pai n ti n g w i t h de mon s

2 painting with dem ons

michael fried
pa i n t i n g
wi t h
dem on s
the art of gerolamo savoldo

reaktion books
for joseph mari oni and frank stella

Published by Reaktion Books Ltd

Unit 32, Waterside
44–48 Wharf Road
London N1 7UX, UK
www.reaktionbooks.co.uk

First published 2021

Copyright © Michael Fried 2021

All rights reserved

No part of this publication may be reproduced, stored in a retrieval system,

or transmitted, in any form or by any means, electronic, mechanical,
photocopying, recording or otherwise, without the prior permission
of the publishers

Printed and bound in China by 1010 Printing International Ltd

A catalogue record for this book is available from the British Library

ISBN 978 1 78914 319 5

Frontispiece Savoldo, Mary Magdalene (illus. 39), detail.

CONTENTS

Introduction 6

pa rt o n e
1 Death of St Peter Martyr 22
2 Hands 44
3 Faces 78
4 Magic 104
5 The Brescia Adoration of the Shepherds 122

pa rt t wo
Breathing the same air 132
Savoldo and the self-portrait 137
Faces, masks, Northern art 142
‘Magic’, ‘inuence’, demons 151
Who, then, was Savoldo? 156
Savoldo and Caravaggio: the inescapable relation 161

Afterword 170

References 172
Acknowledgements 187
Photo Acknowledgements 188
Index 190
 I N T RO D U C T I O N

Wh y a b o o k – not just an article, an entire book – on the sixteenth-

century Italian painter Giovanni Gerolamo Savoldo at this time?
Such a choice may seem surprising. After all, Savoldo is not an artist
with a strong public prole. On the contrary, he has virtually no presence
in general, survey-type accounts of Italian High Renaissance painting,
which understandably dwell on the period’s many major gures – Leonardo,
Michelangelo, Raphael, Correggio, Giovanni Bellini, Mantegna, Giorgione,
Titian, Veronese, Tintoretto and their ilk, all undisputed titans (the list is
hardly exhaustive). Indeed the basic narrative of sixteenth-century art in and
around Savoldo’s native Venice would be essentially the same in his absence.
And yet I think it is fair to say that art historians of the period have long been
aware that his achievement was, to say the least, out of the ordinary and that
his strongest paintings exert a unique force of attraction, even as they would
also acknowledge that his art has never received the concentrated scholarly
and critical attention it deserves.
The present book is a response to this state of affairs. More precisely, it is
a highly personal attempt to engage seriously with Savoldo’s achievement
in the conviction, arrived at over decades, that his is a body of work of sin-
gular distinction and renement (terms that should require no defence even
in our age of crude anti-elitism). For one thing, Savoldo radically reimag-
ines virtually every Christian subject he undertakes, to an extent perhaps
unequalled by any other Italian painter of his time. Thus he depicts a peni-
tent St Jerome performing the traditional action of striking himself on the
1 (facing page) Savoldo,
Portrait of a Man in breast with a stone in a manner different from that of any other treatment
Armour (illus. 20), detail. of the theme; his intensely poetic picturing of Tobias and the Angel is even
8 painting with dem ons

more comprehensively at odds with standard versions of the subject; the

ravishing St Matthew and the Angel puts Caravaggio’s later treatments of the
subject in the shade; and his sublime Magdalenes depict the saint turning to
face the risen Christ, who we are to understand is standing just ‘this’ side of
the picture surface, alongside us, the viewer, an astonishing conception. (In
line with this, Savoldo conceals the Magdalene’s hands under her gleaming
shawl, thereby compelling the viewer to imagine as if in permanent deferral
the gesture with which she will in a moment reach out to touch Christ, as
well as Christ’s response, ‘Noli me tangere’.)
All this barely suggests the imaginative force of many of Savoldo’s
canvases, which are also, it should be said, superbly executed. But the fur-
ther, indeed the deepest interest of his oeuvre for me embraces four issues in
particular.
The rst concerns what has always – or rather, since the pioneering
writing of the great twentieth-century Italian art historian Roberto Longhi
– been seen as Savoldo’s anticipation of signicant aspects of the art of
Caravaggio (for Longhi the decisive gure in the emergence of modern
art). As the author of an ambitious book on Caravaggio (The Moment of
Caravaggio, 2008), I have a large stake in the question of the relationship
between the two painters. In fact I believe the connection is extremely close, for
reasons other than those adduced by Longhi. I will have more to say about
the topic shortly and at greater length towards the close of this book.
A second issue that I nd worth attending to concerns the fact that
self-portraits play a surprisingly large role in Savoldo’s oeuvre, as the late
Creighton Gilbert was the rst to observe. The full signicance of this will
be brought out in the course of what follows, but for a start it supports the
notion of an intimate relationship to Caravaggio, for whom the self-portrait,
explicit and disguised, was nothing less than fundamental (as I demonstrate
in The Moment of Caravaggio).
The third issue that I nd especially compelling turns on Savoldo’s
treatment of hands. Again, the topic was rst broached by Longhi, who
recognized in Savoldo’s depiction of hands one of the keys to his art, though
his observations are, as will be seen, mainly stylistic, having to do with the
presentational logic of Savoldo’s approach. My own line of argument is
different, and here I touch on the principal methodological theme of my
book, namely the need to look at Savoldo’s art empathically, in a spirit of
imaginative identication with the actions of the protagonists, not merely
stylistically, from ‘outside’, as if from a safe distance (‘safe’ in the sense of
 introduction 9

standing beyond the paintings’ various bodily and imaginative solicita-

tions). As I try to show, if the hands are considered in this light, they can be
identied, more or less, with the painter’s left and right hands as he brought
his paintings into being. (In this connection it matters that the hands are
usually shown in intimate proximity to the picture plane.) Such an argument
goes back not only to The Moment of Caravaggio but also to my Courbet’s
Realism (1990), a double afliation I regard as suggesting that the problem-
atic of the inscription of the painter-in-the-act-of-painting that surfaces in
Savoldo’s art turns out to have a rich posterity in the centuries to follow.
(A fascinating canvas by Savoldo, Portrait of a Man in Armour, recognized by
Gilbert as a self-portrait, is crucial in this regard.)
A fourth issue differs in kind from the rst three. Roughly halfway through
my research into Savoldo I became aware, to my considerable surprise, that a
number of his paintings contain images of grotesque or monstrous-seeming
heads and faces as if ‘hidden’ in rocks or, more typically, in the folds, creases,
bulges and depressions of his unusually free-form, often intensely coloris-
tic drapery. (In the previous literature Savoldo’s treatment of drapery has
often been singled out for comment but never on these grounds.) Eventually,
again to my surprise, I came to understand these often grimacing or snarl-
ing physiognomies as essentially ‘demonic’, an identication that will require
extensive discussion further on in this book. Two paintings in which such
heads and faces not only appear but openly declare themselves are distinctly
Northern, more precisely Boschian versions of the Temptation of St Anthony,
which is hardly surprising given the nearness of Venice to the North and in
view also of the presence of works by Bosch in Venetian collections. At this
point the question arises as to whether all or most of the grotesque faces
that appear in Savoldo’s canvases were placed there deliberately by the artist,
and I frankly admit that this turns out to be an extremely difcult question
to resolve. In certain instances, such as the St Anthony panels, the faces were
plainly intended to be seen (they were, so to speak, called for by the subject).
But there are others, for example in the shawls of the Magdalenes or in the
sleeve of the kneeling Joseph in a late Adoration of the Shepherds, that seem
not only anomalous with respect to subject matter but, at a time and place
of religious crisis and widespread pursuit of heresy, extremely dangerous for
the painter. Let me reserve further consideration of this deeply interesting
question for its proper place later in this book.
So much by way of indicating some of the reasons why Savoldo seems to
me to deserve, indeed to require, a serious account of his art. To begin:
10 painting with dem ons

G i ova n n i g e ro la m o s avo l d o , one of the subtlest and most intriguing –

also, I shall suggest, one of the most original and profound – painters of the
Italian Renaissance, most probably was born in Brescia, an important city
in the general dependency of Venice, in the rst half of the 1480s, although
no hard evidence exists as to either his place or date of birth. Nor do we
know when he died, an event usually placed in the late 1540s, by which time
he was referred to by Pietro Aretino in a letter as an ‘excellent old man’
(vecchioni ottimo).1 Indeed there are relatively few documents around which
to construct a reliable outline of his career: among other facts we know that
he was in Parma in 1506 (indicating that he was a practising painter by that
date), and a letter of 1508 from Pietro d’Argenta to Michelangelo’s brother
suggests that he may have been in Florence in 1508, if the letter refers to him
and not to another Brescian painter with a similar name (it makes sense to
think of him as having spent time there). Certain scholars have also posited
an early visit to Rome. For the most part, however, Savoldo seems to have
resided in Venice, in the 1520s and ’30s one of the main centres of ambitious
painting in Italy. There he received commissions for altarpieces in towns
such as Treviso, Pesaro, Verona and Brescia as well as, one surmises, for
the independent paintings on a smaller scale that constitute by far the bulk
of his surviving oeuvre. (The epithet ‘independent’ is a place-marker for a
discussion of the character of those paintings further on in this book. As
will eventually become clear, I am deliberately avoiding what may seem the
obvious descriptive epithet, ‘devotional’.) His former student Paolo Pino, in
an important treatise on painting, Dialogo di pittura (1548), remarks that he
worked for Francesco Sforza, Duke of Milan (probably in the early 1530s),2
while Giorgio Vasari in brief remarks in his Lives of the Painters (1568) reports
having seen ‘four paintings of night and re’ by Savoldo in the Milan mint.3
Three closely related late commissions were for an Adoration of the Shepherds
for the Bargnani chapel in the church of San Barnaba in Venice, today in the
Pinacoteca Tosio Martinengo in Brescia, a work of considerable interest, and
two less impressive variants for San Giobbe in Venice and Santa Maria la
Nova in Terlizzi (all three datable around 1540).
One of the salient facts about Savoldo’s life and art is that in the course
of a long career he produced only a limited body of work; the contrast in
this regard with his Venetian-native contemporary, the prodigious Lorenzo
Lotto, could hardly be more striking. (Lotto, who like Savoldo made his
career largely outside Venice, will be a recurrent term of comparison in
this book. For all the afnities and exchange of inuence between them, the
 introduction 11

differences are even more telling.) Francesco Frangi’s catalogue raisonné

of 1992 lists no more than 47 canvases, some of which are near-repetitions
of previous works, though it should be noted that one of Savoldo’s most
compelling creations, the Death of St Peter Martyr, came to light only in
2001. Another painting, a Crucifixion, very likely the earliest of his surviving
works, today in Monte Carlo, was given to Savoldo two years before.4 Pino
in his treatise also remarks that Savoldo was underappreciated in his lifetime
(‘He has spent his life on few works, and with scant esteem to his name’
[D, p. 304]), and in general one has the impression of a career conducted not
precisely in a minor key but in terms that were strangely personal given the
norms of the age. Or perhaps it would be more accurate to say that Savoldo’s
most acute artistic concerns were not such as to be fully intelligible to his
contemporaries, or for that matter to his modern commentators (I shall try
to make good on this claim in the pages to come). In any case, we are deal-
ing with a restricted corpus, with a handful of modest-sized independent
pictures mainly of religious subjects. Signicantly, the large altarpieces that
survive, such as those alluded to above, are unremarkable, all the more so in
comparison with Lotto’s prolic inventiveness in that domain.
The relative smallness of his oeuvre notwithstanding, starting in the
early twentieth century Savoldo has attracted the attention of some of the
leading art historians of Italian painting, among them Roberto Longhi,
Adolfo and Lionello Venturi, Creighton Gilbert, Mina Gregori and Keith
Christiansen. Gilbert, in fact, made Savoldo his speciality, devoting to the
painter a strongly researched and closely argued doctoral dissertation
(1955), a work that remains the most sustained scholarly engagement with
Savoldo’s art to date. Not surprisingly, given the paucity of documentary
evidence, a number of other writers, Antonio Boschetto prominent among
them, have taken issue both with Gilbert’s proposals for dating and with
his larger sense of Savoldo’s achievement.5 (More broadly, the chronology
of Savoldo’s oeuvre is an area of considerable dispute; the dates given in this
book for individual works represent my best assessment of the arguments of
previous scholars, but we are on uncertain ground.) In addition there took
place in Brescia and Frankfurt in 1990 a major exhibition, Giovanni Gerolamo
Savoldo tra Foppa, Giorgione e Caravaggio, accompanied by an impressive
catalogue with essays by prominent scholars including Gilbert, Frangi,
Gregori and Sybille Ebert-Schifferer.6 No exhibition since that time com-
pares with it for comprehensiveness, though Savoldo, treated as one of the
key precursors of Caravaggio, is discussed both by Keith Christiansen in
12 painting with dem ons

the catalogue for the 1985 exhibition The Age of Caravaggio7 and by Andrea
Bayer in that for another ambitious exhibition, Painters of Reality: The Legacy
of Leonardo and Caravaggio in Lombardy, at the Metropolitan Museum of Art
in 2004.8 (The latter deliberately echoes the name of an important exhibi-
tion, I pittori della realtà in Lombardia, mounted in Milan just over fty years
previously.) Finally, a recent exhibition, Titian e la pittura del cinquecento tra
Venezia e Brescia, at the Museo di Santa Giulia in Brescia in 2018, included
no fewer than eight canvases by Savoldo accompanied by scholarly catalogue
entries for them by Frangi and others.9
As such titles suggest, the overriding tendency in recent scholarship has
been to treat Savoldo situationally, in the rst place as a predecessor to Cara-
vaggio and in the second as a signicant gure in an essentially naturalistic
Lombard tradition, Lombardy being loosely dened historically as includ-
ing Milan, Brescia, Bergamo, Cremona and Mantua (whose courtly culture
set it somewhat aside from larger developments) as well as, stretching its
boundaries, implicating Correggio’s Parma and even Bologna. (The Lom-
bard roots of the Carracci, natives of Bologna, have been a leading theme in
modern scholarship.10) As already mentioned, the by now deeply entrenched
notion that Savoldo looks forward to Caravaggio goes back to the most char-
ismatic presence in Italian art history, Roberto Longhi, who in a series of
inuential articles, notably ‘Quesiti caravaggeschi, II: I precedenti’ (1929),
argued powerfully if succinctly that Savoldo’s treatment of light, along with
the underlying ‘empiricism’ of his art, belonged to a distinctly Lombard tra-
dition that reached its revolutionary climax in Caravaggio’s paintings of the
1590s and the rst decade of the seventeenth century.11 In greater detail,
Longhi contended that there existed a distinct school of Brescia, originat-
ing with Vincenzo Foppa (1430–1515) and giving an honoured place to the
robust sixteenth-century master Alessandro Buonvicino, known as Moretto
(1498–1554), born roughly a generation after Savoldo and a dominant pres-
ence there throughout his career. Naturally enough, subsequent scholars
have modied these views, emphasizing, for example, the likely inuence on
Savoldo of Leonardo da Vinci owing to the latter’s years in Milan; Savoldo’s
growing involvement with Venetian art in the course of his decades-long
residence in that city (emphasis falling on both Giorgione and Titian, among
others); his sometime closeness to Lotto; and his evident engagement with
Northern art, for example that of Bosch, Cranach, Patinir, Dürer, Schon-
gauer and the Flemish generally, in various of his works, most explicitly
in the early Temptations of St Anthony in San Diego and Moscow (assuming
 introduction 13

that both are early) as well as in the Crucifixion in Monte Carlo. In addition,
certain scholars, notably Gilbert, have called into question the sharpness of
the Lombardy-Venice divide basic to Longhi’s arguments – possibly no one
would argue for so strong a divide today. Nevertheless, the Lombard connec-
tion, which is also to say Longhi’s inuence, continues to dominate Savoldo
commentary, as in the exhibition catalogues mentioned above or, to take
a concrete passage of criticism, in the following observations by Sydney J.
Freedberg in his masterly Painting in Italy, 1500–1600, apropos of two early
canvases to be discussed in Chapter Three, Savoldo’s Elijah Fed by Ravens
and SS Paul and Anthony (see illus. 43 and 44):

