understanding
what they say, by not hearing that it is an utterance coming out of their mouths.
(2001, par. 23)

1 The Lessons of Rancière


Rancière in and as the of the
thing that Aristotle would wish to establish as the given starting pointfor poli-
tics. The doubling of indiquer/indique as both the capacity granted to humans
because possess logos and a180 that mark which serves to signify that
have such a capacity this doubling helps Rancière to explore what remains
opaque in AIistotle. Concomitantly, it gives Rancière the leverage to artieulate
his own account of politics. Democratie politics, for Rancière, must concem
our very interpretation of the sign of Aristotle's polities. And this means - in
contrast with the theOly of poli tics grounded on the ontology of language that
l derived from Aristotle above - that the Rancièrean "reading of the sign" must
always precede the very scene that Aristotle invokes. On Rancière's reading,
the relation between the subject and language established by Aristotle has been
undone, or even reversed: rather than starting with a given subject who would
then come to possess language, Rancière has placed the question of language
(of interpretation, of signs) at the beginning.
This radical and provocative reading of the Aristotelian "scene of poli-
tics" (the pre-politieal scene that makes politics possible) is starkly high-
lighted in the "Ten Theses," but Rancière had already offered subtle
indications of that reading in Disagreement. In that book Rancière shows
that we must attend to the primary act of drawing the line between phonë and
logos (and thereby assigning a role to each), and he therefore caUs into ques-
tion the literaI interpretation of Aristotle's phonë/logos distinction. He writes,
"The simplest opposition between logicai animaIs and phonic animaIs is in
no way the given on which politics is th en based" (Rancière 1999: 22). This
means, contra the earlier reading, above, that we cannot simply say that
logos has a part to play in the partage while phonë does not. Bonnie Honig
powerfully illuminates this passage in Rancière when she writes, "Phonë is
the name for the sonorous emissions of the excluded, and logos is the name
claimed by the included for their own sounds" (Honig 2010: 20). Politics
does not predetermine the logos/phonë distinction; rather, politics occurs
when those excluded from the logos lay claim to it. This account of a partage
that divides the sonorous into that which is phonë and that which is logos
points toward a wholly different "political animal" than that described by
Aristotle.
Below, l will show how and why Rancière caUs this politieal animal a "lit-
erary animal." Here 1 focus on Rancière's commentary on Aristotle's own defi-
nition of the political animal as that creature that possesses logos. Rancière
remarks, "In this limpid demonstration several points remain obscure" (1999: 2).
Aristotle would have us believe that the line between mere animality, on the
one hand, and unique humanity, on the other, proves quite stark and always
easily intelligible. However, the sign of the logos, says Rancière, "is one
thing"; the manner in whieh it functions is quite another. For Rancière, to
begin with Aristotle means to raise the very question that Aristotle wishes to

LlTERARITY 1 97
oec1ude: the between and 2; 201
DespiteAristotle's own dec1arations to the eontrary, the Aristotelian animal "is
split up from the very beginning" le: 2; cf. 2003a).9
This notion of the primary split, an originary tear, shakes the Aristotelian
foundations to their very depths. The political animal that grounds the Aristo-
telian model of politics, and that determines the implicit Aristotelian theory of
language, is always already riven, according to Rancière. Aristotle wishes to
make the logos, speech, the determining marker for politics, but as Rancière
says eloquently, "The speech that causes poli tics to exist is the same that
gauges the very gap between speech and the account of if' (1999: 26). Politics
ought properly, for Aristotle, to be about the difference between speech and
the making of mere sounds, but only the logos can distinguish between a po-
litical calI and a cry of pain. Thus, as Rancière famously puts it, "The initial
logos is tainted with a primai)' contradiction" (Rancière 1999: 16).

bjectivation

Rancière's understanding of the "tainted logos" in Disagreement can be full y


elaborated only in the context of a much more important discussion that runs
across a number of his writings, paIticularly those from the 1990s. Here 1 refer
to the political process through which and by which political subjects come
into existence - that is, through which they become visible on the political
stage. This process lies at the center of Rancière's thinking of politics. He
refers to it with the term la subjectivation. The choice of terms - that is, the
naming process itself proves complicated, and it holds a number of implica-
tions for any attempt to come to grips with what Rancière means by both poli-
tics and the subject. To get a grasp of what is at stake here, 1 tirst need to take
a step back to outline a slightly earlier context in which important philosophi-
cal work was done, in French, on the political process of becoming/making
subjects. Of course 1 am speaking once more of the work of Foucault, who
famously grappled with the two-sided nature of subject formation. Put con-
cisely, Foucault demonstrated that to become a subject in the sense of occupy-
ing a space from which one can exercise will and intention becoming a
subject in the sense of having the agentic capacity to act - meant at the very
same time to be subjected to power. Power, as Foucault famously argued,
flows through subject positions; as it does so, it both enables subjects to exer-
cise power and subjects them to that power. Foucault therefore describes
what for now 1 will caU a "subject process" that simultaneously involves, on
the one hand, domination, pas si vit y, and subjugation, and, on the other, au-
tonomy, activity, and resistance. Obviously an enormous amount of critical
attention has been given to this notion in the secondary literature on Foucault,
because this argument is at the crux ofPoucault's works on power, on discursive

1 The Lessons of Rancière


"''-'Ci,",,",,,,,, on and on this literature Foucault
been less than clear when it COlnes to both the terms that Foucault uses (in
French) to describe this process, and the terms that translators use (in English)
to render Foucault's French. And aIl of this matters, will suggest, because
much of this confusion has carried over to work on Rancière.
Luckily, the case of Foucault has been powerfully illuminated and clarified
in an article by Alan Milchman and Alan Rosenberg, where they carefully
explain that Foucault introduced two different tenns to describe and redescribe
the process of subject fOlmation. Milchman and Rosenberg usefully survey
that earlier literature on what 1 have called the "subject process" in Foucault,
and make two convincing arguments. They demonstrate, first, that this "com-
plex of issues ... has received too little attention," and, second, that with re-
spect to these significant issues, sorne of the translations and commentary on
Foucault prove, at best, muddled, and, at worst, wrong (Milchman and Rosen-
berg 2007: 55). In his early writings Foucault used the term assujettissement
to designate the process by which a subject is formed within a nexus of power-
knowledge relations, while at the same time the subject serves as a nodal point
that makes that power-knowledge nexus possible. This point is weIl known in
the English language literature on Foucault and much discussed, perhaps most
famously by Butler (I997b).
The following is less weIl known:

[A]round 1980, in discussing how we constitute ourselves as ethical subjects,


Foucault very deliberately introduces a new terrn: subjectivation. While assu-
jettissement peltains to how one is produced as a subject through the exercise
of power/knowledge ... subjectivation pertains ta the relation of the individual
himlherself; ta the multiple ways in which a self can be constructed on the basis
of what one takes ta be the truth. (Milchman and Rosenberg 2007: 55)

MilToring the broader shift in Foucault's thinking, and the supposed turn from
disciplinary power and institutions to the "care of the self' and the concept of
critique, Foucault transposes the terrns by which he describes the process of
subject formation. As Milchman and Rosenberg indicate, this shift has not
been clearly marked, and its implications not widely discussed in the English
language literature on Foucault. More importantly for my purposes here, in the
absence of specific attention to the clear change of terms - which Milchman
and Rosenberg delineate in their work - a great deal of confusion has perco-
lated. First, and early on, there was little or no consistency or cogency to the
translation of assujettissement. Many translators chose words like subjection,
which emphasized only one side of the dual process that Foucault sought to
describe. Even a close reader of Foucault like Butler, who proved extremely
attentive to the detailed and complex meanings of assujettissement, may have
helped to confuse matters wh en she chose in her discussions of Foucault to
translate Foucault's early terrn, assujettissement, with the direct English cognate

LlTERARITY 1 99
later French 10 This means that Butler's translation

has immediately and decidedly elided the entire distinction that Foucault
draws between assujettissement and subjectivation. And the problem is
exacerbated when English-language literature on the later Foucault writings
fail to make any mention at aIl of his use of subjectivation, or its differences
from assujettissement, and when a variety of terms appear in English - subjec-
tion, subjugation, subjectivation, subjectivization - without a clear sense of
their usage and meaning. For their part, Milchman and Rosenberg favor Niko-
las Rose's translation of assujettissement as "subjectification" as the simplest
term that does not close down Foucault's multiple meanings (Milchman and
Rosenberg 2007: 55). And l concur: to my eye it makes most sense in future
work on Foucault to use "subjectification" for his assujettissement, while using
the same English word, "subjectivation," for the French term subjectivation. 11
What has this to do with Rancière? First of aIl, even if there were no confu-
sions whatsoever, either in Foucault's terminology and translations or in Ran-
cière's own, Foucault's work would still provide a significant context for
understanding what Rancière is up to. At the very least, Rancière seeks to pro-
vide a nontraditional account of the political subject. He decidedly rejects both
the liberal model in which the autonomous subject precedes and serves as the
ground for political action, and the Marxist model (especially in a certain or-
thodox fiavor) of a subject that is fixed in place by a particular economiclsocial
structure. Thus, even if Rancière himself does not mention Foucault, it would
still make a great deal of sense to consider Rancière's arguments in light of
Foucault's, since both thinkers are trying to pro duce a complex account of
subject formation.
Moreover, Rancière's centralline of argument concerning political subject
formation travels on an axis that cuts through Foucault's writings. And Ran-
cière's arguments about the process of producing/becoming political subjects
has been confused in relation to Foucault's work. Let me unpack this last
claim. First, Rancière's most important, sustained, and detailed account of
this process cornes in Disagreement. There, Rancière uses Foucault's later
term, the French subjectivation (Rancière 1995a: 59). Rowever, for Disagree-
ment Rose has translated subjectivation as "subjectification." Renee, in
French, Rancière uses the same term, subjectivation, as Foucault chose in his
later writings, a term that Foucault specifically distinguished from assujet-
tissement; what appears in Rancière's text in English, however, "subjectifica-
tion," is the optimal translation of that earlier term, assujettissement. Thus,
reading the text in English, it is possible to conclude that Rancière is using the
same word in French as Foucault does in his early writings. That is, one might
naturally read "subjectification" as a translation of assujettissement, since
that is what the term renders in the English translation of Foucault's texts. But
this would be a clear mistake: Rancière himself does not use the term assujet-
tissement at aIl. 12

100 1 The Lessons of Rancière


as we saW in some detail in '"-'''''''1-'\.'-''"
choice and usage of and terms over the course of his
been rendered more difficult the fact that he has given many papers in
lish, sorne of which were 1ater translated back into French. this case too, the
issue is confused the fact that in 1991, in a widely circulated essay,
cière refers to this "subject process" in the very title of his essay
1995c). We know now that he willlater use the subjectivation
that it will be translated as "subjectification") but in the earlier text he
himself picked the word "subjectivization." This choice poses obvious
lems, because like his later trans1ator, Rancière here picks an English term that
has been associated with the early work of Foucault that one that has been
used to render Foucault's assujettissement (for example, Nixon 1997). Once
again, when discussing the process of subject formation, Rancière's texts in
English wou1d appear to be referring to assujettissement, despite the fact that
Rancière, in French, will only use the word subjectivation. Rancière's choice
to use "subjectivization" - as his own name for the process by which political
subjects appear - only exacerbates and deepens the potential misunderstand-
ing caused by the Eng1ish translation of Disagreement. A reader coming to
Rancière's work only in English translation would be forgiven for being some-
what 10st.
A dear sense of context (provided by my earlier discussion of Foucault)
and consultation of the later French texts can eliminate most of this confusion.
First, then, we see that in 1995 Rancière consistently uses the French term
subjectivation (1995a), and he confirms this choice when, in 1998, he trans-
lates his 1991 essay back into French, once again using subjectivation (1998a).
Undoubtedly, in his own French texts Rancière wishes to name this pro cess
subjectivation. This therefore is the term that 1 will use throughout my discus-
sion of Rancière in this chapter, and that 1 have already been using in previous
chapters. 1 also standardize this language, using the term in aIl quotations from
Rancière as the translation for his French subjectivation. 13
What then does "subjectivation" mean for Rancière, and what work does it
perform for his thinking ofpolitics? The links to Foucault are not arbitrary and
they must not be 1imited to the use of the same French word, since for Ran-
cière, as for the later Foucault, subjectivation is a pro cess that involves both
actions and agency on the part of a subject, as weIl as the production of a sub-
ject by forces beyond that subject's control. Subjectivation, for Rancière,
names both the process of becoming a subject and the political (and therefore
a1so socioeconomic) process in which such "becoming" cornes about. But
Rancière's account surely cannot and shou1d not be reduced to Foucault's.
Rancière focuses on a "gap" or fissure that p1ays little or no role in Foucault's
work. That is, for Rancière, subjectivation renders visible the gap between
one's identity within the police order (within the distribution of roles, places,
and status) and a certain daim of subjectivity through the action of politics.

LlTERARITY 1
And want to show here that a fuller of can pro-
vide a different angle into Rancière's conception of politics. The two concep-
tions mutually infOllli one another and remain inextricable: subjectivation
makes sense as a political process, and poli tics only takes shape in and
"modes of subjectivation" (Rancière 1999: 32). As have discussed in sorne
detail in the preceding two chapters, we know that Rancière takes the everyday
actions of poli tics and reinterprets them through the logic of the police. This
fundamental point in Rancière's argument will aways raise the same question:
if democracy is not a regime, if poli tics is not a set of procedures, and if poli-
tics has no proper sphere, no specific location, then how do we grasp the logic
of politics? We know that a police order is a particular type of social order, a
particular type of partition/distribution of the sensible. This point will always
lead readers of Rancière to ask: When does a police order shift of its own
accord (because of sorne set of negotiations and compromises from within)
and when is it fundamentally disrupted or altered by politics? ln the following
passage Rancière offers a response to both these questions, a response that,
while not using the word subjectivation, helps to illuminate what he means by
that term:

The stmggle between rieh and poor is not social reality, which politics then
has to deal with. It is the actual institution of politics itself. There is politics
when there is a part of those who have no part [le compte des incomptés thus,
there is politics when there is the count of the unaccounted-for], a part or party
of the pOOl'. Polities does not happen just because the poor oppose the rich. It
is the other way around: polities (that is, the interruption of the simple effects
of domination by the rich) causes the poor to exist as an entity.... Politics
exists when the natural order of domination is intermpted by the institution of
a part of those who have no part. (1999: II)

lt is the very coming into existence of the poor - those who ought not exist in
the standard order of the police, those who ought not count in the bureaucratic
logic - and their emergence on the stage where they previously had no role,
that produces politics. But the coming into being of the poor, of any party that
was not a party to poli tics beforehand - this is what Rancière calls political
subjectivation. "Political subjectivation" is therefore not a type of subjectiva-
tion, since subjectivation only occurs as a process of politics - through a poli ti-
cal moment that recreates or rearranges le partage du sensible. Subjectivation
occurs only through a disruption of the police order; you cannot have subjec-
tivation without this political conftict with the police.
If political subjectivation brings into existence/appearance those who
should not count, then, fundamentally, poli tics is a miscount: a counting of
those who do not count. Poli tics in this peculiarly Rancièrean sense attests to
the limits of counting models: it shows that the math of interest-group plural-
ism withinliberal regimes can never truly add up. Liberal regimes work under

102 1 The lessons of Rancière


the that aIl groups are counted. Â"-UjLL,","V~
ing of dissensus, his concept of disagreement, exposes what we might call the
constitutive outside of those regimes. This outside is revealed at just those mo-
ments when such regimes would invoke their pluralist logic: their "'J',UH.'Uf",

equation is predicated upon certain exclusions, but ones that can never be
named ones that are never visible as exclusions. Poli tics as subjectivation
might best be understood as a process of naming that which cannot be named
because it cannot be seen or heard. "Politics exists because those who have no
right to be counted as speaking beings make themselves of some account"
(Rancière 1999: 27)·
As 1 stressed in the Introduction, as a certain type of counting system, (as
a certain partage du sensible), liberalism is unprepared for the surprise that
is the emergence of a new political subject. In other words, the liberal count-
ing system can never predict this novel political activity in advance; it can
only count that which should be properly counted within the order of the
police. When the "poor," Rancière' s name for the "other" to the liberal police
order, "make themselves of some account," they expose the very mechanism
of that order as deficient, if not delinquent. And as 1 also hinted at in the In-
troduction, one must take some care to be clear about what Rancière means
by "the poor."
The word paar, as Rancière quite carefully uses it, cannot be presumed
in advance to name an extant socioeconomic class. When Rancière says
above that politics "causes the poor to exist," he distinguishes this political
understanding of the poor from that dominant conception of the poor as
produced within most police orders. According to the logic of the police,
the "poor" are already there to be counted; they are those who faB within a
certain box on the grid of income distributions, as measured by the census,
by polling, or by any other techniques of the police. In contrast, when Ran-
cière speaks of "the poor" in the long quote above, he points to a potential
mode of poli tic al subjectivation. The poor in this sense do not exist in the
order of classes priar to the interruption of politics. Thus, in one important
sense, "the poor" do not even exist at aIl. That is, within the police order,
there may be "poor people," but "the poor" is (one of) Rancière's name(s)
for the demas that only emerges when it makes its very claim to be counted.
Such a claim can only be made ta, and at the same time within, a system that
has never before counted them. "The poor" thus cannot be reduced merely
to "people without much money." The latter is a group within the bureau-
cratie, liberal order of the police, while the former is that which explodes
the logic of the police. Moreover, the former only comes into existence
upan such interrupting. The "poor" are only the poor alter they, the pOOf,
have declared they are the poor.
For this reason, Rancière will say that miscounts (i.e., politics) are "the
work of classes that are not classes" (Rancière 1999: 39). Again we see not

LlTERARITY 1 103
but the never
concems itself with subjects alone, but with subjectivation a process in which
subjectivity is produced through the transformation of an iden-
tity (Rancière 1999: 36). The police order already distributes lots in such a way
as to give each person his or her place in the order of things, a place tied to
identity. Political subjectivation is no t, then, the recognition or embrace of
identity, but the disruption of it; it is the production of a space between the
identity of the police order and a new political subjectivity a production that
is, above aIl, political. Subjectivation brings about subjects through the scene
in which poli tics interrupts the police order; the subject does not come before
politics, but rather cornes about through politics - a "coming about" that
cannot occur outside of language.
We are animaIs within language before we lay c1aim to the very power of
speech that would grant us the status of political animaIs. Language precedes
subjects, even if speech marks them as distinct, and this means that language
lays c1aim to us as weIl. Subjectivation names the process whereby those who
are of no account make a c1aim to be counted, thereby disrupting the math of
the police. As Rancière puts it, "The difference that political disorder in scribes
in the police order can thus, at first glance, be expressed as the difference be-
tween subjectivation and identification" (Rancière 1999: 37). In other words,
subjectivation is disidentification; it is the refusaI to not be counted (cf. Ran-
cière 1995c). The disruption of the equation between subjectivation and iden-
tity is the very pro cess that pro duces the fundamental mis cou nt that is politics.
By refusing the given identity of the police order, political subjects lay c1aim
to the fundamental equality that means they too - those who do not count -
must be counted. Subjectivation produces the addition of those who do not
count into the equation, which throws the equation out of balance - leading to
the miscount that is politics for Rancière.
Political subjectivation creates the fundamental wrong of politics, but this
is not to say that pre-given subjects then go on to take up politics. Subjectiva-
tion still rests upon the c1aim to the logos that is the starting point for Aristotle,
but for Rancière, "Parties do not exist prior to the dec1aration of wrong" (Ran-
cière 1999: 39). 1 wish to emphasize the crucial imp0l1ance of the language
used in this c1aim: it indicates that a wrong will be "dec1ared" but without a
ready-made subject to do the dec1aring. That is to say, the dec1aration cannot
be the dec1aration of one of those parties; the dec1aration must precede them.
But this means that language must come "before" the subject, that the subject
cannot preexist the language that "he" speaks. And if the dec1aration preexists
the parties who wouid dec1are it, then we see the impossibility of dismissing
an investigation of those dec1arations - of language, of phrases as a discus-
sion of mere words. In this case, to investigate language is always, potentially,
to analyze modes of subjectivation. It is to ask after that very formation of
parties to the disagreement.

104 1 The Lessons of Rancière


think as that from which (political and
emerge, and within which act and struggle, entails a significant displace-
ment of the Aristotelian conception of man as that being who possesses lan-
guage. Rancière concurs with this line of argument when he insists that
Aristotle's speaking animal is already decentered by a primary split (20IIC:
2). Thus emerges, even in Aristotle's text, the fundamental difference between
"speaking" language and "possessing" it. For Aristotle, this difference can os-
tensibly be located within the animal itself: a mere animal has phonë, but only
anthropos possesses logos. Rancière puts it more provocatively: "Speaking is
not the same as speaking" (201 IC: 2). In other words, the difference between
voice and speech is not given by an ontology; the difference must be read, and
surely only by way of a political reading. We might say that Rancière thereby
brings a third party to the Aristotelean scene of politics: whereas Aristotle has
only the brute animal and the poli tic al animal, Rancière says there must be
someone else to "read the sign" of the logos. But who could read the sign of
the logos other than man, the political animal himself?
Rather than taking this as an infinite regress - wherein we al ways need
"man" to identify "man" by seeing the sign of the logos - Rancière instead
concludes that the original logos is "tainted" in the sense that it is not pure, is
split from itself.14 My own reading in the section above stresses that the logos,
tainted though it may be, can never be held by a subject. This means that if
we are to take take seriously the Rancièrean notion of poli tics as struggle
within a realm of contingency (1999: 16), as contestation of the logic of hier-
archical social order (30), as the verification of the principle of equality (33),
then we must reject any model of language centered on possession by the
subject. This entails an eschewal of any account of the relationship between
the subject and language in which the subject precedes language and pos-
sesses it as a faculty.
To do this, 1 suggest, one most mobilize a conception of language as
never centered on and possessed by the human subject. Such a conception
actually serves to displace and decenter that human subject; it is within
language that the human animal is located, but this location is never a moor-
ing, never an anchoring. This way of thinking about language requires a
concomitant shi ft away from, and a critique of, anthropocentrism. 1 define
anthropocentrism not as an ideology of any sort, but as a characteristic or
aspect of any theory that, explicitly or implicitly, centers on anthropos, the
human animal. To caU an argument or theory "anthropocentric" means to
say that it centers on, revolves around, or channels meaning through the
subject. Further, 1 insist on distinguishing the idea of an anthropocentric
theOl'y or account from three distinct terms with which it shares the closest
family resemblances.

LlTERARITY 1 105
while point to a related and sometimes set of
anthropocentrism and humanism must not be conflated. Unlike anthropocen-
trism, humanism can function as a of ideology - or as a LH""'<-<f-"U
tlewegg!:;r 1977). In making the human being the measure of all value, hu-
manism does much more th an center its account on the human animal; it
judges and evaluates based on the human being. Across the many fomls it
takes, humanism insists on the importance if not uniqueness of "man" (or of
man and woman). Anthropocentrism, in contrast, need not give such norrna-
tive weight to the human being; it need only center its account on the human
animal, but without necessarily weighing values, choices, and options in terms
of "man." Of course, anthropocentric theories will often have a "humanist
bias," a sort of in-built weighting toward humanism, since to see the world
through the position of the human subject is certainly to run the risk of privi-
leging not only that perspective but also that subject itself.
Second, anthropocentrism must not be confused with anthropomorphism;
despite the very similar names, at certain times these two ideas prove to be
utterly at odds with one another, and at other times they produce productive
tensions. Anthropomorphism is a mu ch more widely known and commonly
used term; it names the process by which we project human attributes, human
forms, or human characteristics onto any nonhuman entities - whether they be
inanimate objects, animaIs, or spirits. Just as sorne forms of humanism find
grounding in a broader anthropocentrism, so sorne types of anthropomorphism
are based in an anthropocentric framework. That is, to project hum an charac-
teristics onto nonhuman entities, one often works from a human-centered per-
spective. What we might caU "anthropocentric anthropomorphism" means that
in speaking, acting, thinking, and viewing the world from and as a human
being, 1 see in nonhuman beings my very own characteristics.
However, as Jane Bennett has recently shown, the relationship between an-
thropomorphism and anthropocentrism can work qui te di fferently. Bennett
operates within a very different context than the one 1 am working in here. She
tums her attention to the vibrancy of matter, the vitality of things, and tries to
discem (and caU to the attention of other political theorists) the poweIful and
impOliant agentic forces of matter - of things as diverse as worms, trash, and
electrical grids (Bennett 2009). Like me, Bennett is struggling against anthro-
pocentrism. She worries that the repeated reassertion of anthropocentrism pre-
vents us from seeing nonhuman agency. By centering our political accounts on
the human animal, we frequently make the mistake of thinldng that only the
human animal can act, that no other creature has agency. It is here that our
very distinct projects cross or overlap: even though Bennett does not do so,
one can trace this mistake back to Aristotle. Aristotle's anthropocentrism
means that he not only sees the human animal as the only political animal, but
also attributes to that creature the possession of logos that, almost magically,
gives the human animal this unique power. Within this context, Bennett reveals

106 1 The Lessons of Rancière


the stark between and anthn)pC)ill{)rphi
she deftly shows that the latter can be used not in service of, but rather against
the former. That a bit of intentional anthropomorphizing lets us the
force of nonhuman Anthropomorphizing wind or cicadas
nett' s to the power of the nonhuman, and in so it
actually enables a shift away from anthropocentrism. We resist anthropocen-
trism by tuming to anthropomorphism. As Bennett says, "Anthropomorphiz-
ing ... oddly enough, works against anthropocentrism" (Bennett 2009: 120).
1 do not use Bennett's anthropomorphizing tactics for my project here, but her
work throws into stark relief the differences between those tactics and anthro-
pocentric accounts (and 1 share her resistance to the latter).
FinaIly, anthropocentrism must not be conftated with anthropology, or an-
thropologism. As 1 discuss in more detail in Chapter Four, in his so-called
antihumanist hermeneutic for reading Marx, Althusser made famous the "an-
thropological critique" of the early Marx (the critique of Marx for his "anthro-
pologism"). For Althusser, the problem in the early Marx was the bOlTowing
of an anthropology from Feuerbach, such that rather than conducting an im-
manent critique of the categories of political economy, Marx was reduced to
offering an external and ethical set of cIiticisms (Althusser 1969). It is much
less surprising that Rancière repeats this language in his early contribution to
Lire le Capital and apparently much more remarkable that he echoes it again
in his later cIitique of Plato. But when Rancière accuses Plato of being "the
founder of the anthropological conception of the political, the conception that
identifies politics with the deployment of the properties of a type of man or
a mode of life," he tums the pejorative category of "anthropology" onto
Althusser as weIl (200I: par. 28). The problem with an "anthropological" ac-
count in Rancière's sense is that it uses a theory of "man" in the service of a
philosophy of order: to provide a framework into which each person, with his
or her proper role, can be made/forced to fit. ln this sense, then, anthropology
is a mode of policing. 1 would add that Aristotle can be read as continuing the
"anthropological conception" that Rancière says Plato founds. Aristotle ad-
vances this tradition in another vein: not by giving us "democratic" and "oli-
garchic" types of men associated with their regimes, but by providing us with
the fundamental archetype of man as the political animal.
In my reading, however, Aristotle's possible anthropologism matters much
less than his clear anthropocentIism. AIistotle builds anthropocentIism into
the very core of his politics, by centeIing his theory on speech as the posses-
sion of man. This means not only that his theory is "human centered" in the
general sense of aIl politics starting with the political animal, but also that he
has attempted to contain that force that most threatens to decenter the human
being: language/speech/discourse, i.e., the logos. By putting the logos into the
hands of man, Aristotle's account marks a certain height of anthropocentrism.
In tum, in his inversion of this very relation through his novel and radical

LlTERARITY 1 lOi'
U Rancière brings to bear a direct - even if in his own
!cHVIUV,

hands, tacit - challenge to anthropocentrism. The very dissensus that poli tics
the fact that politics exists, but only exists in impure form - this
Hll·.lV'-'U,,",VLl,

lInIJO~SSll)le the Aristotelian version of anthropocentrism.

in

For these reasons, my focus lies not (with Althusser, and sometimes Rancière)
on humanism or anthropologism, and it lies not (with Bennett) on anthropo-
morphism. 1 am concemed instead with any conception or model of politics
and language that places the presumed human animal at the center, and pulls
political action or linguistic "usage" toward that center (see also Dillon 2003b).
To clarify the stakes of these differences for my broader reading of Rancière
(and my larger arguments), in this section 1 offer a brief reading of a text that
1 see as an exemplary effort to resist the gravitation al pull of anthropocentrism:
Lyotard's The Differend. 1 contend that Lyotard's self-consciously anti-anthro-
pocentric account helps to highlight a number of aspects of language that
remain implicit in Rancière's theory of politics. 1 make this tum to Lyotard
somewhat polemically, given that Rancière has himself been so highly critical
of Lyotard's work. 15 In a sense then, 1 begin with Lyotard in disagreement with
Rancière. However, in doing so, first, 1 actually follow Rancière's "method,"
about which he writes, "Disagreement is not only an object of my theorization.
It is also its method. Addressing an author or a concept first means to me set-
ting the stage for a disagreement, testing an operator of difference" (Rancière
20l lC: 2). And second, 1 do not intend here to bridge the wide gap that Ran-
cière himself has constituted between himself and Lyotard, as 1 neither focus
on nor contest Rancière's central criticisms of Lyotard - none of which have
much if anything to do with language. 16 My goal in this section, then, is cer-
tainly not to offer a refutation of Rancière or to privilege Lyotard; instead, 1
intend to "test an operator of difference," exploring Lyotard's theory of lan-
guage while respecting Rancière's refusaI to produce such a theory.
In the preface of Disagreement, in his first effort to define the term, Ran-
cière writes, "Disagreement is clearly not to do with words alone. It generally
bears on the very situation in which speaking parties find themselves. In this,
disagreement differs from what Lyotard has conceptualized as differend"
(Ranci ère 1999: xi). In the rather unorthodox style of The Differend - Lyotard
himself refers to it as his "naïve ideal ... to attain a zero degree style" the
book opens not with a preface, but with a "reading dossier" (Lyotard 1988:
xiv). If one looks up "thesis" there, one finds a description that would appear
to mesh with Rancière's brief characterization of the book. Lyotard writes,
"There are a number of phrase regimens [the set of mIes by which phrases are
constituted]: reasoning, knowing, describing [etc.]. Phrases from heterogeneous

1 The Lessons of Rancière


"'Al'i·an.A'~" cannot be translated one into the 17

Even though Lyotard insists that the pure translation of the logician - in which
the sense of a within one is transferred to the sense
of a new phrase within a different regimen - proves impossible, this
does not mean that links between phrase regimens cannot, and are not, made.
ln fact, almost none of text focuses on the notion that language
games are heterogeneous that was the thesis of Lyotard' s early cen-
tered as it was on deconstructing the grand meta-narratives of modernity
tard 1984; cf. Lyotard and Thébaud 1985). ln this text, on the contrary,
wishes to insist on both the constancy and the profundity of linking. In the
above quote, then, he goes on to say that phrase regimens "can be linked one
onto the other" (Lyotard 1988: xi, emphasis added, cf. 48-50).
Linking happens. Connections between and among language games are
made aIl the time, despite the fact that no rule of translation can properly and
legitimately govern such linking. Lyotard puts it this way: "To link is neces-
sary; how to link is contingent" (Lyotard 1988: 29). The Differend is not merely
an expansion of the notion that language games are incommensurable; it offers
an analysis of the implications of linking. And it would prove difficult to over-
state the fact that, for Lyotard, the issue of linking cannot be reduced to ques-
tions solely of language (of "words alone"), because linking must always be
thought as a matter of polities. On my interpretation, the important section of
the "reading dossier" is not the "thesis" but the "stakes."18 The stakes of the
investigation of phrase regimens and the genres of discourse that link them are
"to convince the reader ... that thought, cognition, ethics, politics, history, or
being, depending on the case, are in play when one phrase is linked onto an-
other" (Lyotard 1988: xii-xiii, emphasis added). My emphasis on politics
might seem skewed, given Lyotard's long list of items on which linking bears.
But it is Lyotard's emphasis as weIl: "The linking of one phrase onto another
is problematic and ... this problem is the problem of politics" (Lyotard 1988:
xiii). From the fact that pure translation between language games is impossi-
ble, Lyotard draws not the nihilistic conclusion that poli tics disappears or be-
cornes meaningless. Rather, he insists that since linking does and will always
occur in a field in which no link is guaranteed no translation can be under-
written by a meta-rule that would assure its arrivaI at the proper destination -
politics must grapple with the problem of linking.
These "problems of translation," these "miscommunications," entail that
meaning remains unfixed. Therefore the problem of linking bec ornes the pos-
sibility of political speech for those who have no part in the language of poli-
tics. And here Lyotard's arguments circle back around to intersect with
Rancière's polemical reading of Aristotle. The problem oflinking crosses with
the question of disagreement. Poli tics must always, thereby, concern itself
with the problem of language. And by this 1 do not mean "problem" in the
sense of limitations -- as if political problems could be solved merely by more

LITERARITY 1 109
information or dearer Hnes of communication. Rancière and agree in
their refutation of that daim. Instead, politics must be thought, as Foucault
suggests, through the language of problematization (I972; cf. 1997).
And the problem of politics is always a problem of language.
Lyotard makes this case in a powerful mannel' when he draws his own con-
nections (through language) between linking and politics. In the reading dos-
sier, Lyotard says that the "stakes" of his own text are,

[t]o refute the prejudice anchored in the reader by centuries of humanism and
of "human sciences" that there is "man," that there is "language," that the
former makes use of the latter for his own ends, and that if he does not succeed
in attaining these ends, it is for want of good control over language "by means"
of a "better language." (xiii)

Anthropocentrism offers, perhaps, the largest critical target of Lyotard's


entÏre book: although he never centers his argument there, he always returns
to it as a key issue. As 1 have discussed in sorne depth above, and as one can
see from the quote here, the philosophical problem of anthropocentrism
remains intimately bound up with the idea of a given subject that precedes
language, and th en cornes to make use of it as a tool or object. The Aristo-
telian notion of man as the rational animal who possesses the logos - this is
the "prejudice" Lyotard wishes to refute. Specifically, Lyotard hopes to
remove from the legacy of Kant and Wittgenstein the "cumbersome debt" to
what 1 have described as anthropocentrism - a debt centered, says Lyotard,
on the lingering idea of the "use" of language in their works (Lyotard I988:
xiii).19
How, then, does one move beyond this notion of language as the possession
of the ontologically given subject? It is difficult to do so directly without doing
sorne ontological work of one's own, and it therefore seems unsurprising that
Rancière's work sometimes skirts this question by presupposing it has an easy
answer. Lyotard confronts the question head-on. The subject, for Lyotard, is
never simply given. Rather, the subject itself must always be thought as the
product of the discourses of philosophy and the human sciences, a subject
position produced by an entire history. One must begin, then, not with the
subject, but with phrases (another name for what Foucault would call "discur-
sive practices," what Wittgenstein would caB "language games"). The ques-
tion then becomes, how does a particular set of phrases situate a given addressee
or addressor? "Do we, identifiable individuals, x, y, speak phrases or make
silences, in the sense that we would be their authors? Or is it that phrases or
silences take place (happen, come to pass), presenting universes in which indi-
viduals x, y, you, me are situated as the addressors of these phrases or si-
lences?" (Lyotard I988: II). The move from the first rhetorical question to the
second is the move away from anthropocentIism. In that shift Lyotard allows
for the possibility that phrases come before subjects, that the very possibility

no 1 The Lessons of Rancière


of the existence upon the of Dn]~aS{;S
subject finds itself/is found.
This cumbersome formulation itself/is found") seems the way
of avoiding active or passive voice in a language, English, that lacks the middle
voice. To say the subject "finds Ïtself' within language is to a faise sense
of active agency to the subject who acts outside of language. To say the subject
"is found" in language ascribes the same bogus agency to language, as if it
could trap the subject. To put it awkwardly, then, and say that the subject
"finds itselflis found" might suggest that the subject is middle-voiced. It is to
intimate neither that language fixes the position and role of the subject, nor
that the subject controls language.
Still, everything we think we know (the anthropocentrism here, as in so
much of grammar, is obvious) about language leads us into believing that it is
something for our use. Lyotard tums to the idea of phrases and phrasing, and
away from Wittgenstein's language games - the concept from which Lyotard's
work develops - so as to combat this notion. Despite the radical displacement
of the sovereign subject within Wittgenstein's later work, the notion of lan-
guage "use" still allows the pre-given subject to reassert itself. Lyotard writes,
"You don't play around with language ... and in this sense, there are no lan-
guage games" (Lyotard 1988: 137). Language games sound, despite aIl their
radical disruptive entailments, like something that we humans would play, just
as language always sounds like something we could have or hold.
To distance ourselves from this notion, Lyotard drives a wedge between
two different kinds of language. "At bottom, one in general presupposes a
language, a language naturally at peace with itself, 'communicational,' and
perturbed for instance only by the wills, passions and intentions of humans.
Anthropocentrism" (Lyotard 1988: 137). We confuse language with, for ex-
ample, French, but the former simply cannot be reduced to the latter. Language
in this former sense is what makes "a" language in the latter sense possible.
Here Lyotard bonows heavily from Heidegger (197 1), whose ideas on language
have sometimes been ignored because he was said to "mystify" language (e.g.,
Rorty 1991). It requires no faith in mysticism, however, to see that to think
language philosophically is to engage in a task far different from that of lan-
guage professors or linguists (Chambers 2003).
And the stakes, once again, prove quite high. Lyotard makes no effort to
downplay the enormousness of the task before us: "In the matter of language,
the revolution of relativity and of quantum theory remains to be made" (Lyo-
tard 1988: 137). Anthropocentrism reduces aIl language to a language, one that
can be known, mastered, and used for our plll1Joses. To resist anthropocen-
trism means to maintain the gap between languages, to keep open the possibil-
ity that subjectivity itself arises within and through language. Politics appears
in, and tries to preserve, this gap: "Poli tics consists in the fact that language is
not a language" (Lyotard 1988: 138). Politics, Lyotard stresses, cannot be

LlTERARITY 1 111
taken as a genre of discourse that would either be distinct from others, or at-
tempt to rule over them. He shares Rancière's disdain for the reduction of poli-
tics to consensus, what Rancière calls post-democracy (Rancière 1999, chap.
5). Politics bears witness to the differend (Lyotard 1988: 141), where the dif-
ferend "is the unstable state and instant of language wherein something which
must be able to be put into phrases cannot yet be" (Lyotard 1988: 13).
But if we have rejected an anthropocentric view of language, that "some-
thing" that must be put into phrases can itself be a subject. This very possibil-
ity entails that to inquire into the differend of language is to ask after the
potential emergence of the speaking being as such, since the "speaking being"
cannot precede the language in which he or she speaks. This is why the dif-
ferend is not, as Rancière would have it, about "words alone." lt proves impos-
sible to produce, even on Rancière's own telIDs, an argument about only words,
for if the subjects of a dispute can never precede that dispute, then an argument
about words is always already an argument about subjectivation, an argument
about politics.