These works, preceding Savoldo’s likely time of settling in Venice and

following soon upon it, are clearly products of a temperament already
formed. Longer and deeper experience of Venetian art would come to
make important differences in Savoldo’s style, without, however, effecting
an essential change in his attitude towards art. That had been generated
from Savoldo’s Lombard roots, conceiving the rst function of art to be
the reproduction of an ordinary reality; and in this respect Savoldo’s early
paintings represent the fullment of his previous tradition. (PI, p. 225)

This is, one might say, the ground level of the prevailing account of
Savoldo’s style. Freedberg adds to it in two further stages, both of which
would receive widespread assent from commentators to this day, at least up
to a point. First he writes:

To a simplistic and popular descriptive mode he added a renement of

light-searched details that approximates the effects of a Flemish style,
and must be learned in part from it. The human image takes on the
character of still life, but one to be described not just in detail but with
a generalizing and including power of whole plastic form. These large
plastic elements are structurally simple, and they are combined in designs
in which an additive mentality like that of the late Quattrocento détente
style remains dominant. Seen outwardly as still life, and inwardly with
reticent emotion, Savoldo’s persons do not conform to a principle of
modern classicism, and cannot achieve the mobile coexistences we expect
of Cinquecento style. (PI, pp. 225–6)

And then, continuing the same intense paragraph:

Yet in these cautious structures and in this atmosphere of emotional

reserve there is a factor that intensely charges both: the same light that
14 painting with dem ons

in one aspect of its functioning nds out detail in another suffuses sur-
faces with arbitrary strength, reecting brilliances of colour that exceed
our expectation of visual truth. These brilliances, lucent and commanding,
make an effect of illusion still stronger than that of the underlying form,
but they transcend illusion, exalting it beyond verity into an effect of art.
The power of this surcharged colour to transmute descriptions into poetic
meaning is the more compelling from the very contrast of its mutism in
design and emotion that accompanies it. (PI, p. 226)

This is distinguished art writing by any standard. Let me draw out sev-
eral points, starting by observing that Freedberg’s description detects in
Savoldo’s manner a distinct note of archaism in contrast to the uency of
multi-gure relations that Freedberg associates with what he calls classi-
cal style, as practised in contemporary Venice by Titian among others. The
emphasis in Savoldo, in other words, falls strongly on individual gures,
usually no more than a very few in a single canvas, gures who, as Gilbert
and others remark, tend to loom with a certain massiveness and monumen-
tality in the foreground of his compositions. Gilbert applied this observation
to SS Paul and Anthony: ‘The isolated dramatic gure . . . is Savoldo’s concen-
trated theme. The hermits are the rst examples: solid enough to impress by
reality, pensive enough to impress by seriousness’ (G, p. 312).
For Freedberg, as for other commentators, Savoldo’s autograph achieve-
ment concerns the treatment of light and in particular of ‘surcharged’ colour
as revealed by light. This is not yet in play in the Elijah or SS Paul and
Anthony but it is gloriously on view in numerous other works, such as, to
take just one example, the Death of St Peter Martyr, in which the padded crim-
son doublet of Peter’s assassin possesses a coloristic luminosity that almost
seems at odds with the subject: at the very least, both the vividness of hue
and the virtuosic rendering of the doublet’s silken texture and many folds
and indentations attract the eye in a manner that may be felt to consort oddly
with the ostensible violence of the main action. In particular it draws atten-
tion away from the assassin’s face, which in any case is surprisingly impas-
sive, not to say withdrawn. This is not a criticism: the interplay between the
two is gripping, as is the contrast between the dazzling white and black of
Peter’s Dominican dress and the intense red of the doublet.
Freedberg twice suggests that the human gure in Savoldo’s art has
somewhat the character of still life. This is to go rather far – I think too far
– in the direction of stressing the naturalistic object-character of the latter’s
 introduction 15

persons. (Gilbert, only somewhat in contrast, notes a tendency ‘to make the
single gure separate, monumental and compact, to internalize its massive-
ness – to make it portrait-like’ [G, p. 323]. Interestingly, though, Savoldo is
not particularly distinguished as a portraitist – again, the comparison with
Lotto, one of the nest portraitists of the century, is telling.) At the same
time, Freedberg detects a characteristic ‘air of emotional reserve’ and ‘reti-
cence of emotion’, terms that at once shift the discussion into an affective
register, and goes on to note a ‘mutism in design and emotion’ – in short, an
instinctive restraint or holding back – which he sees as contrasting with the
poetic intensity of the light-struck colour and in effect throwing that inten-
sity into even greater relief.
Gilbert, for his part, points to ‘the emotional concentration of Savoldo’s
gures, their quality of reacting with easy bruises to their physical and
mental world’ (G, p. 343) – not a simple remark to cash in terms of the
paintings themselves. But in a later publication, an essay in Italian in the
Savoldo tra Foppa, Giorgione e Caravaggio catalogue, Gilbert goes consider-
ably further. After briey summarizing Savoldo’s preference for construct-
ing gures of impressive density and large bodily architecture (Gilbert’s
term), bearing concentrated expressions and almost always immobile and
isolated, he writes:

Thus was born a gallery of imposing personages who seem to incar-

nate moral and aristocratic stability and an inclination to the intimacy
of thoughts. It’s a matter of solitary personages, sometimes also heroic,
from whom is banished every trace of academicism, who very often bring
with them a sense of artistic heroism, but who rather are interpreted with
a poetic sensibility and a capacity for emotional depth which invites one
to participate in their life . . . They are denitively personages linked to
one another by a profound interior afnity, like those one sometimes
encounters in the work of a great novelist, and yet we cannot say that
that afnity is a matter of literary relations: their characters are dened
uniquely by visual means, in part by gestures, expressions, and costumes,
but also by considerations of form in the most abstract sense of the word.12

I know of no comparable passage elsewhere in the Savoldo literature, but

my impression is that it captures a prevailing intuition that there is some-
thing out of the ordinary about many of Savoldo’s personages and that
that something lies within the realm of subjectivity, that is, in the ‘interior
afnity’ among them, but also that it is not at all easy to specify or otherwise
16 painting with dem ons

get at, largely because the personages’ outward gestures and expressions are
such as to defeat easy verbalization. Note in this connection that Freedberg
imagines a sort of double encounter with Savoldo’s protagonists, whom he
describes as ‘seen outwardly as still life, and inwardly with reticent emotion’
– another remark that is difcult to cash, in this case because it is not quite
clear what ‘seen inwardly’ amounts to. Whose ‘inwardness’ is at stake here,
exactly?
Let me come to my point: the chief limitation of the Savoldo literature
down to the present moment is its preoccupation with style. Gilbert on
Savoldo’s personages is an exception to this, but otherwise stylistic cate-
gories are basic to his approach. An emphasis on style is explicit in Freed-
berg’s summary, which is hardly surprising in view of the fact that his entire
book, a sustained tour de force of art-historical analysis and exposition, is an
attempt to provide a stylistic overview of the decisive century of Italian, and
in a sense of European, painting. Its opening sentences read:

The artistic events that most powerfully determined the history of

sixteenth-century painting took place in the century’s rst two decades
in Florence and Rome, in the time which, implicitly recognizing the stat-
ure of its achievement, we have come to call the High Renaissance. The
most extraordinary intersection of genius art history has known occurred
then and gave form to a style which, again eliciting a term that is a value
judgment, we call ‘classic’ or ‘classical’ – meaning, in its original usage, ‘of
the highest class’. (PI, p. 1)

Freedberg, of course, is today regarded as an unabashed formalist, and there

are few if any practising art historians who would foreground the concept of
style to that extent. (Gilbert, writing in the 1950s, did not hesitate to title a
major section of his dissertation ‘Savoldo’s World of Style’.) But in fact very
nearly the entire secondary literature on Savoldo, as I read it, shares a focus
on style, by which I mean that its concerns are largely taxonomic, deploying
concepts of classical style, archaizing tendencies, Lombard tradition, still life
or portrait-like gures, Leonardesque atmospherics, Giorgionesque visual
poetics, Northern and/or Venetian inuence and so on, and that such con-
cepts mainly refer to surface or external, in any case immediately apprehen-
sible features of his pictures; these are features the nature of which, and to
a large extent their signicance, are taken as self-evident, at any rate to an
informed eye. Jaś Elsner, in attempting to dene ‘style’, writes:
 introduction 17

The basic stylistic reex, then, is the grouping of like with like and the
disjunction of unlikes, on the basis of morphological or formal analysis
(which may amount to the swift taxonomic action of an expert eye, pass-
ing over the material and registering its relationships with a body of other
material stored in the mind).13

So consistent an emphasis on stylistic considerations is somewhat sur-

prising. Elsner also remarks, and as I have just suggested, that the concept
of style, which for decades reigned supreme in art-historical discourse, has
been in eclipse for some time in the face of newer emphases and method-
ologies (Freedberg’s writing is routinely discounted for this reason). But
Savoldo studies, nearly without exception, continue to pursue their interpre-
tively modest ends in a stylistic vein, no doubt because no one has yet come
up with alternative terms of criticism for making sustained sense of his art.
One consequence has been that even as Savoldo’s artistic stature has come
to be acknowledged by scholars and he has featured importantly in exhi-
bitions such as those mentioned above, our understanding of his achieve-
ment may be said to have reached a virtual dead end (no signicant advances
for decades, except for a growing sense of his involvement with Northern
precedents).14 And not only that: despite that wider acknowledgement, and
in effect going along with it, insufciently close attention has been paid to
individual works as if on the assumption, plausible from a stylistic optique,
that most or all of what is worth saying about them has already been said.
As might be imagined, I take strong issue with such a view and the present
book is my attempt to intervene against the status quo as decisively as I can.
Whether the reader will be persuaded by what I have to say remains to be
seen, but at least it will depart from the all too familiar pathways that have
led to the present impasse.
By way of bringing these introductory remarks almost to a close, let
me give one brief example of a major painting the very subject of which
has not been recognized: the superb and poetic Tobias and the Angel in the
Borghese Gallery (mid-1520s; illus. 2). Probably no painting by Savoldo is
today more admired by scholars and curators who concern themselves with
his work. And yet to the best of my knowledge it has never been observed
(until recently, by the present writer) that the action depicted in the painting
consists not, as is always claimed, in the angel Raphael pointing to the sh
emerging from the river in the lower left corner of the canvas, but rather in
the angel’s drawing or summoning the sh out of the water with his extended
18 painting with dem ons

2 Savoldo, Tobias and arm and hand (also of course with his concentrated gaze). There is no prec-
the Angel, mid-1520s, edent for this ‘magnetic’ or quasi-‘magical’ operation in the book of Tobit or
oil on canvas.
indeed, again to the best of my knowledge, in previous art.15 I shall return to
Tobias and the Angel in Chapter Four, in fact the chapter will begin with an
extended encounter with that canvas, but I hope that my brief remarks about
it here will make it seem more plausible to suggest that much, indeed a great
deal, remains to be discovered about Savoldo’s pictorial project, in the rst
place by engaging with his paintings with a new intensity of looking as well
as, equally important, with a more open and empathic sense of imaginative
possibility.
What follows is organized in two unequal parts. Part One (the longer
of the two) comprises ve chapters starting with a close reading of a single
 introduction 19

painting, going on to cover a set of basic themes and issues, and conclud-
ing with a chapter given over to another single work (plus two variants).
Part Two, in unnumbered sections of varied length, goes on from the rst,
returning to those themes and issues by way of amplifying and deepen-
ing them, in part by focusing anew on particular works and motifs, in part
by invoking contextual factors that can be shown to bear intimately on
Savoldo’s art. My approach, in other words, will not be chronological or
indeed systematic in any way. Rather, I shall proceed intuitively, dealing with
works and topics as they present themselves to me, with the aim nally of
doing at least minimal justice to as artistically and intellectually challenging
a body of painting as any of its time.
part o n e
 one • D E AT H O F S T P E T E R M A RT Y R

I n 2 0 0 1 a h i t h e rto unknown Savoldo of extremely high quality, the

Death of St Peter Martyr, surfaced on the art market and was soon purchased
by the Chicago Art Institute, where it hangs today (early 1530s; illus. 8).1
Some background: the historical Peter was a thirteenth-century Dominican
friar and preacher born in Verona and active in Lombardy, often preaching
against heresy (a current concern for the Dominican order in the 1520s and
’30s), who was murdered along with a companion friar in 1252. The decisive
blow was from an axe to his head, and in some of the most famous depictions
of the saint, such as Lorenzo Lotto’s Madonna and Child with St Peter Martyr
(1503; illus. 7), or the same master’s portrait of a Dominican friar as St Peter
Martyr (1548; see illus. 15), the weapon, usually a heavy knife or sword, is
shown embedded in the saint’s skull (it is, in effect, his saintly attribute).
(There will be more on the latter canvas towards the end of this chapter.)
Savoldo’s painting, in contrast, depicts the murderous act, but does so in
a manner that at once links it with other such depictions in contemporary
Venice and its environs and distinguishes it from those depictions in ways
that I shall try to show are revelatory of Savoldo’s unique pictorial vision.
The crucial comparison, as was recognized from the rst, is with Titian’s
monumental altarpiece painted for the basilica of SS Giovanni e Paolo in
3 (previous page) Venice, a work of stupendous dramatic and expressive force (1529; illus. 5).
Savoldo, Portrait of a The original was lost in a re in 1867 and a replacement hangs today in its
Young Man with Flute place in the Chapel of the Rosary (illus. 6). Here is Freedberg on Titian’s
(illus. 37), detail.
painting:
4 (facing page) Savoldo,
Death of St Peter Martyr At an undetermined date between 1525 and 1527 a competition was
(illus. 8), detail. held for an altar painting of the Death of St. Peter Martyr in SS Giovanni
5 (above left) Martino
Rota after Titian,
Martyrdom of St Peter,
c. 1560, engraving.

6 (above right) Carlo

Loth after Titian,
Martyrdom of St Peter,
1691, oil on canvas.

7 (right) Lorenzo Lotto,

Madonna and Child with
St Peter Martyr, 1503,
oil on panel.
 death of st peter martyr 25

e Paolo: Titian won the contest, and in 1528–30 executed the painting 8 Savoldo, Death of
(now lost, and replaced by a copy). Titian’s rivals were Palma [Vecchio] St Peter Martyr, early
1530s, oil on canvas.
and Pordenone [a ‘provincial’ painter of formidable inventiveness and
rude strength]. Palma was no challenge, but Pordenone was, and in terms
that presented to Titian the same problem of a plastic style that he had
faced recurrently since his indirect encounter with the Laocoon and
Michelangelo [indirect because he had not actually visited Florence or
Rome]. In addition, Pordenone was the exponent of the most extreme
dramatic manner in contemporary art. Apparently thinking of Venetian
taste Pordenone moderated his accustomed violence in his design; but
apparently thinking of his competitor, Titian conceived the most radical
26 painting with dem ons

invention he could make, exceeding Pordenone’s precedents in dramatic

force, in urgency of action, and in assertion of an energy of forms expand-
ing in their space. Articulate in form and in expression as Pordenone could
not be, and far surpassing him in his descriptive means, in this picture
Titian generated a vastly higher and more trenchant power. The design in
which the gures act conveys the sense of an explosion: behind the gures
tree-trunks act almost anthropomorphically to extend and elaborate this
effect. The violence of emotion, the intensity of action, and the expansive-
ness of pattern are in a degree which, at least as much as in Pordenone’s or
Correggio’s extreme inventions, suggests the temper of a baroque style.
Yet, more evidently than in them, that end pertains to classicism. The
mode of feeling of the gures is rened towards typology, and their actions
have the structured grace of classical dramatic repertory; their ordering is
in a scheme of counterpoise. In the setting, the explosive impetus of design
is tempered as it rises, then muted, then diffused into a lyrical and tragic
light. (PI, pp. 215–16)2

By now Freedberg’s concern with classical norms should be unsurprising,

but what I want to stress is that his account of Titian’s achievement in his
Death of St Peter Martyr makes much of the competitive framework in which
it was conceived, and of course Savoldo’s canvas, which scholars date to the
early 1530s, must be understood not exactly as in competition with Titian’s
dramatic masterpiece (Savoldo would have been perfectly aware that Titian,
whom he seems to have admired greatly, could not be beaten on that ground)
but nevertheless as offering a radically alternative vision of its subject.
Instead of a towering vertically oriented image, such as Freedberg describes,
Savoldo gives us a much more modest-scaled depiction (115 cm high by 141
cm wide), with two three-quarter-length gures, Peter and his murderer,
in the near foreground and two much smaller ones, Peter’s companion and
his killer, further back in space to the right. The main action consists in the
assassin about to launch a backhanded blow with his dagger, which is to say
that his body is turned partly away from Peter (it is sometimes said that he
turns his back to him, which does not seem right), in such a way that, as was
mentioned in the Introduction, emphasis falls on his crimson doublet and
its folds and creases rather than on his partly obscured and shadowed face.
The backhanded blow is surprising: in sixteenth- and seventeenth-century
paintings it usually precedes an attempt at decapitation with a large sword,
obviously not the case here; the dagger with its curious point seems an odd
 death of st peter martyr 27

weapon to be employed in such a manner. Nor is it entirely clear where the

blow is intended to strike: Peter’s body seems a more likely target than his
head.
As for Peter himself, the painting’s protagonist, he has fallen to his knees –
the killer looms above him – and looks up and to his left, a rapt expression on
his face, while as if reexively raising his left hand before him with its palm
facing the viewer. Signicantly, but it is never mentioned, the hand, exqui-
sitely delineated and painted, occupies almost the very centre of the canvas.
At the same time, Peter’s brilliantly foreshortened right hand, at the lower
left, makes a kind of reexive gesture that conveys a sense of momentariness
and surprise (more will need to be said about both hands shortly, as well as
about the nature and direction of Peter’s gaze). Also at the lower left is a rock
on which are inscribed the letters CR, standing for ‘Credo a Dio’ (impossible
to make out in an illustration), in a reference to the legend of Peter’s dying
act of writing the opening words of the Nicene Creed in his own blood. The
killer’s left hand grips the hilt of a sword still at his side.
The setting is characteristic of Savoldo’s art in that it features a sharp
division between a distant landscape (at the left) and nearer elements – often
rocks but here a stand of trees (at the centre and right). The double murder
takes place in broad sunlight, which casts denite shadows but mainly illu-
minates both the killer’s doublet and Peter’s simple but splendid white robe
or surplice and copious sleeves. Mina Gregori in a short article marking the
painting’s exhibition at a New York art gallery in 2001 writes that

as the light irregularly strikes the man’s doublet it bathes that splendid
surface in sunshine, producing an apparent intensication of color and a
sparkling quality that lends the painting its most sublime note. The calm
placement of light in the friar’s habit and the assassin’s shirt captures the
distinction between the various hues of white according to the way light
falls, and between the luminous and penumbral areas of the picture. (p. 77)

This is acutely observed, but it is characteristic of Gregori’s stylistic pri-

orities that no comparable closeness of observation is directed towards, for
example, Peter’s facial expression.
As Gregori remarks in the same article, another version of the Death of
St Peter Martyr by Palma Vecchio (c. 1529; illus. 9), in the parish church
in Alzano Lombardo, a small town near Bergamo, was clearly present in
Savoldo’s mind when he composed his painting (pp. 75–6). Palma’s work is
a high altarpiece, and Gregori suggests, on the strength of the background
28 painting with dem ons

9 Palma Vecchio, trees, that it probably dates from roughly the moment of the competition,
Martyrdom of St Peter of with which, however, it had nothing to do. In any case, the closeness of the
Verona, c. 1529, oil on
panel. saint’s pose to that of Savoldo’s protagonist is self-evident: Palma’s saint is
on his knees; his right hand points to the famous sentence on the ground,
which he has just written in his own blood; his left hand is very nearly in the
same position as in the Savoldo; and he looks up, presumably towards the
spectacular heavenly display of God the Father and a virtual ring of angels,
one of which, the lowest one at the right, bears the palm branch of martyr-
dom. In line with other depictions of the same subject, a shortish sword is
 death of st peter martyr 29

embedded in Peter’s head, while one of the assassins draws a long sword,
presumably to complete the act of killing.
Obviously, though, the effect of Savoldo’s painting is altogether different
from that of Palma’s, for several mutually reinforcing reasons. First, it radi-
cally undoes the distance between the viewer and the scene of martyrdom:
whereas in Palma’s altarpiece the viewer is imagined as sufciently distant
to see all the gures at full length, Savoldo has moved in for the equivalent of
a close-up (as mentioned, the saint and his killer are depicted three-quarter
length) and the viewer is left to conclude that the saint is on his knees simply
by his position relative to that of his killer.
Second, Savoldo has done away with the sense of crowding and tumult in
Palma’s painting by eliminating the entire heavenly host and concentrating
attention almost exclusively on the two principal gures. In Palma’s can-
vas, in contrast, not one but two assassins directly menace Peter, and the
latter’s eeing companion, only somewhat more distant than he, deects the
viewer’s attention to the right, further lessening the already diffuse effect of
the whole.
Third, as Gregori emphasizes, the close-up framing of the scene allows
Savoldo to pursue his autograph luministic and coloristic ends (never more
brilliantly than here), while at the same time bringing into the sharpest
imaginable focus, thereby granting a wholly new signicance to, in the rst
place, Peter’s facial expression, and in the second his upraised left hand.
Note, by the way, the magnitude of the difference in these regards from
Titian’s altarpiece, in which the dramatic energy of the killing and of Peter’s
companion’s ight and backwards look takes pride of place at the expense of
any but the most ‘typological’ (Freedberg’s term) or indeed ‘theatrical’ (in a
non-pejorative sense) physiognomic and gestural expressiveness; to speak
of all the gures’ actions and gestures as belonging to a classical dramatic
repertory, as Freedberg does, seems exactly right. Savoldo’s painting has
none of this. In fact it could not be more deliberately removed from all such
considerations (from all trace of emphatic rhetoric), a point that contem-
porary viewers could not but have recognized, whether in approval or not.
Put slightly differently, in his Death of St Peter Martyr, knowing himself to
be working in the immediate aftermath of Titian’s titanic achievement, and
having gleaned useful intimations from Palma’s much less powerful but by
no means uninteresting altarpiece, Savoldo found the resources to make a
painting that stands out absolutely among sixteenth-century treatments of
the subject – my task now being to say exactly why and how.
30 painting with dem ons

10 Savoldo, Drawing of To begin with the saint’s face and expression: as Gregori noted, these
a Man’s Head, 1530–35. are closely based on a drawing of a man’s head in Warsaw (early 1530s?;
illus. 10), one of a number of extremely impressive charcoal or pencil draw-
ings by Savoldo, all but one portrait studies of models, all but one of whom
are male. In this case Savoldo had evidently already decided on the basics
of his composition, and had his model – dark-bearded, with rened fea-
tures, seemingly in his thirties – assume the position of the saint’s head and
the direction of his gaze (towards the upper right, the model’s left). Thus
Gregori: ‘In every respect . . . the correspondences are perfect, conrming
the naturalism of Savoldo’s creative process’ (p. 76). But in the rst place
 death of st peter martyr 31

the correspondences aren’t quite perfect, and in the second to speak of natu-
ralism in this connection, while making perfect sense as far as it goes, falls
short of acknowledging the creative thought that went into the conception
of the painting before the artist rst posed his model and began to draw his
features. Gregori knows this, indeed no contemporary scholar knows more
about sixteenth- and seventeenth-century Italian painting and how it came
into being, but once again a prioritizing of stylistic considerations rules out
any intimate imagining of the creative process.
To go beyond such considerations, let us ask a simple question – at what
is Peter gazing? Is the viewer being invited to think of him (to see him) as
gazing at anything in particular? Or is the artist’s point simply that Peter is
in a sort of ecstatic trance, ‘immersed in thoughts of heaven’, as Gilbert puts
it in a short article on the painting?3 My strong intuition, my conviction, is
that the artist intended the saint to be seen as gazing at one or more angels
bringing down to him the palm branch of martyrdom, the traditional Chris-
tian symbol of the triumph of the spirit over the esh. That is why Peter
seems virtually to ignore the assassin, as Gilbert puts it. In this connection it
is obviously not irrelevant that angels bearing a palm branch appear in both
Titian’s and Palma Vecchio’s altarpieces, where in a sense we scarcely notice
them; what Savoldo has done is to confer a wholly new measure of impor-
tance, of virtual ‘presence’, on such an angelic emissary (the compositional
economy of Savoldo’s painting suggesting that there was just one) precisely
by locating angel and palm branch outside (beyond, above) the representa-
tional, and in effect the physical, limits of the canvas, in the direction of the
saint’s enthralled but also, it seems to me, surprised, even astounded, gaze.
To go a bit farther, one might almost say that the main action of the
painting, the action that determines its affective character, is the descent
of the angel, with an extended arm proffering the palm branch. Of course
the assassin is wholly unaware of any of this, which gives added point to
the partial turning of his body away from the saint and also to his con-
tained expression, both artful strokes on Savoldo’s part. Of course, too, the
viewer, while being given all the information needed to understand exactly
what is going on – assuming that he or she is familiar with the convention
of palm branch-bearing angels in paintings of martyrdom, as would beyond
a doubt have been the case for Catholic viewers about 1530 – is at the same
time offered a ravishing distraction in the radiant, light-struck treatment
of the assassin’s gorgeous doublet, as if to underscore the difference, if not
indeed the disparity, between the world of ordinary experience, including
32 painting with dem ons

the experience of painting, and the divine world made momentarily ‘pres-
ent’ by the fact of martyrdom. Scholarly responses to the painting testify to
the appeal of the offer, to the extent that the descent of the angel has been
altogether ignored.
All this makes Savoldo’s canvas a work of the highest fascination. But
there is more to be said about Peter’s gaze, namely that it implies that the
palm branch-bearing angel is to be imagined as located not only above the
depicted scene, which is also to say above the upper framing edge, but also,
even more radically, ‘this’ side of the picture plane – as if in the ‘real’ space
between the painting and the viewer. (I take this to be self-evident, if one
follows the painting’s cues.) As will become clear in the course of this book,
we are not dealing here with an isolated and exceptional feat of compo-
sitional daring. On the contrary, the Death of St Peter Martyr is a particu-
larly gripping example of one of the most profoundly original features of
Savoldo’s art: its implied activation, in certain of his most compelling inven-
tions, of the space before the painting, together with its frequent evocation
of the proximate presence of the viewer, in a manner that goes far beyond
the norms of sixteenth-century painting in the direction of some of the basic
innovations of Caravaggio and his successors around 1600 and after. In this
particular case, the effect, once we allow it to take hold, is at once thrilling
and disconcerting: we suddenly realize that we are imagined as there, not
standing at a ‘safe’ distance before a representation but rather witnessing
at extremely close range a simultaneously dreadful and astonishing event,
though it is also true that no account has been taken of our presence by Peter
or his killer, leaving us unaccounted for, so to speak. At the same time, none
of this is felt to entail a notional dissolving of the limits of the painting (we
are very far from the illusionistic propensities of the full-blown Baroque).
Rather, those limits are made palpable as limits for being activated in these
ways. But this makes it all the harder to imagine Savoldo’s canvas on the side
wall of a chapel in a Dominican church, as Gilbert suggests would have been
its destination (p. 292). Such circumstances inevitably would have militated
against a recognition of the viewer-oriented dynamic I have sought to evoke.
There is still more to be said about the Death of St Peter Martyr, but
before going there I want to leave Savoldo for a moment to consider a major
painting by Caravaggio, one of the two that made his early reputation as a
revolutionary force, the Martyrdom of St Matthew (c. 1599–1600; illus. 11) in
S. Luigi dei Francesi in Rome. I have already noted that, starting with Roberto
Longhi, modern scholars have emphasized Savoldo’s signicance as a key
 death of st peter martyr 33