Literarity

Rancière himself has no qualms about accepting aIl of this. In a sense he


simply buys into what Lyotard calls a "revolution in language" (Lyotard 1988:
137). Rancière is less interested in the revolution in language than in what it
brings about politically. He seeks to move on, to tackle substantive political
problems by articulating a more incisive understanding of language and poli-
tics. He names this conception "literarity." Rancière writes,

1 entirely agree that there is no subject before language or that wou Id merely
use language in an expressive way. 1 clearly stated [in prior works] that a po-
litical subject is in fact a process of [subjectivation], that the "name of the
subject" is only an element of this process and that the "we" who utters a name
and the "we" that it names do not make one identity. So the issue at stake is not
[subjectivation] versus subject. It does not lie in the conception of linguistic
practice. It lies in the conception of the act of interlocution and of its condi-
tions. It lies in the conception of dissensus in which a theory of language is
embedded. (2003c: 8)

1 emphasize the Lyotardian rejection of anthropocentrism and 1 push differend


and disagreement closer together, not to suggest that Rancière's work can be
reduced to Lyotard's frame. Rather, at least on this issue, 1 do not challenge
Rancière's effort to describe his political project as unique (see Rancière
20IIC). Nonetheless, the critique of anthropocentrism forms a crucial back-
ground for grasping the notion of literarity. The "conception of dissensus"
that Rancière articulates must be understood as presupposing that there "is no

112 1 The Lessons of Rancière


before H.4L'F,'-"U.F,'-" lose of what to be
notion (and perhaps Rancière sees as more banal), that language precedes the
subject, would be to fundamentally misread Rancière's thinking of dissensus.
critique of anthropocentrism offers a helpful frame for the
most philosophically radical and politically significant element in Rancière's
work in and around language, literarity.
Rancière does not use the specific term literarity until quite late in his writ-
ings. The word first appears at the end of a book published in French in 1998;
it therefore shows up in Rancière's writings only after the better-known writ-
ings on politics from the mid-1990s. Nonetheless, despite the fact that he does
not use the exact word literarity until 1998, that word itself seems to name a
broader phenomenon that can be traced much farther back into Rancière' s
corpus. As 1 suggested at the outset, 1 contend that "literarity" names the coun-
tervailing force to "police." Before specifying that argument, let me first trace
a few key locations in Rancière's work where we witness the concept in
nascent form. 20
Genealogically, we can trace this emergence back to Rancière's reading of
the Phaedrus. A substantial discussion appears in The Philosopher and His
Poor (2004b [1983]), where Rancière first stresses, perhaps against Derrida,
that Plato does not merely contrast the oral with the written, but distinguishes
"the living discourse of dialectic method [from] the fabricated discourse of
rhetoric" (2004b: 40). Rancière refuses to read Plato as merely privileging
speech over writing (cf. DelTida 198 1; see also Kollias 2007). For Rancière,
the relation between dialectic and rhetoric proves much richer, deeper, and
more complicated than that. What makes the latter discourse, written rhetoric,
mute and dumb - thus, inferior - goes weIl beyond its status as a written mark.
Rather th an focus on the written nature of this ostensibly subordinate dis-
course, Rancière emphasizes its loquacity: it does not know when to shut up;
it goes on and on (see Parker 2004: xvii-xviii). Worse, "it does not know to
whom it is speaking, to whom it should speak, who can and cannot be admitted
to a sharing of the logos" (Rancière 2004b: 40). Written rhetoric thus mns the
risk of failing to honor the distinction between the mere animal and the politi-
cal animal. The philosopher's logos knows when to keep quiet, and when to
silence others, but the "mute discourse" just keeps talking.
A form of this talkativeness also centers Rancière's brief remarks on the
Phaedrus, in The Ignorant Schoolmaster (1991 [1987]):

Understanding must be understood in its true sense: not the derisive power to
unveil things, but the power of translation that makes one speaker confrant
another.... Despite what the Phaedrus teaches us, there are not two kinds of
discourses, one of which cou Id be deprived of the power to "help itself" and be
condemned to stupidly repeat the same thing. AlI words, written or spoken, are
a translation that only takes on meaning in the counter-translation. (63-64)

LITERARITY 1 113
Rancière Plato and also that
speech and writing "are not two kinds of discourses," and thus the former can
certainly never be elevated above the latter. 21 AIl language is translation. Com-
munication, moreover, cannot be reduced to the transmission of meaning be-
tween given subjects, since, for Rancière, meaning arises out of the process of
translation and "counter-translation" (cf. Derrida 1982). The very term coun-
ter-translation suggests to me the interminability of the process of translation
itself. Every translation must be translated again, such that meaning can never
be fixed once and for an. There is an excess of words.
Rancière continues the discussion with a quotation from Joseph Jacotot,
which Rancière uses as indirect commentary on the famous Platonic dialogue
(though there's no evidence that Jacotot meant it that way). Jacotot daims,
"Man is condemned to have feelings and to be silent, or, if he wishes to speak,
to speak indefinitely since he must always rectify by adding or taking away
from what he just said" (Ranci ère I991: 64). Here we see the problem of
rhetoric, the problem of the written mark (that it cannot cease to speak), con-
structed as our problem we cannot cease to speak, because our speech al ways
proves inadequate to, and incommensurable with, our meanings, our feelings.
How to describe this speaking indefinitely? How to name this talking too
much? Rancière's first name for the phenomenon appears obliquely, in that it
describes a slightIy different problem: not the problem of mute discourses that
cannot shut up, but the problem of actors in history, "of a living person who
speaks too much" - too much for the requirements of the historian (Rancière
1994 [1992]: 24)· "Excess of Words" is the titIe to the third chapter of Ran-
cière's The Names of History, a chapter that aIiiculates, again, the predicament
produced by people who speak too much, who cannot stop talking. And here
Rancière uses the term literariness, a concept related to but still distinct from
"li terarity" (Rancière 1994: 76).22
But the "excess of words" is no mere problem or predicament. "Excess of
words" not only describes the fundamental condition of the human animal, but
also constitutes the fundamentally political nature ofthat animal. Contra Aris-
totIe, it is not our "possession" of the logos that enables us to become political
animaIs; it is, instead, the very fact of "literariness" - the fact that we are
always bound up, first, within the circulation of words - that makes us always
and already political. In Disagreement Rancière writes, "The modern political
animal is first a literary animal, caught in the circuit of a literariness that undoes
the relationships between the order of words and the order of bodies that de-
termine the place of each" (Rancière 1999: 37; see Robson 2009). Literarity - a
word that Rancière almost, but not qui te, uses here - is at once ours and not
ours. Literarity makes disagreement possible; it serves as the condition of pos-
sibility for disidentification.
Clearly, then, literarity is a political telm for Rancière, and must be care-
fully distinguished from "literature" and "literariness" (Rancière 2004c; cf.

114 1 The Lessons of Rancière


Rancière 20 Ile; see also Rockhill texts written since
starts to draw these distinctions, and along the way he sharpens both the mean-
ing of literarity and its importance for politics; the "politics of literarity"
to materialize in these texts. Rancière subtitles Flesh ofWords (2oo4a [I998])
"the poli tics of writing," a book built on lengthy discussions of the
modern writers and concerned with arguments about "literature." He distin-
guishes this "interest ... in literature as the specifie case of the art of writing
within a historical system of art," from his arguments about litermity
ci ère, Guénoun, and Kavanagh 2000: I4). In a complementary yet distinct
manner, the politics of literarity manifests itself in the space of poli tics, of
political theory; the political project centers on literarity as a category separate
from both writing and literature. 23 Rancière introduces the actual term literar-
ity toward the end of this book, where we find him once more reading Plato's
critique of writing as a mode of speech inferior to the living logos. And again,
Rancière stresses the disorder created by writing, its ability to confuse the
orders that Plato strives so hard to maintain. In this con tex t, Rancière writes,
"The disordering peculiar to writing confuses this [Platonic/philosophic] hier-
archy, introduces dissonance into the communal symphony" (Rancière 2004a:
103). Writing always sparks a desire, takes its readers to a place that was not
destined, and along the way al ways has the potential to undermine the police
order. "Writing" always erodes Platonic systems; in light of this disordering,
Rancière says, "1 suggest we give this disorder the generic name of literarity"
(Rancière 2004a: 103).24
The explication of this term appears in other places, particularly in inter-
views Rancière gave in 2000. There, Rancière explains that he now uses liter-
arity to name the Platonic problematic concerning living speech and mute
discourse: "1 caU 'literarity' this status of the written word that circulates with-
out a legitimating system defining the relations between the word's emitter and
receiver" (Rancière, Guénoun, and Kavanagh 2000: 7). Literarity thus names
the excess of words, the disordering principle, the power of the dërnos that will
al ways undo Plato's kallipolis. 25 In the same interview, Rancière suggests we
link "literarity" to his thinking of le partage du sensible, thus emphasizing the
ordering/disordering effect of literarity "the distribution of words, time,
space" (8). Though Rancière would surely deny it any sort of ontological
status, literarity in its own way proves more fundamental than any theOl)' of
language, or any theory of man.
The idea of literarity thereby returns us to the question of the poli tic al
animal possessing the (tainted) logos. It is here that we can make sense of
Rancière's most provocative daim: namely, that in the beginning, we are liter-
ary animaIs above aH. The concept of literarity allows Rancière to cut through
his multiple readings of Plato and Aristotle, and to draw to the surface the
stakes for our thinking of both politics and language. Like "disagreement," the
term literarity operates as a powerful polemical tool, and Rancière wields it to

LlTERARITY 1 115
his most The of thus
parallels the politics of disagreement, in that Rancière's alternative to a tradi-
tionai of poli tics" takes shape through his use of and articulation of the
term. In response to an interview question posed to him directly on the ques-
tion of language, Rancière offers the following crucial elaboration:

approach begins from a different reading of Plato's critique of writing.


Here, the central question for me rests upon the political1y fertile potential of
the opposition between two differing accounts of how words circulate. The
"silent" word of writing, according to Plato, is that which will sway no matter
what - making itself equally available both to those entitled to use it and to
those who are not. The availability of a series of words lacking a legitimate
speaker and an equally legitimate interlocutor intenupts Plato's logic of "the
proper" - a logic that requires everyone to be in their proper place, partaking
of their proper affairs. This "excess of words" that l call1iterarity disrupts the
relation between an order of discourse and its social functions. (Rancière and
Panagia 2000: IlS)

Here we can see more sharply why Rancière remains so insistent that his work
has nothing to do with philosophy of language, why he remains so adamant
that his conception of politics neither requires nor implies a theory of lan-
guage. In the terms of literarity, Rancière has little interest in writing as such -
and just as little concern, in this sense, with Plato's condemnation of it. As 1
have shown from the beginning of this section, the question does not directly
address speech or writing per se. The political issue at stake centers on the
availability and accessibility of "writing" to everyone. The excess of words.
Even those who are not entitled to the logos, those whom Plato's "order of
discourse" seeks to relegate to nonspeaking status, even they have access to
"writing." Unlike the dialectic, which is only made available to those who
properly deserve it, "writing" cannot be controlled - it goes places it should
not go, including into the hands of those who should not, according to Plato
(according to the philosopher) wield iL Plato's "logic of the proper" seeks to
establish the order of the city through the order of discourse. But the very
excess of words literarity - always disrupts this ordering. 26
Thus, with a heightened sense of literarity, we can now turn to Rancière's
"conclusion," which turns out to be no conclusion at aH, but rather a return to
the beginning, a return in disagreement. "We can conclude, then, that humans
are political animaIs because they are literary animais: not only in the Aristo-
telian sense of using language in order to discuss questions of justice, but also
bec au se we are confounded by the excess of words in relation to things" (Ran-
cière and Panagia 2000: IlS, emphasis added). Aristotle claims that we are
political animaIs because we have possession of the logos and can use it for
political ends. Rancière does not deny that we do, in fact, use language in this
way - to communicate, to deliberate, to judge. Yet our fundamentally political

116 1 The Lessons of Rancière


nature shines because we are
and caught up in the excess of words, of literarity.
very speaking thwarts any effort (Platonic or
since the speech of "man" attests to the excess of
ing, to the very literarity that makes us political animaIs in the first

If we are to work with Rancière's configuration of democratic politics, pro-


pose that we as readers of Rancière must do much more than re-
peated accounts and appraisals of the terms he proposes politics and police.
Genuine concem with democratic politics (especially over and against liberal
interest-group politics) requires us to attend to the very literarity ofpolitics.ln
a sense, literarity proves constitutive of an untimely Rancièrean politics, but at
the same time, literarity, the excess of words, surely cannot serve as ground for
politics (ontological or otherwise). One way to "attend" to literarity, to take
stock of its force, to mark its effects, is to locate and analyze those time-spaces
in which an excess of words disrupts the link between the order of discourse
and the order of bodies. How is the order of the city, the order of the police,
disrupted through literarity so as to fonn a new partition/distribution of the
sensible? "Disagreement" is clearly just another name, in Rancière's writings,
for democratic politics. Moreover, 1 contend, in an extrapolation from Ran-
cière, that disagreement often occurs through the clash of intersecting discur-
sive practices. The marking and naming of this conflict is another way of
getting at, another way of producing, the excess of words, literarity. To put it
polemically: no disagreement without literarity.
Two of Rancière's favorite examples of politics exemplify the centrality of
literarity to his political project. Both Blanqui's dispute with his judge over the
name proletariat and the May' 68 Paris demonstrators' "wrong" identification
through the claim "we are all Gennan Jews," point toward literarity by illus-
trating a certain kind of battle over the production of a subject position (Ran-
cière 1995c: 66-68; Rancière 1999: 37-38, 59). In both cases politics occurs
in and through language, and politics is at stake through the medium of lan-
guage. At the same time, however, in each ofthese cases we have an individual
or group actively laying claim to fundamental democratic equality by taking
up a particular name themselves. The examples, then, can easily be read in
such a way that they cut against Rancière's central claim that "parties do not
exist" before they declare the wrong that brings them into existence. Given this
potential disparity between the larger argument and the specific examples,
1 now want to ask: What happens when the battle of discursive poli tics cannot
be waged at such an individuallevel when there really is no party prior to the
wrong? What happens when those who have no count are kept from being

LlTERARITY 1
a battle over and excess words that
outside the realm of legibility? As literai)' animaIs, we find ourselves caught
up in the excess of and politics happens too, even before we
emerge as animaIs. We are political, Rancière suggests, because we
find ourselves within literarity. Here 1 contend that Rancière's incisive and
powerful depiction of "polemical scenes" (Rancière 199 1: 4 1) as those spaces
that force the emergence of the contradiction between the logic of the police
and the logic of politics, need not and should not describe only cases in whieh
the parties to the dispute are already apparently present. Sometimes the battle
of democratic politics occurs before the appearance of the subject on that
scene. In such cases, the polemical scenes are themselves constituted and
brought to bear at a different level: namely, at the very level of discourse -
within the excess of words that would precede the emergence (or foreclo-
sure) of the subject.
Thus, 1 conclude my argument about the central importance of literarity to
democratic poli tics by giving an account, a rich description, of one such po-
lemical scene. In this example we see again, as we did in Chapter One, the
importance of immigrants as political actors, as 1 draw the background for this
example from Paul Apostolidis's recent research on immigrant workers, par-
ticularly his work at an Ameriean meatpacking plantY Located in the eastem
p0l1ion of the state of Washington, the plant is owned by a giant corporate
entity, Tysoniiowa Beef Processors, and staffed aimost exclusively by Mexi-
can immigrants (sorne illegal, most now legal), whose lives, work, and struggle
within the union Apostolidis studies in sorne (and sometimes gruesome) detail.
1 want to appropriate certain dimensions of Apostolidis's work in order to map
out a Rancièrean polemical scene. 1 do this by reading the intersection of two
distinct narratives that his work draws to light: first, the narrative of immigra-
tion, and second, the narrative of work and activism. Read through Rancièrean
lenses, Apostolidis's interviews and interactions with these workers makes
palpable the extent to which - as "immigrants" - these Mexicans and Mexican
Americans are of no account. Yet, at times, as "workers" and certainly as "ac-
tivists," they make their democratie claim to be counted. My interpretation
describes, then, the potential emergence of the immigrant-worker-activist as a
democratic subject. My reading of the case tries to highlight the literarity pro-
duced by the excess of words that marks these workers.
Narratives of "the immigrant" have, of course, been central to the building
and consolidation of power within the American nation-state. Those nan'atives
have combined a powerful immigrant xenophilia - for the immigrant whose
ethics of hard work and Old World values fuel the belief in class mobility and
build strong communities - with a toxie xenophobia against the immigrant-as-
other who is a parasite on the system and a roadblock to progressive social
change (Honig 2001: 4-18; cf. Apostolidis 2005: 25; cf. Beltrân 2009: 597).
On my reading, Apostolidis's work raises the following central question: how

n8 1 The Lessons of Rancière


does the
lives and subject formation of the immigrant worker? In the language of Ran-
cière's thinking of politics, this question leads to another: what is the
I-Ive;, ..:nU'H

for the emergence of politics the confrontation between the


order that the immigrant in his or her place, and the
tics, which asserts the fundamental equality of the immigrant who is also a
"worker" and potentially an activist - given these existing discourses?
The immigrant worker cannot attain the "equality of speaking beings"
(Rancière 1999: 33) simpliciter, because the immigrant worker will always be
constrained to speak from within the discourse of immigration. Speaking,
contra Aristotle, does not necessarily mean a realization of equality, if one is
forced to speakfrom a particular location. To speak from a fixed or given loca-
tion within a pariicular distribution of the sensory realm may be not to speak
at aIl. As l discuss in my Afterword, the given partage may render one utterly
unintelligible; as such, there is no position from which one can speak. In other
words, to be marked as a body that fails to possess the logos means to lack the
capacity for poli tics, to find oneself outside the political domain within the
order of the city. One tries to speak, but one can only babble. If there is no
wrong, you are not a party at aIl, because you have no part in the dispute.
In cases such as these - utterly common cases, l would contend the demo-
cratic battle must be fought in a way that shifts discursive practices. Of course,
the shifting of discursive practices is never only that; it is never merely an al-
teration of language. The transformation that goes on within language, by way
of a laying daim to the excess of words, is simultaneously an effort to produce
a new partage; it is an effort to lay daim to the "writing" that circulates every-
where. A new discursive position always remains bound up with a new distri-
bution of the sensory realm; to change "discourse" is not just to alter terms but
to rearrange the world. In the case of immigrant meat packers, this type of
transformative shift would me an to rework the narrative of the immigrant. On
my reading, Apostolidis's research shows that the Mexican immigrants in east-
ern Washington State have appropriated the discourses of immigration so as to
resignify them. Their efforts point toward the possible emergence of the "im-
migrant worker" who is also an activist, a subject position from which a daim
to equality may be heard as speech and not mere babble. Literarity emerges
here when the workers who should not speak, do.
The differences among "immigrant," "worker," and "activist" are thereby
not differences of words alone; they are the very distinction between subject
positions that would entail the space between the logic of the police ("immi-
grant") and the logic of politics ("worker" and/or "activist"). Yet it would be
wrong to think that those employees of the meatpacking plant must identify
with one or another of these subject positions; it would be wrong to think that
they must be confined to one of these positions of enunciation. As Rancière
puts it, "The subject is an outsider or, more, an in-between" (I995c: 67). The

L1TERARITY 1 119
exists somewhere between and The
thus be a place - but also a time, because the in-between subject is a middle-
voiced subject - between that space from which articulation would be
bIe, the subject position, and the one who offers that articulation, the individual.
Arp0.","'r if subjectivation involves a refusaI of identity, agency must lie in the

in-between. Politics emerges through literarity.


Apostolidis shows how the workers meld the narratives of their
own migration, and the unbearable struggles and burden it almost always in-
cludes, with their work within the union at the meatpacking plant (Apostolidis
2005: 29-31). He goes on to show, moreover, that from this newly created
position within language, the Tyson/IBP workers come to take an activist,
democratic stance within the union. Theirs is a democratic politics, one that
emerges as a rejection of standard interest -group liberalism. The intersection
of the immigration narrative with that ofwork pushes them into that in-between
space in which disidentification and political action are possible. The union
members at the Tyson/IBP plant refuse to accept mere wage increases, and
instead argue for better working conditions along with improved access to
health care. Theirs is a democratic assumption/verification of equality (Apos-
tolidis 2005: 12). They speak as political actors to just the extent that they
refuse to merely babble as immigrants, or even as union members (where what
they say might be "heard" but only as phi5në and not as logos).
In making these claims we see the emergence of what Rancière would call
a polemical scene, a scene that concerns the accounting of those who have no
count. Yet this scene cannot be read as a mere assertion of disidentification by
the workers themselves. "Disidentification" is not just "something a subject
does": the process of subjectivation involves disidentification, and therefore
one is not a subject before the disidentification. The power to "disidentify" is
not simply a capacity of the political subject, since there is no subject at this
point. In this sense, literarity makes the miscount (the miscount that is poli-
tics) possible. Therefore, I contend that disidentification - that is, subjectiva-
tion - must instead be understood as a political and literary struggle waged on
a number of fronts: including the local battles at the plant, the constant efforts
by the US government to slow illegal (and legal) immigration, the continued
attempts by Mexicans to migrate, and the shifting discourse of immigration
as it plays out at a local, national, and even globallevel. From this perspective
we can see that a theory of democratic politics, as Rancière defines it, must
remain attentive to that disagreement at the level of discursive practices
Lyotard caUs this the differend - which precedes the final emergence of
politics.

*****
That political moment, that moment when those who are not speaking beings
somehow speak - this moment is only made possible by the prior production

120 1 The Lessons of Rancière


of which can
- "appearing speech" is worth holding on to, since the question of what is
is never far removed from the question of what is see-able. The
moment of hearing noise as speech cannot be distinguished from the moment
of a creature as human. Rancière's work never loses touch with these
temporal and thus he writes, "Poli tics cornes after politics which
itself COlnes after other forms of symbolization of the common and against
them" (2003c: 1-2). To insist on the literarity of the human being means to
reject the notion that the political/apolitical distinction can be settled from the
start. It means to insist that politics is a beginning and not an ending, and that
there can be no prior ground established as a condition of such beginning.
Sorne commentators have taken Rancière's claim about the "rarity" of
politics as an admission of weakness, or an auto-critique. It should be clear
now that 1 read Rancière differently: rather than marking the rarity of politics
as a limitation of the theory, 1 take it as one of Rancière's central insights
into politics. However, to say that the police order typically reigns, or that
the irruption of politics through the interruption of that police order only
occurs under special circumstances, should not lead one to the conclusion
that the work of producing such special circumstances is somehow not po-
litical work. Rancière's politics is not a politics of waiting around for the
event to occur. Rancière's approach proves far too untimely to be assimilated
to a political model of waiting and hoping (cf. Gibson 2005). As Ross puts
it, "Polemics [Rancière's favored approach] after aIl, is just a synonym for
untimeliness" (Ross 2009: 18). Rancière's concepts of politics and police
can be misread as static concepts, as fixed polar opposites, but literarity, the
excess of words, brings in a temporal dimension, an untimely aspect to Ran-
cière's thinking of politics _.- one that prevents such closing down of his ac-
count of police and politics.
Rancière's politics is always and already the politics of literarity: it is a
channeling or harnessing of the very force that can always threaten the police
order, the excess of words. This emphasis on literarity, not merely as another
Rancièrean concept among many, but as a central term in his distinctly demo-
cratic politics, can help to unpack textually and elaborate, both conceptually
and politically, many of the other terms that make up his conceptual vocabu-
lary. In this way, a richer understanding of Rancière' s reading of the Aristote-
lian scene of politics, a clearer grasp of Rancière's thinking of subjectivation,
and a sharper sense of the force of literarity, can aIl be used to tum back to, to
focus and refine, the thinking of politics and police that 1 discussed in the first
two chapters. When shot through with the concept of literarity, when under-
stood against a background of the "excess of words," the stakes of Rancière's
concepts of poli tics and police come into starker relief. At the same time, and
now moving forward, the idea of an excess of words that always thwarts a
police order makes possible a further broadening out of Rancière's writings _.- a

LlTERARITY 1 121
movement away from his own polemical interventions and (as 1 have do ne in
my closing example here) a shift toward concerns that Rancière himself may
not share. Thus, in the next chapter continue the process of broadening out
(and pushing beyond) Rancière's fundamental terms, as turn to a concept - or
perhaps better, a theme - that, like 1iterari ty, has hovered at the margins of
Rancière's work: critique.

122 1 The Lessons of Rancière


Hope can come only From the hopeless.

NY READER WHO PICKS up a text on or about Rancière will quickly and


frequently encounter the word polemical. This book is no exception.
Rancière's overall writings, his specific engagements with other think-
ers, the relationships his writings establish, and his very use of terms and con-
cepts are aIl described - by those who write on Rancière, and occasionally,
though less often, by Rancière himself - as polemical. What do es it me an to
say that Rancière writes polemics, that he writes polemically or that his work
is polemical? Previous chapters have already described Rancière's unique
style, and the word polemical captures mu ch about his sparse, direct engage-
ments. His work can be understood as polemical in the sense that Rancière
seeks to provoke, that he tries to test and probe, and that he refuses to system-
atize. Further, Arditi and Valentine have shown that, when read in a particular
way, polemic or polemicization is another name for what Rancière means by
"disagreement." In this sense, polemic refers not merely to a disagreement
between two given parties over a set of choices for action. Instead, a polemic,
or a scene of polemicization, describes a situation in which, as Rancière puts
it, "The interlocutors both understand and do not understand the same thing
by the same words" (Rancière I999: xi). With the paradoxical formulation
"understand and do not understand," Rancière paints a scene of dispute over
more than just the "choices" to be made by rational actors. In dispute is the
recognized identity of the interlocutors themselves, the intelligibility of the
objects of dispute, and the very ground of argumentation (Arditi and Valentine
I999: 4)·
In Chapter One 1 marked the difference between Rancière's approach to
politics, and that of normative theorists, deliberative democrats, and cer-
tain post-foundationalists. Here 1 would both widen that gap and delineate
its terms: Rancière's polemical approach distinguishes him from others not
as a matter of but substantively as weIl. To say that Rancière
polemieizes is to say something about the way he writes, but also to indicate
sornethHllg about what he to describe the type of work his
does. In this chapter 1 focus on this final dimension, as try to elaborate on
the nature of Rancière's polemic. Polemic, of course, is another name for
critique; a polemic is a particular sort of critique. But what sort? Through a
genealogical reading of Rancière's writings, this chapter traces the contours
of a renewed concept of critique. lt begins a process of rethinking a version
of critieal theory that would be appropriate to, and salient for, contemporary
theory and politics.
Rancière's earliest publication, his contribution to the collective produc-
tion (at least initially) that was Reading Capital, cornes with a rather ungainly
title: "The Concept of Critique and the Critique of Political Economy: From
the 1844 M anuscripts to Capital." This lengthy essay has garnered almost no
cntical attention, due in large part to Rancière's simultaneous decision to
break with Althusser and to take his distance from his own early work. De-
spite its relative neglect, both the title and the substance of this early essay
raise a concern that runs straight through aIl his theoretical work on politics
and that has always been a crucial background topie or issue for Rancière: the
question of critique. Moreover, 1 contend that the question of critique, or what
1 would prefer to calI the problem of critical theory, proves even more acute
today. Given a certain acceptance of post-foundational insights and a certain
diminution of Frankfurt School critical theory at the hands of Habermas's
graduaI but ultimately decisive shift from Marx to Kant, how do we under-
stand, today, what it would me an to formulate a critical theOl"y of X? By this
formula --- "cri tic al theory of X" - 1 want to maintain the otherwise subtle
distinction between a "critique of X" and a "critical theory of X."l The con-
trast signaIs a shift in emphasis away from an idea of critique that might be
used to link up Kant with Foucault (and a whole series of thinkers in between
them) toward a distinct notion of critical theory. For aIl its variety, the Kantian
tradition of critique centers on the work of the subject. Foucault famously
redefined it as follows: "Critique is the movement by which the subject gives
himself the right to question truth on its effects of power and question power
on its discourses of truth" (Foucault 2002b: 194). This chapter focuses more
on rethinking "critical theory" in a sense at one time developed by Marx and
later brought to prominence by Horkheimer. The tradition of critical theory
that runs from Marx through the Frankfurt School has something to say about
taking a system, or structure, or other type of "whole" and putting forth a
critical analysis of it (Horkheimer 1972). Thus, in the formula "critical theOl"y
of X," for Marx, X =society, or in Althusser' s language, X = a social forma-
tion. What, then, does it mean to formulate a critical theOl"y of a particular
social formation?

1 The Lessons of Rancière


The above will my here into a series of texts
but before turning to them want to sorne context both for
my particular readings and for the larger project of reconceptualizing critical
theory. Such a project is clearly mine, not Rancière's; this chapter continues
adding torsional force to Rancière's a force that twists his
toward my ends. Obviously in this case the idea of "critical theOl'y of ... " must
bump up against Rancière's reluctance to produce a theory of anything at aIl
(Rancière 2oo9b). And without doubt, Rancière maintains a certain vigilance
in keeping his work polemical, interventionist, and local, but never analytical,
conservative, or general. While l accept the force of his arguments and the
importance of his interventions, here l ask whether polemicization also de-
pends upon (and needs to work with) a broader concept of critique - a wider
critical apparatus. Do not local interventions somehow need to be linked or
connected, one to another? For me, such linking might be the work of a type
of "theory" that is itself neither grand, nor systematic, nor totalizing. l proceed
on the assumption or hope that there can still be, today, a critical theory that
goes beyond the local without rising to the emptily abstracto Such a thinking of
critical theory need not be at odds with Rancièrean polemicization, but can in
fact run in pm'allel with and complement the latter,
In order both to clarify ternIS and to capture this sense of critique as po-
lemic, l will use the phrase "critical dispositif" throughout this chapter. Ran-
cière repeatedly invokes this phrase in his recent essay on critical thinking
(2007b), and, as is weIl known, the concept of the dispositifhas a long history
in twentieth-century French political thought. The word's meanings in French
are so multiple arrangement, plan, apparatus, ensemble as to make any
simple translation problematic. Foucault famously used the term dispositif in
his attempt to account for the emergence of sexuality in nineteenth-century
Europe. In an interview, he elaborated in depth on the term, saying that dis-
positif captures an array of heterogenous elements within a field of power rela-
tions (along epistemological coordinates) always linked to strategies (Foucault
1980: 194-97). The term proved so important to Foucault's work that Deleuze
wrote an entire essay with the title "What Is a Dispositif?" Deleuze theorizes
the dispositif in a manner that clearly resonates with Rancière's work. Wh en
Deleuze says that dispositifs "are neither subjects nor objects, but regimes
which must be defined from the point of view of the visible and [the sayableJ,"
he might weIl be mistaken for offering an exegesis of le partage du sensible
(Deleuze 1991: r60).2 For my purposes here, the idea of a critical dispositif
helps to capture just that sense in which a critical theory (or critical orienta-
tion) contains a multiplicity of heterogeneous elements, yet is still somehow
held together by the force and strategy of critique. Thus, l will repeatedly refer

CRITIQUE 1 12 5
ta the critical at a more while also the mech-
anisms and logics that animate and mobilize any specifie critîcal dispositif.
am particularly concerned with those critical dispositifs that possess a formaI
structure governed by the specifie logic of inversion.
This brings me to a second context, one marked by the sense that the time
for critical theory has passed, that attaining an "unmystified view" of things
does not lead to any necessary consequences (Sedgwick 2003) or, more gener-
aIl y, that "critique has run out of ste am" (Latour 2004). ln contemporary po-
litical theory today, sorne of the very best writings in the field stake out their
own position by taking their distance from a critical tradition,3 and they do so
precisely by characterizing the critical dispositif as marked by a relation or
structure of inversion. ln her recent attempt to turn theorists' attention to the
vibrance and agency of the nonhuman forces in the world around us, Jane Ben-
nett has argued against the project of demystification that we so often take to
be central to the work of critical theory (Bennett 2009: xiv). Her discussion
serves as a useful jumping off point for me, since she both names and pre-
sumes the understanding of critical theory, the very critical dispositif, that pre-
vails today in contemporary thought, but which is often left implicit. Everyone
knows weIl the form that this critical procedure takes: as a system of demysti-
fication, clitical theory works by way of the classic opposition between ap··
pem'ance and essence - that which appears on the surface is only a falsehood
masking the truth of the essential core. The task of the critical theorist is thus
to perform an operation of inversion, whereby the inner essence will be re-
vealed as truth, and the outer appearance will be unmasked as falsehood. This
sense of critical theOl'y can be found in prominent secondary sources (Hoy and
McCarthy 1994) and in central primary works such as Habermas's Philosophi-
cal Discourse of Modernity (1990: 1 I6).4 My efforts to rethink and perhaps
reinvigorate the idea of a "critical theory of X" - to move toward a new and
distinct critical dispositif operate first and foremost by resisting the critical
apparatus of demystification in terms (at least) as strong as others in the field
have done. ln addition, 1 seek to unbalance permanently the equation that
places critical theory on one side and the logic of inversion and the process of
demystification on the other. My goal, then, is to think a concept of critical
theory that does not rely on the appearance/essence dichotomy, that does not
function by way of inversion, and that does not seek to demystify.
This chapter broaches the following question: what would it mean to do
theory critically, to constitute a new critical dispositif, without relying on a
relation, law, or logic of inversion? My aim is not to offer a history of critical
theory, to trace the rise and faH of the Frankfurt School, or to try to account for
and categorize the numerous historical meanings of "critical theory." As my
discussion of Bennett immediately above indicates, 1 focus instead on the cur-
rent status and the contemporary senses of critical theOl)'. Despite their very
significant differences, these senses of critical theory share a commitment to

126 1 The Lessons of Rancière


the and aIl invoke or on a of
surface and depth. Across their variety then, critique al ways functions by way
of a relation of inversion that either plumbs those or otherwise reveals
a truth that has been masked falsity.5
This chapter builds from the previous ones so as to mobilize read-
ings of Rancière' s work in the service of reconceiving a new cri tic al project - a
critical dispositif that operates without the logic of inversion. As do through-
out the book, here 1 draw widely from Rancière's corpus (often in relation to
my previous discussions), but 1 focus in particular on three locations in Ran-
cière's writings that 1 have not previously engaged with - places where his
work proves particularly salient for the project of rethinking critical theory:6

1. "The Concept of Critique," 19657


ln his initial account of critique in Marx where he works within the
problematic established by Althusser Rancière advances a reading of
Marx as a theorist of inversion. This is not a crude interpretation by any
means (it operates at many complex levels), but nonetheless it clearly
expresses the idea that in his mature works Marx's "critical theory," if
such he has, would be a theory of inversion. This reading proves very
important because it establishes both the concept of critique that Ran-
cière himself will later reject (a point on which 1 wholly agree), while
attributing that concept to the work of Marx (a point on which 1
disagree).8
2. "Archipolitics to Metapolitics," chapter four of Disagreement, 1995
Rancière does not return to self-reflective work on critique or cri tic al
theory for quite sorne time (see below), but 1 will argue here that when
he shifts in the 1990S to the question of poli tics and political philosophy,
Rancière advances his own implicit critical dispositif by mobilizing his
own explicit critique of poli tics and political philosophy. Despite the
emergence of a challenge or resistance to the logic of inversion in Ran-
cière's writings as early as The Ignorant Schoolmaster (1983), and de-
spite the fact that Rancière's critical account of metapolitics itself serves
as a refutation of the logic of inversion, 1 will show that in Disagreement
Rancière's argument silently invokes a very similar logic. 1 will argue,
however, that Rancière's critical dispositif in this text mobilizes a logic
of reversai rather than inversion - a subtle but significant difference.
3. "The Misadventures of Critical Thinking," 2007
Recently Rancière has returned after a long hiatus to take up directly the
question of "critical thinking." In this recently published but already
widely circulating essay (2oo9a), Rancière offers his most polemical de-
nunciation of any critical dispositif that relies on the logic of inversion,
showing not only the foundational flaws in the logic but also its perva-
sive influences - as it underwrites both a certain tradition of critical

CRITIQUE 1 127
on the one hand, and a refutation of that tra(lltlOn, on the
other. Rancière ends this essay with sorne tantalizing hints concerning
the idea of a new critical thinking. Tantalizing though they are,
remain but hints. Here plant the seeds of an argument for a critical
theory beyond inversion, for a new critical dispositif.
Rancière's early text on Marx identifies a critical dispositif centered on and
structured by relations of inversion. His most widely read work on politics chal-
lenges that very dispositif, by way of a polemic that simultaneously draws from
and points beyond the same logic of inversion. And his recent writing on criti-
cal thinking calls for a radical reconceptualization of a new critical dispositif
that does without inversion entirely. Woven together, these three distinct and
varied polemical interventions of Rancière's allow me to raise the following
polemical question for today: what would it mean to "do" criticism or to ad-
vance a "cri tic al theory of X" without falling back on the logic of demystifica-
tion, the logic of inversion? Here 1 offer a series of interpretations of Rancière's
work designed to set the terms for a fuller answer to this question. 9

Critical Theory as Science, Critique as Inversion

1 contend that Rancière's early essay in Reading Capital proves very much
responsive to the question of "critical theory" that 1 have posed above; how-
ever, it would be a serious hermeneutical mistake to take the essayas a general
philosophical account of critique. "The Concept of Critique" cannot be read
outside of a predetermined and thick context: namely, the framework for rein-
terpreting Marx already established by Althusser in the essays that make up
For Marx (1969 [I965]). The seminal' that produced Reading Capital was con-
vened on the principles of interpreting Marx that those earlier essays of Al-
thusser laid out. This is not a projection on my part: Althusser's introduction to
the text makes the importance of this framework c1ear. More importantly, Ran-
cière's own text marks this framework repeatedly, with consistent direct cita-
tions of For Marx and with repeated use of the Althusserian interpretive
structure. This means that "The Concept of Critique" approaches Marx's tex!
with a few already established givens:
I. Periodization. Althusser famously establishes a timeline that breaks
Marx's writings out into four distinct periods: Early Works, I840-44;
Works of the Break, I845; Transitional Works, I845-57; Mature Works,
I857-83. This periodization entails privileging the later writings, espe-
cially Capital, while categorizing the early writings as under the sway of
either Feuerbach or Hegel (or both). And it means that if an author com-
pares a so-called Early text with a so-called Mature text, as Rancière
does in this essay, then he or she is bound to find a radical difference.

128 1 The Lessons of Rancière


2. Given Althusser's supreme emph,lslS
and importance of an epistemological break, works prior to the break
will have to be categorized as caught up within the old eplstemolo,Q~Y
even as they stmggle against it. Althusser's language, this makes the
early works "ideological," whereas the later works can be understood as
"scientific" in the sense that they now operate within the tenns of the
very scientific revolution that the epistemological break makes possi-
ble. lO However, given the very proximity of these texts to the break itself,
they will still sometimes express their science in the old language of
ideology.
3. Antihumanism. In For Marx Althusser establishes a sort of methodologi-
cal mIe for reading Marx's texts. In order to foreground the very peri-
odization and science/ideology distinction that Althusser caUs for, he
insists on taking an antihumanist approach to Marx's writings. In other
words, if we read Marx through the humanist lens, we will fail to see
what is wrong with the early texts (what traps them in the old epistemol-
ogy, what makes them merely ideological despite their insights), and we
will fail to understand what makes the later texts tmly revolutionary. It is
weIl known that Althusser meant to challenge the humanist reading of
Marx that seemed hegemonic in France at the time (and was surely as-
cendant), but it is important to realize that Althusser's own interpretation
of Marx was not merely antihumanist in its final fonn. Antihumanism
constitutes an interpretive mIe for Althusserian Marxism, a rule that
says: whenever you seek an explanation in a text of Marx's, do not locate
it in a theory of human nature, in a pre-given anthropology, or in a con-
cept of the subject. Of course, as we will see, this hermeneutic guideline
sometimes clashes with the given periodization, because the latter re-
quires the reader to locate in Marx's early works traces of an anthropol-
ogy. In other words, the antihumanist rule dictates a uniquely humanist
reading of the early works. 11

This list suggests something of the set structure in which Rancière's essay on
critique operates. And this framework significantly constrains what Rancière
might say about his putative subject matter, "the concept of critique," since aIl
his claims will have to be formed within the language of Althusser's already
established and highly determined discourse. Rancière opens the essay by
laying out the importance of the concept of critique for Marx and choosing to
focus the investigation on two texts (the 1844 Manuscripts and Capital). He
then immediately specifies the general approach that he will take: "In this
study 1 shaH rely on already established theoretical knowledge constituted by
the work of Althusser" (Rancière 1989a: 75). And when he tums to the 1844
Manuscripts, Rancière invokes Althusser's specific thesis: "The critique which
is carried out in the Manuscripts represents the most systematic form of the

CRITIQUE 1 129
r",..,j-hl-,--.-n,-,Ir>n1r'" 01,-,h"""" carried out
on the basis of the Feuerbachian anthropology" (Rancière 1989a: 75). Thus,
Rancière takes it as given that the concept of in the Manuscripts
is an anthropological (a critical term that have tried to parse and
clarify in the previous chapter), and that the philosophical basis for that cri-
tique can be found in Feuerbach.
Given how mu ch Rancière chooses to assume, based upon established "theo-
retical knowledge," wh en it cornes to the title of his essay, his reading of the 1844
M anuscripts serves more to shore up the Althusserian discourse than it does to
investigate or analyze Marx's text. 12 Vnder the banner of the concept of critique,
Rancière offers a rather brief reading of the 1844 Manuscripts that proves
Althusser's point. Rancière shows that the criticalleverage exercised by Marx in
the manuscripts cornes from outside (outside history, outside the structures of
capitalism) and is smuggled or projected into the critical discourse. Marx thus has
an "ideological" concept of critique in the manuscripts, because that concept rests
upon Feuerbach's philosophy of man. Rather than offer a critique of political
economy on its own terms, Marx illegitimately applies to the discourse of politi-
cal economy a concept of man and alienation that cornes from elsewhere (i.e.,
from Feuerbach). Rancière writes, "So the scheme of religious alienation [the
Feuerbachian anthropology of man] has been projected on to the worker-product
relationship [a relationship theorized by political economy]" (Rancière 1989a:
84, emphasis added). We can say, then, that the concept of critique in the 1844
Manuscripts follows the logic that Althusser, elsewhere in the text of Reading
Capital, identifies as the "classical opposition" between appearance and essence
(Althusser and Balibar 2009 [1965]: I23).13 Critique would amount to revealing
the essence of human alienation hidden by the appearance of the freedom of con-
tract and work under the telillS of a capitalist system.
ln one sense, this is a completely standard conception of critique. However,
within the Althusserian framework that Rancière uses here, critique must be a
particular concept in Marx's writing. Rence in this essay Rancière tends to
ftatten and constrain the multiplicity contained by any particular critical dis-
positif. My own reading resists this tendency: thus 1 prefer to talk about the
critical dispositif that Rancière sees in Marx's writing, a dispositif that con-
tains many logics, but which is dominated by the logic of inversion. In a cer-
tain sense the very idea of a singular concept of critique seems untenable;
certainly Rancière will abandon this notion in future work. Nonetheless, given
his reading of Marx, we can safely say that the Marxist critical dispositif op er-
ates according to a traditional structure of demystification by invoking a logic
of inversion. According to this logic, the essence will be revealed by the criti··
cal mechanism. Rancière does not question the dominant structure of this critical
dispositif here in any way, since, working within the Althusserian framework,
he has already rejected this particular application of critique, though not nec-
essarily the dispositif itself, from the outset.