predecessor of the later master, on the strength of his ‘Lombard naturalism’,

his related treatment of light and colour, and his predilection for nocturnal
or penumbral scenes, as in his Magdalenes or the St Matthew and the Angel.
(Longhi also famously focuses on Savoldo’s treatment of his personages’
hands, a topic about which there will be much more to say.) All this makes
perfect sense, but what I want to call attention to in the Martyrdom of St Mat-
thew is something else entirely: Caravaggio’s extraordinary depiction of the
relations among the almost naked executioner holding a sword and rmly
gripping Matthew’s right wrist, the fallen but apparently still unwounded
Matthew and, of particular interest, the winged angel who reaches down
from a strikingly material cloud to extend a palm branch towards Matthew’s
upraised right hand. As has been recognized, the treatment of the executioner
11 Caravaggio,
and Matthew plainly evinces the use of models holding poses. Indeed there
Martyrdom of
is at least the suggestion that the young assassin’s grip around Matthew’s St Matthew, c. 1599–
wrist is less persuasively an image of sudden violence than it is a device for 1600, oil on canvas.
34 painting with dem ons

allowing models for the two gures to maintain a difcult pose long enough
for it to be set down in paint. Put slightly differently, Caravaggio’s decisive
innovation has always been seen as his use of models, as Keith Christiansen
has pointed out: ‘[A]ll of Caravaggio’s contemporaries viewed his practice
of painting from a live model to be the single most outstanding feature of his
work.’4 What this particular gure-group suggests is that Caravaggio not
only wished to insist on that fact; here as elsewhere in his art he was willing,
up to a point, to evoke the original studio situation of models holding poses,
even as he also sought to represent with unprecedented force and realism a
scene of murderous violence.
What particularly interests me, of course, is the presence of the angel
extending a palm branch down towards Matthew’s upraised (or upheld)
hand, at once a traditional feature of depictions of martyrdoms and – in this
particular painting, by virtue of the realistic levelling of the three gures (no
special treatment for the angel beyond giving him a pair of wings) as well
as of the angel’s imminent intrusion into the relations between the assassin
and the saint – a way of casting that feature in a new, defamiliarizing light,
an early example of what would be a major feature of Caravaggio’s art. This
would be the case to the very end: think of the King of the Huns ring an
arrow at point-blank range into St Ursula’s breast in his nal painting, the
Martyrdom of St Ursula. Needless to say, I would like to be able to assert that
Caravaggio was in part inspired to do this by a familiarity with Savoldo’s
canvas with its very different but comparably unconventional treatment of
the palm-branch theme, but there is no solid basis for such a claim. Not
only is nothing known about where the Death of St Peter Martyr originally
hung, for all the widespread assumption on stylistic grounds that Savoldo
signicantly anticipated Caravaggio, we have no hard evidence that actu-
ally places the later master in front of a work by the earlier one. Despite
this, Christiansen has noted that ‘It is signicant that whereas Savoldo’s pic-
tures seem to have made little impression on Venetian painters, they were
the basis of much subsequent painting in Milan, where Caravaggio almost
certainly studied them rst-hand.’5 Mina Gregori agrees, condent that
Caravaggio would have seen Savoldo’s Tobias and the Angel in the Milan mint
in his youth.6 I concur, but like all students of the subject I wish we had
concrete information as to what Caravaggio saw and didn’t see during his
early years. In any case, I nd the relation between the elision and thereby
the foregrounding of the angel in the Death of St Peter Martyr and the
trio of angel, Matthew and assassin in the Martyrdom of St Matthew to be
 death of st peter martyr 35

suggestive of a deeper afnity than simply a stylistic one between the two
painters.

I c om e n ow to another feature of Savoldo’s painting that deserves

further attention: the depiction of Peter’s hands. In an obvious sense, as
Gregori remarks, Savoldo was inuenced in this by Palma Vecchio’s depic-
tion of the saint’s martyrdom in Alzano Lombardo (p. 76), the most obvious
difference being that Palma shows the saint’s right hand pointing at the rst
words of the Credo written in blood on the ground beside him, whereas
in Savoldo’s painting Peter’s right hand merely gestures reexively in a
manner already discussed. But there is a more profound difference between
the two paintings in this regard, I mean the way in which Savoldo’s decision
to pitch the entire scene much nearer to the picture plane, and thereby to
the viewer, gives the saint’s left hand emerging from its sleeve an altogether
more pointed signicance than in the predecessor work. Moreover, the sense
of its signicance is further heightened by the fact, as already noted, that
it now occupies very nearly the exact centre of the canvas, with the result
that it must be accounted one of the painting’s three main foci of pictorial
interest, along with the saint’s facial expression and the killer’s crimson
doublet (or four main foci if one includes, as I have suggested one must, the
descending angel).
The question is what to make of that hand, a question, it seems worth
noting, that so far has escaped being asked. This would be surprising if there
were a larger secondary literature on the painting, which would doubtless
be the case had it not come to light so recently. By this I mean that hands in
Savoldo’s paintings have attracted a great deal of commentary, going back to
Longhi, for whom they constituted one of the signature elements in his art.
(As will become clear, they are a recurrent topic in Gilbert’s dissertation.)
In Chapter Two, taking off from a discussion of the Portrait of a Man in
Armour in the Louvre – in fact, as Gilbert realized, a self-portrait – I shall
engage the larger topic of Savoldo’s treatment of hands. Here, though, I
simply want to focus on Peter’s hand in the Chicago canvas, and use it to
make some basic points about the project of coming to terms with Savoldo’s
highly idiosyncratic achievement.
The rst and most important point is simply to acknowledge the hand’s
particular claim on the viewer’s attention and, going farther, on what I
would like to call his or her capacity for empathic projection, to use a term
36 painting with dem ons

12 (above left) rst put in circulation in a different context by the philosopher Stanley
Caravaggio, Death of the Cavell.7 In my book The Moment of Caravaggio I try to show how in early
Virgin, 1601–2, oil on
canvas.
revolutionary paintings such as Caravaggio’s Penitent Magdalene (1596–7;
illus. 13) and Death of the Virgin (1601–2; illus. 12), a seeming inexpressiveness
13 (above right) – a deliberate avoidance of outward-directed gestures and expressions (of
Caravaggio, Penitent
Magdalene, 1596–7,
the depiction of affetti) – turns out to be perceived, by viewers attuned to
oil on canvas. his art, as a mark or sign of the depth of Mary’s or the grieving disciples’
absorption in thought and feeling. Freedberg’s comment in a 1982 essay
on the painter, that ‘there is no secret of the psyche that Caravaggio cannot
nd out’, bears witness to the efcacy of Caravaggio’s strategy, as well as to
its invisibility as such to his commentators.8 Put more strongly, my claim in
that chapter is that the paintings in question are the scene of a momentous
discovery,
 death of st peter martyr 37

the discovery that a powerful mode of emotional communication can be

actuated by absolutely minimal physiognomic and gestural means. Put
slightly differently, the discovery concerns what viewers, confronted with
certain sorts of outwardly almost wholly inexpressive gures – gures
who are outwardly inexpressive in certain distinctive ways – spontane-
ously do, at least in the Western tradition . . . namely, read that lack of
outward expression as an unmistakable sign of intense inwardness and
sheer depth of feeling, as if in the presence of certain very slight but nev-
ertheless telling visual hints or cues the illusion of absorption, which is
to say the endowing of the gures in question with an imagined inner
life comparable, if not superior, in intensity to the viewer’s own, proves
irresistible. The discovery, in other words, is of the basic truth that human
beings tend strongly to project – that by and large they cannot not proj-
ect – a conviction of inwardness onto, or rather into, painted or sculpted
gures who elicit that act of projection in various barely speciable ways,
which is why the magic of absorption continues more or less unabated to
the present day. (MC, p. 77)

Further on in my book I refer to this phenomenon as ‘empathic projection’,

and I relate it to Cavell’s arguments about scepticism and acknowledgement
– topics there is no need to go into here.
Now my account of the signicance of Peter Martyr’s facial expression
in Savoldo’s canvas, to remain with that remarkable work, does not involve
responding to an instance of expressive minimalism of the Penitent Magdalene
sort (whatever one makes of his state of mind, he cannot be seen as someone
simply holding a pose, which was a risk inherent in such minimalism). The
modality of its expressiveness, however, is as much inward as outward, and
I have also implicitly suggested (I would like to feel I have demonstrated)
that it is only to be understood by an act of something like imaginative
identication on the part of the viewer with the gure of the saint, above all
with the latter’s upward, intense, marvellously nuanced gaze. Put in further
Cavellian terms, what is at stake in one’s encounter with Savoldo’s Death of St
Peter Martyr is nothing less than ‘the discovery of the other, the inwardness
of the other, the beyond of the other’, insofar as it lies within the power of pic-
torial representation to persuasively evoke such a state of affairs.9 In its own
time and place, that power was underwritten by the Catholic faith. What has
replaced this, however, in our time and place, assuming that Cavell’s words
are felt to exert their grip? At any rate, at the risk of getting ahead of myself,
38 painting with dem ons

I am inclined to say that I know of no sixteenth-century Italian painting

– not even Lotto’s portraits, about which there will be more to say – that
outdoes Savoldo’s canvas in this regard.10
Finally, we come to the saint’s left hand (illus. 14), almost at the exact
centre of the canvas. The rst thing to be remarked is that it is an astonish-
ingly subtle, nuanced, feeling entity – one of the most remarkable hands
in Western painting, I want to claim (in my enthusiasm, but I do not think
I overstate the case). The palm faces us, the ngers are slightly separated,
they are all bent to a greater or lesser degree (the forenger least, the middle
nger somewhat more, the next two ngers sufciently so that we glimpse
their ngernails, a descriptive triumph at an intimate scale). A certain
tension seems palpable, but there is not the slightest hint of the sort of man-
nered gesticulation that we nd in Titian’s altarpiece, while the hand as a
whole is the object of an extraordinarily subtle treatment in terms of light
and dark: the palm with its creases is relatively strongly illuminated, the
ngers are delicately shadowed, with light towards their tips, the wrist as
it emerges from Peter’s owing sleeve is in relatively dark shadow. As for
the feeling conveyed, it is hard to pin this down with any precision, but the
hand’s status as a potential focus of what I have been calling empathic pro-
jection (or imaginative identication, but I prefer Cavell’s term) seems to
me undeniable. At any rate, the longer I have spent in front of the canvas on
repeated visits to the Art Institute, the more unwaveringly my attention has
come to rest and dwell on the hand – as if the crucial clue, or one crucial clue,
to the painting’s meaning hovers precisely there.
Having gone this far, I want to take a further step and propose that the
viewer’s response to the hand involves a mimetism (not mimicry) of a very
particular sort. Note, to begin with, that the hand, despite facing the viewer,
in no sense thrusts the viewer away or to even the slightest degree blocks
or resists his or her imaginative access to the painting. On the contrary,
and here I come to my most extreme claim, the hand is felt to offer itself
as a perfect ‘match’ or ‘double’ to the viewer’s hand – specically to his or
her right hand, which is to say to the hand that is in a position to ‘mir-
ror’ the saint’s left hand with perfect naturalness. Indeed I think of the two
hands, the saint’s depicted left hand and the viewer’s actual right one, as
all but coinciding in the near foreground of the painting, very nearly in
the vicinity of the picture plane as such, though precisely what it means
to suggest this, precisely what such ‘coinciding’ involves – obviously it is
imaginative, not physical – is impossible to state with as much clarity as
 death of st peter martyr 39

one would wish. But the force of the suggestion, of the solicitation, on the 14 Savoldo, Death of
part of Peter’s left hand, once it is apprehended with empathic openness, St Peter Martyr (illus. 8),
detail.
and once the viewer allows his or her own bodily feeling to activate the
representation, seems to me undeniable: the ‘beyond’ of the other but also
a certain intimate proximity. Viewed in those terms, the hand is the vital
crux of the composition, a locus of juncture between worlds, one sublunary
and the other divine (the hand will shortly grasp the palm of martyrdom).
For an extended moment in and around Venice (Galileo will not be born
for another thirty years, Descartes for another thirty after that), even as the
Reformation is gathering strength in Northern Europe, followed within
little more than a decade by the summons to the Council of Trent to deal
40 painting with dem ons

with the mounting crisis not just as regards the North but within the Catho-
lic Church itself, these worlds are not yet implacably faced off against each
other, as will eventually be the case.
And what of the saint’s other hand, his right one, in the lower left corner
of the canvas? As was noted earlier, it is shown in foreshortening and as if
caught midway in a grasping gesture, evincing a muscular tension, also per-
haps a sense of distress, that the left hand does not. Because the palm is in
shadow, and also, of course, because of its marginal position, the right hand
is all too easily lost sight of; but the foreshortening is brilliant, the treat-
ment of the ngers with their carefully delineated ngernails once again is
scrupulous, and it makes a ne counterbalance to the upraised left hand once
it is taken in. Moreover, one has the impression that it is ever so slightly
larger than the left hand, as if the right hand were slightly nearer the viewer
than the other. Here one has to recognize the simple fact, invariably ignored,
that the Death of St Peter Martyr was brought into being, that is, painted,
by means of a two-handed sequence of actions: the painter, if he operated
in a traditional fashion, which there is no reason to doubt, wielded a brush in
his right hand and held, or gripped, a palette in his left. My further sugges-
tion is that the saint’s right hand mirrors, or say matches, the painter’s left
hand ‘this’ side of the picture surface. His left hand would have been lower
than his right, so its position in the lower left corner of the painting may be
thought of as ‘matching’ or ‘reecting’ the actual circumstances of the pro-
duction of the painting. Moreover, in the normal course of things it would
for the most part have been nearer the painter than his brush hand. So the
painting’s rst viewer, its rst intimate other, so to speak, would have been
the painter himself.
I shall have more to say about hands in Savoldo’s paintings in Chapter
Two, as well as about the strong element of self-portraiture in his relatively
sparse oeuvre – indeed self-portraiture carrying the implication that the
painter is somehow representing himself in the act of painting, at least up
to a point. But by way of bringing this chapter to a close I want to consider
a painting mentioned much earlier, Lorenzo Lotto’s moving and arresting
A Dominican Friar as St Peter Martyr in the Fogg Art Museum (illus. 15).
In many respects comparing Savoldo to Lotto is bound to be unfair to the
former. Not that Lotto is a canonical master like Titian. On the contrary,
although born in Venice and very likely trained there, he mainly worked in
provincial locales (Treviso, Bergamo, Jesi, Ancona, Loreto), and through-
out his long, diverse and immensely productive career his art was marked
 death of st peter martyr 41

by personal idiosyncrasies, including, as Freedberg writes, ‘an almost mor-

bid hyperphotosensitivity’, by which he means that Lotto ‘extrapolates the
techniques of realism, and in particular its technique of light, to record the
response of a special sensibility, acute and nervous, to visual experience’ (PI,
p. 199). Freedberg also identies two powerful strains in Lotto’s art, the
rst leading to ‘an expressive turbulence in which the toughness and imme-
diacy of forms and feelings are popolano, unstylized, and unidealized’, and the
second, ‘no less irrationally inspired, seek[ing] to translate pathos into an
extreme of grace, conveying the rapture of religious feeling by an articial
beauty given to appearances and forms’ (PI, p. 202). Just these few sentences 15 (below left) Lorenzo
evoke a body of work very different from Savoldo’s, to which it should be Lotto, A Dominican Friar
as St Peter Martyr, 1548,
added that Lotto was also a colourist of rare originality, that his altarpieces
oil on canvas.
in particular reveal an extraordinary richness of visual invention, and that
he was without question a portraitist of genius, in the last two respects leav- 16 (below right)
ing Savoldo far behind. If one compares a characteristic Lotto of the 1520s, Lorenzo Lotto, Virgin
and Child with SS Jerome
for example his Virgin and Child with SS Jerome and Nicholas of Tolentino and Nicholas of Tolentino,
(1523–4; illus. 16) in the Boston Museum of Fine Arts, with the Death of St 1523–4, oil on canvas.
42 painting with dem ons