130 1 The Lessons of Rancière


The dismissal of Marx's method of 0 .... 'Cl,..""" in the
nothing to do with its logical structure and everything to do with the Feuerba-
chian The of critique" in this early text can thus be
labeled ideological Rancière because of its U.U'LLHV!J'V~'--''''''''~'_U "L"'UULU.~'VU'-',
and this fact any possible investigation into the structural fmm that
the critique takes. in many ways Rancière's reading here focuses less
on the critical dispositif as a whole and more on Marx's conceptual frame-
work. Thus, by way of a very traditional reading of the Manuscripts,
Rancière shows that Marx opposes capitalist domination on the grounds that it
has distorted and denaturalized the authentic and true form of man as a labor-
ing animal. Marx grounds the argument upon an external anthropology that
Marx then applies to political economy; critique in this text clearly amounts to
opposing a true essence to the distorted appearance. Rancière sees this mecha-
nism of critique as problematic because within its terms the essence is exter-
nal, but Rancière never questions the logical structure that the dispositif as a
whole relies upon - a logical structure determined by inversion.
He does not do so because when Rancière turns to the "scientijic concept of
critique" as mobilized by Marx in Capital, he also finds there the same logical
structure, one that operates by way of inversion. In other words, where Ran-
cière identifies two distinct concepts of critique (one ideological and one sci-
entific), l would rather describe two similar critical dispositifs, both of which
retain a dominant logic of inversion. Rancière articulates this second version
of critique in Marx through a more detailed and much lengthier reading - this
time, of Capital. And in this second case, Rancière's argument proves to be
quite explicit about the central importance of inversion: he invokes exactly the
language of the "classical opposition" between appearance and essence. After
a long quote from Marx (from chapter 3 of Capital, Volume /), Rancière writes,
"Here Marx distinguishes between two motions: a real motion which is the
movement of value, a movement which is concealed in the repetition and the
process of circulation, and an apparent motion, a movement accredited by ev-
eryday experience, and which presents the inverse of the real motion" (Ran-
cière 1989a: 128). At this point in the argument, and confining ourselves to
their logical form, the two concepts of critique (early and mature) do not look
aIl that different. And as l will now show, despite what Rancière identifies as
significant conceptual differences, the critical dispositifs do prove similar.
Rancière himself would stress that the two types of critique differ in terms
of their "optics" -- that is, the metaphors by which they operate. In the case of
the 1844 Manuscripts, we have two different levels in play, and critique must
enable the movement from one level to the next - a movement from surface to
depth. Moreover, the critique of alienation involves a reconciliation of man
with himself; this entails an overcoming of alienation itself and a process of
synthesis not present in Marx's understanding of critique in Capital. The latter
critique seems to operate on only one level, and thus the "optics" are very

CRITIQU E 1 131
distinct 201 do wish to elide these but
rather to show that, despite them, both critical dispositifs operate according to
the same general logic, one that can be helpfully named a logic of inversion.
the term is not strictly mine; it is put into play in both cases Ran-
cière himself, where he frequently refers to the "law" or "relation" of inversion
(Rancière I989a: 128, I29).
In his reading of the early Marx text, Rancière uses the specific word in-
version to refer to Marx 's appropriation of the Feuerbachian concept of alien-
ation. "Inversion is produced through alienation: the generic life of man
becomes the means of his individual life" (Rancière I989a: 85). In other
words, the problem of inversion emerges because man lives an "inverted"
existence, and only a second inversion can set things right. In the case of the
"scientific" concept of critique, Rancière's language is quite explicit. He ex-
plains that the logic of inversion is itself the logic of capital, so that Marx, in
Capital, exposes this particular law of motion of capital. "The development
of the form of the process is thus governed by the law of inversion; the forms
in which the process of capitalist production presents itself or appears are
rigorously inverted with respect to its inner determination" (Rancière I989a:
129, emphasis added). In other words, capitalism itself inverts essence and
appearance; capitalism makes appear on the outside exactly the inverted form
of its true interior essence. Thus, a critical dispositif that would understand
the se laws of motion, analyze their mechanisms, and expose their effects -
this dispositif depends fundamentally on a logic of inversion. The critique of
capitalism inverts that which capitalism already inverts. The system hides its
own motor, and only a scientific understanding of that motor can grasp the
structure and its effects (Rancière I989a: 145, ISO). Marx's work is "scien-
tific" because it grasps the phenomenal forms in relation to their inner es-
sence (Rancière I989a: 171).
In unpacking these different strands of argument we can therefore conclude
that the second "concept of critique" in Marx's writings - the concept that
emerges only in his mature, post-transitional works - does not differ dramati-
cally from the early concept: both critical dispositifs place a logic of inversion
at their center. In both cases we have a false outward appearance that disguises
the truth of an inner essence. The first concept has two distinct levels, while the
second has only one, and thus the optical metaphors remain distinct. Nonethe-
less, the key difference lies not in the logic of inversion but in the location of
the true essence. In the early works, that essence lies outside of capitalism. In
fact, the essence is a projection that can only be found in an anthropology of
man "- one that will prove to be false, and one that has little to do with either
the capitalist social fonnation or its historical development. In both the early
and mature works we have an essence that we can counter to the mere surface
appearance. The difference is this: in the mature works, that essence is not 10-
cated outside the structure but is precisely internaI to it. The essence can be

'32 1 The Lessons of Rancière


located not it or it, but "'I-",V~'"H,'-''''
already-in-motion process of inversion whereby the essence faits to appear on
the smface. a second inversion the inversion instituted
but the inversion its can that faise surface appear-
ance and reveal the mechanisms at work below.
from Rancière's Althusserian perspective, Marx has two radi-
cally different concepts of critique. One is ideological; the other is scientific.
One is based on an abstract and philosophically dubious anthropology of man;
the other is grounded in a detailed understanding of capitalist structures in
their historical development, so as to grasp the laws of motion of both. One
must find its leverage for critique in nothing else but idealist pnUo,SOIJillCal
principles; the other bases its critical position on the truth of the essence of the
capitalist system, and need not go outside itself for justification. Nonetheless,
when looked at from a slightly different, and not necessarily Althusserian per-
spective, we have something rather different. Rather th an two distinct "con-
cepts" of critique, we have two closely related critical dispositifs - related
because each depends on a logic of inversion. ln both cases Rancière reads a
critical dispositif in Marx that operates according to the terms of what Al-
thusser caBs the "classical conception" of critique, because in both cases true
essences are opposed to faise appearances. ln both cases the critical dispositif
functions by way of an unmasking. Inversion proves central to each, since the
process of inversion will simultaneously reveal and bring to light the inner es-
sence that had been previously hidden by mere outward appearances. The dif-
ference in these two critical dispositifs, according to Rancière's reading, can
be found in the distinct conceptions of structure and not in the logic of inver-
sion itself, which they both share.
Identifying the commonality between these two critical dispositifs proves
significant for a number of reasons. First, it shows that in the effort to maintain
the distinction between an "ideological" early Marx and a "scientific" mature
Marx even within a text devoted to exploring the concept of critique - the
analysis of critical dispositifis gets relegated to secondary status. ln other
words, despite the title, Rancière's early essay is concerned not with two dif-
ferent concepts of critique, but with establishing a key distinction between, on
the one hand, the difference between a challenge to political economy via hu-
manist anthropology, and, on the other, an understanding of capitalism through
an antihumanist structuralism. When it cornes to the critical dispositifs them-
selves (i.e., the putative subject of Rancière's essay, the concepts of critique),
Rancière do es not offer much more than a confirmation of the standard model,
based on an inversion of the appearance/essence dichotomy. Most importantly,
reading Rancière's early essay confirms that, when working within the Althus-
serian structure, Rancière affirms a critical dispositif that depends on the logic
of inversion. Rancière thereby upholds a critical dispositif that looks exactly
like the one he himself williater challenge. 14

CRITIQUE 1
might critique look like in Rancière's own hands? What critical
dispositif will he rely upon as he makes his own critical, polemical interven-
tions? It will be quite sorne time before Rancière explicitly addresses himself
again to the question of critical thinking, but 1 suggest here that we can effec-
tively reread Disagreement as tacitly mobilizing its own critical dispositif. The
core daims of Disagreement are weIl known, they have been thoroughly dis-
cussed in the literature on Rancière in contemporary political theory, and
have already engaged with them repeatedly and often at length in the preced-
ing chapters.!5 Here 1 have no need to rehearse earlier arguments about poli tics
and police. Rather 1 want to emphasize a different point about Rancière's most
widely read and commonly discussed book. As 1 noted in Chapter Two, many
readings of Disagreement never make it past the opening sixt Ypages where we
find Rancière's accounts of police, politics, and "wrong."16 Here 1 wou Id like
to add: this is for good reason. Armed with Rancière's account of police and
his radically new definition of politics, poli tic al theorists can do a lot of work
in terms of either rethinking the meaning of poli tics or mobilizing their own
critiques (Bowman and Stamp 2009). And there is nothing inherently inadmis-
sible about making these moves: as 1 read Rancière, his texts encourage cre-
ative appropriations such as these, because the last thing he intends is to
constmct a comprehensive theory that aIl his readers would be forced to either
refute or work within (Rancière 2009b: 114).17
Nonetheless, Disagreement goes on for three more chapters (after the open-
ing three that are so weIl known for laying out Rancière's novel account of
politics/police), and in those later chapters Rancière does sorne of his own
work with the concepts he has put into play. Moving beyond the conceptual-
ization done in the early pages, Disagreement offers those who continue with
it two powerful polemics: the first against the entire tradition of political phi-
losophy, and the second against the "consensus" politics of the 1980s and
1990s. 1 have already engaged with the second polemic in Chapters One and
Two. Here 1 am interested in pursuing the li ne of argument put into play in
Disagreement's first polemic, particularly since 1 am trying to account in this
chapter for polemic as a particular form of critique. Thus, 1 want to read this
polemic for the type of critical dispositif that it simultaneously relies on and
animates.
Rancière announces the terms of his polemical attack on political philoso-
phy in the opening pages of his fourth chapter of Disagreement. First, he gives
a creative initial definition of political philosophy as "the name of an encoun-
ter" between politics and police. Political philosophy, says Rancière, names
the scandaI of politics, and that scandaI is none other than politics' "lack of
any proper foundations" (Rancière 1999: 6 1). Although he does so quite
quickly and in a densely packed formulation, Rancière provides a radical,

134 1 The Lessons of Rancière


alternative account Socratic emergence of As
the tale, philosophy's inception rests on the very discovery of the scandal of
democracy. Democracy and it operates, without any real grounding.
Philosophy points out this lack of For philosophy cannot
slough off a fundamental encumbrance. Philosophy is post hoc: it ar-
rives on the scene too late. Socrates cornes to philosophize in the Athenian
agora, but this means that democracy is already in place before philosophy.
"The dëmos is already there" (Rancière 1999: 62). But the philosopher (i.e.,
Plato '8 ) recognizes that the dëmos rules without an arehë; there is no principle
of rule that would justify and legitimate democracy. In fact, elsewhere Ran-
cière insists that the rule of democracy is not based on any arehë at aIl; it rests
only on kratos. There is no principle, only a mere prevailing (Rancière 1995b:
94; Chambers 2010b: 65; Vattel' 2012).19 Thus, philosophy exposes fully this
scandaI of democracy, but rather than embrace this as a paradoxical form of
politics (as Rancière himself will do), philosophy, in the hands of Plato, pro-
ceeds to translate the fact of democratic rule into a fatal fiaw, an essential
weakness and limitation. Philosophy turns the "observation of the fact" into "a
diagnosis of an inherent vice" (Ranci ère 1999: 62).
However, if the lack of a proper foundation to polities is itself a failing, it
will be up to political philosophy to save polities from itself, to replace the
an-archic rule of the dëmos with a properly grounded rule. Philosophy must
either find or construct an arehë (and if the latter, it may also be required to
hide or lie about the construction process). Clearly this is what Plato does in
the Republie. In the face of the fateful decision of the Athenian dëmos to act
as an anarchie mob and put Socrates to death, Plato seeks to supplant the
kratos of the dëmos with the arehë of the philosopher king. And if politics
were strictly a matter of ruling, Plato would be doing nothing more than of-
fering an alternative form of politics. However, by showing that polities is not
a matter of ruling but a matter of interrupting rule, Rancière resituates the
entire project of political philosophy, casting it in a completely different light.
On Rancière's account, politieal philosophy is reinterpreted in the most thor-
oughgoing manner. Rather than grounding poli tics in philosophical princi-
pIes, we can, using Rancière's concepts, understand politieal philosophy as
eliminating polities infavor ofpolice. Philosophy is nothing other than polie-
ing, and its effort to ground politics is really nothing else than an effort to
justify one particular police order rather than another. But, no matter the
order, aIl politieal philosophy shares one common feature: the eradication of
politics.
Therefore, Rancière can conclude that "the solution [to the scandaI of de-
mocracy] is to achieve politics by eliminating politics, by achieving philoso-
phy 'in place' of polities" (Ranci ère 1999: 63). However, to achieve poli tics by
achieving philosophy is really not to achieve politics at an, but rather the re-
verse. In his conclusion to this li ne of logic, Rancière explains,

CRITIQU E 1 135
as the of commu-
nit y and the good attached to its nature in place of the distortion of equality as
wrong, means first eliminating the difference between politics and police. The
basis of politics of the philosopher is the identity of the principle of politics as
an activity with that of the police as a way of detemlining the partition of the
perceptible [partage du sensible] that defines the lot of individuals and pmiies.
(Ranci ère 1999: 63)

Put bluntly, what Plato caBs "politics," Rancière caUs "policing." In order to
ground politics on a proper archë, Plato must eliminate the distinction between
politics and police; he must define his concept of poli tics in the very terms of
policing. Along the way, politics itself - as that an-archic principle that would
irrupt within and utterly transform the police order - has disappeared entirely.
Hence we can derive what we might caB the general formula for poli tic al phi-
losophy: the elimination of politics in favor of the (police) order of the
philosopher.
This reading allows Rancière to redescribe the entire project of political
philosophy as antipolitical in the most total and fundamental sense: political
philosophy is the name given for a philosophical project that tries to eliminate
the political moment of dissensus, of disruption and disorder, so as to establish
a stable (and sometimes timeless) social order. In the bulk of chapter four of
Disagreement, Rancière refines this general formula by showing the distinct
ways in which it manifests in canonical political philosophers. Each variety of
the formula has its own category and its own proper name: Plato = archipoli-
tics; Aristotle = parapolitics; Marx = metapolitics. Archipolitics names the
process by which Plato establishes a social order of hierarchy as a total re-
placement for the disordering logic of equality of democracy (Rancière I999:
6S).20 Parapolitics names the process by which both Aristotle and Hobbes (in
their unique ways) try to create an order of philosophy that can contain within
it the very disorder of politics; theirs is still the elimination of poli tics in favor
of police, but they eliminate politics not by excluding it from the order of the
good (as Plato did) but by including it within the very heart of the social order
they construct - and thereby taming it (Rancière I999: 7 2 , 77).
This brings Rancière to his discussion of metapolitics, and it is here that his
account of political philosophy, as a peculiar project of eliminating politics,
intersects with a modern form of critical theory. In his description of metapoli-
tics Rancière caBs on and caUs up a critical dispositif. As Rancière narrates the
tradition of political thought, metapolitics (as the third and final archetype of
political philosophy) arises out of a problem produced by the modern (Hob-
besian) variant of parapolitics. With the emergence of the sovereign people
cornes the concomitant appearance of that other group that would oppose itself
to the sovereign people and deny their sovereignty. They have many names:
the masses, the population, the rabble (Rancière I999: 80). Metapolitics arises

1 The Lessons of Rancière


because so as gap between
the masses. It thus takes on a shape and logic similar to (Rancière repeatedly
calls the relation distinct from, archipolitics. Archipol-
itics announced a gap between true and the of democratic
politics (this is the gap between true and mere doxa, which recurs in
so many of the Platonic dialogues). Archipolitics strives to elirninate
democratic politics and replace it with the order of the philosopher. Metapoli-
tics announces a different gap - not a deficit of justice, but a surplus of injus-
tice. "lt asserts absolute wrong" 1999: 81). Thus, rnetapolitics
operates by way of a very different notion of "truth." As Rancière explains,
"This truth is of a particular kind. It is not [as with archipolitics] sorne idea of
good, justice, the divine cosmos or true equality that would allow a real com-
munit y to be set up in place of the politicallie. The truth of politics is the mani-
festation of its falseness" (Rancière 1999: 8 l -82, emphasis added).
Archipolitics and metapolitics are symmetrical because they both reject
that which is taken for politics in the present, in favor of a social order that they
wou Id establish in the future. But the difference is crucial: Plato announces the
truth of his divine order, whereas Marx announces only the falseness of the
given order. According to Rancière, Marx's truth is thisfalseness, and nothing
else. Truth is not above the fray of politics, attainable only through the march
of philosophy out of the cave and toward the light of the eidos; instead, truth is
found behind or under poli tics, "in what it conceals and exists only to con-
ceal" (Rancière I999: 82). There is no truth other than this falsity. There is no
discovery of truth in and of itself (the very heart of the Socratic/Platonic proj-
ect) but only the repeated revelation of falseness. Perhaps parroting Socrates,
Rancière purports to distinguish between archipolitics and metapolitics in
terms of health: the former offers the medicine that will make the community
truly whole, whereas the latter "presents itself as a symptomology" that always
and unceasingly uncovers only untruth, only disease. Put simply, "Metapoli-
tics is the discourse on the falseness of politics" (Rancière 1999: 82). Meta-
politics is therefore a type of critical dispositif.
Rancière offers a very brief gloss on Marx's "On the Jewish Question" in
an effort to elucidate the logical mechanism that animates metapolitics. While
he frames it, of course, in the language that he has established in this chapter
of his book - so as to demonstrate the specific manner in which Marx's variant
of "political philosophy" seeks, like aIl other variants, the elimination of poli-
tics - we can see at the same time that, as a critical dispositif, metapolitics
participates directly in the project of demystification. The metapolitical dis-
positif depends on its own logic of inversion. Rancière concludes the reading
of Marx as follows: "Politics is the lie about a reality that is called society"
(Rancière 1999: 83). Metapolitics is therefore a philosophical account of poli-
tics that is also in a way the previous archetypes were not .-. a critical theory
of politics as we find it. As a critical dispositif, metapolitics functions according

CRITIQU E 1 137
to a of inversion, because it sees politics as and Meta-
poli tics works by claiming the given politics to be a masking of the truth of the
social that underlies it; the project of political philosophy then becomes one of
unmasking; metapolitical political philosophy obeys a logic of inversion. 21

Metapolitics can clearly be grasped as a specific name for a type of critical


dispositif. And there can be no doubt that Rancière's project in Disagreement
centers on an extensive challenge to an archetypes of political philosophy,
from archipolitics to metapolitics. Thus, unlike his early reading of the con-
cept of critique, Rancière here appears to spurn the project of critical theOl)'.
From affirming a particular critical dispositif Rancière now rejects one (with-
out naming it as such). And unlike the early account of inversion in Marx's
Capital, Rancière now suggests that Marx is nothing more than a political
philosopher himself, and, like aIl the rest, Marx too must be found guilty of
erasing the politics/police distinction -- of eradicating politics in favor of a
foml of philosophical policing. Thus, Rancière clearly distances himself, and
his own account of poli tics, from a critical dispositif that relies on the logic of
inversion. Whatever else Rancière is up to in Disagreement, his assault on
metapolitics indicates that his is not the project of critical theory as tradition-
ally understood, bec au se that project has no choice but to function by way of
the metapoliticallogic of inversion.
Nonetheless, and at exactly the same time, it also seems obvious on just
about any reading of Disagreement that the text itself produces a critical ac-
count of political philosophy. Rancière's challenge to metapolitics must tacitly
presume sorne sort of critical dispositif of its own. Furtherrnore, as 1 suggested
at the beginning of this chapter, "disagreement" names a particular, detemuned
form of confiict, and "polemic" names a definite mobilization of confiict (Ran-
ci ère 1999: x; Arditi 2007; see Bowman 2007). Rancière's account of politics,
his redefinition of police, and his scathing reading of the canon of political
philosophy aIl depend upon and mobilize a certain critical sense. However,
this simple observation has potentially important consequences. Rancière's
rejection of critical theory, as metapolitics, forces a confrontation with the
question of the mechanism by which Rancière takes such distance. Does Ran-
cière have an alternative critical dispositif?
1 want to argue that the key to answering the question, concerning the logic
mobilized by Rancière's critical account of metapolitics, lies in his early defi-
nitional/conceptual work. The opening chapters of Disagreement sometimes
prove difficult to interpret because they appear to be offering rather "neutral"
definitions; it is not always obvious where these definitions are coming from
or what Rancière means to do with them. My experience teaching this text

1 The Lessons of Rancière


others to science C'j-",rla~,t-C' '-'A~-'lHIJLHH-'CJ
point: such students have been trained ta place texts into one of two categodes
normative or non-normative (empirical). Rancière therefore bewilders them:
he is not telling them what to not telling them what is right or what
is just giving them an ideal theOl'y like the Rawls text they read the week
before), and his "non-normative" definitions - of police, of politics, of
wrong - do not sound quite right. Clearly he does not mean to "operationalize"
these "variables," so what is he up to? Of course, it goes without saying that
Rancière would reject this normative/non-normative distinction, and that he
has no desire whatsoever to partieipate in the project of empiricist social sci-
ence. For just this reason, Rancière's writings help to undermine the binary
and to disrupt these categories for my students - one of the many reasons why
1 assign his texts. Nevertheless, 1 contend that these basic interpretive confu-
sions tell us something significant about Rancière's text.
Rancière's polemics against political philosophy and consensus democracy
take shape later in his book by leveraging the "definitions" from earlier in the
work. Further, 1 want to suggest that, rhetorically, Rancière avoids looking like
a traditional critical theorist by way of the bland neutrality of tone and style
that animates the early chapters. 22 Nonetheless, Rancière's definitions of poli-
ties and police, along with the other attendant terms, create the conditions of
possibility for his later critical remarks. His critique of political philosophy
must therefore operate within a critical dispositif of its own. Moreover, 1 assert
here what 1 will demonstrate below: the dominant logie of Rancière 's own
eritieal dispositif rests on a specifie type of inversion, even if Rancière proves
extremely reluctant to present the argument in this manner and even though he
might wish to deny it after the fact.
Looking at the argument as a whole, 1 want to show now that Rancière's
critieal dispositif still utilizes a logic very similar to inversion, but that Ran-
cière's critical operations mobilize the logie in a distinct way. Rather than a
striet logic of inversion, Rancière puts to work a logic of reversal. To make this
case, let me first recount the moves Rancière makes when he glosses the his-
tory of political thought. First, he pinpoints one of the archetypes of political
philosophy; then, he associates this ideal-type with Marx, and he names it
metapolitics. Metapolitics works in the following manner: it says, "What you
see as polities is really a falsehood masking the social truth." There is an inner
truth of man and nature that has been distorted by a polities of the surface.
Metapolitics is political philosophy as critical theory; it works by inverting
truth and falsity (or by showing that a11 is false). This is inversion.
But how does Rancière's own eritieal aeeount of metapolities work? Is
there a logic that mobilizes his critical dispositif? Rancière's critieal work here
does something very similar to that whieh he is challenging. Rancière says, in
effect, "What we take for politics is really just police." And by mistaking a
police order for politics, we inhibit the possibility of any genuine political

CRITIQU E 1 '39
how could we encourage or about democratic mo-
ments of true political dissensus if we continue to think that poli tics is going
on when there is nothing but police actions occurring? Rancière's own
narrative therefore silently invokes an opposition between false appearances
and genuine truth, and it implicitly suggests its own of (re)inversion. After
aIl, as readers of Rancière we would leam to see the metapolitical critique
demystification) not as a project of political emancipation (its surface
-lYJl.UC>IU0l

appearance), but as a project of policing (its underlying truth). Metapolitics


follows the logic of demystification by showing that poli tics has distorted the
natural that lies below it. 23 Thus we can see that Rancière's critical dispositif
works by consistently showing that what we take to be politics is really just
another instance of a different police order. Politics does not institute; it de-
institutes. But this means that the articulation of what metapolitics is and how
it functions (to bring politics to a close or end) would itself seem to follow a
cenain relation of inversion. Consensus politics appears to be the political
answer, but it tums out not to be politics at aIl.
Hence my conclusion that the cIitical account of metapolitics operates by
way of a logic very similar to inversion. However, l need to modify and clarify
that claim by emphasizing a key difference between the inversion of metapoli-
tics and the type of "inversion" invoked by Rancière. The truth in Rancière's
criticallogic is not a truth that wouid underlie politics (as in metapolitics), nor
one that wouid supersede and supplant poli tics (as in archipolitics), but is,
rather, the truth ofpolitics. Thus, Rancière's dispositif does not make a singu-
lar truth shi ne forth (archipolitics) and neither does it discover all to be false
(metapolitics). In Rancière's case, then, through his critical account of meta-
politics, we have a something more Iike a reversaI than an inversion.
Although Rancière does leverage a certain sort of truth/falsity dichotomy in
order to supply critical force to his argument, he never relies on the metaphor
of depth. Since the truth does not lie below, truth must be understood as just as
mu ch a smface phenomenon as falsity. Here Rancière's argument will, for
many readers, clearly call to mind Nietzsche's critique of metaphysical depth,
his deconstruction (avant la lettre, of course) of a "two worlds" view, and his
unique embrace of appearances (Nietzsche 1977). l do not develop these con-
nections, but they indicate something of the plurality of cri tic al theories --
thereby suggesting the extent to which Rancière does not just break with "the"
tradition of critical theory but draws from certain strands of thought while
rejecting others. Critical theory is itself internally contested, and Rancière's
arguments are better read as a poiemicai contestation from within the tradition
than as a radical rejection from without. 24
Where inversion would insist on a hierarchical relationship of power, in
which the terms to be inverted hold normative standing, a reversaI does noth-
ing more than disorganize the terms themseives - it scrambles their current
orientation. ReversaI therefore does not require a commitment to preexisting

140 1 The Lessons of Rancière


criteria would of
distinction between reversaI and inversion makes possible a strong rejection of
the conclusion drawn (2006). accuses Rancière in a sense,
being the ultimate political philosopher, by creating a new and final
of political in the form that names "ultrapolitics."
tics, says is a form of utter depoliticization that cornes about through
the ultimate polarization (in this case of police and politics) 2006: 71).
That is, according to Zizek, by making poli tics pure it becomes nothing at aIl.
But the form of inversion (reversaI) that 1 have identified here do es not work
that way, and as 1 have already shown in Chapter One, the ostensible truth of
poli tics that the Rancièrean "inversion" is meant to reveal is neither a revolu-
tionary anarchist politics, nor a pure political moment, nor the return of the
political. It is a rare and fleeting set of poli tic al moments that can never be
guaranteed in advance, and can almost, as it were, be guaranteed to dissolve
quickly once they have been formed. Therefore, what the reversaI brings to
light is something different in Rancière than it is in the demystifying critical
theory of metapolitics.
We can conclude that Rancière's project in Disagreement (and the other
writings on politics and political philosophy from the 1 990s) vis-à-vis an ac-
count of critical theory, of critical dispositifs, does the following:
1. ln his account of metapolitics, Rancière has definitely rejected the logic
of inversion to which his earlier text on Marx lent at least implicit
support.
2. At the same time, the very critical force of Rancière's account follows its
own specific variant of an inversion logic, which 1 have called "reversaI."
3. However, while the mechanism might be similar, the overall critical ar-
gument that Rancière advances works differently. The difference tums
on Rancière's distinct topography, one that refuses depth.

This outline immediately raises the question of whether Rancière's "inversion


as reversaI," as 1have identified it here, substantively and meaningfully changes
critical theory. Is this subtle difference in logic enough to produce a distinct
critical dispositif, or does it merely add another variant to the broad array of
logics of inversion? 1 contend that while Rancière clearly rejects the logic of
inversion in his account of metapolitics, that very account still remains too
dependent on an implied Iaw of inversion. Where metapolitics posits a natural/
social truth that underlies poli tic al distortion, Rancière sees a contingent social
order that thwarts aIl attempts to naturalize it. This reversaI offers a very dif-
ferent substantive type of inversion, but the overall critical mechanism seems
too dependent on apparent relations of inversion. More significantly, it is not
clear how the 10gic of reversaI accomplishes or makes possible anything other
th an the truth/falsity swap produced by traditional ideology critique. Ran-
cière's argument in chapter four of Disagreement implicitly poses the problem

CRITIQUE 1 14'
of how one might rethink critical it on and
working within a framework outside of the logic of inversion. He therefore
gestures toward a new critical dispositif that his own text does not produce.
Moreover, as 1 try to show in the next section, Rancière's recent, explicit return
to the concept of critical thinking does take up this task.

Critical Theory's "Misadventures"

In his widely circulated 2007 essay, "The Misadventures of Critical Thinking,"


Rancière tackles head-on the question 1 have raised above concerning a new
approach to critical theory.25 In this essay, Rancière not only argues against the
inversion model (as he had in Disagreement) but also gestures toward a dis-
tinct terrain on which to reconsider critical thinking. He therefore suggests
something much more powerful than what he will call "the critique of cri-
tique"; he indicates, on my reading, the possibility for an entirely new form of
critical thinking - a new critical dispositif. 1 also want to emphasize that the
idea of thinking a critical dispositif that does not rely on the logic of inversion
is not a brand new notion for Rancière. In many ways it is a very old one:
almost immediately after his rejection of Althusser, Rancière identified the
problem of inversion - identified it not just in the tradition of critical thinking,
but also in Althusser's project and in Rancière's own writing as weIl.
As 1 noted above, Rancière's essay on "The Concept of Critique" is unique
in that it operates by way of an Althusserianism that Rancière rapidly dis-
tanced himself from, just as the essay itself quickly "disappeared," as it were,
by being dropped entirely from the Italian and English translations of the book.
Ali Rattansi captures the configuration nicely, when he describes the essayas
"finding itself in the unique position of an 'Althusserian' text disavowed by its
author, by Althusser, and by sorne of those who came to be called post-Althus-
serian" (Rattansi I989: 22). Significantly, this "disavowal" was not merely
silent. That is, in addition to making an explicit political break with Althusser
(in rejecting the PCF), and along with announcing his theoretical break in La
leçon d'Althusser (I974), Rancière also commented on his own early, Althus-
serian essay. This commentary came in the form of a preface that Rancière
drafted for the I973 second French edition of Lire le Capital, and to which he
gave the very functional title, "Mode d'emploi pour une réédition de Lire le
Capital" (Mehlman I976: 12).26 The editor of the second edition refused to
include Rancière's "preface," and so Rancière published it in Les temps mod-
ernes instead. In 1976 when Economy and Society published the first English
translation of "The Concept of Critique," they included a translation of Ran-
cière's preface. Hence, the first presentation of Rancière's work from Reading
Capital in English was preceded by Rancière's own distancing of himselffrom
that work. And in those criticisms Rancière makes sorne very important points

142 1 The tessons of Rancière


about the "r.,-, """"",j' of critical
UHUA'--'l"6 take
cades later.
Rancière opens the essay his argument in this text, an
engagement with Althusserian from his remarks on Althuss-
er's "reactionary" polities 1989b: 181; see Rancière 1974a; Ran-
cière 1974b). He then moves on, in a subtle manner, to reject much of the very
foundation of Althusser's interpretive approach to Marx. Rancière now shows
little or no patience with the fundamental Althusserian distinction between
ideology and science. He asks whether it even makes sense to think that Marx
was trying to eliminate ideological elements from his account. Rancière offers
a long guote from Capital (from "The Fetishism of the Commodity" section of
chapter 1), whieh he then follows with this limpid commentary: "so mu ch
ideology in the frontispiece of science" (Rancière 1989b: 186). Rancière's
point, of course, is not to indict Marx with the Althusserian charge of ideology,
but rather to use the example to sweep aside the Althusserian framework. Ran-
cière expresses the point in a concise, apparently banal, and yet utterly power-
fuI formulation, when he writes, "The idea of revolution is fairly ideological"
(Rancière 1989b: 187). In making the se moves Rancière undermines the very
methodologieal structure in which, as he is weIl aware, his own essay in Read-
ing Capital operates. In this way he sUl'ely disavows mu ch of what he argues
there.
However, the real significance of this "preface" lies not in these broad daims
about Althusserian Marxist hermeneuties (nor in Althusserian reactionary poli-
tics), but in the specific argument Rancière makes about the theory offetishism.
On this front, Rancière provides a crucial argument for the project 1am engaged
in here (and to whieh, on and off over his career, Rancière has contributed so
much): how to conceive of a critical dispositif without a logie of inversion un-
dergirding it. Rancière's strongest substantive challenge to his earlier essay
cornes at this juncture, where he argues that he has read the concept of fetishism
far too narrowly. He now shows that within Marx's own text we can find two
dimensions of "the fetishism of commodities." On the one hand we can locate
masking mechanisms and invisible elements, but on the other hand Rancière
also says - and this is what is entirely new in Rancière's later reading that we
can track down a dimension of visibility, of legibility, in which workers (and
capitalists) know very weIl what is going on within capitalism. Rancière argues
that the most problematic aspect of his essay in Lire le Capital was his attempt
to read fetishism narrowly, and therefore, 1 would add, to reproduce a tradi-
tional critical dispositif, one that could operate only by way of inversion. Ran-
cière argues that fetishism in Marx's text always has two dimensions, but his
reading eliminated one of them. "My reading stood on this little stage in which
the cIiticism of 'humanism' ... was concerned entirely with the scientistic por-
trayal of fetishism, i.e. with the representation of a world of agents endosed
within illusion" (Rancière 1989b: 188, emphasis added). Thus, in this short

CRITIQU E 1 143
essay tirst drafted in 1973, Rancière raises the of a cIitical . . u"-'!-'ve'HU
that does not rest on the assumption of illusion, ignorance, or darkness. What
would it mean to think cIitically without the presumption or invocation of "a
darkened mirror to be made c1ean by a critical operation which makes it dec1are
aIl there is to say" (Rancière 1989b: 183).27
It is just this possibility that Rancière bIings to the fore and probes in his
2007 essay on cIitical thinking (and its misadventures). The problem that Ran . .
cière attributes in 1973 to a narrow, scientistic, and ultimately bad reading of
Capital, he will attribute in his 2007 essay to the entire tradition of "critical
thinking." This recent essay contains Rancière's fullest challenge to the logic
of inversion and it embodies his c1earest break from a model of critical theory
as demystification. Despite the thirty-five year gap between them, the two ar-
guments are linked: in both cases Rancière is questioning the logic of inver-
sion and suggesting the possibility that there could be a critical dispositif
without this logic. Moreover, in this recent essay Rancière works with the
language of inversion much more explicitly and much more directly, while he
sets his own thinking on a path very much distinct from that logic. Rancière
opens the essay by telling a familiar tale, similar to the one on demystification
that 1 narrated at the outset of this chapter. He suggests that we aIl already
know that in sorne ways critical theory is exhausted, over and done with. There
is no longer a viable space for the type of criticism that operates by way of, as
he puts it in language that echoes the essay from 1973, the "denunciation of a
bright appearance concealing a dark and solid reality" (Rancière 2007C: 1).
Given this, many commentators have naturally conc1uded that there can be no
future for cri tic al thinking because there is no longer any truth; Rancière, how-
ever, distances himself from such an assertion.
Rancière has little patience for the idea of critical theory being over and
done with. In this essay he contends that the very proclamation of eritieal
theory 's end - based on the assertion that there is no longer a truth that we
could oppose to falsity - depends on implicitly reinvoking and thereby rein-
forcing the very same eritieal dispositif In other words, and to use a language
that is not Rancière's: it is a performative contradiction to state that there is no
truth beneath the lies. But it is not a performative contradiction in the c1assical
epistemological sense of refuting the skeptic - not, that is, because the state-
ment makes a c1aim to truth in the name of disavowing truth. Rather, the per-
formative contradiction emerges because in rejecting the standard logic of
traditional critical dispositifs (i.e., the logic of inversion) one relies on exaetly
the same logie.
Thus, says Rancière, we must offer an accounting of this paraIlellogic if we
ever "hope to engage in a true 'critique of the critique,'" a tum of phrase that 1
am translating here into a call to rethink the meaning of cIitical theory and to
create a new critical dispositif (Rancière 2oo7C: 1). Rancière begins this ac-
cou nt through examples of art, reading them for the logic that they deploy and

144 1 The Lessons of Rancière


""",r,,,,"'+ that caB on. starts with as a
artwork that exposes a hidden truth beneath surface appearances. Even
Rancière suggests, there are two implications of the r.rcH''''''..... ,.,t-:;r> ... •
"[ 1] this is the hidden that you cannot see; you have to be aware of
and behave according to that knowledge. But [2] there is no evidence that the
awareness of a situation bring[s] about the detemlination to change if'
cière 2007c: 2). In other what are the implications of the revelation
produced by critical inversion? We see the putative truth, but what effect does
it have? In suggesting this twofold dimension of the procedure of inversion,
Rancière's arguments here silently invoke his self-critical reading of Marx on
fetishism: on the one hand, the re al it y of capitalism is hidden from workers
and it must be revealed; on the other hand, it is not clear what occurs after such
revelation (nor even that the reality was really ever aIl that hidden). Already,
then, the logic of inversion has been placed in question.
Rancière th en turns to a second artistic example, in the art of Josephine
Meckseper, and it is here that inversion seems to become untenable. As Ran-
cière reads them, Meckseper's works bring together disparate and seemingly
opposed elements in such a way as to show that these very elements "really do
go together" - to indicate that they are a part of one and the same process
(Rancière 2007c: 2). Rather than revealing a deep truth underneath a surface
reality, this "post-critical" (my phrase) work shows that both apparent lies and
apparent truth are connected. ln this way, the work might be taken to make the
same point as post-critical thinkers like BaudriIlard: namely, that there is no
longer a place for reason, for criticallogic. Rancière suggests otherwise:

At first sight we might draw the conclusion that the logic of the critical dis-
positif has been entirely self-cancelled: there is no hidden reality ta unveil, no
feeling of guilt ta arouse. But if it were sa, why keep a dispositif that has no
more relevance? This is why 1 assume that the dispositifitself still works. There
may be no hidden reality behind the curtain. But the logic of the critical mode!
can pelfectly do without it. It is enough that there be something that we cannat
see or don't want ta see. And the very absence of another reality can become
the thing that we are unable or unwilling ta see. (Ranci ère 2007c: 3, emphasis
added)28

According to Rancière, Meckseper's project neither rejects nor abandons the


logic that animates a traditional critical dispositif. It appears at first that Meck-
seper's work overturns the logic of the critical dispositif, but because, as have
already shown, that logic itself is a logic of inversion, "overturning" is merely
another inversion. The work therefore obeys the same logic and upholds a very
similar sort of critical dispositif. Overturning the mechanisms of the critical
dispositif only serves to reestablish them in a different domain ,- one internaI
to the very object of critique. Rancière fOlmuIates this argument concisely: the de-
nunciation of the critical mechanism relies upon the use of that very mechanism.

CRITIQUE 1 '45
This argument be stated at a more I-J~"~~'~~ level: in ,·t.,>1,-~(,'L'n"," a critical dis-
positif based on inversion, that is, in announcing its demi se, we only sustain
that very same critical dispositif by reasserting its centrallogic of inversion.
With this analysis in place, Rancière can offer a powerful challenge to an
those works that would tell us we are beyond critical thinking, that we can do
without any critical dispositif (or, more to the point, that we are beyond the
need for a critical account of capitalism). Hence Rancière's brief but scathing
critique of Peter Sloterdijk's account of modernity (Rancière 2oo9a: 30--3 l,
citing Sloterdijk 2004).29 For Rancière, Sloterdijk's account "sounds as an
appeal to get free from the fOrills and contents of the critical tradition" but the
logic of the argument remains entirely trapped within that tradition (Ranci ère
2007C: 5). Rancière names this mechanism, "ideological inversion," and he
shows that Sloterdijk (and others who would attempt to casually slough off the
weight of the tradition of critical them'Y) has no choice but to remain trapped
within it. While Rancière himself do es not want to return us to an earlier logic
of ideological inversion, to a simpler model of demystification, he reserves his
harshest indictment for this post -critical critique. The earlier "critical proce-
dures were supposed to be means of arousing awareness and energies for a
process of emancipation," but these later procedures are utterly "disconnected
from that horizon of emancipation." As such, "they become tools against any
process or even any dream of emancipation" (Ranci ère 2007c: 5). The denied
yet recycled criticallogic is ultimately conservative.
Herein, l maintain, lies the true target of Rancière's polemical intervention.
It is not traditional critical thinking (or even its misadventures) that Rancière
is after, but rather those who would simply dismiss the entire tradition of cIiti-
cal thinking, almost out of hand, because of its putatively naive reliance on
sorne conception of truth. However, in a manner very consistent with his ear-
lier approaches to an understanding of cIitical dispositifs, Rancière shows that
the real problem lies in the logic of inversion itself. As he phrases it in re-
sponse to interview questions put to him in 2004, "Where one searches for the
hidden beneath the apparent, a position of mastery is established. l have bied
to conceive of a topography that does not presuppose the position of mastery"
(Rancière 2006d: 49). The language of topography illuminates a great deal
about Rancière's argument: it shows that the alternative to a position of mas-
tery depends upon a rejection of hidden depths that would be plumbed or in-
verted. Topographical maps have contours and elevation, but they do not have
hidden depths; the smface vaIies significantly, but nothing lies below the sur-
face (waiting to be found).
Moreover, in the hands of those who would set aside critical thinking, the
logic of inversion takes its most pernicious fOrill. It assumes the shape of what
Rancière caBs "nihilist wisdom," an approach that "pictures the law of domi-
nation as a force that permeates any will to do anything against if' (Rancière
2007C: 6). This is a logic of inversion operating outside the contours of any

1 The Lessons of Rancière


CoV",I1(""T critical its When the of is
challenged by way of its reappropriation, we wind up with the most disabling
and despicable "critique of critique." In the dismissal of critical thinking, nihil-
ist wisdom leads to the following concluding outcome: "The old leftist denun-
ciation of the empire of commodity and images has turned to a form of
melancholic or ironic agreement with their inescapable grip" (Rancière 2007C:
6). Thus, we reconfirm and reify the very logic of inversion that we so glibly
thought we were rejecting, but do so within the terms of an argument that
abandons aIl critical dispositifs. The claim that there is no truth, when uttered
within a framework still imbued by the classical opposition between appear-
ance and essence, winds up ceding politics entirely.
In moving swiftly toward his conclusion, Rancière indicates the extent to
which art has lost its critical edge. He cites Guy Debord's "society of the spec-
tacle" as an example of yet another turn enclosed completely within the logic
of inversion. Once again it is the old (now dated, now out of fashion) claims to
emancipation that have been extinguished: "Any form of emancipation inside
the society of the spectacle is a form of subjection to the spectacle" (Rancière
2007c: l 2). The crux of the problem is that the logic of inversion can never do
more th an reinvert: "The knowledge of the inversion belongs to the inverted
world" (Rancière 2oo7c: I2). There is no such thing as setting matters aright,
of turning things right side up, because what there is is nothing more than in-
version itself. Rancière has shown that the "postmodern turn" only deepens
and intensifies the problem of modernity (a claim that does not deny but rather
reasserts the seriousness of those problems). "There is no shift from modernist
criticism to post-modern nihilism. It is just a question of reading in a reverse
way the same equation" (Ranci ère 2007C: I2),30
1 want to insist on a point that could easily be lost in Rancière's analysis
here: we can neither return to a traditional critical dispositif that centers its
operations on the logic of inversion, nor can we simply escape or abandon the
entire project of cri tic al thinking by rejecting aIl critical dispositifs. For Ran-
cière, it is the possibility of something like "emancipation" that hangs in the
balance. He writes, "The CUITent disconnection between the critical proce-
dures and any perspective of emancipation only reveals the disjunction at the
heart of the critical paradigm. It may make fun of its illusions but it remains
enclosed in its logic" (Ranci ère 2oo7c: I3). In other words, the contemporary
"critique of critique" putatively rejects a traditional vision of emancipation as
the product of inversion, but it has not abandoned that logic of inversion. In the
published version of the essay, Rancière amplifies the point in order to suggest
a different account of emancipation: "We need to return to the original mean-
ing of the word 'emancipation': emergence from a state of minority" (2oo9a:
42). This radical idea of emancipation cuts directly against any vision of social
utopia or philosophy of order, aIl of which depend on a vision "where every-
one is in their place" (Rancière 2oo9a: 42). Rancière here tries to hold on to

CRITIQU E 1 147
of a of errlanlCH)atlon that is not the of the critical
r.r'"ril1r>r

logic of inversion, which itself still "puts people in their place, ev en if it is a


new place (e.g., dictatorship of the proletariat).
The logic of inversion centers traditional critical because it makes
possible the traditional conception of emancipation. Through inversion, a cri-
tique of mere appearances will, in and of itself, lead to "emancipation" by re-
storing a harmony that has been distorted or a unity (of the social whole) that
has been fragmented. In this way, even Plato (but surely neo-Platonists) main-
tains his own version of a critical dispositif dependent on the logic of inversion.
Kantian commitment to enlightenment and pedagogy - "release from self-
incurred tutelage" - invokes the same relations of inversion (Kant 1970: 54). In
the quotes above, then, Rancière emphasizes his own distinct concept of eman-
cipation, one that disconnects the idea of emancipation from any structure of
inversion: emancipation names the break a body makes from any predetermined
roIe assigned to that body (Ranci ère 2oo9a: 42-43). As 1 will discuss below,
this commitment to a different type of emancipation links up Rancière's radical
pedagogy of ignorant schoolmasters with his challenge to aIl critical dispositifs
that rely on inversion: the ignorant schoolmaster undoes the link between inver-
sion (stultification) and emancipation as harmony, by offering the possibility
for emancipation without stultification (see Biesta 2010).31
Rancière's analysis, and my reading of it, continually presses harder on the
one question that matters most: how would we get out of the logic of inver-
sion? What would it mean to think critically - an idea that Rancière never casts
aside - without replicating the logic of inversion, without recreating and reaf-
firming an inverted world? What would a critical dispositif not dependent on
the logic of inversion look like? Rancière ends his essay with a twist of his
own, one that 1 interpret as a response of sorts to these questions. He argues
that rather than attempt to refute a logic of inversion, or to posit a complete
alternative to it, we must instead provide a critical historical analysis of the
development of that logic. In one way, my own work in this chapter can be
understood exactly as an attempt to take Rancière's advice "to re-examine the
genealogy of the concepts and procedures of that logic" (Rancière 2007c: 13).
However, my genealogical work has applied not to Marxism or the Frankfurt.
School, but to Rancière's own corpus. 1 believe that this focus offers greater
criticalleverage to explore Rancière's final suggestion. He argues that aIl tra-
ditional critical dispositifs depend on certain assumptions about health and
healing, and again, while Rancière does not link these to Socrates, the links are
there nonetheless. The logic of inversion, the central opposition between truth
and appearance, resides deep within the structures of critical thinking: "Cri ti-
cal procedures are about healing the disabled, healing those who are unable to
see, unable to understand the meaning of what they see, unable to shift from
knowledge to action" (Rancière 2oo7c: 15). As long as our starting point is the
presumption of blindness, we will be forced, in one way or another, to seek to

1 The Lessons of Rancière


remove the sc ales from the blind, to to U-11JCHU"".,

truth (even if it is the truth of falseness). What else can we do? Rancière sug-
we might find an alternative in a (certain reading) ofhis own
he insists on "disentangling the of emancipation from some 1JA.'-'''UIJIJ\.hH--

tions that the cIitical tradition has uncIitically - borrowed from the
domination." ln contrast, "a true critique of the cIitical tradition might have to
change the very basis of the critical procedures [to reconstitute a critical dis-
positifbased on different logics]." But this would mean to somehow start from
a different assumption entirely. Rancière caUs this "a foolish f..l."'Cl'UU~IJL;'VH
indeed [-] that the disabled are able, that there is no hidden secret of the ma-
chine" (Rancière 2007c: 15, emphasis added).

From Sighted Critical Theory to a Critical Theory


without 81indness

ln this short essay, Rancière has nothing more to say - he leaves off here - and
Rancière has not taken up his own implicit charge (at least as 1 have read it) to
rethink critical theory, to offer genealogical analyses of the tradition of critical
thinking, or to provide a new outline for a distinct sort of critical dispositif -
one that would eschew or escape the logic of inversion. Nor has he do ne much
to explain what it might mean to assume the disabled are able - the incapable,
capable (Rancière 2oo9a: 47). Here 1 want to suggest that my selective reading
of Rancière in this chapter - both in choices of texts and in the interpretations
thereof - indicates the trajectory needed for such a project.
Rancière stresses that we must disaggregate questions of emancipation
from the tradition of critical thinking. Perhaps we should. However, 1 insist
that we cannot abandon the critical tradition. Building from Rancière's own
insights in this essay we see that it is futile and dangerous to attempt to dis-
pense with any and aIl critical dispositifs. The goal cannot be to do without a
critical dispositif; it must be to animate a critical dispositif that is not domi-
nated by the logic of inversion. Rancière's own work gives sustenance to my
argument, since sorne of the most significant claims that Rancière has ad-
vanced in support of a certain kind of emancipation (in the form of democratic
politics) have themselves rested squarely on the mobilization and effects of his
own incipient critical dispositif. Rancière has never explicitly identified with
critical theOl)'; instead, he has either avoided the tradition (in most of his work)
or distanced himself from it explicitly (in the 2007 essay). Yet my reading has
shown, first, that Rancière cannot do without a critical dispositif of sorts, and
second, as he himself says, any move to reject aIl critical dispositifs is doomed
to failure. Rancière's polemical writings invoke and depend upon their own
sense of a critical dispositif. And Rancière himself suggests that we must
somehow renew critical thinking rather than dismissing it. Furthermore, despite

CRITIQUE 1
the nature its pn~sentaltlOln at the end of a
VHJ'IJU'-'U.> think we can
build on Rancière's suggestion of grounding critical thinking on the assump-
tion that the disabled are able.
take this to mean that we would need to conceive of a ctitical project that did
away with the assumption that some are blind and need to be given sight (by the
always enlightening critical theorist), that some are incapable and therefore need
to be guided by masters. 32 Rancière thus gives us a politics of hope, but one quite
distinct from that which I resisted in Chapter One (a pure hope, a dream of per-
fect revolution). For Rancière, as the '68 slogan says directly, can
come only from the hopeless." The Rancièrean reading of the slogan would place
the emphasis on the paradoxical quality of the hopeless being the bearers of hope.
"The hopeless," on this reading, is not an epistemological category and it does
not name a privileged class, because hope is neither knowledge nor faith. 33 To say
that "hope can come only from the hopeless" is to make another foolish assump-
tien about the equality of intelligences. If hope comes not from the masters or
from those who know, but from the hopeless, we can no longer assume that some
possess a critical sight that others lack. Rancière repeatedly shows that emancipa-
tion and democratic politics are both possible only by way of the assumption of
equality. We proceed based on this assumption and we seek to verify it. What if
critical theOl)' did the same thing? What would it mean to build a critical dis-
positif that put into effect a logic of presuppositional equality - a critical dispositif
that sought as its goal a demonstration of that very presupposition? This logic
entails not just assuming that the disabled are able (as the conditions for a critical
theory), but verifying it as weIl (as the resuit of the critical operation).
Let me shift from these rhetorical questions about what such a critical dis-
positif rnight look like, to a concrete (if still very general) example. If we take
capitalism as a social fOlmation that caUs for critique, the task before us would
be to put into motion a critical theory of that social formation that did not ne-
cessitate blindness. This means to advance a critical theory of capitalism that
does not depend on invisible structures - structures intelligible only to the
critic and not to the agents of capitalism. And it means understanding the
agents of capitalism as capable agents 34 (thus doing away with false conscious-
ness). Working out from this simple sketch, a renewed critical theOl)' would
need to answer two central questions:

r) How can we be sure that there is no blindness? That is, what does it mean
to say that there are no invisible structures and no incapable agents?
2) What is the point or uptake of the new critical dispositif, if it is not inver-
sion? That is, what do we get from a critical theory if it is not simply a
revelation of truth?

Rancière's work gives us a powelful and definitely polemical answer to the


first question. How do we know that there is no blindness? We do not know it.

'5° 1 The Lessons of Rancière


we assume and we it. new criti-
cal theOl'y must therefore be an assumption of capability, a presupposition that
structures are not invisible, but can be seen. Here 1 merely echo Rancière's
move with to equality, as should be clear from my numerous
discussions of equality in this book. Rancière says that aIl we can and the
best we can do, with equality, is to assume it and verify it. Equality is neither
a substantive outcome nor a normative ground. Equality must be understood
neither as a good in itself nor as an ideal telos toward which to strive. Thus,
polities does not occur because of equality, nor does politics achieve equality;
instead, poli tics challenges and thwarts the police order. Equality can only
ever be assumed as we approach the world and verified as we witness events
within it (Rancière 199 1: 18). Here I am suggesting that we take just this ap-
proach to Rancière's "foolish assumption" about the blind being sighted, the
sick being healthy. This is the first element of a new critieal dispositif: the
presupposition of equality.
Such is the inspiring assumption that Joseph Jacotot took to teaching. In
many ways Rancière's most important early claims against inversion emerge
in a text that does not really discuss critique, critical theory, or a critical dis-
positif. As 1 discussed briefly in my own Introduction, in The Ignorant School-
master Rancière shows that a traditional model of teaching depends on an
"explicative order" in which the teacher's explanation overcomes the student's
inability to understand (Ranci ère 1991). Throughout The Lessons of Rancière
1 have tried to maintain a certain fidelity to Rancière's radical pedagogy, and
for that reason 1 have resisted or refused to codify a list of statie, formaI, or
timeless "les sons" that one would draw from his texts. Rancière' s novel peda-
gogy must also underwrite and infuse the form of critical theory that 1 gesture
toward and caU for here. Indeed, the Rancièrean pedagogy must make up a
central element of this new critical dispositif. Through its logic of inversion,
traditional critical theory stultifies. Rancière's challenge to the order of expli-
cation and his effort to avoid "enforced stultification," in his very early work
on Jacotot, points the way toward a new fonn of critical thinking - toward a
new critical dispositif that dispenses with the logic of inversion. On the tradi-
tional model, teaching relies on the ignorance of the student; what the teacher
demonstrates to the student is first of aH, and above aIl else, the student's own
ignorance. Traditional teaching then operates according to the same logie of
overturning and demystification - of finding true essences underneath faise
appearances as traditional critical theOl'y.
A critical theory based on the logic of inversion has no choice but to stultify
in just the same way as traditional teaching (Sedgwick 2003: 124). Jacotot's
alternative model of teaching starts with capable students and ignorant school-
masters. What, then, if the critical theorist did not presume that he or she
knows more than the agents within the social formation that he or she studies?
What if the presumption of critical theory were that agents can "see" just as

CRITIQUE 1 151
well as the theorist? a different
education, Rancière can be seen to raise implicitly the question focus on
here: what would it me an to conceive of critical as a that does
not stultify? Rancière's return to these themes is in his recent work
(Rancière 2oo9a). Jacotot asked: What if the student can learn without the
teacher? What if the student can read the book without a translation? And it is
only by making this assumption that Jacotot was able to create a radical and
emancipatory pedagogy. In turn, we can crea te a distinct logic to animate a
new critical dispositif only by overturning the previous assumptions and start-
ing with new ones. The abled nature of the putatively disabled is not some-
thing to be proven with normative logic, nor demonstrated with empirical
evidence. It is an assumption to be made and then to be verified. It is an ap-
proach we take to the production of critical theory. Using Rancière's radical
pedagogy as a guide, we have a clear response to the first question, above.
The second question, concerning effects, proves more challenging and
more subtle. What does a critical theory have to show us, if it is not the truth
beneath the lies? What does a critical dispositif do, if not reveal? What would
be the modus operandi of critical theory if it were not inversion, and what
would be the function of a critical dispositif if it were not unmasking or de-
mystification? In responding to these queries, 1 want to start by emphasizing
that this second question about effects indicates distinctly the gap that sepa-
rates this new type of critical dispositif from any critical theory based on the
logic of inversion. After aIl, the rough equation of critical theory with demys-
tification, or a logic of inversion, has survived for so long and achieved a cer-
tain hegemony because it holds such an obvious and powerful payoff. The
uptake for critical theory as inversion is obvious. Critical theory as demystifi-
cation offers truth, enlightenment, clarity. It bursts the bonds of ideology and
undoes false consciousness. It frees, it liberates, and it provides the solid epis-
temological grounds for change, for revolution. If we give up aIl that - which
to sorne extent we must when we reject the logic of inversion on which aB of
that is based - what then are we left with? If not consciousness and revolution,
then what?
Traditionally, critical theory brings sight to the blind; it illuminates what is
dark; it is sighted. The alternative begins not with the sight of the theorist, but
with the absence of blindness. To unpack this sense of a critical theory without
blindness, and in an effort to clarify what a new critical dispositif (one not
premised on darkness) might do, let me propose a distinction between "the
hidden" and "the unintelligible." According to a critical logic of inversion,
there are hidden mechanisms within any system or social order. What is hidden
is exactly a deep truth that cannot be seen because of faise smface appear-
ances. Critical theory, according to this line of thinking, must find and locate
the hidden or the invisible; it brings the hidden to light and reveals the truth

'52 1 The Lessons of Rancière


inversion.
l-hr,r.llrrh that
but cannot be seen.
contrast, in my of Rancièrean democratic and as
discuss in the find an element or term that must not be con-
ftated or confused with the hidden. Rancière gives this entity many names: the
dëmos, the pOOl', une part des sans-part -- a part without part, a part with no
part (Rancière 1995a: 3 caU it "the unintelligible." The unintelligible is not
a hidden part of the police order. Within the police order we already have,
plainly visible, entities such as the marginalized, the oppressed, or "the
in the common socioeconomic sense. Groups such as these have a part in the
police order, though surely it is a marginalized or dominated part. But the part
des sans-part cannot be contained by or included within the police order; it is
the part that has no part within the police order. It is the part that only cornes
to be through politics. Thus, the unintelligible is not hidden and for that reason
unseeable; rather, the unintelligible is not there at aU. A critical theory of a
social formation must produce a certain sort of "awareness" (for lack of a
better word) of the part that has no part, but it does not unmask the sans-part
or make it visible. Only poli tics can do that. A renewed critical theory would
have to be a critical theory attuned to the poli tics of unintelligibility. This
analysis yields the second element in a new critical dispositif: the logie of
unintelligibility.
A new critical dispositif must therefore never lose sight of Rancière's sense
of le partage du sensible. Politics makes possible a new partage. 1 am describ-
ing a critical dispositif that would operate with the sense that there is always
sorne partage that makes the visible visible, and that renders the unintelligible
unintelligible. Only politics can produce a new intelligibility where before
there had been unintelligibility.35 But even as a new partage produces a new
order of intelligibility, the unintelligible can never be eliminated - neither
through the incorporation of liberal politics nor through the illumination of
traditional critical theOl)'. The critical dispositif divides up and thereby cap-
tures/constitutes reality. It does not opera te by way of a representation of real-
ity, and for this reason it can escape the logic of mimesis that is the necessary
basis for critique as inversion (cf. Rancière 201 le). Here it seems worth restat-
ing an otherwise obvious point: the Rancièrean logic of political dissensus is
nothing like the logic of inversion.
In a traditional critical dispositif, the logic of inversion points toward the
production of a concrete form of knowledge: the deep truth will be revealed.
In contradistinction, a new critical dispositif - one that operates according to
the logic of intelligibility eschews the goal of knowledge production. Better
put, it sees the idea that the task of thinking is to create knowledge as neither
necessary to a critical dispositif nor adequate to democratic politics. Knowl-
edge and understanding are not the goals of this critical dispositif, because
poli tics is not about sense-making; it is about dissensus, about disagreement.

CRITIQU E 1 '53
in this context, when he eXl)lamS what is
inadequate about the translation of mésentente as "disagreement": "an entente
is an understanding, and més-entente implies at once a disagreement and a
m.isunderstanding" (Panagia 2006: 89, emphasis added). And the mésentente
for Rancière is not a misunderstanding that can be "c1eared up" or overcome
through more communication (cf. Rancière 2005c). Mésentente names a con-
stitutive misunderstanding: "not the conflict of one who says white and an-
other who says black [but] the conftict between one who says white and another
who also says white but does not understand the same thing by it" (Rancière
1999: x). This famous passage from the preface to La mésentente serves as the
inspiration for Arditi and Valentine's concept of polemicization with which
1 began this chapter and therefore lies at the core of any Rancièrean thinking
of the critical dispositif (I999: 4; cf. Panagia 2006: 89).
The passage also helps to emphasize the extent to which "making sense"
and "understanding" are projects of the police order. ln his effort to rethink
critique through a reconstructive reading of Habermas, Nikolas Kompridis rel-
egates his own critical project to the terms of the police, when he argues that
"critical theory needs above aU to understand" (2006: 29) and that the new
project of critique must be to disc10se meaning (2006: 255, cf. 262). ln stark
contrast, the critical dispositif that 1 am describing rejects the idea of taking
understanding as a goal. The aim of such a critical dispositifhas to be a politics
beyond sense-making, since the police order can easily accommodate such a
goal; indeed, the logic of inversion works perfectly weIl within the terms of
the police order's partition of the sensible. Disagreement and dissensus there-
fore provide the third key element within the ensemble of a new critical
dispositif. 36
Let me further indicate something of the shape of the idea of this new criti-
cal dispositif by returning to Foucault. At the beginning of this chapter 1 sug-'
gested a need to part ways with a Foucauldian rethinking of critique (as a
practice of the self on the self) in order to refashion critical theory (as a project
that offers a critical account of a social formation). Here 1 return to Foucault in
an effort neither to join my project back up with his nor to offer criticism of his
work. Rather, 1 wish, briefty, to mark sorne of the overlaps and sorne of the
distances between the conception of critical theory that 1 am gesturing toward
here, on the one hand, and the well-known work of genealogy, the "history of
the present," and the historical "problematizations" for which Foucault and
Foucauldians are rightly famous, on the other. The type of critical dispositif
that 1 seek to animate here shares with Foucault an understanding of systems
and structures as open and porous. That is, earlier forms of critical theory often
developed a critical theory of a system or structure that relied on the notion of
that system being closed. In fact, the critical dispositif would often presume
that the structure itself could be taken as a total, sealed object. In contra st with
this sort of approach, my sense of a "critical theOl)' of X" requires a conception

154 1 The Lessons of Rancière


of open to a F-,VLL'-'U.~Vf;;"'-""U
sense of a "structure" is therefore more poststructuralist
and my of a resonates more
with recent work on than it does with neo-Marxist or Frank-
furt School approaches (Connolly 2010).
I\/lr\1"p,,,,,~,," this Rancièrean critical dispositif certainly shares with Foucaul-
dian genealogy the effort to render contingent that which might otherwise feel
most natural. Both Foucault's form of critique and the concept of cIitical
theory that 1 am seeking to build here require a heightened, trained, and per-
haps Nietzschean, historical sense, and both function by way of a certain de-
naturalization. However, there is a tendency within some strands ofFoucauldian
thought (or at least within sorne narrow readings of Foucault) to rest with the
establishment of historical contingency. aim is not to impugn Foucault's
own account, but simply to insist that a new cIitical dispositif must go beyond
contingency, and it is on this point that critical theOl'y and certain variants of
genealogy must be distinguished.
In other words, the critical dispositif 1 am both outlining and calling for
cannot be subsumed under the heading of genealogy. First, genealogy can
sometimes become overly reliant on its own logic of inversion, which says:
you think your terms of self-understanding are true, but they are mere surface
appearances and underlying them is the true essence - that essence is nothing
other th an the truth of history as a motor force that renders aIl contingent. On
this line of thinking, contingency in history itself becomes the only truth that
underlies aIl historical appearancesY Again, on the whole, Foucault studi-
ously avoided reaching this general conclusion from his particular genealogical
analyses, but this notion has been drawn from the idea of genealogy - particularly
when genealogy is thought through as a social science "methodology" especially
by critics (Habermas 1990 : 277).
Second, the type of critical dispositif that 1 am describing here would ana-
lyze a given social formation not in an eff0l1 to as sert sorne singular truth of it,
nor in an attempt to find one particular "fact" about it. The critical theory of a
social formation must remain local and fragmentary. One cannot derive a gen-
eral "critical theory of society" - for example, one that "explains society" as
class conflict. There is no such general theory. However, and related to the first
point, above, to say that a critical analysis is local and situated is to describe
one characteristic of this critical dispositif; it is not to say that the content of
the theory can be reduced to an assertion of situatedness. Thus, to produce a
critical theOl"y of US capitalism during and since the Great Recession of 2008
would mean to go beyond both (a) explaining it as a symptom of univers al
capitalist forces; and (b) explaining it as a situated "product of history." Ac-
cording to the critical dispositif that 1 am defending here, a critical theory of
any pal1icular social formation must locate particular mechanisms and forces
and then describe their particular effects. Marx illustrates this point quite

CRITIQUE 1 155
in "Introduction." There he argues that the of clas-
sical poli tic al economy cannot stop with, or amount to, the idea of historiciz-
concepts; nor can it historicize capitalism
'-'V'~ULlH"" it to be a contingent social formation 1996). Marx
denies neither of these arguments, but they are not his primai)' claims. The
historically contingent nature of capitalism - and those concepts used to un-
derstand it matters much less than the force and effects of the system itself.
It is to the question of how a capitalist system works, and not that of where it
cornes from, that IvIarx applies himself most assiduously in Capital.
Thus, my account of a critical dispositif is committed te the possibility of
producing a "critical theory of X" that is not a singular refutation or rejection
of X and is at the same time not merely an effort to demonstrate the contin-
gency of for example, a critical theory of capitalism need not and
must not be the same as a critique of capitalism, premised upon the invisibility
of alienation as the true, essential core of capitalism (going on "behind the
backs" of workers).38 A critical theory of capitalism could still be produced out
of the assumption that there is no hidden truth of capitalism, but this means it
would be incumbent on such a theory to show how capitalism operates, how it
works, and what its effects are. And a demonstration of how capitalism func-
tions might also serve to denaturalize, render contingent, and call into question
sorne of the mechanisms of capitalism (perhaps even the system as a whole).
It would do so, however, not by presuming that there is a secret truth to be re-
vealed, but rather by indicating the way a complex system functions.
Perhaps, therefore, a renewed form of critical theory, based on foolish as-
sumptions, could offer insights and possible paths toward political action and
political change, while eschewing the dream of unshakeable epistemological
grounds and rejecting the goal of a guaranteed road to revolution. A cri tic al
dispositif that refuses to impose blindness, that refuses to operate by way of
inversion, might produce its own form of the very Socratic wisdom that Plato
(as the chief architect of a theory based on blindness) so long ago jettisoned:
the wisdom to know that we do not know, that there are no guaranteed out-
cornes or locations of pure epistemological surety. For Socrates, su ch wisdom
was a caU for action. A new critical dispositif could therefore reject the dream
of guaranteed revolution without thereby abdicating democratic politics.

1 The Lessons of Rancière


R

T M1GHT SEEM TO go without saying, but 1 will say it anyway: The Lessons
of Rancière cannot close with an enumeration of, or a report on, those les-
sons. This book must end without a moral to the story. In the final two chap-
ters of the book 1 have started to push Rancière's conception of politics past
itself, to gesture in directions that frequently exceed Rancière's concems and
sometimes work against his intentions. In this way, 1 have already drawn out
the implications and significance of the work 1 do in the book as a whole.
However, trying to come to a specific set of conclusions at the end would only
serve to undercut the very reading of Rancière's writings that this book has
offered.
ln the last lines of the opening chapter of The Ignorant Schoolmaster, Ran-
cière suggests an important first conclusion from the ex ample of Jacotot. In a
formulation often overlooked, Rancière writes, "Whoever teaches without
emancipating stultifies. And whoever emancipates doesn't have to worry about
what the emancipated person learns. He will learn what he wants, nothing
maybe" (Rancière 1991: 18, emphasis added). Radical pedagogy does not
mean verifying what the student has leamed; it means verifying equality, and
nothing more. The conclusion to the lesson cannot attempt to shore up what
has been leamed without either reinforcing earlier stultification, or, worse, in-
troducing a stultifying element to a lesson that had been free of it. What mat-
ters is not endings; rather, "The circle of emancipation must be begun"
(Rancière 1991: 16). Here, then, 1 end this book on Rancière's "lessons" by
beginning a new argument about democratic politics. 1 bring the book to a
close with a set of my own redefinitions that produce their own attendant po-
lemical claims.
My primary polemical claim can be captured by the phrase, a "queer
poli tics of the democratic miscount." This phrase holds together two inter-
connected contend that the most salient and
powerful version of queer politics proves to be a Rancièrean democratic
politics. That Rancière's conception of democratic politics illuminates
what is distinct about queer poli tics vis-à-vis lesbian and gay identity poli-
tics. Second, perhaps more contentiously, claim that Rancière's miscount
proves to be a queer miscount. Over the past two decades queer theory has
advanced and developed a relational understanding of identity, a conception
of political subjectivity not rooted in any preexisting or essential features of
the self, but derived in practice from norms and politics. This positional and
post hoc sense of the queer subject brings into stark relief Rancière's under-
standing of subjectivation, particularly his sense that the subject is an "in-
between" and that the political agent is not simply an agent who enacts
politics, but an agent brought about by politics. It is in these senses, l will
suggest, that we might simultaneously describe democracy as both a poli-
tics of the miscount and a queer politics.
In advancing these polemical claims, l neither reject nor depart from the
arguments and readings of Rancière that make up this book. Quite the con-
trary: l maintain that Rancière's own claims and contentions often gesture
in the direction that l go here, and that these final arguments of mine remain
consistent with the interpretations l have offered from the beginning of the
book. Moreover, Rancière's writings, l insist, can be marshaled in the ser-
vice of ends other than his own, and to make this sort of hermeneutic claim
is not at aIl to violate the spirit of Rancière's writings, but exactly to work
within it. Nonetheless, there can be no doubt that Rancière himself does not
take up the work l do below. While l suggest powerful affinities between
queer theory and Rancière's thinking of democratic politics, Rancière has
explicitly rejected queer theory, and subtly resisted aligning his work with
that field (Davis 2010: 88-90; Davis 2009; Chambers and O'Rourke
200 9).
The potential tension between Rancière's arguments about queer theory
and my reading here should not be a shock, as this tension is but a part of the
larger torsional force l have been applying to Rancière's work. It makes sense
to maximize that force here in the Afterword. Being, literaIly, the word that
cornes after the word, the Afterword should be thought through Rancièrean
tenus as "une part des sans-part." The Afterword must be a part of the book,
yet somehow not a part of the book. The appearance of the Afterword is the
emergence, but never quite the inclusion, of the part that has no part. This
means, also, that any Afterword nms the risk of being one argument too many;
it is the location of the "one-too-many." In this sense, though, nothing could be
more "faithful" to Rancière's thought than the Afterword, since the "one too
many," Rancière says, "constitutes the dëmos as such" (Rancière 1995b: 94; cf.
Rancière 1999: 118). In adding one argument too many, the Afterword threat-

1 The Lessons of Rancière


ens to say too as the
about a democratic miscount. 1

*****
We're here!
We 're queer/
Get used to it. 2

How can we read this slogan and chant, popularized first by Nation and
used widely throughout the 1990S during marches, protests, demonstrations,
and other events? How can we hear the words and what might they tell us
about both so-called queer theory and about democratic politics, today? It is
easy enough to dismiss the idea that this well··known slogan of queer activism
does any genuine poli tic al work. 3 To neutralize the chant we need only read it
through the given terms of today's interest group and identity politics. The
logic of liberalism tells us that rallies and marches need slogans and chants.
And Queer Nation came up with a catchy one: it is easy to remember, easy to
yell, and mixes just the right doses of political force and humor. Liberallogic
tells us there is no need to go much beyond this analysis - nothing else to see
here - since the Queer Nation chant resonates closely with the most popular pro·,
test chant of them all: the one that uses the formula, "What do we want? _ _ !
When do we want it? __ !" The first blank can be filled by aIl sorts of de-
mands: peace, equality, freedom, equal pay, shorter hours, health care, etc. The
second blank, almost invariably, is occupied by the word now, which serves to
intensify and add urgency to the demand. 4 One can complete the logic quite
easily: just as women and blacks demanded their rights, so gays and lesbians
demand theirs. This move situates the chant squarely within the terms of
rights-based (identity-based) liberalism.
But how might we read the chant so as to pull it away from liberalism and
toward democratic politics? In the case of this chant, what would it mean to
continue the effort of this book to prize apart liberalism and democratic poli-
tics? largue that this sort of Rancièrean move produces a reading of the chant
that dovetails with a specifically queer interpretation of the chant. Bath moves
push the Queer Nation chant away from the standard demonstration chant.
While the logic that links the Queer Nation chant to "What do we want?" has
a certain obvious coherence, it proceeds only on the basis of an almost com-
plete disregard for the content of the chant. If we look closely at the Queer
Nation chant, we must notice the stark difference between the two: most im-
portantly, we find no demand here. There is nothing that these queers "want"
and thus no timeline for their wanting it. There are no specific claims whatso-
ever being made in this chant. This is not, in short, a claim for inclusion (Jagose
1996: 112). While the chant does identify sorne hard-to··describe subject- "we're

Afterword 1 159
- Rancière's would alert to the way in which this
subject do es not seem to preexist the chanting itself. The "wrong" being de-
clared by the chant may be the very thing that brings the into exis-
tence. And while the chant does provide a geo-temporal location - "we're
here" - Rancière would alert us to the ambivalent or perhaps paradoxical
nature of this "here." "Here" denotes both the geographical and political space
(at this march, on these streets, in this polity) and also echoes the temporal
"now" of earlier prote st chants. However, it does not situate or locate the citi-
zen in a political district, nor project the subject into a state of nature. Finally,
the last line does not fit the script at aIl. Rather than, "'Ve're here; we're
queer; give us our rights," rather than "We're here; we're queer; we demand
equality," we find nothing at aIl claimed by this "queer" subject, and nothing
at aU demanded from the other. Get used to it.
However, this is surely not to say that the statement "Get used to it" does
nothing. Indeed, 1 will argue here that it does a great deal more than the claims
or demands of interest-group liberalism. Rather than making a claim of inclu-
sion, rather than requesting something of the dominant power structure, "Get
used to if' declares a powerful distance between those who occupy the domi-
nant position and those who chant the slogan, and it refuses an opportunity to
close that gap. It signaIs a resistance by those on the margins to move to the
center. "Get used to it" is also, notably, an imperative sentence; it tells others
what they have to do. And the changes within the norms that are announced by
the imperative must be brought about by those who occupy the center of the
norm. It asserts that those on the margins will continue to be who they are -
namely, queer - and argues that any alterations will have to come about by way
of broader changes to the norm. This is therefore a potentially subversive
claim, since it refuses to reify or even respect heteronormativity.
Here, then, 1 read the Queer Nation slogan as tapping into a number of the
most crucial elements of queer theory; simultaneously, 1 also read it as captur-
ing fundamental elements of Rancière's thinking of democratic politics. Let
me take each dimension of this double reading in turn, before 1 then combine
them. When the term queer first emerged in the early I990s, both its meaning
and its significance pivoted on the difference between queer, on the one hand,
and a fixed, given (gay) identity, on the other. 1 place gay in parentheses here
because one of the texts that played a crucial role in establishing "queer theory"
was Judith Butler's Gender Trouble, a work that did not address itself directly
to a lesbian and gay audience or subject, and that made no explicit effort to
reconsider lesbian and gay sexual identity.5 Moreover, Butler's work, like the
similarly influential books by Sedgwick and Halperin - all of which were pub-
lished in the same year, I990 - made no use of the term queer (Butler I999
[1990]; Sedgwick 1990; Halperin I990).
Butler wrote as a feminist, but she did so in order to challenge the hetero-
sexist assumptions buiIt into second-wave feminism's commitment to the

160 1 The Lessons of Rancière


of women. 6 And her argmnem Dr()ce!;;de~d
way of a deconstruction of the sexlgender distinction. To be clear, a "decon-
struction" does not level or erase the distinction; it is not a destruction that
would tear down what has been built. the deconstruction shows how
the construction works. It reveals that there is construction going on in the tirst
place. In the case of sex/gender, this means exploring the mechanism through
which the distinction is produced and maintained. Thus, it was not that Butler
wished to retum to that "prefeminist" time in which there was no difference
between sex and gender (this is what a literaI destruction or erasure of the sex/
gender distinction would accomplish). Rather, she sought to challenge the idea
that the difference between sex and gender mapped on to a difference between
nature and culture. Butler countered the typical if cormnonly tacit narrative
that quietly attributed to sex the capacity to serve as a natural foundation from
which contingent gender would then develop - a development carried out
through politics, through culture, and through historical variation. Butler's
most powerful set of rhetorical questions took this form: how do we get at this
idea of natural sex? Do we have any access to "sex itself," or does "sex" only
emerge within and through the discourses of gender? Thus, must we say not
just that gender is socially constructed, but that the very distinction between
sex and gender is itself constructed?
It is specifically in the content of Butler's deconstruction of sex/gender
that we locate its significance for queer theOl·Y. The distinction between sex
and gender is constructed, and the binary of gender difference is maintained,
through a series of regulatory norms and mechanisms that Butler names "the
heterosexual matrix," and which today we might refer to as heteronormativ-
ity.7 Binary gender gets produced in the way that it does only because of a
primary presumption of heterosexual desire that lies at the centre of the
matrix. And heteronormativity is just another name for heterosexuality when
it functions as a normative and normalizing force (O'Rourke 2005). More-
over, the problem of heteronormativity remains irreducible to the problem of
homophobia. Putting the psychological problems of homophobia to one side,
the political problems of homophobia can be adequately dealt with using a
combination of identity theory and liberal political theory. A theory of lesbian
and gay identity would serve to locate and specify that "minority group," (i.e.,
lesbians and gay men) that would be subject to the threat of homophobia.
Liberalism would offer a theOl'y of minority rights and equality before the law
designed to avert or lessen acts of discrimination or violence against such a
minority group. But none of this would necessarily challenge or offer resis-
tance to heteronormativity. Moreover, the effects of heteronormativity cannot
be reduced to the idea of a homophobic discrimination against lesbians and
gays. The entire liberai approach starts with the idea that there is a given and
known subject of discrimination or oppression. A queer approach challenges
just this liberal assumption.