Peter Martyr, it can be hard to see what common ground, beyond their shared
religiosity and a feeling for intense colour, exists between the two painters.
Both were keenly interested in Northern art, however, and it is also true, as
Bernard Berenson was the rst to recognize, that Savoldo in the late 1520s
and early 1530s was inuenced by Lotto. No doubt inuence also owed
the other way. Nevertheless, the difference between their most characteristic
productions could not be more striking.
But there is this: one central issue in my reading of Savoldo’s Death of
St Peter Martyr concerns its relation to the viewer, who I have tried to show
is implicated in the most intimate terms in the gradual bringing forth or
progressive revelation of the painting’s meaning. While I would argue that
nothing quite like that is to be found in Lotto’s oeuvre, it is also true that
his portraits have always been found remarkable for their apparent depth
of psychological penetration, which is also to say for the forcefulness with
which they inect a basic portrait convention – the sitter looking directly
out of the painting – with the further implication that the viewer is addressed
or, perhaps more accurately, engaged by the sitter in a manner that calls for
a response, if only that of seeking actively to interpret the sitter’s mood or
state of mind. This is often one of sadness or melancholy, since something
appears plainly to weigh on the minds and indeed the bodies of certain of
these, especially his older male subjects.11 To cite Freedberg one more time,
referring to the Fogg painting (see illus. 15):

One of the last portraits we possess, a Monk as St. Peter Martyr (c. 1548–9),
half-portrait, half-icon, is a powerful afrmation of a simple and intense
presence; it is not only by an accident of subject matter that we nd here,
as in the last religious paintings, that art has become the servant of belief.
(PI, p. 207)

Freedberg, like Berenson before him and other commentators since, plau-
sibly sees Lotto’s art in the context of the religious turmoil of his time.12
Needless to say, I am not suggesting that Lotto’s canvas refers back to
Savoldo’s of at least twenty years before. But it is by no means impossible
that Lotto would have known the earlier work: the two friars are not devoid
of all facial resemblance. In Lotto’s picture, too, Peter’s hands play a conspic-
uous role, albeit not one keyed to the act of painting (the right hand pointing
downwards, here too at the rst sentence of the Credo), while a palm branch
of martyrdom has been placed partly behind Lotto’s Peter to the right. (The
Credo lying off-canvas recalls the off-canvas angel bringing the palm of
 death of st peter martyr 43

martyrdom to the saint in the Chicago canvas.) Lotto’s Peter, a cleaver rest-
ing in his skull, also bears a sword or dagger that has been plunged into his
upper left breast, and there is no reason to doubt that the viewer is meant to
imagine a blow struck from in front. But that is also – may I say, suggestively,
or does this go too far? – more or less where we might imagine it having
been planted by a backhanded blow of the kind depicted in the Death of St
Peter Martyr. In any case, the two paintings are in dialogue with one another,
whether intended to be so by Lotto or not.
 two • HANDS

I n the moment of caravaggio I begin by suggesting that an aston-

ishing early Caravaggio, his Boy Bitten by a Lizard (c. 1595–6; illus. 18), may
be seen as a disguised right-angle mirror-reected and mirror-reversed
self-portrait.1 By right-angle mirror-reection I refer to a studio set-up (or
dispositif ) according to which the canvas on which the self-portrait is to be
made is placed at a right angle to the mirror in which the (in the interest of
simplicity, male) painter will study his own features; in effect the painter in
the act of painting pivots his attention back and forth between the reected
image and the canvas on which he is working, a process that of course can-
not actually be depicted in the nal painting. But as I show by surveying a
wide range of examples, such self-portraits can easily be recognized by the
way in which the painter’s gaze out of the painting – apparently at the viewer
but ‘originally’ at his own image in a mirror – is invariably at odds with the
orientation of his body, which, again ‘originally’, was turned towards the
canvas on its easel. When the painting is completed it in effect takes the
place of the mirror. Thus in the Boy Bitten by a Lizard my claim is that the
boy’s upraised left hand was ‘originally’ the painter’s right hand wielding
a brush and his left hand being bitten by a lizard his left hand holding a
palette, the entire ctive scenario, with its strong sense of surprise (hence
instantaneousness), being a way of disguising the basic arrangement I have
just outlined as well as, by so doing, putting out of mind the protracted
17 (facing page) process of making the painting.
Savoldo, Penitent
St Jerome (illus. 33), A further claim, going beyond Caravaggio, concerned mirror-reection as
detail. such, which by its nature ‘reverses’ right and left. Thus a right-handed (male)
46 painting with dem ons

18 (above left) painter’s reection in a mirror will show him with the brush in his left hand;
Caravaggio, Boy Bitten but in fact it turns out that for several centuries self-portraits that depicted
by a Lizard, c. 1595–6,
oil on canvas. the artist holding his tools or in the act of painting (most did not) presented
a normalized image – brush in right hand, palette in left – that aimed to be
19 (above right) faithful to the reality of the case rather than to the reection in the mirror.
Annibale Carracci,
Self-portrait with Other
There are a few exceptions, notably Annibale Carracci’s marvellous Self-
Figures, c. 1585–90, portrait with Other Figures (c. 1585–90; illus. 19) in the Brera, but basically the
oil on canvas. normalized status quo prevailed until around 1860, when quite suddenly and
with no formal acknowledgement of the change, painters everywhere began
to be faithful to the reversed image, which is to say to what they saw in the
mirror rather than what they knew to be true.2 In the book I also show how
a number of pre-1860 self-portraits hint at mirror-reversal without actually
spelling it out; nothing could be more surprising than that the entire issue
largely escaped art-historical notice for so long.
Signicantly, an impressive painting by Savoldo belongs squarely to this
problematic: the Portrait of a Man in Armour (c. 1525; illus. 20).3 For a long
 hands 47

time it was considered a portrait of the French military commander Gaston 20 Savoldo, Portrait
de Foix, duc de Nemours, but in fact there is little doubt that we are dealing of a Man in Armour
(previously known as
here with a self-portrait, as Creighton Gilbert appears to have been the rst Gaston de Foix), c. 1525,
to recognize,and it is this identication that allows us to understand it as the oil on canvas.
conceptual and pictorial tour de force that it is.4
First, though, it is necessary to say something about a famous work, albeit
no longer extant (and perhaps simply legendary): a painting by Savoldo’s
Venetian predecessor Giorgione as described by Savoldo’s former student
Paolo Pino. In an often-cited passage in his Dialogo di pittura of 1548, a text
we shall return to, Pino writes that Giorgione

painted a picture of an armed St George, standing and leaning on the shaft

of a spear, with his feet at the very edge of a limpid and clear pool – which
[pool] was transxed by the entire gure, foreshortened as far as the
48 painting with dem ons

crown of the head; in addition, he had feigned a mirror leaning against a

tree trunk, in which the entire gure was reected from the back and one
side. He depicted a second mirror opposite this, in which was visible the
entire other side of St George. And this he did in support [of the argu-
ment] that a painter can show an entire gure at a single glance, which a
sculptor cannot. (D, p. 367)5

The last reference is to what has come to be known as the paragone, a com-
petition between artistic media, in this case between painting and sculpture,
which by its very nature shows gures in the round. This was a major theme
in sixteenth-century writing about the arts.6
As regards Savoldo’s canvas, the paragone issue is doubtless relevant,
but in The Moment of Caravaggio I go on to argue, on the strength of its
identication as a self-portrait, that questions of right/left mirror-reversal
are the focus of its operations:

In fact, Savoldo’s canvas provides a precocious precedent for Caravaggio,

in that I understand the respective position of the artist-sitter’s right
and left arms and hands as miming in reverse those of the artist-viewer
as he worked on the painting (as if the artist-sitter’s extended left arm
were holding a brush and the bent right arm were gripping a palette) – in
reverse, because, this being a self-portrait, Savoldo would presumably have
been portraying his image in a mirror (i.e., the entire scene represented
in the painting should be understood as a mirror-image). In addition,
Savoldo depicts two other mirrors, one of which, the large mirror that
occupies the upper right quadrant of the canvas, reverses back the implied
mirror reversal of the artist-sitter’s arms and hands. Thus, in that mirror
the artist-sitter’s right arm is shown extended and the left not, which is
what in actual fact would have been the case (the artist-viewer’s right
hand wielding a brush, the bent left arm holding a palette). The genius
of this construction is that it explicitly presents the mirror not simply
as an indispensable element in a paragone between painting and sculp-
ture but also, equally important, as a device or technology for reversing
right and left; this in turn shows that a heightened consciousness of the
workings of reversal was in place as early as the second half of the 1520s.
(MC, p. 13)

Consistent with this last claim, in Chapter One I concluded my reading

of Savoldo’s Death of St Peter Martyr by invoking the viewer’s imagination –
 hands 49

his or her empathic projection – of something like a virtual merger between

his or her right hand and Peter’s centrally disposed left one, and I further
suggest that, inasmuch as the rst viewer would have been Savoldo him-
self, we are entitled in this case to think of the notion of merger as evok-
ing, implicitly, the act of painting the picture. Which is to say that there
exists a deeper afnity between the Death of St Peter Martyr and the Portrait
of a Man in Armour than the absence of formal resemblance would seem
to indicate.7
The wider stakes of these observations emerge when it is recognized, as I
stress in The Moment of Caravaggio, how deeply involved with self-portrayal
Caravaggio’s paintings often are. Specically, Caravaggio is thought to have
depicted himself on a number of occasions: in addition to the Boy Bitten by
a Lizard (if I am right about that), in the early Bacchino Malato (c. 1593–4;
illus. 21); as the musician at the right holding a cornetto in The Musicians
(c. 1595–6; illus. 22) (seemingly a right-angle mirror-representation); as an 21 (below left)
Caravaggio, Bacchino
anxiously eeing bravo in the Martyrdom of St Matthew (1599–1600; illus.
Malato, c. 1593–4, oil on
23); as a man holding aloft a lantern in the Taking of Christ (1602; illus. 24); canvas.
in the severed head of Goliath in the magnicent David with the Head of
Goliath (1606; illus. 25); and as a looker-on straining to witness the tragic 22 (below right)
Caravaggio,
events unfolding before him in Caravaggio’s valedictory masterpiece, the The Musicians, 1595–6,
Martyrdom of St Ursula (1610; illus. 26). There also appears to be a minuscule oil on canvas.
50 painting with dem ons

reection of a painter at his easel in the wine ask of the Ufzi Bacchus
(c. 1596–7), but that is probably too uncertain to count.
In Caravaggio’s case this has everything to do with the modality of his
realism, which I see as anchored not merely in his personal experience,
as is often rightly said, but more precisely in a profound sense of his own
embodiment, his worldly existence as a corporeal being, a crucial dimen-
sion of which involved the making of paintings. (I think of this as correcting
the traditional characterization of Caravaggio as concerned above all with
delity to ocular perception.) So for example Caravaggio’s pictures char-
acteristically thematize the issue of bodily orientation, by which I mean the
basic fact that a painter at work on a painting faces towards it, or say into
it, while the painting itself, the canvas, faces outwards. This fact is made
all but explicit in The Musicians by virtue of the juxtaposition of the gure
23 (below left) studying a score in the right foreground, who has been depicted largely from
Caravaggio, Martyrdom
the rear, with the central musician tuning a lute, who is shown gazing melt-
of St Matthew (illus. 11),
detail. ingly out of the painting as if seeking to meet and hold the viewer’s gaze. As
already mentioned, the cornetto player, a self-portrait, evokes a right-angle
24 (below right) mirror-reection dispositif, though of course the painting as a whole cannot
Caravaggio, Taking
of Christ, 1602, oil on be imagined as having been produced in such a fashion; in other words we
canvas. are dealing here with an intensely self-referential work of art. More broadly,
 hands 51

25 (above left)
Caravaggio, David with
the Head of Goliath, 1606,
oil on canvas.

26 (above right)
Caravaggio, Martyrdom
of St Ursula, 1610, oil on
canvas.