Afterword 1 161
can thus see a certain convergence in the of Rancière and
Butler, in that they both consistently thwart this liberal assumption and they
both attempt to think the remainder (that which can never be recuperated by
interest groups) of the liberal, identity-based approach (see I993). Here
the relevance of Butler's famous early work on sex/gender becomes clear in
relation to Rancière's understanding of politics. Butler shows most powerfully
in Gender Trouble that the category of "woman" cannot be presumed in ad-
vance. Therefore a feminist theory cannot stubbornly insist that poli tics only
cornes after the subject. If the category of woman - and even the "experience"
of woman only emerges within the terms of politics, feminism must centrally
concern itself with the production of that category (Scott I99I).
Queer theory, 1 would insist, sUl'ely cannot be reduced to a mere analogy to
feminism. Yet there can be no doubt that the third-wave feminist critique of
second-wave feminism parallels and illuminates the queer critique of lesbian
and gay identity politics. Whereas lesbian and gay both name identities based
upon sexual orientation, queer points to no such fixed position. Lesbian and
gay (and also, in their own way, transgender and bisexual) are identities; queer
is a relationality. That is, queer describes a particular, relative position in rela-
tion to norms of sexuality. There is therefore nothing fixed, nothing permanent
about queerness; it is always context-dependent (although heteronOlmativity
aimost always makes up a significant part of the context in contemporary
cases).
To draw out the substantive links between Rancière's account and the proj-
ect of queer the ory and politics requires engaging and activating the paradoxi-
cal qualities of Rancière's account of politics particularly in terms of two
claims that 1 have discussed at greater length in Chapter Three. First, there is
the assertion that "parties do not exist prior to a declaration of wrong" (Ran-
cière I999: 39). Second, there is the claim that the original Aristotelian logos
remains tainted (Rancière 1999: I6). Contra Aristotie (or at Ieast sorne read-
ings of him) the question of speech and voice cannot be determined by a pre-
politicai account. It can only be detelmined politically, through the act of
hearing, of recognition. This is a politics of (in)audibility - those that one
hears and those that one does not hear. It is a politics of (in)visibility - "those
that one sees and those that one does not see" (Rancière I999: 22). Logos
cannot be taken as that tool that makes politics possible, because only the
logos itself can provide a "hearing" for any political articulation. Rancière puts
it this way: "The speech that causes politics to exist is the same that gauges the
very gap between speech and the account of it" (Rancière I999: 26). Aristot-
le's logos is thus not a ground, but a paradox. And Rancière's politics, like a
queer politics, proves to be both paradoxical and groundiess.
Rancière's rereading raises the question of where the "account" will come
from. How do we read, hear, or see a creature as human or animal? If we cannot
make the political/nonpolitical distinction prior to politics, the difference will

162 1 The Lessons of Rancière


emerge within And for this means that ac-
count only cornes about through disagreement, through the staging of a confiict,
through the dedaration of a wrong. Thinking Rancière's politics as a queer poli-
ties throws a new light on his central interpretation of the tainted logos and the
Aristotelian scene of polities. At the same time, it also places into much starker
relief the radical daim that l commented on at some length in the Introduction:
"Parties do not exist prior to the confiict they name and in whieh they are counted
as parties" (1999: 27). Only a political confiict can detennine the parties to the
confiict, but this means that there are no parties prior to the confiiet.
As have explained in previous chapters, instead ofresolving this paradox,8
Rancière repeatedly restates and reanimates it, by demonstrating the ultimately
contingent basis of polities: "Politics exists simply because no social order is
based on nature, no divine law regulates human society" (1999: 16). This pas-
sage from Disagreement is echoed in both earlier and later writings where
Rancière explains - through a reading of Plato's Laws that l discussed in
Chapter One - the difference between democracy and aU other political fOnTIS.
In that canonized text, Plato lists seven different "titles" to mIe. Rancière fo-
cuses his reading on the seventh, final, and apparently "extra" title: "a title that
is not a title, and that ... is nevertheless considered to be the most just." It is the
drawing of lots, the principle of randomness as the principle of mIe (Ranci ère
2006c: 40). This final claim to mIe has no basis in a princip le at aIl. To reiter-
ate, it is not an archë (a principle of mIe), but a kratos (a mere prevailing)
(Rancière 1995b: 94; cf. Vatter 2012). Rancière's argument here crosses with
Butler's queer deconstmction of sexlgender. Just as Butler refuses to allow the
contingency of gender to be grounded in the soli dit Yof sex, so Rancière resists
the temptation to give politics any ontological grounding. Like Rancière,
Butler toyed with the tactic - especiaUy early on in her writings - of eschew-
ing aIl ontology (Chambers and Carver 2008).9 At the "bottom" of both ac-
counts we find no bedrock; we find nothing but contingency. Yet for both
thinkers contingency is not conclusion. Neither author celebrates contingency
for its own sake and neither suggests that contingency is somehow an "answer."
Both Rancière and Butler make the move l called for in Chapter Four of refus-
ing to seule for, or with, contingency. What we might caB the "fact" of contin-
gency does not console or reconcile; rather, contingency demands responsible
action, since social orders - contingent though they may be are also hierar-
chical, exclusionary, and often violent.
And Rancière also exposes the full force of such a paradoxical poli tics by
insisting that politics is not the articulation of demands, not the claims of inter-
est groups. Politics is not the announcement or claim of identities (LGBT or
any other) by pre-given parties. Politics is the declaration of wrongs, the staging
of disagreements that serve to constitute the very parties of polities. This is why
he can write that "parties do not exist prior to the declaration of wrong" (1999:
39). Logically, the declaration of the wrong must precede those parties who

Afterword 1
would declare it. The declaration cornes before exist. have this
is not mere double-speak. Rancière uses this paradoxieal language in order to
insist upon his deconstructive reading of Aristotle. We will know if the chant we
hear is the slogan of a political group, if we hear that chant as a political
articulation and not as mere babbling. And the slogan itself, that which declares
the wrong, will also constitute the party to the wrong. But the process cannot be
guaranteed in advance, and Rancière's polities leaves ample room for the pos-
sibility of failure. This means that sometimes, perhaps often, a slogan or chant
constitutes nothing more than mere noise (phëmë). Sometimes, however, a claim
is heard as a political claim (as logos), and it thereby serves to articulate a
wrong. When this occurs, at just this poli tic al moment, we have the miscount
that, for Rancière, defines democratic polities.
Rancière's understanding of democracy as a miscount helps to highlight the
specificity of queer theory, while queer theory's understanding of identity and
nonTIS shows why Rancièrean politics always proves to be, over and beyond his
own intentions, queer politics. Rancière's approach resonates most strongly
with queer theory on "the question of identity."l0 Rancière's radical and polemi-
cal rendering of the original Aristotelian political scene can be used to throw
into stark relief the distinction between a lesbian and gay identity politics, on
the one hand, and a queer politics, on the other. As l have argued previously, a
gay identity can be established through an expressive act the declaration of
coming out, literalIy saying, "1 am gay." But a queer identity can only be articu-
lated within the context of particular norms, particular sets of power relations -
that is, within a determinate political context (Chambers 2003). If, as Rancière
insists, the wrong is itself that which constitutes the party that would declare it -.
and if both the constitution of the party and the dedaration of the wrong come
about through the action of politics - then the polities involved here can never
be based on identity. 1dentity - and with it interests, demands, and daims - can
never precede the politieal dedaration of a wrong.
Identity is therefore relational for Rancière, just as it is in que el' theory, a
key point that Rancière renders lucid when he calls the subject an "in-between"
(Rancière 1995c: 67). l quoted this passage in Chapter Three, but its full sig-
nificance can be brought to light by the context of queer theory. To calI the
subject an "in-between" is to speak as if the subject is not a thing, but a space
between things. But this is exactly one of the senses of the statement "1 am
queer." To say l am queer is not to identify with a given identity, but to disiden-
tify. It is to give an account of my positionality with respect to norms. Just as
queer "identity" must always be grasped in relation to norms (and therefore as
an "identity" only in a unique sense), so Rancière thinks identity in relation to
the political moment. And this means that the subject as in-between must be
grasped not merely in the spatial sense but in its temporal dimension as weIl.
The in-between subject is much like the middle-voiced subject that l discussed
in Chapter Three: a subject that finds itself/is found at both spatial and temporal

1 The Lessons of Rancière


interstices. The that queer
betweenness and an out-of-timeness. Rancière's
both have a certain liminality; both are
the same time, both Rancière and queer have ways
for an understanding of subjects and identities as fixed (or as HV:U-~''-'H.tUV:lH.U
This dimension of both theoretical modes of engagement proves important for
the simple reason that fixed conceptions of identity remain dominant
in theoretical articulations but in historical configurations. while queer
theory affirms a relational understanding of identity, it can still COlmnreflen.d
and work with (even as it resists) an understanding of lesbian and gay
as articulated on a model of fixed sexual orientation. And Rancière's under-
standing of the police makes space for a conception of social identity as given,
fixed, and hierarchical. Sometimes the pOOl' is the part des sans-part, while
sometimes the poor is an economic category of policing.
In this sense we might say that the presumption of equality that is demon-
strated through political action amounts to a queering of the police order. The
political moment, in Rancière's account of politics and police, is above aIl a
queer moment. The torsional force applied by politics to the police order is a
force of queering. ll Therefore, queer politics offers one provocative and
productive way of understanding what Rancière me ans by an impure politics.
l began my reading of Rancière with an elaborated defense of his claim that
there can be no such thing as "pure" poli tics (Rancière 20 Ile: 3), and surely
the struggle against purity also illuminates the struggle of queer politics. There
can be no pure queerness, and a queer politics will always be a poli tics of
impurity.
Rancière's argument also highlights the difference between a (mis)reading
of the Queer Nation chant as a set of demands by a given interest group, and
an interpretation of this slogan as a queer political articulation - a declaration
of wrong that bIings into existence the queer party to politics. The Queer
Nation chant must be seen as the articulation of political voice by the voiceless - a
declaration of a wrong that brings about a party that does not exist beforehand.
The slogan must therefore be heard as a resistance to heteronormativity, given
that heteronormativity would, on first reading, render queerness invisible or
impossible. "We're here! We're queer!" makes the invisible or impossible
queers, visible. But it does not bIing out the visibility ofthis new human animal
with logos merely to have that animal articulate a demand; rather, it insists on
a certain distance from the norm - get used to it - and thereby refuses to be
absorbed within the terms of that norm.
Of course, one could plausibly argue that "getting used to il" names a pro-
cess of normalization, wherein that which was once "at odds with the normal"
becomes less so for the very reason that we have become "used to it." But
l would insist that, when chanted on the streets, the illocutionary force of the
phrase resists this process of normalization. "Get used to it" is not a call for

Afterword 1
as
P'fir',,'Hl"V/,t-U,'? but rather an insistence that the normal
will persist. After aIl, what "we," as those who hear the chant, are supposed to
used to" is precisely the fact that "they" are queer - not that are like
us but that never will be. What " (those who used to are
announcing is our deviance from the normal, our distance from "them" (those
who hear the chant and who are thought to occupy a place closer to the median
point on the normal curve). If queer is that which resists normativity, 12 getting
used to it must mean not normalization but a persistence of queerness.
There is therefore something about the Queer Nation chant that do es not
add up. It introduces a new term into the equation, but without balancing out
the equation. In fact, the chant establishes a new variable in such a way as to
throw the equation out of balance. For Rancière, su ch "fuzzy math" is the
very stuff of democracy. Democracy, we might say, is the regime that cannot
count properly. This is what makes democracy a space or moment of impro-
priety, and it is also why democracy, in truth, is not a regime at aIl (Rancière
2006c: 69-73). Rancière describes the democratic subject (thus, the political
subject par excellence) as "le compte des incomptés" - the count of the unac-
counted-for. Democracy is a miscount because democratic poli tics only
comes about when those who have no part in the social order stake a claim
and take a part within it. "There is politics" argues Ranci ère, only "when
there is a part of those who have no part, a part or party of the poor" (Rancière
1999: II).
1 once glossed this argument by calling democratic politics "the taking-part
of those who have no part" (Chambers 2005: par. 1). This translation has the
benefit of expressing the point in its properly paradoxical foml (those who
have no part, take part), but it flirts with the danger of overstating the willed
participation of a party prior to politics. ln other words, that earlier construc-
tion comes perilously close to the liberal interest-group model of politics that
1 have been at pains throughout this Afterword and throughout this book - to
distinguish from Rancièrean politics. As Rancière stresses, the "party of the
poor" that has a part only when "there is politics" - this party does not initiate
political action but rather is brought about by political action. In other words it
is only politics "that causes the pOOl' to exist as an entity" (Rancière 1999: II,
emphasis added). And this paradoxical formulation can be elaborated (if never
quite "explained") wh en we see that unlike any other political system, democ-
racy involves a fonn of mIe in which there is no title to mIe. It is not aristoc-
racy, rule by the best; it is not oligarchy, rule by the rich; instead, democracy
is mIe by anyone at aIl. The title to rule in democracy is the lack of any title
whatsoever (Rancière 2006c: 41; cf. Rancière 1995b: 94). But this is why de-
mocracy always involves a miscount: It always amounts to "counting" those
who do not, who ought not count.
This cannot be a simple addition, an inclusion of those who were once
excluded. Democracy is not a second count, but a fundamental miscount, and

166 1 The Lessons of Rancière


this democratic be reduced to a
sion. This is a point that Rancière himself has strongly insÎsted upon in clari-
fying remarks he has made in relation to his earlier work. Thus, he
"For my part, tried to conceptualize democratic practice as the LHU'VLLiJ~A.VH
of the part of those who have no part - which does not mean the 'excluded'
but anybody or whoever" (Ranci ère 2007a: 99, emphasis added; also quoted
in Stamp 2009: II). The second phrase, which marks the gap between his
sense of democratic politics and a politics of inclusion, proves essential to
Rancière. Thus, in a separate text from the same time period, Rancière pro-
ductively clarifies and strengthens his sense of "miscount" (and shows why
it must be distinguished from "more counting"): called [emancipation
through dissensus] the count of the uncounted, the part of those who have no
part. It was sometimes misunderstood as the part of the excluded" (Ranci ère
2007c: 15, emphasis added). To include the excluded would be merely to
count differently; it would not amount to a "fundamental miscount" (Ran-
cière 1999: 6). For this reason, Rancière describes the proletariat - the ex-
cluded, the po or as "the class of the uncounted that only exists in the very
declaration [account] in which they are counted as those of no account"
(Rancière 1999: 38). This "miscount" is therefore not merely a failure to
count properly, and it is surely not that which caUs for a recount. The mis-
count names an irreducible remainder; it points to a persistent unaccounted-
for within any count. 13
The miscount demands a more rigorous understanding of "the problems" of
democracy. Democracy cannot solve aIl problems merely through inclusion or
recognition (cf. Deranty 2003a). The struggle against oppression will surely
be an important one, but democratic poli tics both precedes and exceeds the
problem of oppression. 14 Butler frames these issues using the language of
nonns. Norms, Butler shows, render sorne lives (sorne genders, sorne sexuali-
ties, sorne races, sorne nationalities) legible and intelligible. And certain nomlS
create a zone of indiscernibility that go es beyond a question of recognition.
Butler calls this "unintelligibility." She writes,

To be oppressed means that you already exist as a subject of sorne kind, you
are there as a visible and oppressed other for the master subject. ... To be op-
pressed you must first become intelligible. To find that you are fundamentally
unintelligible (indeed, that the laws of culture and language find you to be an
impossibility) is to find that you have not yet achieved access to the human, to
find yourself speaking only and always as if you were human, but with the
sense that you are not ... (Butler 2004: 30, emphasis added)

We can easily draw the connection that Butler leaves implicit here: to find
yom'self rendered unintelligible is sUl'ely to find yourself in a queer relation to
dominant nOffilS, To be unintelligible means to exist in a zone of inaudible,
invisible marginality, such that nonns of gender and sexuality make one illeg-

Afterword 1
ible. of con-
text of theorizing marginalized genders and sexualities, particularly
transgenderism. Yet her arguments here also fit weIl with the broader and more
abstract frame in which Rancière theorizes politics, and Butler has '-'"'+'<-<u,-"....,u
her theory to think about "rogue viewpoints" that are rendered unthinkable and
unspeakable by norms of legitimacy as weIl as by govemmental policy (Butler
2009: 795). In the rare moments when politics does occur, the conflictual con-
junction of the logic of equality with the logic of domination serves, on the one
hand, to render that order of domination visible and, on the other, to expose
(as intelligible) the very subject of politics that had previously remained
unintelligible. But, put in the language of queer theory, this means that politics
both exposes the norm and questions its dominance in the name of that which
it would make queer. As Rancière says, poli tics occurs through the democratic
miscount. We can add: poli tics occurs when the unintelligible make them-
selves/are made intelligible.
His insistence on the miscount makes Rancière's a very queer thinking of
politics (even if he would never put it that way). By refusing any conflation of
le compte des incomptés with the excluded, the marginalized, or the victim-
ized, Rancière consistently queers democracy. By saying that Rancière "queers
democracy" I point to two related phenomena. First, Rancière's fidelity to the
sense of dissensus and to the possibility of disagreement - thought as a situa-
tion of conflict not over the object of speech but "over what speaking means"
(Rancière 1999: xi) - is also a fidelity to a certain queerness, a commitment to
a marginality that cannot merely be included within the dominant frame of the
CUITent police order. Second, the distinction between the queer or the unintel-
ligible (thought as une part des sans-part), on the one hand, and the marginal-
ized or excluded, on the other, hinges on the difference between a liberal police
order and a possible democratic politics. The democratic miscount is a queer
form of counting and a queer form of politics.

*****
Rancière says that to teach without emancipating is to stultify - thus implying
that to teach without stultifying is to emancipate. And if the teacher creates the
conditions under which the student emancipates himself (no teacher ever ac-
tively emancipates a student), it is of no concem to the former what the latter
go es on to do (Ranci ère 1991: 18). Something similar must be said about both
"the lessons of Rancière" and The Lessons of Rancière. Rancière will not tell
his readers what to conclude nor what to do with his arguments, and neither
will 1. Rancière aptly chastises Plato for his belief that the circulation of words
can be controlled, that we can keep them out of the "wrong" hands or contain
their meaning. There is always an excess of words, and they are always
beyond our control. In this Afterword I attest, as really any afterword must, to
that "excess of words" that is literarity, and which makes democratic poli tics

168 1 The Lessons of Rancière


dismiss "an of words" as
words" as Rancière puts it, "nothing but words." But the critique of words for
being just words is really no critique at and to parrot that critique is to
futilely the miscount, the power of literarity, the very force of fJ'V.L.L~H","'.

It is not necessarily scandalous to hear it said that words are merely words. To
dismiss the fantasies of the word made flesh ... , to know that words are merely
words ... , can help us arrive at a better understanding of how words and
images, stories and pelformances, can change something of the world we live
in. (Ranci ère 2009a: 22-23)

If words change the world it is not because those words draw maps for readers
to follow, or provide models for readers to build; it is not because those words
explicate. Ifwords change the world it is because they emancipate; it is because
they caB on that power of literarity that is the excess of words thereby bringing
about a new configuration of the world (le partage du sensible).

Afterword 1
N

Introduction
I. Rancière gave his first book the title La leçon d'Althusser (Rancière 1974a). The
book is widely referred to for marking Rancière's famous "break" with Althusser, but
little attention has been given to the specificity of Rancière's title. While it may also
allude to Althusser's larger "teaching," the singular Lesson in question in the title is Al-
thusser's "Reply to John Lewis" (Althusser 1976). Rancière interprets this text - along
with Althusser's short early essay "Student Problems" (Althusser 2011) - as a "les son in
orthodoxy" (Rancière 201 ra: xxiii). His critique of Althusser starts with a set ofreadings
of Marx that show the gap between Marx's own texts and the "M-L" figure that oc cu pies
Althusser's "lesson" to John Lewis. This makes Althusser's "les son" nothing less than a
les son of mastery - one that establishes the mastery of "M-L" and of Marxist orthodoxy,
while constituting the blindness of John Lewis and his common sense. For more on the
critique of Althusser over questions of mastery, see the recent exchange in Radical Phi-
losophy convened in response to the long-awaited translation of Rancière's first book
(Rancière 20IIa; Montag 2011; Brown 20II). BothAlain Badiou and Slavoj Zizek have
previously used slight variations of the title 1 have chosen for my book .- Badiou's in
French, and Zizek's in the singular (Badiou 2009; Zizek 2006).1 briefiy address the rela-
tion between Badiou and Rancière at various points in the book, and 1 formulate an indi-
rect response to Zizek's critique in the first two chapters. Here 1 would stress that neither
of these authors shows any interest at aIl in the questions of lessons, of mastery, and of
radical pedagogy that prove central to Rancière's work and my reading of it. My choice
of title is therefore designed both to resonate with Rancière's own book titles and to fore-
ground the importance of these issues to both Rancière's writings and to my project
here.
2. Two very fine introductory texts on Rancière's writings have recently appeared,
both of which provide a broad overview of Rancière's writing, proceeding chronologi-
cally across his corpus and both raise just this problem. Tanke responds by saying that
his book offers a "reading" of Rancière rather than an "exegesis"; this, he suggests, allows
him to provide commentary and arguments on Rancière's oeuvre without "undermining
the supposition of equality" (Tanke 20 II: 1). For his part, Davis asserts that his "book
about Rancière cannot with any confidence claim to be Rancierian"; Davis has a different
goal, that of making possible an engagement with "Rancière's singular project" by pro-
viding the needed context (Davis 20IO: xi). In the text above and in the chapters that
follow implement my own "solution" to the dilemma of reading Rancière without stul-
tifying. By declining the temptation to produce an "overview" or "introduction" to Ran-
cière's work, by rejecting a chronological approach, and by refusing to discuss aU of
Rancière's writings systematically, my goal is to produce a book that is less "on Ran-
cière" and more "Rancièrean."
3. In Chapter One 1 offer sorne brief comparisons and contrasts between Arendtian
and Rancièrean accounts of politics. Arendt's commitment to a politics of natality is in a
certain way a commitment to a politics of novelty, of newness, and creation. This sort of
project has affinities with the Rancièrean politics of surprise that 1 describe in the text.
However, 1 do not think it amounts to the same sort of genuine surprise that Rancière
believes in. As 1 show in Chapter One, despite their commonalities, one fundamental
point separates them: for Rancière there is no proper sphere of the political.
4. In this, liberal theorists and social scientists display classic symptoms of para-
noia. As D. A. Miller writes: "surprise ... is precisely what the paranoid seeks to elimi-
nate" (Miller 1988: 164; quoted in Sedgwick 2003: 130). Sedgwick also shows that a
certain sort of paranoid structure, and a resistance to surprise, inhabits and determines
the traditional approaches to critical theory, which she identifies with Ricoeur's phrase
"the hermeneutics of suspicion." ln Chapter Four 1 outline an alternative critical dis-
positif, one that differs in many ways from Sedgwick's "reparative reading" but that
shares with her both the rejection of demystification as stultifying and an openness
to surprise.
5. In my frequent references to "liberalism," 1 mean to call attention to a broad cluster
of concepts including the idea of government based on consent that protects the rights and
property of individuals conceived as autonomous agents. It is this core of liberalism that
matters to me - below 1 call this "doctrinalliberalism" especially as it is connected to a
political philosophy premised on a certain order and hierarchy. 1 neither deny nor mask
over the great variety of liberalisms and the significant extent of tension between and
among them. 1 also do not mean to imply that seventeenth-century liberalism can compre-
hend aIl of the politics or political philosophy of today. Twenty-first-century neoliberal-
ism looks very different from seventeenth- or eighteenth-century liberalism, and the
Lockean state of nature theory looks very different from the Rawlsian original position
theory. That said, for my limited purposes here, two points will suffice. First, a variant of
Lockean liberalism (with the same core terms of equality, rights, and limited government
based on consent) appears to provide the only terms of challenge or resistance to the
growing hegemony of twenty-first-century economic neoliberalism. The former is just as
impoverished as the latter is dominating. Second, the explicit effort to create a political
philosophy of order marks Rawls (e.g., "the well ordered society") even more perspicu-
ously than Locke. For the links between neoconservative thought and neoliberalism, see
Brown (2006).
6. This is not to say that social scientists themselves will not often produce unex-
pected or unanticipated results. But those results will still have to be surprising, or not,
within the terms of order laid out by the social scientific approach; the "surprise" can

1]2 1 The Lessons of Rancière


appear within the grid of inte1ligibility. contrast, and as will
show, the "surprise" of politics for Rancière involves a dimension of unintelligibility, of
an appearance outside the grid, or a disruption of the grid. The social scientific insistence
that there can be no surprises therefore stands in direct opposition to my Rancièrean as-
sertion that polities is always a surprise.
7. Here 1 refer to the now common wisdom that events in Tunisia had a catalytic
effect in Egypt.
8. 1 do not mean to equate the military with social science as a who le, but merely to
suggest that the approach that the military takes in trying to prediet social unrest is un-
doubtedly a social scientific approach. And therefore this quotation serves to substantiate
my daim in the text about the effort expended trying to predict unrest.
9. By choosing the Egyptian revolution as an example to help me work through the
politics of surprise, 1 do not, by any means, wish to reduce Rancière's conception of poli-
tics to revolutionary politics. As 1 show in Chapter Two, Rancière's conception of politics
proves to be both surprising and scandalously impure, and for this reason among others
one cannot equate Rancièrean politics with revolution.
10. A certain "politics of surprise" is but one of the multiple dimensions to Jane Ben-
nett's recent work on "vital materialism." Although Bennett parts ways with Rancière,
she sees in his conception of politics the capacity for surprise that animates ber project.
In this context, she helpfully glosses Latour's phrase, "the slight surprise of action," by
c1aiming that poUlics is "irreducible to the sum of propositions of even an ontologically
plural public" because, in Latour's words, ''There are events. 1 Hever act; 1 am always
slightly surprised by what 1 do" (Bennett 2009: 103; quoting Latour 1999: 281). In this
context, see also Timothy Morton's theorization of the "surprising surprise ... a surprise
that remains a surprise," a formulation that c1early exceeds any social scientific concep-
tion of future events (Morton 2007: 75).
II. Egypt surely also had its share of small-scale demonstrations, sometimes sup-
pressed, along with more than its share of police brutality. When 1 say in the text that
there "appeared to be no conflict," 1 mean specifically that there appeared to be no fun da-
mental political conflict in the Rancièrean sense that 1 elaborate below (BBC 2010).
12. At the time ofthis writing in February 2012, there is clearly more fear than hope
concerning Egypt's political future, but none of this changes my interpretation of the
events of early 20 II. A moment of democratic politics, in Rancière's sense, can provide
no firm guarantees for the future. The politics of surprise marks not just revolutions but
their aftermaths.
13. Unless explicitly noted, as here, throughout the book aIl emphasis in quotations
is original.
14. Coming at the issue from a different angle, Foucault's lectures on biopolitics
show c1early that, historically, there has always been a gap between these logics: "De-
mocracy and the Rule of law have not necessarily been liberal, and nor has liberalism
been necessarily democratic" (Foucault 2008: 32 1; also quoted in Repo 20II: 19). This
point is sometimes lost on those who mount theoretical defenses of "liberal democracy."
15. In the preface to Disagreement, Rancière himself signaIs the gap between his
conception of democratic politics and that of Habermas, but readers have tended to focus
on a specifie critique of discourse ethics and not on the broader point about conceptions
of politics (Rancière 1999: x).

Notes 1 173
r6. The contradiction in terms "police politics" must be kept distinct from the idea
of the "politics of the police;' which 1 discuss in detail in Chapter Two. As 1 show there,
while policing must never be confused with or directly associated with politics, there are
still political implications and effects of various police orders.
17. 1 call Schmitt's critique "infamous" because his criticisms ofWeimer-era liberal
democracy prove so difficult to dissociate from his own support of the Nazi regime and
his later refusai to participate in denazification. This is just one reason why, in clarifying
my argument about the difference between democracy and IiberaIism, Mouffe's mobili-
zation of Schmitt's work makes for a much better comparison than wou1d any direct
comparison between Rancière and Schmitt. As a distinct contrast 1 would also note that
Schmitt's concept of "the political" offers a perfect example of just the sort of poIiticaI
onto1ogy that Rancière rejects - as 1 discuss below (see Marder 20IO).
18. 1 am not suggesting that commenters on an Internet article are representative of
anything at a11, nor that these arguments should be given any particu1ar weight. These
responses simply suggest how hard it sometimes is, within the prevailing common sense,
to see a difference between the democratic and the given liberal political order.
19. For Iiberalism, the question of how to implement political rule always comes
after the liberal framework has been established. Like Locke's discussion of political in-
stitutions late in the Second Treatise (r988), the question of age requirements for office-
holding would thus be a secondary topic of debate, regarding the issue of how to uphold
Iiberal rights and the sovereignty of the people while still making it possible for govern-
ment to protect life, liberty, and property.
20. This discussion also hints at the key differences in political rhetorics between norrna-
tive liberal theory, on the one hand, and Rancièrean democratic polemics, on the other, in that
here we also see two distinct ways to think about the relation between a rule and an instance
of it. Liberal theory does not typically use examples to show or embody how something
works; rather, liberal theory itself produces rules, and examples are meant to illustrate a rule
that is already there, to provide pro of ofthe rule's existence. In contrast, in my discussion here
1 am using the examp1e as an analogy of the general process by which liberalism operates - to
make that process clearer and easier to grasp. (See Finlayson and Atkins forthcoming).
2 r. As 1 show in Chapter Two, while it remains crucial to see the 10gica1leverage that
this move gives Rancière - by, in short, freeing him to rethink poli tics - one shou1d not
conclude that Rancière means nothing more by police than "what we usually mean by
politics." Rancière's concept of la police has a longer history and a greater conceptual
reach than that notion wou1d indicate, and that concept does a great dea1 of important
work in his writings.
22. 1 do not engage in specific detai1 with Rancière's writings on literature and
cinema. Further, while the concept of le partage du sensible lies at the very heart of aIl
Rancière's writings - and 1 therefore address it in numerous locations throughout the
book - 1 do not offer an explicit treatment of Rancière's contribution to art and aesthetics.
Crucially, Rancière himse1f does not consider "aesthetics" a separate category from poIi-
tics, as he says, "Politics is aesthetics in that it makes visible what had been excluded
from a perceptua1 field, and that it makes audible what had been inaudible" (Rancière
2004b: 226; Rancière 2005a; see a1so Ferris 2009). In trying to unravel the meaning of
this claim in political theory, Davide Panagia's work proves exemplary (Panagia 2006;
Panagia 2009, which quotes the line above in its first footnote, r65).

174 1 The Lessons of Rancière


23. Rancière's sense of or therefore
from Foucault's account of the "polemicist" who privileges his own position, refuses dia-
logue, and denies the rights of his interlocutors (Foucault I99T lO9). Rancière and Fou-
cault share the refusaI of mastery that Foucault criticizes in his remarks on polemic, and
Rancière, while never identifying as a polemicist uses the word polemic quite differently,
as explain in more detail in Chapter Four.
24. Rancière shares with Foucault a commitment to "think otherwise." And Foucault
1S one of the very few thinkers that Rancière cites approvingly, one of the only thinkers
that Rancière allows his work to link up with or even rely upon. The case of Foucault
therefore provides an important exception to Rancière's general tendency to disavow
links to other thinkers (Rancière 20I la: 158; Rancière 201 lb: 30).
2S. By "ethical tum" Rancière refers to a phenomenon in the treatment ofboth aesthet-
ics and politics that he sees as paralleling the "retum of the political" in political philosophy.
In bis sense, the ethical tum has little to do with allowing moral judgments to determine
political action. Rather, the "reign of ethics ... signifies the constitution of an indistinct
sphere where not only is the specificity of political and artistic practices dissolved" but also
where the traditional moral distinction between is and ought has been lost (Ranci ère 2006a:
2). Rancière associates this term with the "infiniticization" of the Other, of exceptionality,
especially in the work of Lyotard (Rancière 20 II c: II; cf. Lyotard 1990). Thus, as with
ontology, the "ethical tum" evacuates the specificity of politics as dissensus.
26. 1 focus on Connolly's arguments because he has made a sustained case for ontol-
ogy over the years and because the contrast with Rancière's position proves so stark in his
case. Stephen White has advanced and sustained his own project on ontology in political
them'y, defending a particular and nuanced version, weak ontology (White 1997; 2000;
200S). White thus agrees with Connolly, contra Rancière, that political theory cannot l'id
itself of ontological commitments. Here 1 do not address the subtleties ofWhite's account
(and the distinction he draws between weak and strong ontology) since my goal is to
throw into relief the terms of Rancière's radical rejection of ontology.
27. In her reading of Foucault (a key source for Connolly) Johanna Oksala makes a
related and complementary point: "Ontology is politics that has forgotten itself' (Oksala
2010: 44S). On ber reading, Foucault shows that politics cannot be derived from ontol-
ogy, but rather that "the ontological order of things is itself the outcome of a political
struggle" (Oksala 20 lO: 464). Oksala thereby affirms the inescapability of the ontopoliti-
cal, but from a distinct angle.
28. This leads to what 1 myself see as a fundamental impasse between Rancière's
approach and that of Connolly and others. 1 should emphasize here, however, that 1 con-
trast Connolly's approach with Rancière's for the sole and limited purpose of clarifying
and highlighting Rancière's unique position on ontology. Thus, 1 do not intend for my
in-text discussion of these two thinkers to suggest that there are only two positions on
ontology in the field of contemporary theory; instead, it makes more sense to define that
debate as multipolar, rather than bipolar. The Hardt and Negri position that Rancière
dismisses looks a lot more like what Connolly calls "heavy ontology" (Connolly 20lO:
126) than it resembles Connolly's own efforts to render existential faiths (which partake
of metaphysics and ontology) politically contestable (Connolly 1999: 8; Connolly 20lO:
r68). In fact, Connolly contrasts his sense of the ontopolitical with ontological traditions -
within a chapter titled "Nothing Is Fundamental" (Connolly 1995: S). Thus, it seems

Notes 1 175
that Connolly would not recognize his own approach to ontology in Rancière's
critique, and that there are at least three positions at stake -- heavy ontology, ontopolitical
interpretation, and Rancière's rejection of aU ontology. Nevertheless, the wide gap be-
tween Connolly's insistence on the ineluctability of ontology, on the one hand, and Ran-
cière's own insistence on eschewing all ontology, on the other, helpfully serves my
purpose in the text of drawing out the contours and stakes of Rancière's overall approach
to ontology.
29. Chapter Three not only explores Rancière's conception of subjectivation in more
depth but also provides a much fuller explanation of the use and meaning of the tenTI in
English and th en explores the question of translation from Rancière's French.
30. See Chapter Four for a much fuller discussion of the uniquE: publishing history
of Lire le Capital, including Rancière's contribution, which was excluded from English
translations of the work. Chapter Four also includes a close reading of this text and a
discussion of its significance for Rancière's broader corpus. Rancière's single essay first
appeared in English translation broken into separate pieces, in various journals in the
1970s. A complete, translated version of the essay did not appear until 1989, but this ver-
sion was published in a little-read and little-known edited volume that is now out of print
and very hard to find: hence my claim in the text that it has not been properly published,
since it surely is not yet widely available to Rancière's very large English-speaking
audience.
3 1. One way to understand the complicated and difficult reception of Rancière by
British and North American audiences is by emphasizing that Rancière does not produce
arguments in the same way as so many political theorists do, and therefore Rancière's
writing often does not fit the established reading habits of these audiences. Understood in
a traditional way, an argument presupposes a set of criteria for validation of that argu-
ment, but Rancière insists that politics undoes and denies any sort of pre-established cri-
teria. In this sense we might call "argumentation" a process of the police, whereas
polemicization is a process of politics. Polemics is an alternative genre to argument, one
that intervenes, affirms and resists, but does so without presurning the validity of a prior
set of conditions that would legitimate political claims. Polemics is thus itself a form of
"part-taking." In this context, 1 first wish to stress that Rancière's polemics, and his writ-
ings in a more general sense, can also be understood to offer their own sort of arguments -
so long as we are willing to broaden our conception of what constitutes an argument.
Nonetheless, since my goal is not to recreate Rancière's polemical engagements - a proj-
ect that wou Id be doomed to failure in any case, given the conjectural nature of those very
engagements - my style is not Rancière's. Those that try to write about Rancière by writ-
ing like Rancière often do very Httle to open their readers to an encounter with his work.
Indeed, their style ends up being more isolating and alienating than Rancière's own;
mimicking his style does litt le to allow Rancière's thinking to reach a broader audience.
Since 1 do not consistently polemicize the way that Rancière does, there are times when
my own approach may even appear to be at odds with his. In other words, 1 often make
arguments where Rancière refuses to do so; sometimes 1 even engage in exposition or
exegesis. My goal is to animate Rancière's polernics, to illurninate them, and to transform
them for my own ends. My audience is also different from Rancière's. Rancière's writ-
ings do not speak to a pre-given discipline; rather, they contest, perhaps even exp Iode, the
bounds of disciplinarity; his work thrusts toward anti-disciplinarity (Mowitt 1992) or

1 The Lessons of Rancière


alter-disciplinarity (Bowman 2008b). too, wish to to a broad and inter-
disciplinary audience, but at the same time also write to and for an audience of political
theorists. bet in this book is that the occasional tensions between my more argumen-
tative style and the polemics of Rancière can prove productive and will therefore pay off.
wish to facilitate both dialogue and understanding, on the one hand, and dissensus and
misunderstanding (meséntente), on the other.
32. As discuss in Chapter Two, Rancière caUs politics a rare, sporadic, temporary
and unpredictable event (Ranci ère I999: 17). This emphasis on the rarity of politics must
be paired with a heavy dose of skepticism toward claims that "nothing happened." "Noth-
ing happened," has, in many quarters in France and elsewhere, come to be the dominant
interpretation of May '68: the event that was a nonevent. What should we say about May
'68? That nothing happened in May '68 (Ross 2002: 66). To take up Rancière's under-
standing of politics means to commit to the idea that in the claim that "nothing happened"
something else (though surely not politics) very important is happening. The insistence
that nothing happened is an effort not just to say that nothing changed, but also an effort
to enact this nothing, to put everything back in its proper place. Some commentators are
already preparing the ground for offering a similar interpretation of January 201 l and the
Arab Spring as events in which "nothing happened." In this context, Ross's account of
May '68 serves as a healthy warning.
33. ln this way Rancière's thinking of history and the event resonates much more
with Foucault's understanding of his own books as emerging from archives and then
disappearing through their own explosions as "object-events" (see Huffer 2009: xi).
34. One might go so far as to say that, in a certain sense, Hobbes is both the first and
last author in the liberal tradition to spell out his argument for equality. Like Hobbes,
Locke also characterizes human beings in a state of nature as free and equal, but Locke's
argument for equality -- humans are an species of the same "rank" and God has not placed
one above the other - looks a lot fuzzier than Hobbes' s powerful imagery of the small and
the weak dropping big rocks onto the heads of the big and the strong. Later in the Second
Treatise Locke even withdraws part of his own argument in favor of equality (Locke
1988: 304) (see Carver 2006, and Carver 2004, chapter 7, for further discussion).
35. Another way to capture this line of thinking would be to describe equality as a
liberal conceit. Liberal theory operates with equality as a fixed point, a given, yet one
somehow also gets the sense that even liberals know the world is marked not by equality
but by radical and almost comprehensive inequality.
36. 1 owe this formulation to an anonymous reviewer.
37. On the importance of logic (thought in a way quite distinct from analytic phi-
losophy) to aIl of Rancière's work, see Tanke (20II: 9).
38. Rancière's emphasis on the "anyone and everyone" should not be mistaken for a
move of abstraction or a generalizing tendency. Rancière does not see individuals as re-
placeable, in the sense that for Rawls we do not need real individuals in the original posi-
tion, just their generic "representatives." The "anyone" for Rancière is always a someone.
As 1 read hi m, Rancière's point in referring to the "anyone at aIl" it not to generalize but
to democratize. Anyone has a title to rule - because democratic rule is based on no title at
all but someone will always rule. The fact that the anyone is always a specific individual
someone also poses problems for the use of gender-neutrallanguage, since to refer always
to a plural "they" or an apparently ungendered "he or she" (in an effort to maintain gender

Notes 1 177
neutrality) would be to substitute a generic individual for the definite individual to which
Rancière is always referring. Faced with this difficulty, will occasionally alternate be-
tween a concrete "he" and a concrete "she," as do in the discussion of radical pedagogy
in the text below.
39. Throughout this section 1 provide a particular reading of "liberalism" and liberai
theory designed to clarify the contours of Rancière's conception of equality and at the
same rime to mark the stakes of my own argument for democratic poli tics. We might
helpfully name this variant of liberalism, and its approach to seventeenth-century authors
such as Hobbes and Locke, "doctrinalliberalism." It is a widely recognized account of
liberalism, one that resonates with les sons given each year in history of political thought
courses, and one that dovetails with arguments made by Rawls and other important con-
temporary liberal theorists. Nonetheless, one couid certainly reread authors like Hobbes
and Locke against this very doctrinalliberalism for which their texts have been appropri-
ated by contemporary liberai thinkers. Such an interpretation might actually complement
Rancière's thinking of equality and democracy. On this reading, one would first empha-
size the rhetorical and to some extent practicai force of Hobbes's and Locke's arguments
about equality: namely, to tear down any naturalization of aristocracy and politicai role of
"birth." Equality was therefore not so substantive for Hobbes and Locke as it is for doc-
trinalliberalism. Second, their arguments for equality were ultimately designed to legiti-
mate inequality, since it was obvious to Hobbes and Locke that some are legitimated to
rule others. Equality was therefore not so central for Hobbes and Locke as it is for doc-
trinalliberalism. Third, the Rancièrean sense of politics as "calling a system of domina-
tion into question" looks a great deal more like what Hobbes and Locke were up to
politically when they wrote their texts, that is, prior to the constitution of "liberalism" as
a doctrine. This means that Hobbes' and Locke's arguments, as originally put, and when
thought as explicit political interventions (which they were) prove significantly more
Rancièrean than the decontextualized version out of which liberal "theory" has been con-
structed as "what liberalism is" - that is, modern doctrinalliberalism. Finally, one might
add that just as May '68 was an irruption of politics in Rancière's sense, so were the
events in Europe of 1848, events led by "revolutionary liberals" - a term that makes no
sense when understood through doctrinal liberalism, but is nonetheless historically ac-
curate. When read this way, of course, Hobbes and Locke would, quite simply, not be
liberals in the current doctrinal sense. For just this reason, while 1 am sympathetic to this
more historically situated and also more radical reading of Hobbes and Locke, 1 stand
behind my own effort to contrast liberalism in the doctrinal sense with Rancière's account
of equality. Moreover, the hegemony of doctrinal liberalism minoritizes the historical
reading of Hobbes and Locke. On this issue, 1 owe a debt to Terrell Carver for extended
engagement, deep insights, and a few turns of phrase.

Chapter One
1. ln this chapter 1 cite a large number of the early texts on Rancière and poli tics, but
1 do not discuss a few key essays that should be mentioned here: Michael Dillon (2003a),
Kirstie McClure (2003), Aamir Mufti (2003), and Paolo Palladino and Tiago Moreira
(2006). At the time of this writing, a wealth of literature on Rancière has either just
emerged or is about to appear, including special issues of journals and secondary source
books on Rancière's work. It goes without saying that it is impossible to cite aU of this

1 The Lessons of Rancière


literature, but at various places in the book point the reader to sorne of the major books,
including monographs by (2008, 201Oa), Davis (2010) and 1'anke (201 n, and edited
volumes by Deranty (20IO) and Bowman and Stamp (201 n.
2. This is the first of many places where my reading of Rancière might appear to
intersect with the thought of Alain Badiou. In this case, Badiou and Rancière share a
harsh condemnation of "political philosophy." For Rancière this judgment reflects a fun-
damental conflict between philosophy and politics (and Rancière sides with politics in
this dispute), whereas for Badiou the problem of "political philosophy" consists in a
misunderstanding of the proper l'ole of philosophy in relation to politics (Badiou 2005:
118). In discussing the ide a of "political difference" in the text, below, one could also
broach an analysis of the shared but distinct special status that Rancière and Badiou both
give to la politique. On this point, see Marchart's very helpful chapter on Badiou, specifi-
cally the argument concerning Badiou and Rancière's shared "reversaI" of the privileging
of "the political" over "politics" (Marchart 2007: 119-20) within Marchart's broader
exploration of political difference. In general, further exploration of the sirnilarities, and
radical differences, between Badiou and Rancière would take me too far afield within this
chapter, but 1 would suggest that Marchart's analysis cornbined with the exegesis 1 offer
here (along, of course, with Badiou's own chapters on Rancière in Metapolitics) might
help set the stage for future work on Rancière and Badiou. On Badiou and Rancière, see
also the work of Todd May (201Ob) and Nina Power (2009). Ultirnately, however, 1 must
also stress that in my efforts to mobilize Rancière's thought for the sake of rethinking
democratic politics, 1 simply do not see the benefits of extended discussion of Badiou.
Indeed, 1 take Badiou's own words as evidence here. Contrasting his own work with
Rancière, and in the face of Rancière's commitment to the equality of intelligences,
Badiou writes, "1 believe that my hypothesis is, simply put, aristocratie" (Badiou 2009: 37).
Badiou's Platonism will always be anathema to a Rancièrean approach.
3. The "an-archic disorder of politics," the way in which politics for Rancière always
remains bound up with a dis-ordering principle, must be sharply distinguished from a
political theory of anarchism - a theory that wou Id try to ontologize that principle and
make it serve as a ground for its own political theory. Chapter Two casts these differences
in starker light.
4. 1 call this a vertical organization to express Rancière's insistence on thinking
social order as an order of hierarchy and domination (see Bennett 2009). One might
wonder whether "police order" describes any social order, whether the tenn is broad
enough to include spontaneous and horizontal ordering. Rancière simply does not ad-
dress these issues; for him, the social order is always stmctured hierarchically. But we
might go on to say, within Rancièrean terms, that the horizontal dimensions of a social
order are produced by politics.
5. 1 take up a more detailed engagement with le partage du sensible in Chapter Two.
For a more intricate discussion of translation vis-à-vis this phrase, see footnote 3 in that
chapter.
6. In Chapter Two and Chapter Four, 1 make clear my own resistance to and rejection
of this "normative grounds" line of argument, which 1 think fails to find its target in either
Foucault or Rancière. More significantly, in Chapter Four 1 offer an expanded reading of
Rancière's writings in the service of contributing to a new thinking of critical theory. My
point in the text ab ove is much more delimited: merely to wield Rancière's definition of

Notes 1 '79
politics for argumentative leverage (as a tool of critique) is to read his project quite nar-
row ly and to limit the overall import of his work. Thus, this first "use" of Rancière' s defi-
nition of politics is surely "critical work," but only in a relatively thin sense that must be
distinguished from the idea of a developed critical the ory.
7. As just one example, Markell demonstrates that the territorial reading of The
Human Condition leads to a conftation of the "philosophical representation of work" and
"the phenomena Îtself' (Markell zo II: 35). In this argument 1 hear important echoes of
Rancière's critique of the tradition of political philosophy, as 1 will discuss in the next
chapter.
8. It is also worth nothing that the text where Honig caUs on Rancière the most, De-
mOCl'acy and the Foreigner (Z001), is also the text where she relies on Arendt the least.
9. For non-native French speakers such as myself, 1 would interrupt here to coyer
the basics: this "distinction," in French, is tirst of an nothing more than the difference
between feminine and masculine - and thus becomes no difference at aH when translated
into the "genderless" English language. In regular French usage, la politique connotes
more of the everyday business of politics (e.g. "department politics" as used in English)
while le politique suggests something broader, more systematic, more philosophical.
IO. And it does seem to be Zizek's claim - and not a misreading by Marchart as
Zizek opens his reading of Rancière in The Ticklish Subject by referring to "la politique!
police" as a singular entity that would be "perturbed" by a "political mode of rebellion"
(Zizek 1999: 17 Z ).
II. In his translator's introduction to The Politics of Aesthetics, Gabriel Rockhill
do es make reference, if only in passing, to a distinction between "politics" and "the po-
litical," but he gives no French translation of those terms in his English introduction, and
his translation of the French marks no distinction either (Rockhill 2006: 3). 1 will return
to this point in the text below.
12. Tanke follows Deranty entirely here, when he says, "The political [le politique]
is this third space of contestation, an indeterminate and always shifting meeting point of
the police and politics" (Tanke 201 1: 5 1, emphasis added). In making this claim, Tanke
provides no citation of any texts at an - neither those of Rancière nor of anyone else. A
few lines later, on a different point, Tanke does cite Deranty.
13. Or better, the distinction emerges from the French translations of those English
texts, as 1 exp Iain in more detail in the text below.
14. All translations from this I998a edition are mine.
15. The French translation leaves out these first two paragraphs, and begins instead
with Rancière simply asking, of himself and the reader, "Qu'est-ce que le politique, nous
est-il demande?" (Rancière I998a: 83).
16. It seems plausible to make this move, but technically it is surely a mistranslation,
since any English-to-French dictionary will give la politique as the translation for policy.
Here we see even more starkly that in this lecture Rancière proposes differences that
really hoid only in English. A direct translation back into French of the English Rancière
uses would give us the "difference" between la politique and le politique.
17. Even here, where Rancière explicitly clarifies what might be at stake for him in
the difference between la politique and le politique, we still see nothing like the stark
"political difference" (as Marchart tinds in so many other authors) nor a clear delineation
of three terms as suggested in Deranty's work.