27 (left) Caravaggio,
Judith and Holofernes,
c. 1599, oil on canvas.

a contrast between gures facing into the painting and others seemingly
addressing the viewer will be prominent in the work of Caravaggesque art-
ists like Bartolomeo Manfredi, Valentin de Boulogne and Nicolas Régnier,
while central gures depicted from the rear, also in my view keyed to the
orientation of the painter, will play a crucial role in Gustave Courbet’s
Realist canvases of the late 1840s and ’50s.8 In other works by Caravaggio
one or another version of the painter-painting relationship may be under-
stood as having been transposed into a lateral axis, the Judith and Holofernes
(c. 1599; illus. 27), Taking of Christ and Martyrdom of St Ursula being
52 painting with dem ons

particularly rich and complex examples of this sort of structure. In still

others, and also in the Judith and Holofernes, the motif of decapitation
provides thematic reinforcement for the project of making a particular
sort of painting – the full-blown gallery picture – that will present itself as
severed from the context of its exhibition as well as from the very activity
that brought it into being, or so I claim. There will be more on this towards
the close of Part Two.
The question now is what bearing all this has on a consideration of Savoldo,
whose art cannot be regarded as bodily in the ways just summarized. But
Savoldo – like Caravaggio, indeed like Courbet – does make surprising use
of the self-portrait: no fewer than four paintings in his surviving oeuvre
of less than fty, as well as a highly nished drawing, have been identied
as such (initially by Gilbert). And my thought – my suggestion – is, rst,
that this deserves to be regarded as a matter of some importance (as Gilbert
implies, though he cannot say why), and second, that it precisely indicates a
deep afnity with Caravaggio, one not based on considerations of style of the
sort that have dominated the secondary literature down to the present. For
in Savoldo’s art what anticipates and up to a point stands in for the sense of
bodiliness that plays so central a role in Caravaggio’s oeuvre is the depiction
of hands.

I n de e d th e d e pic t io n of hands is a central crux, in certain respects

the central crux, in Savoldo’s art – at least that is a basic premise of this
book. Signicantly, it is a topic that caught the notice of Roberto Longhi in
his rst, brief, characteristically penetrating remarks about the painter in
his 1917 article ‘Cose Bresciane del Cinquecento’ (see below), and it recurs
continually in Gilbert’s dissertation, where he both pays tribute to Longhi
and adds various observations of his own, none of which, in my view, quite
captures the complexity of the pictures he discusses.9
For a start, there are three paintings that deserve particular attention.
The rst and almost certainly the earliest is the so-called Prophet (or Apos-
tle) in the Kunsthistorisches Museum in Vienna (early 1520s; illus. 28), per-
suasively identied by Gilbert as a self-portrait on the basis of the sitter’s
resemblance to certain other gures in Savoldo’s oeuvre and also on the
strength of the fact that the scroll held in the gure’s left hand, which ordi-
narily would bear the personage’s name, instead bears Savoldo’s signature
(of which there are very few).10 The second, a later work, is the so-called
 hands 53

Evangelist in a private collection in Milan (mid-1520s?; illus. 29); the sitter 28 Savoldo, Prophet
bears a loose resemblance to the Vienna prophet – they are both young- (or Apostle), early 1520s,
oil on poplar.
ish men with strong physiques, brown hair and beards, and piercing gazes
– but no one has suggested that the ‘evangelist’ (identied as such on the
strength of the open book behind him to the right) is a self-portrait, which
54 painting with dem ons

29 Savoldo, Evangelist, seems to me correct – but the situation is perhaps more interesting than
mid-1520s?, oil on panel. this suggests, as will emerge shortly.11 The third work is the so-called
Self-portrait in the Costume of St Jerome in a private collection in Bergamo
(c. 1530; illus. 30).12 As Alberto Veca remarks, the sitter, whom he identies
as the painter, is much younger than traditional images of the saint (includ-
ing one by Savoldo, as we shall see), but he is shown gazing at a crucix with
a skull at its foot (‘a common formula’, as Gilbert notes), and then there is the
coincidence of names between the saint and the painter, which although not
decisive is nevertheless suggestive.13 Gilbert agrees that the features of the
 hands 55

sitter are consistent with those in the other self-portraits previously assem- 30 Savoldo, Self-portrait
bled by him, but, somewhat surprisingly, says nothing about the fact that in in the Costume of
St Jerome, c. 1530, oil on
this case the sitter does not meet the viewer’s gaze but rather is turned so canvas.
that one sees only his left prole.
In all three paintings hands play a perspicuous role. The crucial statement
about hands in Savoldo’s paintings is made by Longhi in 1917 in two bril-
liant paragraphs apropos the Milan Evangelist:
56 painting with dem ons

You can sense it’s Savoldo right away; if nothing were left but his ‘hand’,
that would be enough to recognize him. And then there’s the melted
stuff of his red cassock, studied in the usual key of folds where Savoldo
delighted in solving certain of his queries about form; and there’s also –
and it’s a shame that it was cut out of the photograph – a bubble of light
that wanders over the edge of the book and gives back [restores] freedom
and unanticipated life [vivacity] to the entire luminous composition.
But we must return to that hand, which grabs the eye in a formal com-
plexity that must be untangled like the knot of Solomon; to think that
this is some sort of perspective game would be a bit supercial, since it’s
a choice both too bony and too eshy to be enclosed in a few planes, in a
perspective prism. Actually, it both hides and reveals a simpler and more
profound pictorial anagram; it’s a moment, an impression of a hand xed
[frozen] with wondrous perspicacity by a mirror placed at the strangest
[possible] point; it’s a cursive [feat of] foreshortening, which although it
is rendered in paint that is still ‘quattrocento’, anticipates the swift hand of
Caravaggio, and perhaps of Degas.14

The basic idea appears to be that the hand is no mere perspectival feat of
strength, though it is also surely that, but rather that in its esh-and-blood
form it conveys a sense of having been rapidly seized by a swiftly moving pic-
torial act, as if on the basis of its image in a mirror. That is, Longhi’s remarks
draw attention to the ‘cursive’ act of painting itself (or seeing-and-painting),
an act that in his account looks forwards to Caravaggio or even to Degas (in
1917 the equation of Impressionism with modernity as such was extremely
strong). Gilbert will cite Longhi, as I have said, approving particularly of the
latter’s emphasis on ‘studies of form’, seeing in the hands

the classic expression of Savoldo’s gift of materiality and dignity to each

element of the gure. Indeed, the attitude that produces this kind of hand
appears to be the same that produces this kind of color . . . Both evince the
slow and sober construction of mass and reality in each major section of
the gure contemplated, undeniable and brilliant, neutral but powerful.
The hands, stretching forward, not perspective games but caught in light,
live in themselves, betoken the energy of the gure by its gesture, and
create a breadth of space to give it elbow room. (G, pp. 346–7)

Although Gilbert does not say so, such a reading is at odds with Longhi’s
stress on speed and cursiveness, but the difference between them is perhaps
 hands 57

less important than their shared focus on the hands as what Longhi strik-
ingly called ‘pictorial anagrams’ for Savoldo’s vision and practice. Put
another way, one can see and appreciate the force of both their emphases,
the rst on the singleness and, so to speak, the suddenness with which the
hands are grasped as functional entities, the second on slow and deliberate
construction (the rst on the hand in its knot-like complexity, the second on
the gure as a whole installed in its space). (Gilbert does say that the hands
‘live in themselves’, but leaves unclear what he means by that.) What I now
want to suggest is that both accounts, for all their considerable interest, fall
short of recognizing an important aspect of what the three paintings offer
to be seen.
This is probably clearest in the case of the Prophet or Apostle, which at once
presents itself as a portrait – one believes instantly in the ‘reality’ of the sit-
ter – with three major foci of interest: rst, the sitter’s gaze, which engages
directly and as if interrogatively the viewer’s own (the sitter’s expression,
full of unspecied emotion, along with the turn of his head towards the
viewer, undergird the force of the encounter); second, the sitter’s left hand
rmly but lightly holding the smallish rolled scroll mentioned earlier; and
third, his right hand, which seems to be caught mid-gesture, palm down,
ngers separate and as if about to grasp something but without any clear
purpose (in fact it is strikingly similar to the protagonist’s right hand in the
Death of St Peter Martyr). In addition there is the sitter’s garment, a blouse
with wide sleeves of an almost garish green material and strong white high-
lights suggesting bright illumination, the contrasting dark background
giving the blouse an incandescent sheen.
In short, the viewer instinctively understands that he or she is being chal-
lenged by the painting to come to terms with its complex assertion of human
and material presence. My suggestion, as in the case of the Death of St Peter
Martyr, is that it is not enough to respond to the hands as both Longhi and
Gilbert do, each in his own way, as fascinating entities viewed as it were
objectively from outside, at a near but disengaged distance. Rather, I take the
hands as inviting what in Chapter One I called both imaginative identication
and empathic projection, which is to say a kind of heightened phenomeno-
logical or kinaesthetic awareness of the hands as experienced by the subject
as living corporeal organs, respectively doing the work of holding the scroll
and of reaching or grasping (to what end is unclear, at least at rst). And as
in Chapter One, but with added emphasis in view of the fact that we are in
this case unmistakably dealing with a self-portrait, I also suggest that we are
58 painting with dem ons

entitled to go on from there to regard the hands as displaced representations

of the painter’s hands at work on the painting, taking the sitter’s left hand (to
begin with, at any rate) as a representation of the painter’s left hand holding
a palette and his right hand as a representation of the painter’s right hand
wielding a brush. Of course, that is to take the painting as a whole as what I
have described as a normalized self-portrait (brush in right hand, palette in
left, assuming the painter is right-handed, as Savoldo appears to have been).
But is there not also a sense in which, in the act of projecting empathically,
the viewer may be led to identify the prophet’s or apostle’s left hand with his
or her right hand and, even more naturally, the apostle’s or prophet’s right
hand with his or her left hand (the ‘match’ between the latter two being par-
ticularly close, if one gives it a chance) – in other words, in a ‘mirroring’ rela-
tion to left and right within the painting? Needless to say, we are on inrm
ground here, Savoldo’s treatment of the hands leaving open the question of
the relation of left and right within the painting to left and right outside it
(that is, ‘this’ side of the picture plane). But it seems to me beyond question
that the viewer’s hands are, in effect, solicited, imaginatively mobilized, in
something like these terms.
The Milan Evangelist makes this point even more succinctly by virtue of
depicting only one hand, the sitter’s left, which again is largely open (we see
the palm) even as it seems to reach strongly towards the picture plane as if to
grasp something ‘this’ side of the latter. One sees Longhi’s point: the hand is
too vigorous and carnal to be subjected merely to the laws of perspective; but
his counter-suggestion that it appears miraculously captured in a mirror fac-
ing the sitter’s hand (presumably at close range) is highly problematic: how
is one to imagine this? Indeed even if one takes the ‘owner’ of the hand to be
the painter, the mirrored left hand would then represent the painter’s right,
which is to say that it is that hand that would be doing the painting; once
again, how would this work? The idea, for all its attractiveness, seems not
quite thought through. But what seems clear, at least to me, is that however
this was accomplished, the sitter’s left hand ‘matches up’ most persuasively
with the viewer’s, hence also with the painter’s, right hand, a relationship
that recalls that of the viewer’s right hand to the friar’s upraised left hand in
the Death of St Peter, as analysed in Chapter One (in that sense being a ‘mir-
roring’ relationship).
The third painting mentioned above, the St Jerome in Bergamo, departs
from the Vienna and Milan canvases by virtue of the subject having been
depicted in prole, facing left, with both hands in view: his right hand at the
 hands 59

lower left, his left hand, seemingly extended towards the picture plane, at
the lower right. As already mentioned, the sitter gazes at a crucix set up
in front of him and slightly to his right. There is nothing natural or com-
fortable about such an arrangement, with the result that the picture is less
immediately compelling than the other two – one particularly misses the
confrontational gazes – but in a certain sense it is the most ambitious, as it
were intellectually, of the three. Simply put, I suggest that the composition
represents an attempt by Savoldo to make a self-portrait based all but explic-
itly on the right-angle paradigm, which he interprets in a particularly chal-
lenging way. The ‘classic’ self-portrait based on right-angle mirror reection
shows the painter facing directly out of the painting, as if directly at the
viewer, for the simple reason that the painting portrays what the painter
saw in the mirror, which in effect hung where the nished canvas now hangs
(with the canvas-in-progress at right angles to it, often just outside the lim-
its of the canvas itself). Annibale’s Self-portrait with Figures, in which a slice
of the canvas viewed from the side is included, exemplies this structure
with perfect clarity. But what Savoldo has sought to do, if I am right, is to
imply that the sitter is gazing at his reection in a mirror within the painting
(symbolized by the crucix, or perhaps both crucix and skull together),
while holding his palette in his left hand and reaching towards the canvas
with his right or brush-hand, both of which have been reversed right and left
by their reection in an actual mirror, though of course the implied mirror
in the painting is not in the correct position to bring this about. Never-
theless, the sense of the sitter’s extended left arm and hand at the lower
right somehow representing the painter’s right arm and hand at work on the
canvas is surprisingly strong. I realize, how could I not, that such a reading
of the Bergamo canvas may not compel immediate conviction; what would
help is to insert it among all the other right-angle self-portraits I discuss
in Chapter One of The Moment of Caravaggio: it would still stand out as
anomalous, but its structure of intentions would be more apparent. (Anoma-
lous, in part, because of its early date: Annibale’s masterly group portrait and
Caravaggio’s displaced versions of the right-angle dispositif belong to a
distinctly later moment.)
A fourth painting, denitely not a self-portrait, of keen interest to Gilbert
is the Portrait of a Young Man (St John the Evangelist?) in the Borghese Gal-
lery (mid-1520s; illus. 31),15 the association with the saint being based on
the resemblance between this gure and that of St John in the Lamentation
over the Dead Christ, formerly in the Kaiser-Friedrich Museum in Berlin and
 hands 61

destroyed in the Second World War (c. 1515–20; illus. 32).16 Here we are 32 Savoldo,
shown the protagonist rather closely cropped by the smallish canvas; the Lamentation over the
Dead Christ, c. 1515–20,
sitter’s right hand is off-canvas to the bottom left but his left hand, seem- oil on canvas.
ingly resting on a woven gold belt or scarf, thrusts prominently towards us,
the palm (which we mostly do not see) facing upwards, the ngers loosely
bent, and as usual the perspective is compelling and the ngers themselves
shapely. Gilbert, impressed by the hand, characterizes it ‘as a sort of vivid
still life, in which the varied interrelations of parts in space are clearly
reported and interesting’ (G, p. 110). (He also sees the hand as a variation on 31 (facing page)
Savoldo, Portrait of a
Christ’s right hand in the Lamentation.) As for the young man himself, his
Young Man (St John the
features could not be more rened: his gaze travels off-canvas to the right Evangelist?), mid-1520s,
(in effect before him), his skin is delicate, his lips rosy, his cheeks slightly oil on canvas.
62 painting with dem ons

ushed; he appears thoughtful, perhaps because of what he is seeing, but the

viewer has no idea what that might be. The treatment of light and shadow
is exquisite, characteristic of Savoldo at his most luministically nuanced.
The question, our question, concerns the meaning of the hand, assuming
we are not satised by Gilbert’s still-life metaphor or the association with
the Lamentation. Again, I see it as inviting empathic projection, but precisely
how this plays out in left/right terms remains unclear, as does whether or
not the viewer is invited to imagine the sensation of the back of the young
man’s hand against the woven gold belt or scarf on which it partly rests. It is
tempting to relate the hand to the painter’s left hand holding a palette, but
there is no clear warrant for doing so; perhaps all one can say (by my lights)
is that once again the viewer is invited, at least to some degree, to participate
imaginatively in the carnal life of the hand, which therefore emerges as fully
as important to one’s experience of the painting as the youth’s meticulously
rendered head.