180 1 The Lessons of Rancière


18. For invaluable assistance with the translation of Marchart's thank
Anne Kantel.
19. Rancière' s original English text (I 995c) uses the word "field" - translated as
terrain in the French text (1998a) and Terrain in Marchart's German (2010).
20. Marchart himself do es not make note of the fact that the I998 French text is a
translation of the earlier English lecture, Bor does he emphasize, as will below, the fact
that Rancière's key texts on politics l'rom the I990S l'ail either to refer to le politique or to
introduce a third term.
1. Rancière elaborates on this point in his recent third-person essay: "Most of those
who conceptualize politics today do it on the basis of a general theory of the subject, if
not on the basis of a general ontology. But Rancière argues that he cannot make any de-
duction l'rom a theory of being as being to the understanding of politics, art or literature.
The reason, he says, is that he knows nothing about what being as being may be. That's
why he had to manage with his own resources which are not that much. Since he cannot
deduce politics l'rom any ontological principle, he chose to investigate it out of its limits,
he means out of the situations in which its birth or its disappearance are staged" (Rancière
2009b: I17)·
22. This doubling may disappear in English translations if and when they lose a
sense of any subtle differences between la politique and le politique - when the act of
translation turns these two terms into the oneness of the English "poli tics." But it would
be just as wrong for English translators to construct a third term, that is, "the political,"
and insist on its appearance whenever Rancière writes le politique. This reading is further
supported by the fact that Rancière's own references to a distinction emerge only in Eng-
lish. Rancière himself thereby suggests a doubling of la politique much more than he
points to a creation of three tenns, à la Deranty.
23. Notice that on Rancière's reading, "the drawing of lots" is not a procedure or
mechanism for implementing democratic princip les or maintaining a democratic regime.
Consistent with Rancière's arguments elsewhere that democracy is not a regime at aIl
(Rancière 2006c: 7I), he here refers to the drawing of lots as just another name for de-
mocracy itself.
24. For calling my attention to Rancière's essay on Derrida and pushing me to
answer the question of what to make of the term political when and where it does appear
in Rancière's writings, 1 am very grateful to Patchen Markell but 1 owe him a much
bigger debt for the arguments of this paragraph.
25. For this reason, there would be no reason to go back and retranslate Rancière's
writings on politics from the mid- 1990s, rewriting them in Bloomian fashion, rendering
"politics" for each and every appearance of la politique and "the political" for aH in-
stances of le politique. (And the fact that Deranty himself often refers to la politique as
"the political" would undoubtedly trouble such a project.) For his part, Rancière made
little or nothing of that difference at the time of writing those texts. But even if he might
make something of it now, English translators would still be right, in a certain sense, not
to substitute for Rancière's French a coherent system of three English terms (police, poli-
tics, the political). Current English translations of both Disagreement and the "Ten
Theses" fail to mark for the English reader any potential distinctions between le politique
and la politique, but this is not a failure to be fixed. Even Davide Panagia, who otherwise
reads Rancière more subtly and deftly than anyone, pays no heed to the difference

Notes 1 181
between la and le politique in his retranslation of "The Ten Theses"
and Panagia 2001). Perhaps it would make sense to leave la politique and le politique in
French (at least in parentheses) in future translations of Rancière's work.
26. In this vein, Deleuze's project to articulate a non-dialectical opposition in Ni-
etzsche proves exemplary, and, of course, in many ways the entire Derridean project
devotes itself to a thinking of non-dialectical difference, that is, différance (Deleuze
1983; Derrida 1982). Numerous commentators on Rancière have either addressed the
issue of "dialectical thinking" in his work directly (Thomson 2003), applied dialectics to
Rancière (Zizek 1999), or offered a putatively Rancièrean understanding of politics that
obeyed its own dialectical logic (May 2008). For the purposes of my specific argument
here, the question of "dialectics" is not the essential one, since what matters, ultimately,
is whether or not politics can be rendered pure, or if politics is that which always renders
itself (through its doubling) and its other, impure. A dialectical overcoming produces a
purified result, so in that sense a dialectical approach to Rancière would arrive at the
same problematic telos of a purified politics. For the sake of clarity my argument here
largely avoids the language of dialectics. 1 address this argument from a different angle
in Chapter Two.
27. On impure politics, see David Kaplan's notion of "impure democracy" in the
work of Paul Ricoeur (Kaplan 2008: 207; cf. Marchart 2007).
28. In Chapter Three 1 offer an in-depth discussion of Rancière's notion of the
"excess of words," including both a genealogy of the phrase's emergence in his writing
and an argument about its relation to the concept of "literarity." At this point, 1 should
mention only that the notion of an "excess of words" is neither an abstraction nor a eu-
phemism for Rancière. The phrase points to two distinct but related claims that Rancière
makes across his writings: first, a general claim about political philosophers' repeated
attempts to eliminate the excess of words (going back to Plato's efforts to regulate dis-
course and banish the poets and sophists); and second, a specific historical claim. In this
second sense "excess of words" names a phenomenon produced by eighteenth-century
democratic revolutions whereby names proliferate without titles to fix them in place.
Democracy offers unruliness, then, because in it names circulate without any authority
(whether epistemological or political) to pin them down to a fixed identity. "Excess of
words" therefore offers a different way of expressing the impurity of politics: namely, the
impropriety of anyone and everyone counting as a citizen, of having "citizen" itself count
as a proper title. This note owes a deep debt to an anonymous reviewer.
29. In the Afterword to the English translation of The Philosopher and His POOl;
Rancière refers to his own "theory of politics." To my knowledge, this is the only place
Rancière uses such language in a positive sense, though even here he complicates it sig-
nificantly, saying that it "moved considerably away from what is generally understood by
that name" (Rancière 2004b: 225).
30. It is also why the moment a putatively democratic project becomes purely self-
referential- concerned only with those who struggle and not with the stmggle itself - is
the moment that it is no longer "political" in Rancière's sense (Thomson 2003: 17).

Two
1. In this chapter 1 stage a detailed engagement with, and offer a close reading of,
May's central book on Rancière (May 2008), but May's other writings on Rancière (May

182 1 The Lessons of Rancière


20 l oa; 20 lob) and his on anarchism also prove to
his project and therefore to my encounter with his work here.
2. and Muhle are the only commentators know of who spend any time on the
links between Rancière's concept of "police" and Foucault's work on governmentality.
notes the cormection but does not place much weight on it in his reading of Ranci ère.
Muhle emphasizes the connection much more and elaborates on its importance, and her
unpublished reading has proved to be a great resource for my own argument.
3. Oliver Davis goes against the grain of almost all other translations of Rancière to
make a strong case for rendering le partage du sensible as the partition or distribution of
the sensory, rather than the sensible. Davis argues that Rancière uses the phrase le part-
age du sensible to make the point that not everything is available to sensation. A prior
partage divides, carves up, and shares out the sensory realm. Because what is available to
sensation has been limited by a partage, we emphasize the sensible realm - the realm in
which sensation can occur. However, Davis contends that "as a matter of logic this re-
strictive work is already performed, in the expression le partage du sensible and the
concept it designates, by the noun le partage (or its verb partager), such that to try to
perform it again by opting for 'sensible' over 'sensory' is to try to say the same thing
twice" (Davis 2010: 180). 1 think Davis makes a compelling case. However, it is not
completely clear to me whether the redundancy Davis wishes to avoid, "the distribution
of distributed meaning," is not one that Rancière, in his paradoxical style, actually wishes
to embrace (Davis 2010: 179; quoting Rancière et al 2009: 159). In any case, rather than
weighing in on this debate over translation (and definitively deciding to translate the
phrase one particular way or another), 1 choose to use a variety of terms in English (dis-
tribution, sharing, partition, cutting, for partage; sensory and sensible for sensible) and as
often as not to simply use the French. Clearly the phrase le partage du sensible proves so
subtle and multivalent that no singular translation will do, and surely none will be the
definitively "correct" one.
4. In the Afterword 1 address this point in much more detail and by way of the spe-
cifie example of queer politics.
5. Even though "state apparatus" is obviously an Althusserian term, Rancière ne-
glects (or refuses) to cite Althusser on this point. In "Ten Theses," on the other hand, in
the context of rejecting the notion that the police primarily interpellates subjects, Ran-
cière does cite Althusser (2001: par. 22).
6. As 1 will argue in much more detail in Chapter Four, taken cumulatively these five
points add up in such a way as to indicate that a challenge to the police cannot rely on the
standard criticallogic of inversion. The police order's partage du sensible cannot be re-
sisted by turning it upside down, but only through internaI dismption, that is, dissensus.
7. This fact seems significant in light ofthe relative consistency of Rancière's thought
across his more than fort y years of writing. Many of Rancière's key tenus remain with
him from his early archivaI work, through his writings on politics and philosophy in the
1990s, and all the way on to his work on the politics of aesthetics over the last ten years.
It is striking then that "the police" do es not appear in Aux bords du politique, the collec-
tion of essays first published in I992, and it hardly appears at all in La haine de la dé-
mocratie (2005b), first published in 2005. Only in the two major political works published
in the middle of the I990S - La mésentente (I995) and "Dix theses" (I997) does Ran-
cière offer any serious attention to the concept of "the police." One might argue that

Notes 1
le partage du sensible la police in Rancière's later the idea
of a police order but on a broader level. But that only lends more credence to the idea that
"the police" serves a particular purpose in terms of Rancière's political intervention in the
mid-1990s.
8. In this context, however, would insist on not conftating Rancière's analysis of
post-democracy with his articulation of archipolitics, parapolities, and metapolitics. The
difference matters because it is a difference in both objects and levels of analysis. Ran-
cière's analysis of archipolitics (for ex ample) provides a critique of the project of poli tic al
philosophy, whereas his analysis of post-democracy provides a critique of contemporary
politics. The latter therefore constitutes a political intervention in a sense different from
the former. As 1 show in the text below, 1 think May sometimes elides these differences.
9. Another way of putting this would be to say that May did not discover anarchism
through a reading of Rancière. Quite the contrary, May was already a committed, well-
known, and forceful defender of anarchist thought before he came to Rancière's writings
and folded Rancière's thought into May's own anarchist project. See May (I994); see
also Ward (1982).
10. For a fuIler reading of Rancière's arguments about archipolitics, parapolities,
and metapolities, see my discussion in Chapter Four.
1 1. Zizek outlines the paradox of this dialectieallogie and accuses Rancière himself
of falling prey to it (a faIse accusation, according to my argument, and as 1 show below).
Alex Thomson had earlier worked through a similar logie in Rancière's work and arrived
at more subtle and productive conclusions. See Thomson (2003), Zizek (2006); cf. Valen-
tine (200S).
12. Although 1 present this reading as a critique of May, May himself might freely
admit to the difference between his position and Rancière's, on this point. May departs
from Rancière's notion that politics begins with a wrong, with what Deranty helpfully
caUs an "ontological torsion" (Deranty 2003a: par. S). Explicitly ftagging his claim as an
argument against Rancière, May claims that poli tics "is not necessarily antagonistie"
(May 2008: SI-S2). May must, in a way, mark this difference from Rancière, because
Rancière's politics always remains impure, always occurs in medias res. In contras t, and
as 1 exp Iain through my reading of May in the text below, anarchist politics must be self-
contained, self-referential and sui generis. May sees in this difference a need to tweak
Rancière; 1 see in this difference the very reason not to read Rancière as May does. In a
later text May explicitly states that Rancièrean polities "is not simply an affair of self-
involvement" - a distinction that does not seem so clear in May's earlier book (May 2009:
1I6).
I3. May would insist on the distinction between his conception of anarchism and
"revolutionary anarchism," and would even suggest (rightly, in my opinion) that Ran-
cière's thought undermines the traditional distinction between reform and revolution
(May 2012). My point in the text is neither to pigeonhole May's thought within the tradi-
tion of revolutionary anarchism, nor to enter into debates about the nature of anarchist
thought. My disagreement with May centers on the idea of politics as autochthonous and
self-sustaining. It is the idea that poli tics concems only itself that 1 think takes May far
afield from what 1 see as Rancière's more radical conception of democracy (and in so
doing, this account of politics seems to imply an attendant sense of "revolution," even if
that is not May's stated position within anarchist thought).

1 The Lessons of Rancière


14. frames his second - wherein he offers his detailed of
politics and police and provides the majority of passages on which 1 rely in my reading
here - with the story of Clemson, South Carolina. discusses citizen mobilization in the
wake of the death of a young African American man who was killed when he was hit by
a car driven a white male student at Clemson. In the aftermath, himself worked
for two years as a community organizer with a group trying to improve relations between
the police force and the African American community. The group eventually ran two
candidates for city council, both of whom lost (with low African American voter turnout).
In concluding the chapter, notes that the group won certain concessions from the
police force. But May's own assessment of the results of his two years of work is deci-
sive: "Politics did not happen at Clemson" (May 2008: 75). For a more detailed account
of this case, see May (2009).
15. This argument also reveals another problematic dimension of the title to May's
book: "creating equality" has no part in "the political thought of Jacques Rancière."
16. While Zizek (2006) proposes this lack of encounter as a critique of Rancière and
while Thomson (2003) delineates it as a potential trap for Rancière, it should be noted
that an important body of commentators directly contests this line of reading, arguing
instead, as 1 have in Chapter One, that Rancière's politics can never be pure (see Muhle
2007; Panagia 2006; Rockhill 2006).
17. By insisting that we never live in democracies and never will, 1 take seriously and
extend Rancière's claim that democracy is not a regime or an exercise of power. In this
vein, my argument for a Rancièrean conception of democracy could also serve as a partial
response to Jodi Dean's recent critique of the celebration of democracy on the US 1eft. As
she nicely puts it, "Calling for democracy, leftists fail to emphasize the divisions neces-
sary for politics" (Dean 2009a: 76). 1 am extremely sympathetic to Dean's argument,
particularly as she applies it forcefully to the conception of democracy advanced by de-
liberative theorists. As Dean shows, the idea of more democracy, as merely the increased
circulation of opinions, is entirely compatible with, and entirely recuperable by, neoliber-
alism. Despite all this, none of the "invocations of democracy" that Dean discusses bear
any resemblance to the Rancièrean account of democracy. Indeed, one central motivating
force for my argument in this book that is, my effort to use Rancière's writings to drive
a wedge between liberalism and democracy - is to offer an alternative account of demo-
cratic politics to the one Dean appropriately and powerfully criticizes. Rancière plays no
central role in Dean's book, but at times in the book (and elsewhere) Dean does suggest
that Rancière's project can also be dismissed on similar grounds to those that Dean ef-
fectively marshals against deliberative democrats. This is the point at which 1 dissent
from Dean's account. After aIl, Rancière's account of democracy as dissensus surely
cannot be accused of "fail[ing] to emphasize ... divisions," and Rancière's own critique of
Habermasian conceptions of communication and democracy helpfully complements
Dean's account. My general point is a simple one: there is a way to defend democratic
politics against neoliberalism. Furthermore, if secondarily, Dean's major points of con-
tention with Rancière rest on questionable readings. She criticizes Rancière directly by
rejecting two claims that she attributes to him: tirst, "The claim that we are in a post-
political time, that poli tics has been foreclosed ... is childishly petulant" (Dean 2009b:
23; cf. Dean 2oo9a: 12); and second, "Rancière's claim that 'the state today legitimizes
itself by declaring that poli tics is impossible' simply does not apply to the United States

Notes 1
Dost-SeDtemtJer II'' (Dean 2009a: I4). the case of the first critique, Dean has attrib-
uted to Rancière an argument that is not his own, bur rather is the object of his critique.
"Post-democracy" (Rancière's term) is not a regime we live in, but a critical conceptual-
ization of the very goal of "consensus democracy" (Ranci ère 1999: lO2). In the case of
the second critique, Dean quotes Rancière's description of early I990S France and then
asserts strongly that it does not capture late 2000S USA. Her statement is certainly true,
but 1 am not sure it has the critieal force she would appear to attribute to it. In not taking
Rancière's conception of democratic polities seriously, Dean may miss out on a potential
ally in her battle against neoliberal "communicative capitalism."
18. For reminding me of the connotations of the word struggle in traditional ac-
counts, and thereby helping to clarify and refine the important senSè of "struggle" in
Rancière's understanding of democratie politics, 1 owe a debt to Benjamin Arditi.

Three
1. Reading Oliver Davis's helpful summary of the work done in the I970S by Ran-
cière and his collaborators at Les révoltes logiques, one can see clearly that Rancière's
work made a certain turn to language, to words, to logic, and to a polemical sense of argu·,
ment from very early on. As Davis puts it, reflecting on the meaning of the title of the
journal (borrowed from a Rimbaud poem): '''Logiques' reflects less the inexorability of
spontaneous resistance and points more to the words, the language, involved in that resis-
tance" (Davis 20lO: 39-40). Rancière himself formulates the point succinctly: "What is
called rebellion or revoit is also a scene of speech and reasons" (Rancière 20r rb: lo).
This chapter develops the account of literarity within the context of philosophieal debates
over language, but 1 should stress at the outset that the question of literarity, of the excess
of words, always exceeds the often narrower terms of language or discourse. Rancière's
is a polities of the unsayable and the inaudible, and also of the invisible and unpresentable
(Panagia 2006: 88). In this chapter 1 center my account on the former only as a matter of
fOCllS and never to the exclusion of the latter.
2. This maneuver must not turn into an ontologizing move. The idea is not to ontolo-
gize something in Rancière but to get at sometlzing that Rancière himself would never get
at, to put his arguments into motion toward ends he does not conceive.
3. Sinclair's translation stands out from others in its unique rendering of the distinction
between phonë and logos as the difference between expressing and indicating; only Sin-
clair's translation uses the word express to describe whatphonë does (Aristotle 1981). Rack-
ham, Barker, and Everson aU choose the word indicate for the role of phOnë, while they
differ on the work done by logos: Rackham uses indicate a second time; Barker pieks de-
clare; and Everson goes with setforth (Aristotle 1944; Aristotle 1958; Aristotle 1996).
4. Because of this, as 1 show below, it turns out that a fairly literaI translation of Ar-
istotle's Greek into English looks very similar to a translation of Rancière's French trans-
lation back into English. And both ofthese versions diverge significantly from the Sinclair
version used by Rose.
5. 1 calI the translation "mine" because 1 have chosen the final rendering and because
1 take full responsibility for any mistakes, but for making this translation possible 1 am
deeply indebted to the indefatigable work of Nathan Gies.
6. Immediately after the quote, Rancière digresses very briefly, noting that these
famous words of Aristotle have quite obviously been read by many political philosophers

186 1 The Lessons of Rancière


as the first to delineate a political human nature. Rancière chastises
Hobbes for dismissing this notion and Strauss for seeking to restore it; instead, he sug-
gests that we look more closely at the argument from which the so-called theory ofhuman
nature 1S derived. To do so means to offer a reading of both the logic and the implications
of these lines of Aristotle the crux of this first chapter of his book (Rancière I999: 2).
7. In ordinary usage the primary meaning of the French sensible is "sensitive" in
English, yet there are no translators or readers of Rancière, to my knowledge, who sug-
gest "sensitive" as the best rendering of le sensible. Even authors such as Davis (201O)
and Panagia (201O), who both remain highly attentive to issues of translation, particularly
the question of le partage du sensible, make no mention of the meaning of sensible as
"sensitive." Yves Citton does caU Rancière to task for his critique of Negri's idea of a
"sensitive community" [une communauté sensible], pointing out that for Rancière any
community must always be une communauté sensible, and Citton probably offers the
fullest account of the multivalent meaning of sensible in relation to Rancière's project
(Citton 2009: 125). l do not advance this discussion further, but only mention it here as
one other dimension of the poUtics of translation in Rancière's work. 1 am very grateful
to Bonnie Honig for calling my attention to this matter.
8. Here Rancière uses signe, not indice, in his discussion of Aristotle, a move that
might lend some support to Rose's choice of "sign" for indice to the extent that translat-
ing both French terms with the English "sign" establishes a certain continuity between
these two discussions (Ranci ère I998a).
9. Given that Habermas explicitly adopts an Aristotelian conception of the human
animal using language for political ends, Rancière's critique of Habermas offers yet an-
other clue to his fugitive theory of language. Rancière explicitly rejects, and in stringent
terms, both Habermas's theory of discourse ethics and the model of language that it im-
plies. The Habermasian form of "political rationality" and its search for consensus "is a
bit too quick to take as read what is in fact in question" (Ranci ère 1999: 44). In any dia-
logue that can properly be called "political" in Rancière's sense, the very object of dis-
pute and the parties to the dispute must themselves be in question (Rancière 1999: 55).
Subjectivity itself, Rancière often emphasizes, must be polemicized (Arditi and Valentine
1999; Rancière 2003a). Habermas simply takes the subjects and objects of dispute as
given; his theory can therefore operate only within the logic of the police that has already
presupposed what politics places in doubt - that which politics contests. The scene of
politics may always intervene, as a rather unwanted guest, in Habermas's theory of dis-
course ethics, such that the sphere of communicative action that Habermas wants so
dearly to protect from the realm of strategie action will always find itself rent through and
through by the irruption of politics (see Deranty 2003b).
1O. While praising Butler's understanding of Foucault's meaning of assujettisse-
ment, Milchman and Rosenberg criticize Butler for "her translation of assujettissement as
'subjection'" (Milchman and Rosenberg 200?: 63). 1 completely concur with their criti-
cisms of subjection as a rendering of Foucault's early French term. However, in this case-
and unlike the case of Robert Hurley's translation (Foucault 1978) the criticisms prove
misplaced; Butler simply does not translate assujettissement as "subjection." In her intro-
duction, just two pages before the location from which Milchman and Rosenberg draw
their lengthy quotations, Butler writes: "'subjectivation' Ca translation of the French
assujetissement [sic])" (Butler I997b: II). She recurs to this usage at the end of ter

Notes 1
introduction mentions it again in her next chapter and then she uses it repeat-
edly and explicitly (more than a dozen times) in her discussion of Foucault (84'-149).
can only assume that Milchman and Rosenberg have mistaken Butler's broader project
in the book, which involves what she calls "theories of subjection, with her specifie
rendering of Foucault's assujettissement. As mention in the text, her term of choice
does, in fact, prove problematic, but not for the reasons that Milchman and Rosenberg
in their otherwise invaluable study.
II. One other wrinkle worth noting: of the more than a dozen French dictionaries
(hat 1 have consulted, none contains an entry for "subjectivation," leading me to conclude
that the word is most likely a coinage (by Foucault or someone else) derived from the
French words subjectif and/or subjection. Milchman and Rosenberg make no explicit
argument for how to translate the French subjectivation into English, but they implicitly
support the case 1 make in the text above by frequently using the English "subjectivation"
as a direct translation of the French word (Milchman and Rosenberg 2007: 56).
I2. 1 made this mistake in an early article on Rancière (Chambers 2005).
r3. Sorne commentators/translators in English have muddled what in Rancière's
French might otherwise be clear. Thus, in the widely consulted "Glossary of Technical
Terms" that appears as an appendix to The PoUtics of Aesthetics, Rockhill translates
subjectivation as "subjectivization," listing "subjectification" and "subjectivation" as
alternate translations (Rockhill2006: 92). This move cornes close to collapsing a number
of potentially important distinctions, especially when one considers the connotations
that sorne of those terms have vis-à-vis Foucault's work. In recently released secondary
works on Rancière by Davis (2010) and Tanke (2011), there is an apparent consensus,
which 1 support, for the English subjectivation. However, neither of these authors says
anything about the context, and neither mentions the fact that Rancière's subjectivation
has been translated many other ways (even, in an important sense, by Rancière himself).
Moreover, Tanke's discussion of subjectivation explicitly turns to Foucault, but Tanke
fails to mention that while Rancière uses but the one French term in his publications,
Foucault actually introduces subjectivation in order to mark a distance from his earlier
assujettissement.
14. This line of argument from Rancière's famous 1995 book Disagreement con-
nects quite clearly, if indirectly, to his earliest work at Les révoltes logiques: "There is no
voice of the people. There are fragmented, polemical, voices which split the identity they
put forward every time they speak" (quoted in Davis 2010: 41).
15. Rancière has articulated a number of important and powerful criticisms of Lyo-
tard's worle. In particular, he challenges Lyotard's opposition between the "grand narra-
tives" of modernity and the proliferation of language games in postmodernity. Perhaps
most significantly, Rancière opposes Lyotard's move toward an aesthetics without eman-
cipation and the construction of a "univers al victim" (Rancière 1995c: 64; see also Ran-
cière 1999: 39; cf. Rancière 201 IC). This means that Rancière targets both the earlier
(I984) and later (I990) Lyotard texts and does not focus on the work that is central to my
reading here (Lyotard I988). For a more detailed comparison of différend and mésen-
tente, see Déotte (2004).
16. As Rancière indicates in a footnote to the now published version of "The Think-
ing of Dissensus," 1 gave a paper in 2003 that did, in fact, try to demonstrate, at turns,
either the ontological proximity between Lyotard and Rancière on language, or the extent

1 The Lessons of Rancière


to which Rancière's account relied too on Hs Aristotelian (and anthn)pc)ce:ntrlc
starting point. As to the former point, throughout this book try to follow Rancière in his
eschewal of ontology; thus make no effort to connect Rancière and Lyotard by way of a
theory of language, since for these purposes take Rancière at his word that he has no
such theory of language. As to the latter point, as say in the text here, Rancière's "begin-
ning" with Aristotle is always a beginning in disagreement. His tum to the Aristotelian
logos is always a tum to a tainted logos. criticisms from almost a decade ago simply
will not hold. With the aid of hindsight, however, might cite the utterly confused and
misleading presentation of Rancière's initial reading of Aristotle (in the English transla-
tion of Disagreement) as mitigating circumstances.
17. This thesis could safely be titled the "heterogeneity of language games," but,
of course, Lyotard do es not stop here. He certainly do es not draw from this starting
point the idea of "the incommunicability of languages, an impossibility of understand-
ing" or the "inescapability" of language games, as Rancière might have it (1999: 50;
2003a: par. 20).
18. In the section titled "Reader," Lyotard points out that one may choose to use this
reading dossier to "'talk about the book' without having read it" (xii). At the risk of pro-
testing too much, 1 feel compelled to emphasize that 1 wish to use the "thesis" and "stakes"
headings of the reading dossier only to sharpen the contrasts with the cursory reading that
Rancière offers, not merely to offer my own cursory reading.
19. Lyotard himself does not, of course, make the same set of fine-grained distinc-
tions that 1 have made between anthropocentrism, humanism, and anthropomorphism. In
fact, in his own polemical formulations, Lyotard tends to lump aIl these terms together so
as to build up a large target for his critique: he refers to all three terms, somewhat inter-
changeably. 1 contend, nonetheless, that the force of Lyotard's argument appHes specifi-
caIly to anthropocentrism as 1 have described it in the text above. Moreover, the insights
of Lyotard's argument can best be drawn out by reading his work as a description of, and
challenge to, anthropocentrism. Thus, 1 occasionally "translate" Lyotard's 100ser use of
terms as 1 try to draw a more careful set of distinctions, but these translations do not
change the overall force or contour of Lyotard's arguments.
20. The method 1 follow here, of locating the concept prior to its naming, is not at aIl
one 1 wou Id commit to for all thinkers, or even for all concepts developed by this particu-
lar thinker, Rancière. 1 heed Quentin Skinner's waming about the myth of prolepsis
(Skinner 1988). However, in this case it seems to me that the evidence lies on the side of
the notion that the idea clearly preceded the coining of the term, as Rancière talks about
something that looks very much like "literarity" long before he used that word.
2 I. Rancière, like Derrida, wishes to contest Plato's claim in the Phaedrus that writ·
ing is inferior to speech. Despite his own deconstructive moves with Saussure, Derrida,
however, insists on maintaining Saussure's distinction between langue and parole with-
out tuming that distinction into a hierarchy (Derrida 1982). Rancière's claim against
Plato that "there are not two kinds of discourses" would appear to collapse that distinc-
tion, but 1 am not so certain one should draw su ch a strong conclusion from such a smaIl
piece of the text.
22. In a much later essay titled "The Politics of Literature," Rancière looks back at
his own texts to show the early emergence of this idea of "literariness" and to implicitly
clarify its relation to "literarity" (Rancière 2004c).

Notes 1
23. Rancière will go on to argue that literarity forms the centrallink between poUtics
and literature (2004a: I08).
24. have tried to be very clear in this discussion to distinguish between, on the one
hand, when the idea of literarity emerges and takes shape in Rancière's writings, and, on
the other, when the specific term itself first appears. To summarize, the term literariness
appears earliest; it is then followed by a very particular use of the word literary (in the
sense of the political animal as a "literary animal"); and this is followed in tum by the
coinage of the word literarity (at the end of Flesh ofWords). In ber otherwise helpful sum-
mary chapter on Rancière's concept of literarity, Alison Ross runs the risk of confusing
matters greatly when she says directly, "The concept of 'literarity' first appears in the 1995
book - La mésentente" (Ross 20 IO: 136). Ross then goes on to summarize the arguments
in Disagreement that do, 1 agree, help to ftesh out what Rancière means by literarity. How-
ever, her presentation implies that La mésentente actually uses the tenn littérarité. This is
simply not true. Rockhill's glossary entry for "literarity" indicates, in outline, a similar
genealogy as the one 1 work through in the text (Rockhil12006: 87; cf. Rockhill 2006: 5).
25 We probably should not forget -- though Rancière does not bother to remind us -
that Plato himself took the risk of the excess of words by deciding, unlike his teacher
Socrates, to write his philosophy as dialogues. Thus the threat to his kallipolis, to his
entire philosophy of order, is a threat that he himself puts into play. Thanks to Bonnie
Honig for reminding me of this important point.
26. 1 have consistently framed this discussion in tenns of Rancière's critique ofPlato,
but it could also be expressed in terms of his consistent admiration for a writer like Flau-
bert. Rather than denying the excess of words, Flaubert's literary style depends upon em-
bracing a certain literarity: "His very refusaI to entrust literature with any message
whatsoever was considered to be evidence of democratic equality" (Rancière 2006d: 14).
27. Here 1 focus on one piece of Apostolidis's early published research on immigrant
workers (Apostolidis 2005); he has recently published a much larger body of research in
book form (Apostolidis 2010).

Chapter Four
1. In so doing 1 also mean to lay out a Iogic quite distinct from the conservative one
that Berlant and Wamer criticize in their important "What Does Queer Theory Teach Us
about X?" (1995).
2. Read in the context of the partage, we might say that a dispositifis a mechanism
that apprehends a particular distribution of the sens ory realm. Thought in this sense, a
dispositif would not represent reality, but would capture it. This particularly Rancièrean
reading of the dispositif points already toward my conclusion in this chapter: a critical
dispositif beyond inversion. 1 will take these points up again in the final section; at this
juncture 1 wish only to delineate the tenns through which 1 will address a variety of no-
tions of critique.
3. In other words, much work today rightly resists the idea that a theory of the politi-
cal must be a critical theory, and 1 join my colleagues in insisting that not all political
theory needs to be critical in the sense 1 want to articulate here. Creative, imaginative, and
very positive reconsiderations and reenvisionings of political theory prove absolutely cru-
cial to the possibility of a renewed sense of politics - to new political possibilities. For
just two of the numerous recent examples of such work, see Shapiro (2009) and Connolly

190 1 The Lessons of Rancière


(2010). is in)1'''opr"ti"p that be more than but it must also include
UlllHY,''','.

analysis.
4. Bennett also makes a not-so-common argument when she suggests that the critical
logic of inversion "presumes that at the heart of any event or process lies a human agency
that has illicitly been projected into things" (Bennett 2009: xiv). Thus, for Bennett, the
logical inversion performed by the critical theorist operates by way of, and smuggles in,
a certain form of anthropocentrism - because the essence that inversion uncovers is
always ultimately human will and human agency (a human essence). As l discussed in
Chapter Three, Bennett and 1 share a desire to challenge anthropocentrism. More signifi-
cantly, in making this argument Bennett also echoes (though again, in a distinct key) the
Althusserian critique of the humanist reading of Marx; according to Althusser, Marx's
early writings always smuggle in an elicit anthropology that serves as an ethical ground
for critique. Bennett's methodological anti-anthropocentrism and Althusser's method-
ological antihumanism thus share important affinities, in that both authors argue that in
order to understand structures, processes, and (ultimately) politics aright, we must reject
the tendency to trace their origins and causes back ta human sources. This means that the
Marx that Bennett gestures toward in her brief account of demystification (a humanist),
and the Marx that Althusser wants to bring to light (perhaps not an antihumanist, but
certainly a non··humanist) - these are not at all the same Marx. At stake between them is
the possibility of a non-anthropocentric reading of Marx; l do not pursue such a reading
in this chapter, but 1 do try to keep such a possibility alive.
5. To the extent that one could trace the current dominant connotations of "critical
theory" (as Bennett and others use it) back to specifie authors and texts, it seems that the
most prevalent understanding today fits better with Horkheimer's early work - especially
his famous, and widely read within social science circles, essays on ''Traditional and Criti-
cal Theory" and "The End of Reason" (Horkheimer l 972) - than with, for example, that
of Adorno, who was already struggling (as so many others have) to move past the logic of
inversion (Adorno 2005).1 offer this sense of the dominant understanding of critique so as
to provide contrast for my own tum away from this understanding. However, 1 have no
wish to reify the concept of critique that 1 try to resist here. 1 recognize the great diversity
of efforts to articulate a critical theory - and realize that the logic of inversion that 1 chal-
lenge is not the only way of thinking critique in social and political theory today - just as
1 recognize that mine is not the only project designed to rethink critique. For one promi-
nent and important example of rethinking critique, and a source for further overviews of
the contemporary state of criticism and critical theory, see Kompridis (2006).
6. 1 address these three locations in chronological order, but this move has nothing to
do with progressivism (either in Rancière's work or in my account). Rancière's polemics
have a punctual temporality that undermines any sort of linearity. Each piece of writing
1 address contains an engagement by Rancière with a particular conjuncture, and my own
account of these three pieces produces its own encounter with the current conjuncture.
Thus, there is no inevitability about the development of Rancière's thinking; such prog-
ress that there has been was never guaranteed in advance (progressivism); and there is
absolutely no end point or final state to it (teleology). Rancière's "ending," as 1 show, is
surely no more than a beginning to the task of rethinking critical theory.
7. For clarity and simplicity, in this list 1 will give the translated English title coupled
with the original publication date (in French for aU but the last text, which was drafted in

Notes 1 19'
English). The strange timing of the translation of Rancière's work into English has had a
number of interesting consequences for the reception and understanding of Rancière's
writings by an English-speaking audience. For more on this matter see the Introduction
(see also Chambers 20roa).
8. In addition to many other scattered references to Marx throughout his writings,
in my move from point l to point 2, l skip over entirely Rancière's critique of Marxism
in The Philosopher and His Poor (1983). This move probably demands a few words of
explanation. As 1 read his discussion of Marx in that text, the key issue for Rancière is the
concept of the proletariat as the "pure proletariat," kept in its place within the social order.
Rancière reads Marxism as an inverted form of Platonism: where Plato kept the shoe-
maker in his place so as not to disrupt the Platonic order, so Marxism keeps the artisan in
his place so as to make a revolution possible. In laying out this argument, Rancière does
quote Marx (Rancière 2004b: 80), but he offers a very selective and highly edited selec-
tion from Marx's text; 1 would interpret the full passage much differently. In any case, the
book does not really set out to offer a reading of Marx, but rather to show that both Pla-
tonism and Marxism can be seen to participate in the same logic of the philosopher. This
is an argument about Platonism and about Marxism (especially a certain Marxist sociol-
ogy), about a philosophy of order, and about sociological accounts of class, but it is not
reall y a reading of Marx.
9. Although it falls weIl beyond the scope of either this chapter or this book, 1 also
hint at the idea that such an answer might lie in a different understanding of Marx than
that provided either by traditional Marxism or by Rancière's own critique of Marx: one in
which Marx himself rejects the classical opposition between essence and appearance, one
in which Marx's critique of capitalism operates not by inversion (or demystification) but
according to a distinct logic of denaturalization. Such a reading would have to ask, in
Rancièrean fashion: What if Marx himself made the very assumption that Rancière says,
with genuine praise, is "foolish"? What if Marx privileged the proletariat not because (as
so many Marxists after Marx have often suggested) they had he Id sorne special knowl-
edge, not because they occupied a unique epistemological space, but sim ply because they
were less disabled than they were taken to be?
10. Crucially, then, such texts are not called "scientific" because they are empirical
in nature or because they operate by way of empiricist epistemology. Althusser uses the
word science in the way that his contemporary French historians of science used it (he is
concerned with scientific revolutions in the Kuhnian sense of paradigm changes) and not
in the way that twentieth-century North American social scientists used it. Empiricism is
always Althusser's ultimate opponent.
II. As a corollary, this means that when used rigorously, the antihumanist methodol-
ogy might enable an anti-anthropological reading of the so-called early works. Thus,
Althusserianism breaks with the hermeneutics of antihumanism in order to denigrate the
early works (Althusser and Balibar 2009 [1965]).
12. One finds other places in the text of the essay where Rancière reads Marx at a
very close textuallevel, and it is actually here - deep inside the text, as it were - that one
locates Rancièrean insights that often break out of and exceed the overall Althusserian
framework. As one example 1 would point to the powerful illumination of what Rancière
calls the "impossible equation" in Marx, and in capitalism, of "x commodity A = y com-
modity BOO (Rancière I989a: I06-ro).