At th i s po i n t it will help to turn to a rather different-seeming painting,

which nevertheless bears a deep structural relation to the works we have
just considered: the magnicent Penitent St Jerome in the National Gallery,
London (1527–30; illus. 33).17 In a sense, the treatment of the subject is
traditional or at least consistent with other pictures of the saint physi-
cally belabouring himself in sixteenth-century Venetian or Brescian art, for
example in works by Montagna, Lotto, Cima da Conegliano, Romanino and
others; as Hans Belting and others have remarked, paintings of St Jerome
either beating himself or simply reading or writing in the wilderness are
among the most numerous in Venetian Renaissance art.18 In this case, the
grey-bearded saint, wearing a loose red-purple robe, kneels in a mountain
landscape before a stone ledge (an improvised table) and a stone wall with
a bit of foliage at the top, slightly further away than the ledge, all the while
gazing at the gure of Christ on a modest-sized crucix at the upper right
and about to strike himself in the chest with a stone clutched in his right
hand. Below the crucix, propped open on the ledge facing the saint, is a
book that presumably represents the Bible he has been translating into Latin;
no writing materials are in view, however, and of course the idea of a bound
volume is anachronistic. Just in front of the book, also resting on the ledge, is
the saint’s left hand, about which there will be more to say. Behind the saint,
in the middle distance, there is a view, or more precisely a glimpse, of the
 hands 63

bank of the Fondamente Nuove in Venice with the Church of SS Giovanni e 33 Savoldo, Penitent
Paolo. Much further away blue mountains rise against the backdrop of what St Jerome, 1527–30,
oil on canvas.
is usually described as a dawn sky. The luminous sky and darkish clouds,
although an area of some repaint, are superb.
What is of particular interest in the present context, however, is the
hands, in the rst place the right one clutching a stone at the end of a mus-
cular right arm that has been depicted in the act of preparing to strike the
stone against the saint’s chest. The arm itself is not silhouetted against the
background or the sky, or for that matter against the chest, all familiar dis-
positions, but rather shown in what Gilbert rightly describes as ‘virtuoso’
perspective foreshortening against the saint’s body (G, p. 105), a highly origi-
nal treatment that in fact compels a far more focused act of attention on
64 painting with dem ons

the part of the viewer to take its measure. And in the second place there
is the saint’s left hand, which does not so much rest on the ledge, as I just
said it did, as rise from it by an action of the wrist, the hand itself making a
by now familiar grasping gesture with open palm and bent ngers. This is
another knockdown instance of perspective rendering, familiar by now but
also infused with an unusual degree of muscular tension. In other words, the
hands bear a close relation to those in the paintings we have just considered,
in particular the Vienna and Bergamo self-portraits. My suggestion, or fur-
ther suggestion, is that while the St Jerome is not exactly what I have been
calling a disguised or displaced self-portrait, the two-handed operation that
it depicts, the decisive ‘actor’ being the right hand and arm, is not without
some resemblance to the two-handed operation of working on a painting,
and indeed it is tempting in this connection to think of the open volume
with its carefully delineated two-column pages as equivalent to a canvas,
and to think of the crucix, the focus of the saint’s strained attention, as
. . . what exactly? The subject of the saint’s ‘painting’? Or indeed a mirror,
taking up the analogy put forward in the Bergamo Self-portrait as St Jerome?
This would make the London St Jerome a kind of disguised or, perhaps
better, displaced self-portrait after all. Needless to say, these last suggestions
are nothing if not speculative, but I mean them to imply that the Penitent
St Jerome, like the Death of St Peter Martyr – a painting with which it has
more than a little in common (for one thing, both involve blows) – calls for
a sustained effort of empathic attention on the part of the viewer, an effort
that in this case requires the guidance of familiarity with the other Savoldos
we have considered to nd its way to even the tentative proposals advanced
above. One more small point: note the afnity between the grasping left
hand and the dead-seeming broken branch with smaller twigs in the lower
right corner of the canvas, an afnity that I will shortly describe as one of
auto-mimesis.
Other paintings by Savoldo conspicuously involving hands are less com-
plex. For example, the early Brera altarpiece depicting the Madonna and
Christ Child and Two Angels in Glory with SS Peter, Dominic, Paul and Jerome
34 (facing page)
(1524–5; illus. 34), originally made for the church of San Domenico in Pesaro,
Savoldo, Madonna and
Christ Child and Two appears conventional in most respects.19 As mentioned earlier, there are few
Angels in Glory with altarpieces in Savoldo’s oeuvre and it seems he did not favour them, as Freed-
SS Peter, Dominic, Paul berg suggests in his description of the Pesaro altarpiece as ‘heavy in form
and Jerome (the Pesaro
Altarpiece), 1524–5, oil almost to inertness, sombrely powerful in colour, and insisting on veristic
on panel. presence’ (PI, p. 226); again Freedberg’s still-life metaphor is never far away.
66 painting with dem ons

35 Savoldo, Madonna But once alerted to the signicance of Savoldo’s hands the viewer cannot fail
and Christ Child and to notice that each of the four saints is doing something precise and nuanced
Two Angels in Glory with
SS Peter, Dominic, Paul in that regard (illus. 35). From right to left, Jerome with his powerful left
and Jerome (illus. 34), hand supports an open book partly resting on his thigh, while his right hand
detail. grasps a length of red drapery crossing the same thigh (the action of the
right hand has no necessary role apart from exhibiting itself as such); Paul,
to his right (our left) and standing farther back in space, holds a volume to
his breast with his left hand and rather elegantly rests his right hand on
the pommel of a long sword; Dominic, in prole and looking up towards
the Madonna and child, holds open a smaller book with his left hand (we
see mainly his long, elegant ngers), while with his right hand, ring nger
exed and partly elevated, he seems to mark his place; nally, Peter, the one
saint making eye contact with the viewer, holds a book to his body with his
left hand while seemingly proffering his keys to the viewer with his right,
 hands 67

another small tour de force of foreshortening, in a gesture that comes closer

than any other in the picture to the ones we have been analysing. There is,
I think, no deeper content to any of these actions, no meaning of the kind
we have seen in play in the self-portraits and related works. But Savoldo’s
investment in the saints’ hands is palpable and sets the Brera altarpiece apart
from similar works by contemporary masters.
Then there is the exquisite Portrait of a Young Man with Flute in the Pina-
coteca Tosio Martinengo in Brescia (1525–30; illus. 37), briey analysed
by Keith Christiansen in the seminal 1985 The Age of Caravaggio exhibition
catalogue, in which he compares Savoldo’s canvas with respect to its lighting
(from a single source), spatial illusionism and basic naturalism to Caravag-
gio’s Lute Player in the Hermitage.20 Christiansen remarks only briey on
the ute player’s hands, but Gilbert in his dissertation writes, ‘The hands,
playing the instrument, are beautifully expressive of this action – they not
only have the tentative holding and touching effect of the other hands of
this time, but a detail like the shadow of the ngers on the palm adds to
the melodic evocation’ (G, p. 125), perhaps because the nger-shadows have
somewhat the look of musical notation. Gilbert’s remarks come close to what
I wish to suggest, namely that we are dealing in this canvas not only with
perhaps the most ne-grained and persuasive transcription of optical effects
in all Savoldo’s oeuvre (note in particular the fold-bearing scrap of paper
covered with musical notation afxed to the wall to the left, not to mention
the delicate play of light and shadow on the young man’s face), but also with
an extraordinarily nuanced evocation of the player’s bodily awareness of the
ngertips of his two hands resting precisely on the different holes of his
instrument (the shadow of the ngers heightening the effect). The musical
notation, both on the scrap of paper and on the pages of the book open before
the musician, only strengthens the point.
Or consider the somewhat confusingly titled Shepherd with a Flute in the
Getty Museum (c. 1535; illus. 36). Very simply, this shows a youngish man
in a wide-brimmed hat that casts a shadow over his eyes (as in the Portrait
of a Young Man with Flute) and with a slight moustache and beard. He is
seated, although we are not shown what he is sitting on, while seemingly
making a quiet pointing or ‘presenting’ gesture with his open right hand. In
traditional scholarship he is described as dressed as a shepherd, but Andrea
Bayer notes that ‘his face is rened and his shirt collar a creamy white’.21
Indeed Bayer remarks that the gure has been interpreted as an allegorical
portrait of a patrician disguised as a shepherd, which seems plausible. In any
 hands 69

case, his gesture appears to point towards a rustic scene in the right middle 37 Savoldo, Portrait of
distance that Bayer describes as including ‘a bagpiper, a ock of sheep, farm a Young Man with Flute,
c. 1525, oil on canvas.
buildings (one taken from a print by Dürer), and an imposing, ruined clas-
sical building’. But the fascinating hand is the left one, which does double
duty of a pictorially perplexing sort, at least I nd it initially perplexing:
the left hand most obviously loosely holds the end of a longish staff with
its bark still on, the hand resting on the staff, and the staff, although this
detail is easy to miss, extending between the shepherd’s forenger and
middle nger. Moreover, the left hand also holds a small ute or recorder,
or at least the mouthpiece of one, the disjunction between the staff and the
musical instrument being obscured by the hand itself. Bayer once more pro-
vides more details: ‘Interpretations focusing on [the shepherd’s] recorder –
36 (facing page)
we see its pple and three nger-holes – stress that this was the instrument Savoldo, Shepherd with a
par excellence of the Arcadian shepherd.’ And she cites an earlier scholar Flute, 1535, oil on canvas.
 hands 71

as pointing out that ‘the deliberate juxtaposition of the staff and recorder in
the shepherd’s hand alludes to the natural, rough origins of the instrument,
further associating it with the rustic ways of Arcadia’ (PR, p. 139).
All this is interesting and might conceivably bear on Savoldo’s intentions,
but for me the painting’s visual crux is precisely the hidden discontinuity
between the staff and the ute (a thought reinforced by the composition,
which conspicuously isolates the rather large left hand and its contents).
Resolving the crux involves not simply understanding intellectually, in effect
from outside, that there are two items in play and not merely one, most
obviously the staff, but also coming to access – empathically to ‘share’ – the
shepherd’s bodily awareness of the separate items in his left hand, a truly
Savoldesque touch (no pun intended). (His awareness, too, of the shaft run-
ning between his ngers.) Seen in this light, the ‘presenting’ gesture might
have an additional valence, as if instructing the viewer to look more closely
or say more feelingly than he or she might normally do.22
Another painting to discuss in the light of these considerations is the
Portrait of a Gentleman as St George in the National Gallery of Art in Wash-
ington, dc (1530s; illus. 38).23 A youngish man with dark brown hair, a
moustache and a small well-trimmed beard gazes at the viewer while mak-
ing a characteristic, which is also to say not quite readable, gesture with
his upraised right hand and resting his left hand on a staff (belonging to a
spear?), one end of which clearly rests on the ground below the picture-edge
to the lower right. Also in the right-hand sector of the canvas, in the middle
distance above the hand on its staff, is a small image of St George on horse-
back ghting the dragon.
Characteristically, too, the subject’s right hand is located almost exactly at
the centre of the painting, much as in the Death of St Peter Martyr the saint’s
left hand is all but centred (as is the shepherd’s right hand in the Shepherd
with a Flute even though, as we have seen, it is less important than the left).
Equally to the point, in a much less elaborate variant of the role of armour
and reections in the Portrait of a Man in Armour, Savoldo has contrived to
show us a partial reection of the young man’s right hand in his gleaming
breastplate as a left hand turned away from the viewer, as if to facilitate an
act of imaginative identication with that reection (at least in principle,
since in fact the latter is too small and indeed obscure to function effec-
tively in those terms, but the use of mirror-reversal to turn left into right is 38 (facing page)
Savoldo, Portrait of a
typical of the painter). Another feature worth noting is the drapery and its Gentleman as St George,
deeply shadowed folds, especially on the young man’s left sleeve and across 1530s, oil on canvas.
72 painting with dem ons

the lower portion of the picture; there will be more on Savoldo’s treatment
of drapery in Chapter Three. It should also be said that considered simply
as a portrait, as a pictorial rendering of a distinct personality, the National
Gallery canvas is undistinguished: that simply isn’t one of Savoldo’s aims.