192 1 The Lessons of Rancière


13. It is standard to reference the English translation of as
a "co-authored" work by Althusser and Balibar. However, as 1 have shown elsewhere, the
book is certainly not a work by two authors: it was original1y a collaborative project,
produced jointly by five different authors; it was later transformed, through editorial deci-
si ons and reception history, into something like a single-authored work by Althusser, with
Balibar's name curiously attached to it (Chambers 20I 1). Thus, contrary to the sense
given by the proper parenthetical cites to "Althusser and Balibar," all of my references
here are strictly to Althusser's contributions to the work.
14. This sort of challenge emerges almost immediately in Rancière's writings in his
unique, self-critical preface to his own essay. 1 discuss this preface not here, but rather,
out of chronological order, when 1 show in a later section of the text below that the pref-
ace contains important traces of Rancière's later critique of inversion.
15. Much of that literature has also been discussed in previous chapters, but to col-
late a few key references here: Arditi and Valentine 1999; Honig 2001; Deranty 2oo3a;
Deranty 2003b; Panagia 2001; Panagia 2006; May 2008; Chambers 2005; Chambers
20rob; Davis 20lO; Tanke 2011.
16. "Wrong" is both a key term in Rancière's account of politics, and the title of the
second chapter of Disagreement. 1 first pointed to the importance of this concept in my
Introduction, and throughout the book 1 have engaged with many of the ideas and argu-
ments that surround and link up with this term. However, none of my chapters address the
concept directly. 1 would define wrong as that which is manifested when the logic of
equality disrupts the logic of domination. Rancière puts it this way: "Wrong is simply the
mode of subjectivation in which the assertion of equality takes its political shape." Thus,
a "wrong" emerges not because injustice has been done to a minority identity; it has noth-
ing whatsoever to do with victimization. Rather, the conflict between equality and domi-
nation manifests itself in the form of "wrong," which is a form of subjectivation as
disidentification (Rancière I999: 35, 39)·
17. In this context it is important to note that Disagreement proves to be the least
polemical of Rancière's major works. Compared to the bulk of his other writings, Dis-
agreement reads much more like an epistemological or ontological work (hence the ten-
dency for sorne interpreters to take Disagreement as proposing a political ontology,
des pite Rancière's clear insistence that he rejects aIl ontology). Indeed, as an anonymous
reviewer suggested to me, Disagreement reads "almost as if [Rancière] is consciously
writing for an audience that he knows will not have access to the resonances of his polem-
ics (we are dealing here with someone who is highly attuned to affect and sensibility)."
Regardless of Rancière's intentions, it is clear that Disagreement was significantly better-
received by an Anglo-analytic audience already enmeshed in debates over deliberation
and traditional democratic theory. My deepest thanks go to the anonymous reviewer who
both prompted me and (in the quoted words above) helped me to make this point.
18. It is not completely clear whether Rancière means to distinguish Socrates from
Plato or to conflate them. Regardless of his intentions, 1 wish to mark a separation be-
tween the philosophizing activity of Socrates and the philosophical project of Plato.
Rancière's critique of political philosophy appHes with force to Plato, since the reorder-
ing project that Plato takes up cornes only after the dis-order of the dëmos has led to the
death of Socrates. But it would be a mistake, 1 contend, to project the philosophical
project of Plato back on to Socrates. Socrates, that is, may weIl have been quite content

Notes 1 193
to philosophize post hoc, and perhaps even to understand philosophy as a project com-
patible with the an-archic politics of the dêmas. Socratic elenchus may be much more
untimely than Platonic political philosophy.
19. This point powerfully separates Rancière's thinking of democracy from that of
Chantal Mouffe, thereby illuminating a contrast l drew in the Introduction. Mouffe de-
fines democracy as a form of rule, but for her this means specificaIly "the principle of the
sovereignty of the people" (Mouffe 2000: 2). From the perspective l have developed here,
such a move amounts to an inept and untenable attempt to transform the kratas of democ-
racy into an archê (Rancière 2006c: 76).
20. While Rancière takes perhaps the most radical and polemical approach to the
tradition of political thought imaginable, in his own way he agrees with the most conser-
vative defenders of that tradition: Plato is the most important figure in it.
2 I. Rancière goes on to say more about metapolitics, and thereby about critical
theory as inversion. It is not simply that the social is the truth underlying and masked
by its distortion: the logical structure of a metapolitical critique entails that the social
cornes to function not just as truth, but as the untruth of the political. Rancière shows
how the concept of class, in Marx and after, oscillates: sometimes functioning just like
an archipolitical truth, yet often operating according to a "nihilism of the falseness of
aIl things" (Rancière I999: 84). The metapolitical inversion can be understood as either
a revelation of the truth that underlies falsity, or merely the revelation of falsity as aIl
that there is.
22. Alain Badiou calls this Rancière's "median" style (Badiou 2009: 40). Perhaps by
"median" he means to distinguish Rancière from the "zero degree style" that Lyotard
takes up, as l discussed in Chapter Three (Lyotard I988: xiv). In any case, and to be clear,
my point is not that Rancière's style allows him to smuggle in a so-called normative posi-
tion, but that the style makes it harder to discern the logic of his critical project. l reject
Habermas's account of Foucault as a "cryptonormativist" because l think Foucault's work
undoes the normative/non-normative distinction on which Habermas bases his argument.
1 would refuse the application of this label to Rancière on similar grounds (Habermas
I990: 284)·
23. This interpretation can be bolstered by reading further into Rancière's text: in
chapter five Rancière turns to an ev en more explicit critique, this time of "consensus de-
mocracy." Here Rancière quite clearly attempts to demonstrate that what appears to be the
ultimate political solution (third way social democracy) turns out, in reality, not to be
politics at aIl. Consensus politics only appears political on the surface; in truth, it is noth-
ing but a new order of police. l have engaged more fully with these critiques in Chapter
One and Chapter Two.
24. Along this line of thinking, one could also argue that there are different types
of inversion within the complex history of critical theory, but for purposes of clarity in
my reading of Rancière here, 1 will maintain the distinction between inversion and
reversaI.
25. In 2007 and 2008, Rancière gave a number of talks at various universities in the
USA, aU of which shared the title "The Misadventures of Critical Thinking"; at the time
of this writing, sorne video recordings of these talks are available on YouTube. The Dart-
mouth undergraduate philosophy journal, Aparia, published a transcript of the lecture he
gave there (Rancière 2007b). Rancière also shared written drafts of the talk with students,

194 1 The Lessons of Rancière


scholars, and other audience members, and some of those have now been pvtpn!;'1u,"'I"
circulated. In 2009 the essay was finally published, as chapter two of The Emancipated
Spectator (Rancière 2oo9a). Given some of the significant changes in language from the
earlier draft to the published version, my citations here are mostly to a draft circulated at
the University of California at Irvine in late 2007 (Ranci ère 2007c).
26. Unlike the Italian and English translations, this second French edition did in-
clude aH the original contributions. However, this edition changed the order so as to
mirror the Italian and English editions; thus, Althusser's and Balibar's essays are at the
front, with Rancière's essay pu shed to the middle. This means that even in the French
version a certain sense of the book as primarily "Althusser's" was retained from the trans-
lated editions. In 1997 a posthumous edition of Althusser's works led to the creation of a
French edition that finally restored the original nmning order (Elliott 1998).
27. Here readers can see clearly Rancière's own identification of the optical meta-
phors that drive Marx's early version of critique, and to a certain extent, Rancière's
own.
28. In the published (2009a) version of the essay, the French dispositif has been
variously translated as "procedure" or "paradigm." 1 pre fer the earlier version in which
Rancière, writing in English, kept the French term in place. 1 do so not only for con-
sistency with my own argument here, but also because the regular use of the term
helps to demonstrate one of Rancière's central points in this particular essay: namely,
that the critical dispositif runs through both earlier accounts of critical thinking and
later denunciations of it. The same "critical dispositif' connects the oider "criticai
theory" formulations with the later "postmodern" putative turns away from those for-
mulations. IncidentaUy, in the new and carefully thought out translation of La leçon
d'Althusser, Rancière's new English preface also leaves dispositifuntranslated (Ran-
cière 201 la: xvi).
29. Here 1 cite the published version of Rancière's essay, as it contains a proper cita-
tion to Sloterdijk.
30. While at first glance they might sound somewhat similar, there is a crucial dif-
ference between this "reading in a reverse way" that Rancière criticizes as a conserva-
tive move, and the notion of literarity - an "excess of words" that 1 discussed at length
in Chapter Three. On my reading, Rancière's critique of Debord (and others) pivots on
showing that the logic of inversion has not been challenged: there is no force outside
the inverted world of appearance and essence, only an hourglass being repeatedly
turned over. In contrast, the democratic power of literarity caUs on the force of an
excess of words that can disrupt a given distribution of the sensory realm, and can
thereby repartition it. The postmodern turn, Rancière shows in this case, do es nothing
more than twist within the same partition/distribution ofthe sensible. Rancière's argu-
ments concerning literarity and le partage du sensible might thus help us to reconsider
long-standing debates in contemporary theory over the politics of resignification, be-
cause resignification can be understood as a move that calls on the excess of words so
as to pro duce a new configuration of the sensory realm (see Butler 1997a; Chambers
and Carver 2008).
31. Thanks to an anonymous reviewer for help in specifying Rancière's unique
thinking of emancipation and thereby allowing me to develop these crucial linkages be-
tween his radical pedagogy and lùs reading of critical tlùnking.

Notes 1 195
32. This move in the project takes it in a direction distinct from other contemporary
efforts to rethink "ideology critique" (see Laclau 2006; cf. Laclau 1996; 1994;
Lievens 2012).
33. This reading of hope resonates with that of Cornel West's vision of a
politics of hope that would actually "be betrayed by naive projections of a better future
that ignore the necessity of doing the real work" (West 2011).
34. Todd May has suggested that one can eschew the assumption of blind agents -
that one can affirm the Rancièrean equality of intelligence - without abandoning the idea
that sometimes such agents are "mistaken." In this way the (critical) theorist attempts to
show people something they did not see before, without presuming that they were blind
and now have sight. To me, the distinction makes intuitive sense and does not obviously
contradict anything that Rancière says about the abled and the disabled. However, l am
leery of the need to hold on to "mistaken" agents, as it seems to flirt with the danger of
sneaking in an "explicative order" through the back door. Surely students are sometimes
mistaken, but as 1 read Rancière, to make the assumption of equality is to require a radical
pedagogy in which teachers never presume mistaken students (May 20 11). Indeed, in
Rancière's radical pedagogy, teachers need know nothing. (This may weIl make Socrates
the first ignorant schoolmaster, and Plato the first stultifier.)
35. Deleuze's account of the "new" in relation to dispositifs helps shed light on what
remains critical in the critical dispositif that 1 am outlining here. Deleuze says, "We
belong to social apparatuses and act within them" (Deleuze 1991: 164), a description that
again makes dispositif resonate with le partage du sensible. To take up a critical orienta-
tion to the police order in which we find ourselves does not, therefore, require a logic of
inversion, because the critical dimension need not concern fundamental truths masked by
lies, but rather can hinge on the dimension of newness, of becoming, of change that trans-
forms and ruptures the police order.
36. It goes almost without saying that the police order can co-opt any polities, not
just one of "sense-making." 1 am showing here, however, that the polities of "sense-
making" has already adapted itself to an order of the police in which an can be known, aIl
can be seen, aIl can be understood. For pressing me on Rancière's rejection of under-
standing and sense-making, and for reminding me of the richer meaning of la mésentente,
1 again thank an anonymous reviewer.
37. David Couzens Hoy provides an excellent example to illustrate my pejorative
description in the text: Hoy affirms just the account that 1 here resist. He reads genealogy
as a project of critical theory committed to the repeated assertion of historical contin-
gency. His persuasively written and widely read account of critical the ory thus serves as
strong support for my argument that today the tradition of critieal theory and the broad
idea of a critical dispositif is often conflated with a generalized conception of genealogy
as a project that renders historically contingent anything that would appear ahistorical,
universal, or natural (Ho Y and McCarthy I994, especiaIly chapter 6). In tlùs context it is
helpful to cite Gabriel Rockhill's introduction to the recently translated Mute Speech,
where he argues that Rancière's own contribution to critical thinking "only approxi-
mates . .. radical historicism" (Rockhill 20 II: 5).
38. As a critique of capitalism, Marx (or better, Marxism) is often understood to
assert that capitalism alienates the worker by separating him from the object of his labor
(and from labor itself, and from his fellow laborers, and from his species being). Capitalism

1 The Lessons of Rancière


must be rejected because of alienation. There is a violence or injury done
a violence that makes capitalism wrong, and contained in the critique is a vision of what
wou Id be right. AlI of this belongs to the traditional, yet stultifying version of critical
theory as inversion. Nonetheless, a new critical dispositif need not abandon In-
stead, it would need to reread Marx in the way suggest here. should also note that while
the "behind the backs" line appears literally in Capital (in three places), use the phrase
here to refer to the general idea that has been taken up by certain Marxists, that is, as a
metaphoric stand-in for the notion that there are secret and invisible mechanisms within
capitalism. Therefore, 1 contend that Capital itself can be read against this idea that has
been generalized out of a few Hnes within Capital. One can also show that Marx's own
use of the lines "behind the backs" is not meant to convey secret or invisible mechanisms
but simply to indicate that various conditions of labor and the laboring process are pro-
duced through - as Marx says the first time he uses the phrase "a social process" that no
one individual or group of individuals has pre-planned or chosen (Marx I977: I35; see
also 20I, 485).

Afterword
1. Thanks go to Nathan Gies for helping me to think the connection between After-
word and miscount.
2. 1 use the slogan as a prompt for pulling together sorne of the strands of the book,
and thus 1 offer my own reading of these eight words. 1 am by no means the first to do so.
Düttmann (1997) provides a detailed and lengthy exegesis in the context of theorizing the
politics of recognition, while Arditi and Valentine (1999) draw from Düttmann's own
work to help them render their crucial distinction between polemic and polemicization.
Both texts offer rich accounts from which 1 see no need to take a critical distance. None-
theless, my emphasis remains distinct; Arditi and Valentine interpret Düttmann (rightly, 1
think) as arguing that "on the surface, this is a demand for recognition" (Arditi and Val-
entine 1999: 1). But 1 would contend that even "on the surface" the Queer Nation chant
resists this reading. Düttmann explores a certain negative dialectics of recognition through
his extended meditations on the chant, but on my account the slogan itself refuses the dia-
lectical game of recognition in the first place. To read it otherwise is to run the risk of
reducing it to the standard formula of minority identity politics, as 1 discuss in detail in
the text below.
3. The dismissal can operate on a more general level, as weIl, particularly since it
proves easy to argue, and even simpler just to presume, that the realms of academic
theory and direct action on the street remain irreconcilably distant. It is child's play to
offer proof for this sort of daim: essays and arguments published in journals and books
are self-evidently not the same as the passing of legislation, the signing of laws, the issu-
ing of executive orders, or the judgment of judges. Of course, this sort of argument only
shows how potentially significant Rancière's thinking of politics can be; it does so be-
cause this very list of activities is not at aU what Rancière means by politics. Moreover,
what 1 am tempted to caU the "academic politicallament" the enunciation of the worry
that writing and argument never change anything - always seems to constrict the realm of
the political and to severely constrain the concept of change. After aU, who wants to cede
poli tics to politicians? If they are our only hope, there is not much hope at aIl. Moreover,
are not the best politicians the ones who refuse just this very move - the ones who think

Notes 1
that does have something to do with the And if professional politicians
can refuse a radical theory/practice dichotomy, perhaps so-called theorists can resist it as
weIl.
4. Matt (2008) provides an interesting, hands-on illustration of the tactical
uses of this slogan. The strategic arguments about using the chant provide good evidence
for my reading of its politics in the form of demands for equal rights and inclusion.
5. With these claims 1 emphasize the original appearance of the book in 1990 and
attempt to reconstruct a context that is easily lost sight of given the impact of the book.
That is, the effects of the book make it easy to rewrite its origin story. Nonetheless, in
1990 queer theory did not exist as a field and Butler's text was not written as a part of
lesbian and gay studies. Sedgwick's Epistemology of the Closet (1990) was published
earlier that year, but Butler's book was already in press. Moreover, Sedgwick's book was
not itself an intentional contribution to "queer theory." Theil' books - along with Halp-
erin's text from the same year, and others from much earlier (e.g., Foucault 1978) -- were
only retroactively claimed as "canonical texts" in the coming years, when works by de
Lauretis (1991), Warner (I993), and others helped (still mostly unintentionally) to form
a field around the name "queer theory." To say all this is not to deny or downplay the
importance of Gender Trouble in helping to radically reconceive gender and sexual iden-
tity, in helping to bring into existence, and then serve as a founding text for, the very
field of queer theory. It is merely to emphasize, as Butler herself do es in the I999 pref-
ace, that the book was not written with these goals in mind, and the context in which we
now situate it proves very different from the one in which it was produced (Butler
I999)·
6. Given the increasing confusion surrounding these terms today, it may be helpful
if 1 clarify my limited use of the language of "waves" of feminism. 1 refer to the "second
wave" to describe the women's movements and feminist theories of the I960s and I970s.
In the second wave, political struggles for the right to work, the right to fair pay and equal
treatment, reproductive rights, protection from rape and domestic abuse, challenges to
sexual harassment and many other goals, were loosely associated with daily practices
such as consciousness-raising and were variously linked with theoretical work concem-
ing the unique experiences of women, the epistemological primacy of women, or the
particular values and ethos of women. Nothing about the second wave has come to an
end; hence the metaphor of waves. 1 associate thinkers like Dworkin, MacKinnon, and
Harding with the second wave. In the 1980s and 1990s, a third wave of feminism took
shape and emerged in its own form, out of elements that were already a part of the second
wave. Like the second, the third wave proves very diverse, but it loosely congeals around
certain critiques of the ide a of a unified "experience" of woman, of the epistemological
primacy of women, or of the notion that politics can begin only after the subject of woman
has been established. Early lesbian feminism, Black feminist thought, pro-sex feminism,
and Butler's work aH have affinities in this sense, and many of these ideas can be traced
back to the 1970S (see, for example, Lorde 1984 [1979]). For this reason, referring to the
"waves" by time periods is always a very rough approximation: what matters most in
distinguishing the waves are the ideas and arguments, which historically are themselves
mixed up (like water in the ocean). More significantly, 1 do not use the label "third wave"
to define generational differences, and 1 entirely reject any linking of third wave feminism
with "post-feminism" (this last move is what forces some feminists to try to identify

1 The Lessons of Rancière


thinkers like Butler as "second wave feminists" so can hang on to
in their thought). In a later text, Butler describes herself as "a latecomer to the second
wave, and perhaps she feels compelled to make such a remark so as to interrupt the faIse
eqllation between so-called post-feminism and the third wave (Butler 2004:
7. In previous work 1 have made the full case for reading Butler's concept of the
"heterosexual matrix" through the later-developed language of heteronormativity. ln
Gender Trouble Butler shows that it is the heterosexual matrix that binds together sex,
gender, and sexual desire, and she articulates the workings of this matrix through the
complementary concepts of "regulatory practices" and "gender intelligibility." argue
that these concepts can be fteshed out with the language of heteronormativity, as coined
by Warner (1993). See Chambers 2007; cf. Chambers and Carver 2008.
8. The insistence on resolving the paradox always diminishes any effort to think
beyond inclusion; such insistence reduces the project to a politics of inclusion. Butler and
Rancière converge at exactly the paradoxical formulation of politics thal animates Spiv-
ak's famous, and broadly misread, essay "Can the Subaltern Speak. 7" (Spivak. 1988).
Spivak. was formulating a critique of the politics of inclusion (within the context of post-
colonialism), but her question-and-answer title led to a misreading of her work within the
tenns of a politics of inclusion. Butler and Rancière formulate their own paradoxes. For
Rancière, we might say that the paradox amounts to "counting" those who do not count;
for Butler, we might say it involves the production of a politics of intelligibility that re-
fuses the possibility that all can be intelligible. 1 discuss the substance of these two points
in greater detail in the text below. This sense of paradoxical poli tics is deftly expressed in
one of the more famous May '68 slogans: "Be realistic - demand the impossible." By
reading the slogan as capturing a paradoxical sense of politics, 1 resist the tendency to
turn "the impossible" into the utopian. For a challenge to this utopian idea of "the impos-
sible" - that is, a resistance to the idea of "demanding the impossible" as demanding
something radically other - see Paul Bowman's rich and powerful critique (Bowman
2008a: 98).
9. In her more recent writings Butler appears to have turned away from her earlier,
more totalizing rejections of ontology. Indeed, she seems to have embraced a certain on-
tology of human vulnerability, a turn that may result in a new humanism (see Honig
20IO).
IO. Rancière's early essay on subjectivation (1995c) was first published in a collec-
tion called The Identity in Question.
IL In making this claim 1 am of course applying my own torsionalforce to Rancière,
and throughout my reading here, any ostensible equations that 1 suggest between Ran-
cière and queer theOl'y are themselves polemical equations. In other words, my goal is not
to reduce Rancière to queer theory (or vice versa) but to pluralize and contest both
terrains.
12. Queer resists not just heteronormativity but al! normativity, and in many con-
texts this means that queer stands opposed to homonormativity as well perhaps even
more so.
13. Butler has her own version of "miscount" in her thinking of unintelligibility, a
thinking that refuses the notion that "the unintelligible" can sim ply be made intelligible.
1 refer to just this sense of unintelligibility in Chapter Four, where 1 insist that no partage
can eradicate the unintelligible. The partage produces a new order of intelligibility, but it

Notes 1 199
does not simply or directly make visible what before had been invisible before.
the invisible visible is the project of tradition al critical thinking and of linear politics.
14. In many cases the "struggle against oppression" will best be waged within the
terms of the police order - that is, by improving those tenus and that order. To say this, as
have stressed through the book, i8 not to denigrate such work, but merely to retain a
clear sense of the difference between such work, on the one hand, and the action of demo-
cratic polities, on the other.

200 1 The Lessons of Rancière


Adorno, Theodor (2005). Minima Moralia: Refiections on a Damaged Life. Translated by
Dennis Redmond. Available from: http://www.marxists.org/reference/archive/
adorno!I95 1/mm/index.htm (Accessed: July 7, 2011).
Althusser, Louis (1965). Lire le Capital. Paris: Maspero.
Althusser, Louis (1969). For Marx. London and New York: Verso.
Althusser, Louis (197I). "Ideology and Ideological State Apparatuses (Notes toward an
Investigation)," in Lenin and Philosophy, and Other Essays. New York and London:
Monthly Review Press; 127-86.
Althusser, Louis (1976). Essays in Self-Criticism. Translated by Grahame Locke. London:
New Left Books.
Althusser, Louis (2011). "Student Problems." Translated by Dick Bateman. Radical
Philosophy, 170: 11-15·
Althusser, Louis and Étienne Balibar (2009). Reading Capital. Translated by Ben
Brewster. London and New York: Verso.
Apostolidis, Paul (2005). "Hegemony and Hamburger: Migration Narratives and
Democratie Unionism among Mexican Meatpackers in the US West." Political
Research Quarterly, 58-4: 647-58,
Apostolidis, Paul (2010). Breaks in the Chain: What Immigrant Workers Can Teach
America about Democracy. Minneapolis: University of Minnesota Press.
Arditi, Benjamin (2007). Politics on the Edges of Liberalism: Difference, Populism,
Revolution, Agitation. Edinburgh: Edinburgh University Press.
Arditi, Benjamin (20rr). "Fidelity to Disagreement: Jacques Rancière and Politics."
Paper given at the conference Jacques Rancière: Politics andAesthetics, Northwestern
University, April 22.
Arditi, Benjamin and Jeremy Valentine (1999). Polemicization: The Contingency of the
Commonplace. Edinburgh: Edinburgh University Press.
Arendt, Hannah (1958). The Human Condition. Chicago: University of Chicago Press.
Arendt, Hannah (2004). The Origins ofTotalitarianism. New York: Schocken Books.
Aristotle (1944). Politics. Translated by Harris Rackham. Cambridge, MA: Harvard
University Press.
Aristotle (1957). Politiea. Edited by W. Ross, Oxford: Clarendon Press.
Aristotle (1958). Polities. Translated by Ernest Barker. New York and Oxford: Oxford
University Press.
Aristotle (1981). The Polities. Translated by T. Sinclair. London and New York:
Penguin Books.
Aristotle (r996). The PoUties and the Constitution of Athens. Translated by Stephen
Everson. Cambridge: Cambridge University Press.
Badiou, Alain (2005). Metapolities. Translated by Jason Barker. London and New York:
Verso.
Badiou, Alain (2009). "The Lessons of Jacques Rancière: Knowledge and Power after the
Storm," in Philip Watts and Gabriel Rockhill (eds), Jacques Rancière: History,
Polities, Aesthetics. Durham: Duke University Press; 30-54.
BBC (20ro). "Egyptian Policemen Charged over Khaled Said Death" (Online). Available
from: http://www.bbc.co.uk/news/ro47672o (Accessed: July 7,201 r).
Beltrân, Cristina (2009). "Going Public: Hannah Arendt, Immigrant Action, and the
Space of Appearance." PoUtical Theory, 37-5: 595-622.
Bennett, Jane (2009). Vibrant Matter: A Political Ecology of Things. Durham: Duke
University Press.
Berlant, Lauren and Michael Warner (r995). "What Does Queer Theory Teach Us about
X?" PMLA, l ro·3: 343-49.
Biesta, Gert (20ro). "A New Logic of Emancipation: The Methodology of Jacques
Rancière." Educational TheOly, 60. 1: 39-59.
Bowman, Paul (forthcoming). "Sick Man of Trans-Asia: Bruce Lee and Queer Cultural
Translation," in Sudeep Dasgupta and Mireille D. Rosselo (eds), What's Queer about
Europe? New York: Fordham University Press.
Bowman, Paul (2007). "This Disagreement is Not One: The Populisms of L ac1 au ,
Rancière, and Arditi." Social Semiotics, 17·4: 539-45.
Bowman, Paul (2008a). Deconstructing Popular Culture. New York: Palgrave.
Bowman, Paul (2008b). "Alterdisciplinarity." Culture, TheOly and Critique. 49.r:
93-1 ro.
Bowman, Paul and Richard Stamp (2009). "Jacques Rancière: In Disagreement." Parallax,
r5.3: I-2.
Bowman, Paul and Richard Stamp, eds (20I r). Reading Rancière. London and New
York: Continuum.
Brown, Nathan (201 r). "Althusser's Lesson, Rancière's Error and the Real Movement of
History." Radical Philosophy, 170: 16-24.
Brown, Wendy (2005). Edgework: Critical Essays on Knowledge and PoUtics. Princeton:
Princeton University Press.
Brown, Wendy (2006). "American Nightmare: Neoliberalism, Neoconservatism, and
De-Democratization." Political Theory, 34.6: 690-7I4.
Butler, Judith (I997a). Excitable Speech: A Politics of the Performative. London and New
York: Routledge.
Butler, Judith (1997b). The Psychic Life of Power. Stanford: Stanford University Press.
Butler, Judith (1999). Gender Trouble. London and New York: Routledge.
Butler, Judith (2004). Undoing Gender. London and New York: Routledge.
Butler, Judith (2009). "Critique, Dissent, Disciplinarity." CrUicallnquiry, 35-4: 773-97.

202 1 The Lessons of Rancière


Tenell (2004). in Political Manchester and New York: Manchester
University Press.
Carver, TerreIl (2006). "Rhetoric and Fantasy Revisited: to Zerilli's
'Philosophy's Gaudy Dress.' " European Journal of Political
Chambers, Samuel (2003). Untimely Politics. Edinburgh and New York: Edinburgh
University Press and New York University Press.
Chambers, Samuel (2005). "The Politics of Literarity." Theory & Event, 8.3: n.p.
Chambers, Samuel (2007). "Normative Violence after 9/II: Rereading the Politics of
Gender Trouble." New Political Science, 29.1: 43-60.
Chambers, Samuel (20IOa). "Jacques Rancière," in Jon Simons (ed), From Agamben
Contemporary Critical Theorists. Edinburgh: Edinburgh University Press; 194--209.
Chambers, Samuel (20IOb). "Police and Oligarchy," in Jean-Phillipe Deranty (ed),
Jacques Rancière: Key Concepts. London: Acumen; 57-68.
Chambers, Samuel (201 I). "Untimely Politics avant la lettre: The Temporality of Social
Formations." Time and Society, 20.2: 199-225.
Chambers, Samuel and Terrell Carver (2008). Judith Butler and Political Theory:
Troubling PoUtics. London and New York: Routledge.
Chambers, Samuel and Michael O'Rourke (2009). "Jacques Rancière on the Shores of
Queer Theory." borderlands, 8.2: n.p.
Citton, Yves (2009). "Political Agency and the Ambivalence of the Sensible," in Gabriel
Rockhill and Philip Watts (eds), Jacques Rancière: History, Politics, Aesthetics.
Durham: Duke University Press; 120-39.
Connolly, William (I987). PoUtics and Ambiguity. Madison: University of Wisconsin Press.
Connolly, William (I995). The Ethos of Pluralization. Minneapolis: University of
Minnesota Press.
Connolly, William (I999). Why 1 Am Not a Secularist. Minneapolis: University of
Minnesota Press.
Connolly, William (20ro). A World of Becoming. Durham: Duke University Press.
Davis, Oliver (2009). "Rancière and Queer Theory: On Irritable Attachment." border-
lands, 8.2: n.p.
Davis, Oliver (20IO). Jacques Rancière. Cambridge: Polity.
Dean, Jodi (2009a). Democracy and Other Neoliberal Fantasies: Communicative
Capitalism and Left Politics. Durham: Duke University Press.
Dean, Jodi (2009b). "Politics without Politics." Parallax, 15.3: 20-36.
de Lauretis, Teresa (r99r). "Queer Theory: Lesbian and Gay Sexualities." differences
3.2: iii-xviii.
Deleuze, Gilles (I983). Nietzsche and Philosophy. Translated by Hugh Tomlinson. New
York: Columbia University Press.
Deleuze, Gilles (I99r). "What Is a Dispositif?" in Timothy Armstrong (ed), Michel
Foucault: Philosopher. New York: Routledge.
Déotte, Jean-Louis (2004). "The Differences Between Rancière's Mésentente (Political
Disagreement) and Lyotard's Différend." SubStance, I03. 1: 77-90.
Deranty, Jean-Philippe (2oo3a). "Jacques Rancière's Contribution to the Ethics of
Recognition." Political Theory, 3 I. r: 136-56.
Deranty, Jean-Philippe (2003b). "Rancière and Contemporary Political Ontology."
Theory & Event, 6,4: n.p.

Works Cited 1 203


ed (2010). Rancière: Key London: Acumen.
Derrida, Jacques (1981). Dissemination. Translated by Barbara Johnson. Chicago:
of Chicago Press.
Derrida, Jacques (1982). Margins of Philosophy. Translated Alan Bass. Chicago:
University of Chicago Press.
Derrida, Jacques (1994). Specters of Marx: The State of the Debt, the Work of Mourning, and
the New International. Translated by Peggy Kamuf. London and New York: Routledge.
Dillon, Michael (2oo3a). "(De)void of Politics?: A Response to Jacques Rancière's Ten
Theses on Politics." TheOly & Event, 6-4: n.p.
Dillon, Michael (2003b). "The Ignorant Statesman: Philosophy, Pedagogy and Politics."
Unpublished manuscript presented at the conference "Fidelity to the Disagreement,"
Goldsmiths College, London, September 17.
Düttmann, Alexander Garcia (1997). "The Culture of Polemic: Misrecognizing
Recognition." Radical Philosophy, 8 1: 27-34.
Elli ott , Gregory (1998). "Ghostlier Demarcations: On the Posthumous Edition of
Althusser's Writings." Radical Philosophy, 90: 20-32.
Fenis, David (2009). "Politics after Aesthetics: Disagreeing with Rancière." Pa rallax ,
I5·3: 37-49·
Finlayson, Alan and Judi Atkins (forthcoming). "'A 40-year-old black man made the
point to me': Anecdotes, Everyday Knowledge and the Performance of Leadership in
British Politics." Political Studies.
Foucault, Michel (1972). The Archeology of Knowledge. Translated by Alan Sheridan.
New York: Pantheon Books.
Foucault, Michel (1978). The HistOly of Sexuality, Volume 1: An Introduction. Translated
by Robert Hurley. New York: Vintage.
Foucault, Michel (I980). Powel/Knowledge: Selected Interviews and Other Writings.
Colin Gordon (ed). New York: Pantheon.
Foucault, Michel (1984). The Foucault Reader. Paul Rabinow (ed). New York: Pantheon
Books.
Foucault, Michel (1997). "Polemics, Politics, and Problematizations: An Interview with
Michel Foucault," in Paul Rabinow (ed), Ethics: Subjectivity and Truth; Essential
Works of Michel Foucault, 1954-1984. New York: The New Press; 111-19.
Foucault, Michel (2002a). "Omnes et Singulatim: Toward a Critique ofPolitical Reason,"
in James D. Faubion (ed), Power: Essential Works of Michel Foucault 1954-1984.
New York: The New Press; 298-325.
Foucault, Michel (2002b). "What Is Critique?" in David Ingram (ed), The Political.
London and New York: Routledge; 191-211.
Foucault, Michel (2008). The Birth of Biopolitics: Lectures at the Collège de France,
1978-1979. New York: Palgrave.
Gibson, Andrew (2005). "The Unfinished Song: Intermittency and Melancholy in
Rancière." Paragraph, 28.1: 61-76.
Habermas, Jürgen (1990). The Philosophical Discourse of Modemity. Translated by
Frederick Lawrence. Cambridge, MA: MIT Press.
Hallward, Peter (2009). "Staging Equality: Rancière's Theatrocracy and the Limits of
Anarchie Equality," in Gabriel Rockhill and Philip Watts (eds), Jacques Rancière:
History, PoUtics, Aesthetics. Durham: Duke University Press; 140-57.

1 The Lessons of Rancière


David (1990). One Hundred Years of Homo'se.~;uatlty And Other on
Greek Love. London and New York: Routledge.
Richard (20 II). "Theater and Democratic Thought: Arendt to Rancière. Critical
lnquiry, 37·3: 545-7 2 .
Heidegger, Martin (1971). On the to Language. Translated by Peter Hertz. San
Francisco: Harper & Row.
Heidegger, Martin (1977). "Letter on Humanism," in David Farrell Krell (ed), Basic
Writings. San Francisco: Harper & Row; 213-65.
Hobbes, Thomas (1994). Leviathan. Edwin Curley (ed). Indianapolis: Hackett Publishing
Company.
Holman, Christopher (2011). "Dialectics and Distinction: Reconsidering HannahArendt's
Critique of Marx." ContemporCllY Political Theory, IO.3: 332-53.
Honig, Bonnie (1993). Political Theory and the Displacement of PoUtics. Ithaca and
London: Cornell University Press.
Honig, Bonnie (2001). Democracy and the Foreigner. Princeton: Princeton University
Press.
Honig, Bonnie (20IO). "Antigone's Two Laws: Greek Tragedy and the Politics of
Humanism." New LiterCllY His tory , 41. I: I-33.
Horkheimer, Max (1972). Critical Theory: Selected Essays. New York: Herder and
Herder.
Hoy, David Couzens, and Thomas McCarthy (1994). Critical Themy. Oxford:
Blackwell.
Huffer, Lynne (2009). Mad For Foucault: Rethinking the Foundations of Queer Theory.
New York: Columbia University Press.
Ieven, Bram (2009). "Heteroreductives: Rancière's Disagreement with Ontology."
Parallax, 15.3: 50-62.
Ingram, James (2008). "What Is a 'Right to Have Rights'? Three Images of the Politics
of Human Rights." American Political Science Review, I02.04: 40I-16.
Jagose, Annamarie (1996). Queer Them"y: An Introduction. New York: New York
University Press.
Kant, lmmanuel (1970). Political Writings. Edited by H. S. Reiss. Cambridge: Cambridge
University Press.
Kaplan, David (2008). "Ricoeur's Critical Theory," in David Kaplan (ed), Reading
Ricoeur. Albany: SUNY Press; 197-212.
Ko lli as , Hector (2007). "Taking Sides: Jacques Rancière and Agonistic Literature."
Paragraph, 30.2: 82-97.
Kompridis, Nikolas (2006). Critique and Disclosure: Critical Theory between Past and
Future. Cambridge, MA: MIT Press.
Kropotkin, Petr (1995). '''Anarchism,' from The Encyclopedia Britannica," in Marshall
Shatz (ed), The Conquest of Bread and Other Anarchist Writings. Cambridge:
Cambridge University Press; 233-47.
Laclau, Ernesto (1996). Emancipation(s). London and New York: Verso.
Laclau, Ernesto (2006). "Ideology and Post-Marxism." Journal of Politicalldeologies,
11.2: I03-I4.
Lacoue-Labarthe, Philippe and Jean-Luc Nancy (1997). Retreating the Political. Simon
Sparks (ed). London and New York: Routledge.

Works Cited 1 205


Latour, Bruno (1999). Pandora 's on the of Science Studies.
Cambridge, MA: Harvard University Press.
Latour, Bruno (2004). "Why Has Critique Run Out of Steam? From Matters of Fact to
Matters of Concern." Criticallnquiry, 30: 225-48.
Lievens, Matthias (2012). "Ideology Critique and the Political: Towards a Schmittian
Perspective on Ideology." COlltempOrm)1 Political Theory, II.4-
Locke, John (1988). Two Treatises of Govermnent. Peter Laslett (ed). Cambridge:
Cambridge University Press.
Lorde, Audre (1984). "The Master's Tools Will Never Dismantle the Master's House," in
Sister Outsider: Essays and Speeches. Berkeley: Crossings Press; 1 IO-I3.
Lyotard, Jean-François (I984). The Postmodem Condition: A Report on Knowledge.
Translated by Geoff Bennington and Brian Massumi. Minneapolis: University of
Minnesota Press.
Lyotard, Jean-François (I988). The Differend: Phrases in Dispute. Translated by Georges
Van Den Abbeele. Minneapolis: University of Minnesota Press.
Lyotard, Jean-·François (1990). Heidegger and "the Jews." Translated by Andreas Miche]
and Mark Roberts. Minneapolis: University of Minnesota Press.
Lyotard, Jean-François and Jean-Loup Thébaud (1985). Just Gaming. Translated by Wlad
Godzich. Minneapolis: University of Minnesota Press.
Marchart, Oliver (2007). Post-Foundational Political Thought: Political Difference in
Nancy, Lefort, Badiou and Laclau. Edinburgh: Edinburgh University Press.
Marchart, Oliver (2010). Die politische Differenz: Zum Denken des PoUtischen bei Nancy,
Lefort, Badiou, Laclau und Agamben. Berlin: Suhrkamp.
Marder, Michael (201O). Groundless Existence: The Political Ontology of Carl Schmitt.
London: Continuum.
Markell, Patchen (2006). "Ontology, Recognition, and Politics: A Reply." PoUty, 38.1:
28-39·
Markell, Patchen (2011). "Arendt's Work: On the Architecture of The Human Condition."
College Literature, 38.1: 15-44.
Marx, Karl (1977). Capital: A Critique of Po litica 1Economy. Translated by Ben Fowkes.
New York: Vintage.
Marx, Karl (1988). Economie and Philosophic Manuscripts of 1844. Translated by Martin
Milligan. Amherst, NY: Prometheus Books.
Marx, Karl (1996). '''Introduction' to the Grundrisse," in Terrell Carver (ed), Marx: Later
Political Writings. Cambridge: Cambridge University Press; 128-57.
May, Todd (1994). The Political Philosophy of Poststructuralist Anarchism. University
Park: Pennsylvania State University Press.
May, Todd (2008). The Po litica 1 Thought of Jacques Rancière: Creating Equality.
Edinburgh: Edinburgh University Press.
May, Todd (2009). "Rancière in South Carolina," in Gabriel Rockhill and Philip Watts
(eds), Jacques Rancière: History, PoUtics, Aesthetics. Durham: Duke University
Press; 105-19.
May, Todd (20IOa). Contempormy Political Movements and the Thought of Jacques
Rancière: Equality in Action. Edinburgh: Edinburgh University Press.
May, Todd (20IOb). "Thinking the Break: Rancière, Badiou, and the Return of a Politics
of Resistance." Comparative and Continental Philosophy, 1.2: 253-68.

206 1 The Lessons of Rancière


May, Todd (20 In. "Rancière, Anarchism, Governance." at the conference
Jacques Rancière: Politics and Aesthetics, Northwestern University, April 22.
Todd (2012). Personal communications with the author.
McClure, Kirstie (2003). "Disconnections, Connections, and Questions: Reflections on
Jacques Rancière's 'Ten Theses on Politics.'" Theory & Event, 6-4: n.p.
Mehlman, Jeffrey (1976). "The Case of Marx in France." Diacritics, 6-4: 10-18.
Milchman, Alan and Alan Rosenberg (2007). "The Aesthetic and Ascetic Dimensions of
an Ethics of Self-fashioning: Nietzsche and Foucault." Parrhesia, 2: 44-65.
Montag, Warren (2011). "Introduction to Althusser's 'Student Problems.'" Radical
Philosophy, 170: 8-IO.
Morton, Timothy (2007). Ecology without Nature: Rethinking Environmental Aesthetics.
Cambridge, MA: Harvard University Press.
Mouffe, Chantal (2000). The Democratic Paradox. London: Verso.
Mowitt, John (1992). Text: The Genealogy of an Antidisciplinm:v abject. Durham: Duke
University Press.
Mufti, Aamir (2003). "Reading Jacques Rancière's 'Ten Theses on Politics': After
September l rth." Them)' & Event, 6-4: n.p.
Muhle, Maria (2007). "Politics, Police, and Power between Foucault and Rancière."
Unpublished manuscript.
Nietzsche, Friedrich (1977). The Portable Nietzsche. Translated by Walter Kaufmann.
New York: Penguin.
Nixon, Sean (1997). "Exhibiting Masculinity," in Stuart Hall (ed), Representation:
Cultural Representations and Signifying Practices. London and Thousand Oaks:
SAGE Publications; 291-336.
Noyes, Matt (2008). "What Do We Want? When Do We Want It?" (Online). Available
from: http://www.re.rollingearth.orgl?q=node/I 3 l (Accessed: October 29, 2008).
Oksala, Johanna (20 IO). "Foucault's Politicization of Ontology." Continental Philosophy
Review, 43-4: 445-66.
O'Rourke, Michael (2005). "On the Eve of a Queer-Straight Future: Notes Toward an
Antinormative Heteroerotic." Feminism & Psychology, 15. 1: II 1-16.
Palladino, Paolo and Tiago Moreira (2006). "On Silence and the Constitution of the
Political Community." Theory & Event, 9.2: n.p.
Panagia, Davide (2001). "Ceci 12 'est pas un argument: An Introduction to the Ten Theses
on Politics." Theory & Event, 5.3: n.p.
Panagia, Davide (2006). l1w Poetics of Political Thinking. Durham: Duke University
Press.
Panagia, Davide (2009). The Political Life of Sensation. Durham: Duke University Press.
Panagia, Davide (20IO). "The Sharing of the Sensible," in Jean-Philippe Deranty (ed),
Jacques Rancière: Key Concepts. London: Acumen; 95-103.
Parker, Andrew (2004). "Mimesis and the Division of Labor:' introduction to Jacques
Rancière, The Philosopher and His POOl'. Durham: Duke University Press; ix-xx.
Parker, Andrew (2007). "Impossible Speech Acts," in Martin McQuillan (ed), The PoUtics
of Deconstruction: Jacques Derrida and the Other of Philosophy. London: Pluto
Press; 66-80.
Power, Nina (2009). "Which Equality? Badiou and Rancière in Light of Ludwig
Feuerbach." Parallax, 15.3: 63--80.

Works Cited 1 20]


Rancière, Jacques (I974a). La leçon d'Althusser. Paris: Gallimard.
Rancière, Jacques (I974b). "On the Theory of Ideology." Radical Philosophy, T 2-I5.
Rancière, Jacques (I989a). "The Concept of 'Critique' and the 'Critique of Political
Economy' (From the Manuscripts of 1844 to Capital)," in Ali Rattansi (ed), Ideology,
Method, and Marx: Essays from Economy and Society. London and New York:
Routledge; 74--180.
Rancière, Jacques (I989b). "How to Use Lire le Capital," in Ali Rattansi (ed), Ideology,
Method, and Marx: Essays from Economy and Society. London and New York:
Routledge; 181 -89.
Rancière, Jacques (I989c). The Nights of Labor: The Workers' Dream in Nineteenth.,
Century France. Translated by John Drury. Philadelphia: Temple University
Press.
Rancière, Jacques (I99I). The Ignorant Sehoolmaster: Five Lessons in Intellectual
Emancipation. Translated by Kristin Ross. Stanford: Stanford University Press.
Rancière, Jacques (I994). The Names ofHistory: On the Poetics ofKnowledge. Translated
by Hassan Melehy. Minneapolis: University of Minnesota Press.
Rancière, Jacques (I99Sa). La mésentente: politique et philosophie. Paris: Galilée.
Rancière, Jacques (I99Sb). On the Shores of Politics. Translated by Liz Heron. London
and New York: Verso.
Rancière, Jacques (I99Sc). "Politics, Identification, Subjectivization," in John Rajchman
(ed), The Identity in Question. London and New York: Routledge; 63--72.
Rancière, Jacques (I998a). Aux bords du politique. Second ed. Paris: La Fabrique.
Rancière, Jacques (1998b). "The Cause of the Other." Parallax, 4·2: 2S-34.
Rancière, Jacques (1999). Disagreement: PoUties and Philosophy. Translated by Julie
Rose. Minneapolis: University of Minnesota Press.
Rancière, Jacques (2001). "Ten Theses on Politics." Translated by Rachel Bowlby and
Davide Panagia. Theory & Event, S.3: n.p.
Rancière, Jacques (2003a). "Comment and Responses." TheOfy & Event, 6.4: n.p.
Rancière, Jacques (2003b). Short Voyages to the Land of the People. Translated by James
B. Swenson. Stanford: Stanford University Press.
Rancière, Jacques (2oo3c). "The Thinking of Dissensus: Politics and Aesthetics."
Unpublished manuscript written in response to papers given at the conference
"Fidelity to the Disagreement," Goldsmiths College, London, September 17.
Rancière, Jacques (2004a). The Flesh of Words: The PoUtics of Writing. Translated by
Charlotte Mandell. Stanford: Stanford University Press.
Rancière, Jacques (2004b). The Philosopher and His POOl'. Translated by John Drury,
Corinne Oster and Andrew Parker. Durham: Duke University Press.
Ràncière, Jacques (2004c). "The Politics of Literature." SubStance, 33.r: I0-24.
Rancière, Jacques (2004d). "Who Is the Subject of the Rights of ManT' South Atlantic
Quarterly, 103.2-3: 297-310.
Rancière, Jacques (200sa). "From Politics toAesthetics?" Paragraph, 28.1: 13-2S.
Rancière, Jacques (200Sb). La haine de la démocratie. Paris: La Fabrique.
Rancière, Jacques (200SC). "Literary Misunderstanding." Paragraph, 28.2: 9I-I03.
Rancière, Jacques (2006a). "The Ethical Tum of Aesthetics and Politics." Critical
Horizons, 7.r: 1-20.
Rancière, Jacques (2006b). Film Fables. Translated by Emiliano Battista. New York:
Berg.

208 1 The Lessons of Rancière


Rancière, (2oo6c). Hatred J01"1/1r','ru'" Translated Steve Corcoran.
and New York: Verso.
Rancière, Jacques (2006d). The Politics of Aesthetics: The Distribution of the Sensible.
Translated by Gabriel Rockhill. London and New York: Continuum.
Rancière, Jacques (2007a). "Does Democracy Mean Something?" in Costas Douzinas
(ed), Adieu Derrida. New York: Palgrave Macmillan; 84--IOO.
Rancière, Jacques (2007b). "The Misadventures ofCritical Thinking." Aporia, 24.2: 22-32.
Rancière, Jacques (2007c). "The Misadventures of Critieal Thinking." Unpublished
manuscript.
Rancière, Jacques (2007d). "What Does It Mean to be Un?" Continuum, 21.4: 559-69.
Ranci ère, Jacques (2oo9a). The Emancipated Speetator. Translated by Gregory Elliott.
London and New York: Verso.
Rancière, Jacques (2009b). "A Few Remarks on the Method of Jacques Rancière."
Parallax, IS.3: 114-23.
Rancière, Jacques (2009c). The Future of the Image. Translated by Gregory Elliott.
London and New York: Verso.
Rancière, Jacques (2009d). "The Method of Equality: An Answer to Sorne Questions," in
Gabriel Rockhill and Philip Watts (eds), Jacques Rancière: History, Politics,
Aesthetics. Durham: Duke University Press; 273-88.
Rancière, Jacques (20Ioa). Chronicles of Consensua! Times. Translated by Steven
COl'coran. London and New York: Continuum.
Rancière, Jacques (20IOb). Dissensus: On Politics and Aesthetics. Edited and Translated
by Steven Corcoran. London and New York: Continuum.
Rancière, Jacques (201 la). Althusser's Lesson. Translated by Emiliano Battista. London
and New York: Continuum.
Rancière, Jacques (20IIb). Staging the People: The Proletarian and His Double.
Translated by David Fernbach. London and New York: Verso.
Rancière, Jacques (201 IC). "The Thinking of Dissensus: Poli tics and Aesthetics," in Paul
Bowman and Richard Stamp (eds), Reading Rancière. London and New York:
Continuum; 1-17.
Rancière, Jacques (201 Id), Personal communications with the author.
Rancière, Jacques (201 le). Mute Speech: Literature, Critical TheO/y, and PoUties. New
York: Columbia University Press.
Rancière, Jacques and Anne Marie Oliver (2008). ''Aestheties against Incarnation."
CriticalInquiry, 3S. 1: I72-90.
Rancière, Jacques and Sudeep Dasgupta (2008). "Art Is Going Elsewhere and Politics
Has to Catch It: An Interview with Jacques Rancière." Krisis, 9.1: 70--76.
Rancière, Jacques, Solange Guénoun and James Kavanagh (2000). "Literature, Polities,
Aesthetics: Approaches to Democratie Disagreement." SubStance, 29.2: 3-24.
Rancière, Jacques and Davide Pan agi a (2000). "Dissenting Words: A Conversation with
Jacques Rancière." Diacritics, 30.2: 113-26.
Rancière, Jacques, et al. (1977). "Editorial." Les révoltes logiques, Issue S: 3--6.
Rancière, Jacques, et al. (2008). "Democracy, Anarchism and Radical Politics Today: An
Interview with Jacques Rancière." Anarchist Studies, 16.2: 173-8S.
Rancière, Jacques, et al. (2009). "Politique de l'indétermination esthétique," in Jan Volker
and Frank Ruda (eds), Jacques Rancière et la politique de l'esthétique. Paris: Éditions
des Archives Contemporaines; IS7-7S.

Works Cited 1
Rattansi,Ali, ed (1989). ldeology, Method, and Marx: Economy
London and New York: Routledge.
Rawls, John (1971). A TheOly of Justice. Cambridge, MA: Harvard University Press.
Rawls, John (1993). PoUtical Liberalism. New York: Columbia University Press.
Repo, Jemima (2011). The Biopolitics ofGender. PhD Dissertation. Helsinki: University
of Helsinki.
Robson, Mark (2009). "'A Literary Animal': Rancière, Derrida, and the Literature of
Democracy." Parallax, 15.3: 88-IOI.
Rockhill, Gabriel (2004). "The SHent Revolution." SubStance, 33. 1: 54-76.
Rockhill, Gabriel (2006). "Preface, Introduction, and Glossary," in Gabriel Rockhill (ed),
The Poliûcs of Aesthetics. London and New York: Continuum; vii-x; 1-6; 80-93.
Rockhill, Gabriel (2011). "Through the Looking Glass: The Subversion of Modernist
Doxa," introduction to Mute Speech: Literature, Critieal Theory, and PoUties.
New York: Columbia University Press.
Rorty, Richard (1991). Essays on Heidegger and Others: Philosophical Papers, Volume
2. Cambridge: Cambridge University Press.
Ross, Alison (20 IO). "Expressivity, Literarity, Mute Speech," in Paul Bowman and Richard
Stamp (eds), Reading Raneière. London and New York: Continuum; 133-50.
Ross, Kristin (2002). May '68 and Its Afterlives. Chicago: University of Chicago Press.
Ross, Kristin (2009). "Historicizing Untimeliness," in Philip Watts and Gabriel Rockhill
(eds), Jacques Rancière: His to ry, PoUtics, Aestheties. Durham: Duke University
Press; 15-29.
Schaap, Andrew (2011). "En acting the Right to Have Rights: Jacques Rancière's Critique
of Hannah Arendt." European Journal of Political TheOly, 10.1: 22-45.
Schmitt, Carl (1988). The Crisis of Parliamentmy Democracy. Translated by Ellen
Kennedy. Cambridge, MA: The MIT Press.
Scott, Joan (1991). "The Evidence of Experience." Criticallnquiry, 17-4: 773-97.
Sedgwick, Eve Kosofsky (1990). Epistemology of the Closet. Berkeley: University of
California Press.
Sedgwick, Eve Kosofsky (2003). Touching Feeling: Affect, Pedagogy, Pelformativity.
Durham: Duke University Press.
Seery, John (201 la). Too Young to Run? A Proposal for an Age Amendment to the U.S.
Constitution. University Park: Pennsylvania State University Press.
Seery, John (20 lIb). "Wh y Should Holding Office Be for Old People Only?" Salon
(Online). Available from: http://www.salon.comlnews/politics/wacroomI201 1/06/
26/john_seery_age/index.html (Accessed: July 7,2011).
Shachtman, Noah (20Il). "Pentagon's Prediction Software Didn't Spot Egypt Unrest"
(Online). Available from: http://www.wired.comldangerrooml201 l/02/pentagon-
predict-egypt-unresU (Accessed: June 16,2011).
Shapiro, Michael (2009). Cinematie Geopolitics. New York: Taylor & Francis.
Skinner, Quentin (1988). "Meaning and Understanding in the History ofIdeas," in James
Tully (ed), Meaning and Context: Quentin Skinner and His Critics. Princeton:
Princeton University Press; 29-67.
Sloterdijk, Peter (2004). Spharen 1//: Schtiwne. Berlin: Suhrkamp.
Spivak, Gayatri Chakravorty (1988). "Can the Subaltern Speak?" in Cary Nelson and
Lawrence Grossberg (eds), Marxism and the Interpretation of Culture. Bloomington:
Indiana University Press; 271-313.

210 1 The Lessons of Rancière


Richard (2009). "The Torsion of Politics and t'nenclshlp in Derrida, Foucault and
Rancière. borderlands, 8.2: n.p.
Strauss, Leo (1959). What Is Political Philosophy? And Other Studies. Glencoe, Free
Press.
Tanke, Joseph (201 r). Jacques Rancière: An Introduction. London and New York:
Continuum.
Thomson, Alex (2003). "Re-placing the Opposition: Rancière and Derrida." Unpublished
manuscript presented at the conference "Fidelity to the Disagreement," Goldsmiths
College, London, September 17.
Valentine, Jeremy (2005). "Rancière and Contemporary Political Problems." Paragraph,
28. I: 46-60.
Vattel', Miguel (2008). "Book Review" of Todd May's The Political Thought of Jacques
Rancière: Creating Equality. Notre Dame Philosophical Reviews (Online). Available
from: http://ndpr.nd.edu/review.cfm?id=I5405 (Accessed June 12,2011).
Vatter, Miguel (20I2). "The Quarrel between Populism and Republicanism: Machiavelli
and the Antinomies of Plebeian Politics." Contemporary Political Theory, I I.3:
24 2- 63.
Ward, Colin (I982). Anarchy in Action. London: Freedom Press.
Warner, Michael, ed (1993). Feal' of a Queer Planet: Queer Polilics and Social Theory.
Minneapolis: University of Minnesota Press.
Watkins, Robert (20lO). "Book Review" of Davide Panagia's The Political Life of
Sensation. Perspectives on Politics, 8.2: 665-66.
West, Cornel (2011). "Hope can be betrayed." (Twitter update). http://twitter.com/#!/
CornelWestistatus/9I615996804202498 (Accessed July 20, 20II).
White, Stephen (1997). "Weak Ontology and Liberal Political Reflection." Political
Theory, 25-4: 502- 23.
White, Stephen (2000). Sustaining Affirmation: The Strengths of Weak Ont%gy and
Political TheOlY. Princeton: Princeton University Press.
White, Stephen (2005). "Weak Ontology: Genealogy and Critical Issues." The Hedgehog
Review, 22 June: 11-25.
Wittgenstein, Ludwig (1972). On Certainty. Translated by Denis Paul and G. E. M.
Anscombe. New York: Harper Torchbook.
Zivi, Karen (2012). Making Rights Claims: A Practice of Democratie Citizenship.
New York: Oxford University Press.
Zizek, Slavoj (1994). Mapping Ideology. London: Verso.
Zizek, Slavoj (I999). The Ticklish Subject: The Absent Centre of Political Ontology.
London and New York: Verso.
Zizek, Slavoj (2006). "The Lesson of Rancière," in Gabriel Rockhill (ed), The Politics of
Aesthetics. London and New York: Continuum; 69-79.

Works Cited 1 211


"A Few Remarks on the Method of Althusser's Lesson (Rancière), 142,
Jacques Rancière" (R anci ère ), 3-4, 171n1,195n28
181n21 anarchism
Adorno, Theodor, 191 n5 distinguished from an-archie poUties,
aesthetics 79-80,179n3
cinema, 4, 174n22 equality in, 67, 77, 78, 79, 80, 82,
and critique, 144-45, 147 83-84, 196n34
ethies in, 175n25 as framework for interpreting Rancière,
literature, 91,114-15, 174n22, 181n21, 45, 75-84, 184n9
189n22, 190n23, 190n26 hopein, 63
politics as, 174n22 and normative grounds, 80-81, 85,
andtheory,4,181n21 179n3
See also literarity; partage du sensible as ontology, 179n3
age restrictions for holding office, 12-14, and police orders, 75-84
59,174n19 as pure polities, 34, 40, 45, 67-68, 76,
Althusser, Louis 82-84, 184n12
on anthropologism, 107, 128-33, and revolution, 78, 83, 141, 184nl3
191n4,192n11 stmggle in, 85-86
framework for interpreting Marx, 5, 17, animaIs. See nonhumans
107,128-33,142-43, 171n1, 191n4, anthropocentrism
192nnll-12 and anthropologism, 107, 191n4
on interpellation, 24, 183n5 and anthropomorphism, 106-7
and police, 24, 65, 72, 183n5 definition of, 105
Rancière's break with, 22, 35--36, 107, and humanism, 106, 189n 19
124, 142, 171n1 and Kant, 110
Reading Capital, 35-36, 124, 127, and language, 34,90,96, 105, 110-12,
128-33,142-43, 176n30, 192n11, 188-89n16
193nl3,195n26 and literarity, 112-13
and science, 129, 143, 192nlO and Lyotard, 89-90, 108-l3, 120,
on social formations, 65, 124 188-89n16, 189n19
students of, 35-36, 62 and Marx, 191 n4
anthropocentrism (continuee!) See also On the Shores of Politics;
and ontology, 96 translation: of Rancière's works
and Wittgenstein, 110-11 written in English
See a/so language: as tool or
possession Badiou, Alain, 51, 62, 171nl, 179n2,
anthropologism, 107, 128-33, 191n4, 194n22
192nl1 Balibar, Étienne, 62, 193n 13, 195n26
anthropomorphism, 106-7 Baudrillard, Jean, 145
antihumanism. See anthropologism; Beltrân, Cristina, 5, 39, 46-48, 118
humanism Bennett, Jane, 106-7, 126, 173n10,
Apostolidis, Paul, 35, 118-20 179n4,191nn4-5
Arab Spring. See Egyptian revolution Berlant, Lauren, 190n 1
(20 Il) and Arab Spring Blanqui, Auguste, 117
archë,33,39,49,59-60,66,79,135-36, blindness. See cri tic al dispositif: and
163,166, 179n3, 193-4nn18-19 inversion
archipolitics, 77, 127, 136-37, 140, Bowman, Paul, 89,134,138, 176-77n31,
184n8, 194n21 199n8
See also political philosophy Butler, Judith
archives. See history and feminism, 160-62, 198n6
Arditi, Benjamin, 20, 123, 154, 187n9, on Foucault, 99-100, 187n10
197n2 on intelligibility, 167-68, 199nn7-8,
Arendt, Hannah 199n13
Arendtian framework vs. works of, 39, and literarity, 195n30
46, 180n7 and ontology, 163, 199n9
interpretations of Rancière based on on queer politics, 160-63, 167-68,
works of, 32, 39-40, 45-49, 180n8 198n5, 199nn7--8
and language, 96 on subjectivation, 99-100, 162, 187n 10
and "political difference," 51
as political philosopher, 39-40, 74 Capital (Marx), 128-33, 138, 143-44,
and political space, 57, 69 196-97n38
and political subjects, 17 capitalism, 86, 130-33, 143, 145, 146,
on surprise, 5, 39, 46, 172n3 150,155-56, 185-86nI7, 192n9,
Aristotle 196n38
anthropocentrism of, 96, 105, 106, Carver, Terrell, 177n34, 178n39
107-8, 109, 110 complexity theory, 155
anthropologism of, 107 "Concept of Critique, The" (Rancière), 22,
and language, 34-35, 89-98, 104, 105, 36, 107, 124, 127, 128-33, 142-43,
106, 109-10, 114, 115-16, 119, 162, 176n30, 192n12, 195n26
164, 186n6, 187n9, 188-89n16 Connolly, William, 7, 18--21, 155,
and ontology, 65-66, 188-89n 16 175nn26-28, 190n3
as political philosopher, 39, 74, 136 consensus
translations of, 90,92-95, 186n3 and critique of political philosophy,
See also language: as tool or 184n8
possession; phonë/logos distinction distinct from democratic politics, 9
Aux bords du politique (Rancière, second in Habermas, 187n9
edition), 54-56 as police order, 33,44,67,68, 73-75

214 1 The Lessons of Rancière


as post-democracy, 73-75,112, Debord, 147, 1951130
185n17, 194n23 Deleuze, Gilles, 25, 125, 182n26,
reversaI in Rancière's critique of, 134, 196n35
139-40 deliberative theory. See Habermas
See also liberalism dëmos. See political subjects; poor, the;
critical dispositif subjectivation
and dissensus, 153-54 demystification. See critical dispositif:
and equality, 150-52 and inversion
and intelligibility, 152-54 Deranty, Jean-Philippe, 32, 50-58, 60-62,
and inversion, 36, 126-33, 137-56, 65-66,80,88, 180n12, 181n22,
172n4, 183n6, 190n2, 191n4, 191n5, 184n12, 187n9
192n9, 194n21, 194n23, 195n27, Derrida, Jacques, 8, 59,113-14, 182n26,
195n30, 196n34 189n21
and partage du sensible, 125, 153-54, dialectics
190n2, 196n35 and equality, 50, 53, 83-85
and polemicization, 124-25, 175n23 as framework for interpreting Rancière,
and radical pedagogy, 148, 151-52, 182n26,184n11
172n4,196n34 and queer politics, 197n2
and reversaI, 139-41 Rancière's critique of, 61-62
See aiso dispositif See a/so purity of politics: in "three-
criticai theory term" model
alternatives to, 190n3 differend, 108-12, 120, 188n15
in Althusserian Marxism, 128-33 Dillon, Michael, 108, 178n1
of capitalism, 130-33, 143, 145, 146, Disagreement (Rancière)
150, 155-6, 185-86n17 192n9, critique in, 127, 134-42, 154
196n38 and language, 108, 188n14
and critical dispositif, 125 liberalism in, 173n15
distinguished from Kantian tradition of literarity in, 114, 190n24
critique, 124 and ontology, 74, 193n17
and Foucault, 44, 124, 154-55 police in, 41, 68-75, 79, 183n7
Frankfurt Schooi tradition of, 124, 155, and "the political," 53-54, 56
191n5 reading of Aristotle's Polilies in,
in Habermas, 124, 126, 154 34-35,89,91-98
and history, 132-33, 155-56, 196n37 reception of, 38, 74,138-39, 173,
and Marx, 107, 124, 128-33, 136-38, 193n17
155-56, 191n4, 192n9, 194n21, See also translation: of La mésentente
195n27, 196n38 discourse ethics. See Habermas
as metapolitics, 136-38 dispositif, 125-26, 190n2, 195n28,
and ontology, 36 196n35
and police orders, 154, 183n6, 196n36 See also cri tic al dispositif; critical
"post-critical," 144-47, 195n3 theory
as a tool, 44, 179-80n6 distribution of the sensible. See partage
du sensible
Davis, Oliver, 158, 171-72n2, 183n3, doubling. See impurity of politics;
186n1, 187n7, 188n13 supplement
Dean, Jodi, 185n17 Düttmann, Alexander Garcia, 197n2

Index 1 21 5
education. See pedagogy Foucault, Michel
Egyptian revolution (2011) and Arab and Butler, 99-100, 187nlO
Spring, 7-9, 173n7, 173n9, on critique, 44, 124, 154-55
173nnll-12, 177n32, 177n33 on dispositif, 125
emancipation, 5, 54, 146, 147-49, 157, and liberalism, 173n14
167,168-69,188n15 and normative grounds, 44, 179n6,
equality 194n22
in anarchism, 67, 77, 78, 79, 80, 82, and ontology, 175n27
83-84, 196n34 and polemics, 175n23
and critique, 36, 150--52, 196n34 and police, 42, 70, 76, 183n2
dialectical account of, 50, 53, and queer theory, 198n5
83-85 Rancière's affinities with, 42,69--70,
and hope, 150 175nn23-24,194n22
in the liberaI tradition, 13,25-27, °
on subject formation, 17, 89, 98-1 1,
28-29,159,161, 177nn34-35, 187nlO, 188nl1, 188n13
178n39,198n4 Frankfurt School, 124, 155, 191n5
and May '68,24,29, 117, 150
and normative grounds, 13,26-27, gay and lesbian politics. See queer
63-64,80-81,85,150-52 politics: distinct from identity
and ontology, 25-27, 29, 80-81, 85 politics
and pedagogy, 4,30-31, 151--52, Gender Trouble (Butler), 160-62, 198n5,
196n34 199n7
and subjectivation, 29-30, 43
verification of, 10,25, 27-30, 63-64, Habermas, Jürgen
66,80,119,150--52 on critique, 124, 126, 154
ethics, 18, 107, 175n25, 191n4 on language, 90, 187n9
event and "liberal democracy," 11-12,
in Deleuze, 25 173n15,185n17
Egyptian revolution (2011) and Arab on normative grounds, 80, 194n22
Spring as, 7-9, 177n32 Hatred of Democracy (Rancière), 74, 75,
in Foucault, 177n33 183n7
and history, 21, 24-25 Hegel, Georg Wilhelm Friedrich, 40, 50,
in liberalism, 25, 178n39 57, 128
May '68 as, 24, 177n32 See also dialectics
politics as rare, 8, 28, 66, 121, 141, Heidegger, Martin, 106, 111
168, 177n32 heteronormativity, 160--61, 165-66,
Revolutions of 1848 as, 178n39 167-68, 199n7, 199n12
and surprise, 5, 8, 173nlO hierarchy. See police
See also history; surprise; temporality history
excess of words. See literarity and critique of capitalism, 132-33,
explication. See pedagogy; stultification 155-56, 1961137
and event, 5, 8, 23-25, 1771133
feminism, 160-62, 198n6 and Foucault, 177n33
Feuerbach, Ludwig, 128, 130-32 as location, 23, 25, 31--32
Flesh ofWords, The (Rancière), 115, and ontology, 19,21,45
190n24 of Rancière's publications in
For Marx (Althusser), 128-29 translation, 21-23

216 1 The Lessons of Rancière


Hobbes, Thomas, 26-27, 29, 136, 177n34, and queer 37, 167-68
178n39,186--87n6 and social science, 6
See also liberalism and surprise, 172-73n6
Bonnie, 47-48, 180n8, 199n9 and untimeliness, 9, 199-200n13
hope, 63, 123, 150, 196n33 See also partage du sensible
Horkheimer, Max, 124, 191n5 interests and interest group politics, 8, 9,
"How to Use Lire le Capital" (Rancière), 33,67,72-73,102-3,117,120,
142-44 159-66 passim
David Couzens, 126, 196n37 See also consensus; identity; liberalism
humanism, 106, 129, 143, 191n4, 192n11,
199n9 J acotot, Joseph, 30-31, 114,
See also anthropologism 151-52, 157

identity Kant, Immanuel, 40, 110, 124, 148


and literarity, 182n28 Kompridis, Nikolas, 154, 191n5
and polemicization, 120, 123, 154 kratos, 59, 135, 163, 194
and queer politics, 159-68 passim,
197n2 la police. See police; translation
°
and subjectivation, 1 1, 104, 112, la politique. See translation
119-20,164-65, 193n16, 199nlO Laclau, Ernesto, 51, 196n32
See also liberalism Lacoue-Labarthe, Philippe, 51, 53
Ieven, Bram, 19, 21 language
Ignorant Schoolmaster, The (Rancière), in Arendt, 96
31,113-14,127,151,157 in Habermas, 90, 187n9
immigrants, 46, 118--20 and intelligibility, 96-97, 118·-19
impurity of politics in Lyotard, 89-90, 108-13, 120,
and dialetics, 182n26 188n16, 189n19
and equality, 28-29 and ontology, 35,90,96-97, 186n2,
and language, 41, 58, 63,97-98, 105, 188n16
108,115, 182n28, 188n14 and partage du sensible, 95, 96-98,
and paradox, 49-50,57,61,85 119
and police, 32-33,49--50, 62-64, 65, in Plato, 113-14, 115-17, 189n21,
84-86, 182n27 190nn25-26
and queer politics, 165-66 and revoIt, 186n1
and revolution, 173n9 and subjectivation, 34-35, 104-5,
as supplementation, 40-41, 58-63, 117-19,188n14
182n26 as tool or possession, 34, 93-97,
See also paradox; purity of politics 105,106,107,109-10,114, 116,
intelligibility 187n9
in Butler, 167-68, 199nn7-8, 199n13 See also anthropocentrism; Aristotle:
and critical dispositif, 152-54 and language; literarity; phonëllogos
distinguished from hiddenness, 152-53, distinction; translation
199n13 le politique. See political, the;
andlanguage,96-97,118-19 translation
and paradox, 199n8 Les révoltes logiques (journal), 4-5, 22,
and partage du sensible, 70, 199n13 186n1, 188n14
and polemicization, 123 les sons. See pedagogy

Index 1 21 7
liberalism non-anthropocentric view of language
and deliberative democracy, 11-12, in, 89-90, 108-13, 120, 188n16,
173n15,185n17 189n19
democratic poli tics distinguished from, Rancière's criticisms of, 175n25,
10-14,17,159,160,166,168, 188n15,189nn17-18
173nnl4-15, 174nn17-18, 174n20,
185n17 Manicheanism. See impurity of politics;
doctrinal, 172n5, 178n39 pu rit Y of politics
and equality, 13,25-27,28-29, 159, Marchart, Oliver, 51-52, 56, 179n2,
161, 177nn34-35, 178n39, 198n4 180nl0, 180n17, 181nn19-20
and event, 25, 178n39 Markell, Patchen, 19,46, 180n7
and gay and lesbian identity, 159-65 Marx, Karl
passim and anthropologism, 107, 128-33,
and literarity, 90, 117 191n4,192n11
neoliberalism distinguished from, 172n5 Marxism vs. works of, 171n1,
as police order, 10-11, 14,29,67, 172n5 192nn8-9
political subjects in, 16--17,26,29, as political philosopher, 39, 77,
100, 102--3 194n21
and queer politics, 37, 159-68 passim, and social order, 65
197n2, 199n8 See also critical theory
and revolution, 178n39 Marxism
and surprise, 6-·8, 103, 172n4 and anthropocentrism, 191 n4
See also consensus; Habermas; Hobbes; hopein,63
identity; interests and interest group "the poor" in, 5, 17, 100, 192n8
politics; Locke; Rawls struggle in, 5, 85-86
literarity See also Althusser; critical theory
and anthropocentrism, 112--13 mastery,4, 13,24,30-31, 111, 146, 150,
and Butler, 195n30 171n1,175n23
exceedslanguage, 186n1 See also pedagogy
genealogy of the concept, 113-14, May '68,23-25,29,71,117,150,
189n20, 189n22, 190n24 177n32, 178n39, 199n8
and impure politics, 41,63, 182n28 May '68 and its Afterlives (Ross), 23-24,
and liberalism, 90, 117 29
and literature, 91, 114-15, 189n22, May, Todd, 34,45,67-68, 75-87, 179n2,
190n23, 190n26 183n2, 184nn8-9, 184nn12-13,
and partage du sensible, 115-16, 169, 185n14-15, 196n34
195n30 McClure, Kirstie, 178n 1
and pedagogy, 168--69 metaphysics. See ontology
and police orders, 88, 113, 121 metapolitics, 127, 136-39, 140-41,
and subjectivation, 114-21 184n8, 194n21
as untimely, 117, 120-21 See also political philosophy
See also language "Mis ad ventures of Critical Thinking,
Locke, John, 6, 17,26-27, 172n5, The" (Rancière), 127, 142, 144-49,
174n19, 177n34, 178n39 194n25
See also liberalism miscount
logos. Seelanguage and Arendtian framework, 47
Lyotard, Jean-François consensus as suppression of, 73

218 1 The Lessons of Rancière


and 120 Rancière's of, 14,
and queer politics, 36-37, 157-68 17-21,44
and subjectivation, 102-4 strong vs. weak, 175n26,
as untimely, 10, 102-4 175n28
Mouffe, Chantal, 11-12,51, 174n17, and subjectivation, 20
194n19 order. See police
Muhle, Maria, 42-43, 49, 70, 183n2,
185n16 Panagia, Davide, 71,154, 174n22,
181·-82n25,187n7
Names of HistOl)~ The (Rancière), 114 paradox
Nancy, Jean-Luc, 51, 53, 61 and democracy, 11-12,59,83,85,
neoliberalism, 172n5 162-64, 166
See also consensus; interests and and hope, 150
interest groups; liberalism and impossibility, 12, 16, 84, 199n8
Nietzsche, Friedrich, 140, 155, 182n26 and impure politics, 49-50, 57,
nonhumans, 34, 93-98, 105-7, 113, 61,85
162-63 and ontology, 21
normative grounds and partage du sensible, 183n3
and anarchism, 80-81, 85, 179n3 and polemicization, 123
and criticai reversaI, 140-41, 139, and untimeliness, 121
194n22 parapolitics, 29, 136, 184n8
and equality, 13,26-27,63-64,80-81, See also political philosophy
85, 150-52 partage du sensible
and Foucault, 44, 179n6, 194n22 and critical dispositif, 125, 153-54,
and Rancière's politics, 63 190n2, 196n35
See also ontology and intelligibility, 70, 199n 13
and language, 95, 96-98, 119
Oksala, Johanna, 175n27 and literarity, 115-16, 169, 195n30
On the Shores of PoUties (Rancière), 74, and subjectivation, 102-·3
75, 183n7 translation of, 70, 183n3, 187n7
See also Aux bords du politique See also intelligibility; police
(Rancière, second edition) parties. See political subjects; poor, the;
ontology subjectivation
and anarchism, 179n3 pedagogy
and anthropocentrism, 96 and Althusser, 171n1
in Butler, 163, 199n9 and critique, 148, 151--52, 172n4,
and critical theory, 36 196n34
and Disagreement, 193n17 and equality, 4, 30-31, 151-52,
and ethics, 175n25 196n34
and Foucault, 175n27 and literarity, 168-69
humanism as, 106 and ontology, 20-21
and language, 35, 90, 96-97, 186n2, radical vs. traditional, 4, 30-31, 151-52
188n16 See also stultification
and police orders, 65-66 Philosopher and His POOl; The (Rancière),
and "the political," 174n17 113, 182n29, 192n8
and political subjects, 181 n21 phfmëllogos distinction, 34, 92-98, 105,
and queer politics, 163 120, 186n3-4

Index 1 21 9
Plato as pure category, 34, 62, 67-68
anthropologism of, 107 and queer politics, 165
on democracy, 59, 74, 135, 163 and subjectivation, 43, 101, 117-21
on language, 113--14, 115-17, 189n21, supplement excluded by, 40-41,
190nn25-26 43, 72
and pedagogy, 196n34 Zizek's interpretation of, 52
as political philosopher, 39, 76-77, See alsa poli tics of the police
135-37, 192n8, 193n18, 194n20 political, the
pluralism. See interests and interest group as doubling, 58-63, 181n22
politics and ontology, 174n 17
polemicization opposed to police and politics in
and critical theory, 124-25, 175n23 Rancière's sense, 52, 180nlO
distinct from argumentation, politics as Rancière's term for, 51
176-77n31 Rancière's use of the term, 53-56,
and intelligibility, 123 59-60
and untimeliness, 191 n6 in Schmitt, 174n 17
police as theme in twentieth-century political
and Althusser, 24, 65, 72, 183n5 thought, 51-52
and anarchism, 7 S.-84 as "third term" in Rancière's thought,
and anthropologism, 107 32-33,40-41,45,50,53-62,
better and worse forms of, 10, 72, 83--85, 180n12, 180n17, 181n20,
200n14 181n22, 181n25
concept as intervention, 73-75, 139, political difference. See Marchart;
183-84n7 political, the
consensus as, 33, 44, 67, 68, 73-75 political philosophy, 39-40, 61, 74,
and critical theory, 154, 183n6, 196n36 76-77,88,127,134-42,179n2,
in Foucault, 42, 70, 76, 183n2 180n7, 182n28, 184n8, 192n8,
genealogy of the term, 183n7 193n18,194n19
as hierarchieal order, 28, 33-34, See alsa arehipolities; parapolities;
42-43,65-66,179n4 metapolities; and specifie authars
irreducible to law enforeement, 42, politieal subjects
70-71 in Arendt, 17
liberalism as, 10-11, 14,29,67, 172n5 and feminism, 160-62, 198n6
and literarity, 88, 113,117-19,121 in Foueauldian aecounts, 17
and May '68, 24 in liberal theory, 16-17, 26, 29, 100,
as neutral term, 72, 83 102--3
and ontology, 65-66 in Marxist theories, 5, 17, 100, 192n8
and "ordinary polities," 14,39,41,68- and ontology, 181n2l
69,174n2l politieal the ory
as partage du sensible, 42,60,70-72, and politicians, 197n3
76, 102-3, 183n6, 183-84n7 Rancière's resistance to, 3--5, 14,32,
and polemicization, 176-77n31 63-64, 182n29
and politieal philosophy, 134-36 reception of Rancière in discipline of,
politics as dismption of, 8-9, 28, 38, 38,41-42,44-57,74
43,86-87 See alsa specifie authars
polities blended with, 49-50, 61-64, Palitical Thaught af Jacques Rancière,
65,80,82,84-85 The (May), 75-84, 184n14-15

220 1 The Lessons of Rancière


Polities (Aristotle), 34-35, 89--98 and intel1igibility, 37, 167-68
"Politics, Identification, Subjectivization" and normativity, 160-61, 165-66, 167-
(Rancière), 54-56, 101 68, 199n7, 199n12
See also translation and ontology, 163
Polities of Aestheties, The (Rancière), and police, 165
180nll, 188n12 Rancière's rejection of, 158
"Politics of Literature, The" (Rancière), and subjectivation, 159-60, 162-63,
189n22 164
politics of the police, 33, 34, 62-64, 65, and torsion, 165
84-87,174nI6
See also police Raw1s, John, 11-12, 14, 17,26--27,29,
poor, the 38,80, 172n5, 177n38, 178n39
in Marxism, 5,17,100, 192n8 Reading Capital (Althusser et al)
as police category, 103, 153, 165 Rancière's contribution to, 22, 36, 107,
and political philosophy, 5, 115, l35, 124, 127, 128-33, 142-43, 176n30,
192n8,193n18 192n12, 195n26
as politieal subject, 9, 17,37, 102-3, translation of, 35-36, 142, 176n30,
150, 153, 165-67 193nl3,195n26
post-democracy. See consensus recognition, 50, 197n2
Post-Foundational Political Thought regime
(Marchart),51-52 democracy not a, 9-10, 12, 14,68, 102,
proletariat. See poor, the 166, 181n23, 185-86n17
purity of polities See a/so police
in anarchism, 34, 40, 45, 67-68, 76, remainder. See impurity of politics;
82-84, 184n12 supplement
in Arendtian framework, 32, 39-40, revolution
45-49,57,172n3 and anarchism, 78,83, 141, 184n13
and concept of equality, 29 and liberalism, 178n39
and dialetics, 182n26 and literarity, 182n28
in politics and police relation, 1° politics not reducible to, 69, 173n9
and "the poor" as a category, 5, 192n8 and reform, 184n 13
and struggle, 182n30 and surprise, 7, 173nll
in "three-term" model, 32-33,40--41, See also event
45,50,53-62,83-85,180nI2, Ricoeur, Paul, 51, 172n4, 182n27
180n17, 181n20, 181n22, 181n25 Rockhill, Gabriel, 33, 180n11, 190n24,
196n37
Queer Nation, 159, 197n2 Ross, Kristin, 5, 8, 23-24, 29, 89, 121,
queer polities 177n32
and Butler, 160-63, 167-68, 198n5,
199nn7-8 Saussure, Ferdinand de, 114, 189n21
as democratic miscount, 36--37, Schaap, Andrew, 39,48
157-68 Schmitt, Carl, 51, 174n17
development of, 198n5 Sedgwick, Eve Kosofsky, 160, 172n4,
distinct from liberai identity politics, 198n5
37,159-68 passim, 197n2, 199n8 Seery, John, 12-13
and feminism, 160-62 sexlgender, 161-62, 163, 167-68
as impure polities, 165-66 Skinner, Quentin, 189n20

Index 1 221
social body/formation/order/whole. See supplement, 40-41, 43,58-63, 72,
police 182n26, 183n3
social science, 6-8, 139, 172n4, 172n6, See also impurity of politics; paradox
173n8,192nlO surprise
Socrates in Arendt, 5, 39, 46, 172n3
and critique, 148, 156, 196n34 and event, 5-8, 9, 173nlO
and political philosophy, 39, 135, 137, and intelligibility, 172-73n6
190n25, 193n18 and liberalism, 6-8, 103, 172n4
speech. See language; phonë/logos and revolution, 7, 173n9, 173n Il
distinction and social science, 6-8, 172n4, 172n6,
Spivak, Gayatri Chakravorty, 199n8 173nlO
state of nature, 6, 16-17,27, 172n5,
177n34 Tanke, Joseph, 71, 171-,-72n2, 180n12,
Strauss, Leo, 39, 186-87n6 188n13
struggle, 5, 29, 81, 83, 85-86, 182n30 teaching. See pedagogy
stultification temporality, 8, 15-17,31-32
and critique, 151-52, 172n4, 196n34 See also event; history; untimeliness
and emancipation, 148, 157, 168 "Ten Theses on Politics" (Rancière), 42,
as interpretive challenge, 30-31, 74,96-97, 183n5, 183n7
171-72nnl-2 See also translation: of "Dix theses sur
and police, 66 la politique"
See also pedagogy Thomson, Alex, 85, 184n11, 184n16
subject. See political subjects; torsion
subj ecti vation as queering, 165
subjectivation as reading practice, 58, 88-89, 95, 125,
in Butler, 99-100, 162, 187nlO 158, 199n11
compared with other terms for subject wrong as, 57-58, 88, 184n12
formation, 98-101 translation
as disidentification, 49, 104, 120, of assujettissement, 89,99-101,
193n16 187nlO
and equality, 29-30, 43 and counter-translation, 113-14
in Foucault, 89, 98-101, 187nlO, of dispositif, 125-26, 195n28
188n11,188n13 of "Dix thèses sur la politique," 53,
and identity, 101, 104, 112, 119-20, 181-82n25
164-65, 193n16, 199n10 history of Rancière's publications in,
andlanguage,34-35,104-5 21-23,41
and literarity, 117-21 ofindice, 95, 187n8
and ontology, 20 of indiquer, 92-95
and partage du sensible, 102-3 of la mésentente (concept), 154
and police orders, 43,101,117-21 of La mésentente (Rancière), 35,
and queer politics, 159-60, 162-63, 53-54,91-95,100-1, 181n25,
164 186n4, 188-89n16
translation of, 89,98-101, 188nlO, of [apolitique, 51-54, 58, 180n16,
188n13 180n9, 181n2~ 181n25
See also political subjects; pOOl', the; of le politique, 51-63, 58, 180n16,
untimeliness 180n9, 181n22, 181n25

222 1 The Lessons of Rancière


as linking, 109 and 117,120-21
of Lire le Capital (Althusser et al), and police, 71
35-36,142, 176n30, 193n13, and Rancière's trajectory, 191n6
195n26 of subjectivation, 8-10, 15-17,43,
of manifester, 92-95 102-4, 111-12, 159--60, 162-65
of partage du sensible, 70, 183n3,
187n7 Valentine, Jeremy, 187n9, 197n2
of phfmë and logos, 92-94, 186n3 Vatter, Miguel, 75, 79-80
of Politics (Aristotle), 90, 92-95, voice. See phonë/logos distinction
186n3-4
of Politics of Aesthetics, The Warner, Michael, 190n 1, 198115, 199n 7
(Rancière), 180n11 White, Stephen, 19, 175n26
of Rancière's works written in English, Wittgenstein, Ludwig, 110-11
54-56,101, 180n13, 180nn15-16, wrong
181n19 and subjectivation, 16-17, 162-64,
of subjectivation, 89, 98-101, 176n29, 193n16
188n10,188n13 as torsion, 53,57-58,88, 184n12
as verification of equality, 193n 16
ultrapolitics, 39, 62, 83, 141
untimeliness Zizek, Slavoj, 39,52,54,62,83, 141,
and intelligibility, 9, 199-200n13 17h11, 180nlO, 184nll, 185n16

Index 1 223

You might also like