O n e las t g ro u p of paintings is pertinent in this connection, three nearly

identical Magdalenes in London, Berlin and Los Angeles (1530s; illus. 39,
40, 41).24 For present purposes they are interchangeable, so I will restrict
my present remarks to the London canvas, which is universally agreed to
be particularly strong. As it happens, the Magdalenes as a group (with spe-
cial emphasis on the London version) are the subject of an important article
by Mary Pardo, ‘The Subject of Savoldo’s Magdalene ’ (1989), the rich and
detailed argument of which I won’t try to summarize here.25 The crucial point
is that Savoldo’s paintings depict the early-morning moment (or moments,
as Pardo suggests) when the Magdalene, having previously discovered that
Christ’s tomb was empty, returns there alone having told John and Peter
that Christ’s body was missing. Pardo notes that this was ‘late enough in
the narrative, incidentally, for dawn to have broken’ (p. 73), in reference to
the sunrise on the horizon in the paintings. A key to the identication of the
woman as the Magdalene is the ointment jar on a small ledge at the bottom
left of the London and Los Angeles canvases. Pardo then quotes the source
of the subject as it appears in John 20 :11–16:

11. But Mary stood at the sepulchre, without, weeping. Now as she was
weeping, she stooped down, and looked into the sepulchre.
12. And she saw two angels in white sitting, one at the head, and one at the
feet, where the body of Jesus had been laid.
13. They say to her: Woman, why weepest thou? She saith to them, Because
39 (facing page) they have taken away my Lord, and I know not where they have laid him.
Savoldo, Mary 14. And when she had thus said, she turned herself back, and saw Jesus
Magdalene, 1530s, oil on
standing, and she knew not that it was Jesus.
canvas.
15. Jesus saith to her: Woman, why weepest thou, whom seekest thou?
40 (page 74) Savoldo, She, thinking that it was the gardener, saith to him: Sir, if thou hast taken
Mary Magdalene, 1530s, him hence, tell me where thou hast laid him, and I will take him away.
oil on canvas.
16. Jesus saith to her: Mary. She, turning, saith to him, Rabboni.
41 (page 75) Savoldo,
Mary Magdalene, 1530s, In Pardo’s account the paintings depict both Mary turning from the
oil on canvas. tomb, a rst moment when she takes Jesus for a gardener, and the almost
76 painting with dem ons

immediately subsequent moment of revelation, leading the viewer, as Pardo

puts it, to imagine ‘the temporal interval they bracket, in the course of which
a mere gardener at the saint’s back metamorphoses into a resurrected deity’
(p. 73). Seen in these terms, the extraordinary radiance that plays over the
highly reective surface of Mary’s silver-coloured shawl, in the London
canvas, the focus of Pardo’s remarks (but the shawls in the other versions
are also reective in just this way), has its source in the risen Christ, who
would, Pardo suggests on the basis of the distribution of light and shadow,
therefore appear to be standing ‘this’ side of the picture surface and just
slightly to the viewer’s right, assuming the viewer to be centred before the
image (both Christ and the viewer are situated very near the picture surface,
needless to say). As for the alluring Magdalene herself, head tilted and eyes in
shadow, she turns her beautiful face towards the viewer (and towards Christ),
with an unreadable expression that leaves open the possibility that she has
been weeping, as described in the biblical text. Pardo further proposes that
Mary’s gesture of raising her cloaked right hand to her face ‘clearly derive[s]
from a well-established convention for weeping mourners at the scene of
Christ’s death, but here the wrapped st is not pressed to the eyes, and the
saint gazes out as if acknowledging a momentary interruption’ (p. 73).
In all three paintings the Magdalene’s shawl covers her entire upper body,
except for the oval of her face and a glimpse of her left hand that is easy to
miss, wrapping the shawl around her at the lower left. Commentators uni-
versally have praised this altogether unusual piece of costuming as a remark-
able invention, and in fact the shawls differ both in colour and in the pattern
of their creases and folds from one Magdalene to the next (more on this in
Chapter Three). But what has not been considered is the bearing of this
extraordinary artistic decision on the topic of this chapter, Savoldo’s attitude
towards hands and their depiction.
What I mean will be instantly clear in the light of the next two verses
in John:

17. Jesus saith to her: Do not touch me, for I am not yet ascended to my
Father. But go to my brethren, and say to them: I ascend to my Father and
to your Father, to my God and your God.
18. Mary Magdalene cometh and telleth his disciples: I have seen the
Lord, and these things he said to me.

‘Do not touch me’ (Noli me tangere), a traditional theme in Christian art,
is the subject of Titian’s great painting of 1511–14, today in the National
 hands 77

Gallery, London. Here the gure of Jesus, carrying a gardener’s tool, bends
gracefully away from the kneeling Magdalene’s attempt to reach out and
touch him.26 Savoldo would have known Titian’s canvas, of course, and even
apart from that he would have been acutely aware that according to the bibli-
cal text his Magdalene was just a moment away from reaching through the
picture plane towards Jesus with her hand or hands. So the stroke of genius
of Savoldo’s Magdalenes, keeping Mary’s hands out of sight and indeed
replacing them by the visually ravishing shawls with their dazzlingly exe-
cuted, brilliantly arbitrary-seeming patterns of highlights, creases, folds and
shadows – including folds around the ngers – treated differently in each
of the versions, has the effect of deferring that moment, as it were perma-
nently, and therefore also of deferring any possible invitation to the viewer to
perform the sort of act of empathic projection or quasi-kinaesthetic imagina-
tive identication that I have associated with the treatment of hands in the
extreme foreground of the other paintings discussed in this chapter. And yet
this feat of deferral also has the effect of giving extraordinary salience to the
hands, whose precise conguration under the shawls the viewer can only
guess at. Not for nothing did Vasari characterize Savoldo as ‘fanciful and
artful’ in Pardo’s translation (p. 69).
 thre e • FAC E S

Two o f s avo l d o ’s early paintings, Elijah Fed by Ravens in the National

Gallery of Art, Washington, DC (1512–15; illus. 43) and SS Paul and Anthony
in the Gallerie dell’Accademia, Venice (c. 1516; illus. 44), give a clear idea
of the state of his art around the middle of the second decade of the cen-
tury, prior to the various works we have considered.1 What is noteworthy,
all commentators agree, is an ‘archaizing’ quality, looking back to the previ-
ous century; there is no hint yet of the inuence of Giorgione or the young
Titian, or indeed the aged Giovanni Bellini (himself a gure spanning two
worlds, the late quattrocento and the new century). Instead the person-
ages are grasped in their stolidity and massiveness – Freedberg’s still-life
metaphor takes hold precisely here – ‘struck by light but not penetrated’, in
Gilbert’s helpful phrase (G, p. 299), with strong, slightly abstracted contours,
and are provided with just enough ambient space to contain them. In both
works, however, we glimpse a distant landscape towards the edge of the pic-
ture: at the left in the Elijah, at the right in SS Paul and Anthony. In the sky
42 (facing page) at the left in the Elijah we are also given a distant glimpse of the prophet
Savoldo, St Matthew
being carried to heaven in a chariot pulled by horses, observed by a group of
and the Angel (illus. 53),
detail. gures on the ground: this sort of detail, a celestial event that seems almost
to burst through from another world, recurs elsewhere in Savoldo’s art. Nor
43 (page 80) Savoldo,
is there uidity of movement; on the contrary, in both pictures the gures
Elijah Fed by Ravens,
1512–15, oil on panel exhibit a certain rigidity, almost as if they were sculptures, with an emphasis
transferred to canvas. on joints, as Gilbert also remarks (G, p. 87). Elijah sits before a rocky outcrop
with his legs somewhat apart, his bare left foot planted on the ground (as
44 (page 81) Savoldo,
SS Paul and Anthony, is the other foot, but we see only a bit of it), his left hand on his left knee,
c. 1516, oil on canvas. supporting his bearded head with his right hand as he looks up towards his
82 painting with dem ons

left (our right), where a raven bearing food in its beak perches on a twisted
branch. In the Venice painting the two hermit saints sit before a similar
outcrop, Anthony on the left with his arms and hands crossed on his breast
(right over left), Paul on the right with his hands brought together in prayer
(not quite clasped, with the ngertips lightly touching). This is a traditional
contrast of prayerful gestures that will turn up again in Savoldo’s oeuvre.
Like Elijah, Anthony, with a balding head and white beard and moustache,
looks up towards a raven bearing food; while Paul’s gaze seems to have no
precise object, his face is more nearly turned towards the viewer and his
expression with eyebrows raised and eyes wide open carries a plangency that
draws the viewer’s attention. In both paintings the gures are ruddy, almost
45 Savoldo, Elijah
Fed by Ravens (illus. 43), terracotta, in hue, adding to one’s sense of their solidity. Drapery plays an
detail. important role in both, as it will continue to do throughout Savoldo’s career:
Elijah wears a luminous blue robe with
deep folds and a raspberry-red shawl with
a fur reverse and white highlights around
his shoulders; while Paul and Anthony
are in browns and greys respectively, with
Anthony wearing a robe, a darker cloak
and what seems to be a fallen hood of a
slightly different hue. In this canvas, too,
the fall of the cloth over both saints’ lower
bodies is marked by powerful folds.
It is the Elijah that rst attracted my
particular attention, for a simple reason. In
the lower left corner, several modest-size
brownish rocks rise vertically just this side
of the larger outcrop (illus. 45). When one
looks closely at these, in particular the rock
nearest the picture surface and the smaller
grey rock behind it, one becomes aware of
. . . heads and faces: at least two, somewhat
abstract and skull-like, on the nearest rock
and a third, especially clear-cut, on the
grey rock. Indeed the grey rock seems to
consist solely in a Leonardesque grinning
head facing downwards. I say ‘one becomes
aware’, but in fact the heads and faces have
 faces 83

never been remarked, so far as I can tell; I nd that surprising, obviously,
but will refrain from speculating as to why and how they have been missed.
What instead I want to stress, to begin with, is that they are an extreme
expression of a tendency in Savoldo’s art towards what I will call, for want
of a better term, internal mimesis, or auto-mimesis – a tendency that, as
will emerge, takes a number of different forms in his work throughout his
career (in Chapter Two I called attention to an instance of this in the London
Penitent St Jerome).
Nor are the heads the only instance of such mimesis in the Elijah.
Consider the striking afnity, as it seems to me, between the raven on its
branch and Elijah’s upper body, in particular the way in which the reddish
shawl tied around his shoulders may be seen as miming the raven’s wings
(not that we exactly see the raven’s right wing, which seems as though it
should be outspread but may not be), or indeed vice versa. There is also the
further respect in which Elijah’s arm and hand supporting his head bear an
analogy with the branch on which the raven is perched (the raven’s body
plus the bread in its beak would therefore be approximately matched with
Elijah’s head). At another scale, is there not a rough accord between, on the
one hand, the divided shawl and the raven’s wings and, on the other, the left/
right division of the leaves in the small plant at the lower right? For that
matter, is there not a sense in which the rocky mass with horizontal stria-
tions to the right of and slightly above Elijah’s head may be seen, or should
I say imagined, as a large head-like structure in its own right looking back
down at him? (Compare the rocky mass covered at the top with hair-like
vegetation confronting at close range, as if head to head, the bearded saint
in the London St Jerome.) And is there not a hint of something like inter-
nal mimesis in the relationship between Elijah’s powerful left hand with its
strong ngers sculpted by light and his equally impressive left foot almost
directly beneath it? I am not quite claiming that empathic projection is
invited by the treatment of the hand, or for that matter the foot, but the sheer
pictorial authority of both points towards future developments.
In SS Paul and Anthony, in which another raven brings sustenance to the
two hermit saints, there is at rst glance nothing quite like the heads and
faces at the bottom left of the Elijah. But the dark rocky mass directly beyond
Anthony’s largely bald head not only frames but in a general way repeats or
mimes that head, while in the case of Paul (on the right) the dark grey rocks
just beyond him and to his left (our right) are divided in two vertically, as
are the similar rocks just behind his head, a division I see as thematically
84 painting with dem ons

linked to the not-quite clasping of his hands. In this picture, too, full-blown
empathic projection on the part of the viewer is not elicited, though one
can sense, knowing what will happen, that it is not far off. But there is a
strong and subtle sense at once of contrast and of interrelation between the
two saints: Anthony in his more elaborate robe, cloak and hood, Paul in his
simpler, rougher garment leaving his arms bare; Anthony’s feet as if deco-
rously out of sight, Paul’s naked feet in plain view; Anthony in prole and
with his eyes in shadow distinctly composed and collected (within himself,
one might say, even as he gazes calmly towards the raven), Paul, his face illu-
minated, in a more labile state, as if in spiritual distress (in the brief narrative
of his life in the Golden Legend he was near death when Anthony visited him).
But the strongest contrast between the two saints concerns their respec-
tive hands and arms, Anthony’s crossed over his chest signifying contain-
ment, Paul’s hands not quite clasped in prayer, his ngertips barely touch-
ing. My sense is that the viewer registers the contrast, and what makes it
all the more telling is the way in which the parallelism of Anthony’s legs
beneath his drapery invites comparison with Paul’s hands, while the almost-
crossing folds of Paul’s garment over his thighs express a similar relation to
Anthony’s arms, which is to say that the contrast between their actions also
provides the basis for a chiasmatic mimetic relationship between the two
gures. Another mimetic element is the palm tree, a source of food for Paul,
rising just beyond them, at once separating and sheltering them.
Three further observations are apropos before moving on. The rst is
that one has the denite impression that both saints are in effect portraits
of specic persons, by which I do not mean to suggest that the painting was
commissioned with that end in view, but rather that already by this time
Savoldo very likely proceeded by rst making portrait drawings of models
whom he thought suited his subject matter, and then developing these into
the gures in the nal work (a ‘naturalistic’ practice). On the strength of
SS Paul and Anthony, it is clear that Savoldo’s skill in this regard was already
considerable. The point is worth stressing because the portrait as a genre
was not a speciality of his as it was of his great contemporary Lotto, but his
drawings of individual heads are masterly.
The second point is a different form of observation: if one looks closely at
Paul’s garment one sees that it has not been painted a single colour (a con-
tinuous brown, which is the impression it conveys at a distance). Instead
it comprises a plethora of small, squarish areas within which a number of
parallel brushstrokes in brown overlay a somewhat lighter background, the
 faces 85

strokes in each area running at right angles to the strokes in the adjacent
ones (see illus. 74). One result of this is that the garment also conveys a feel-
ing of texture, consistent with the Golden Legend claim that it was woven
from palm leaves, as opposed to the smoother-seeming robe, cloak and hood
of St Anthony. But what I want to emphasize is that all this becomes evident,
and becomes fully visible, only at extremely close range, within a few feet
or less of the canvas. In fact it is quite easy to miss, although an unusually
informed viewer would have known to look for it. This leads me to go fur-
ther and remark that Savoldo’s paintings often posit precisely such a physi-
cally proximate relation to the viewer, as for example in the Death of St Peter
Martyr with its suggestion of the angel descending with the palm of martyr-
dom, not only from above the upper framing edge but also just ‘this’ side of
the picture surface. More broadly, the very operation of empathic projection
focusing on the pictorial subject’s hands, as I have described it at work both
in the Death of St Peter Martyr and the self-portraits and other paintings
discussed in Chapter Two, requires that the representations in question be
pitched seemingly in close proximity to the viewer – closer, certainly, than is
normally the case in the art of Savoldo’s contemporaries. Whether this basic
fact had something to do with his relative lack of success, as related by Paolo
Pino and as suggested by his limited oeuvre, it is impossible to know.
Finally, something should be said about the treatment of drapery in the
painting as a whole, both because it is unusual in its own right and because
it too anticipates future developments. In fact commentators have always
been aware that Savoldo’s highly idiosyncratic use of drapery is one of the
hallmarks of his art, Longhi early on seeing in it a precedent for Caravag-
gio’s.2 For his part, Freedberg notes Savoldo’s propensity for ‘an exaggerat-
edly abundant drapery, strong in colour and brightly lit, disposed around
plastically large but inert forms’ (PI, p. 227). For the moment, though, I
want simply to acknowledge the unusual forcefulness as well as the seem-
ing arbitrariness of the treatment of folds and the like in both the Wash-
ington and Venice paintings. Look in particular at the folds in the drapery
covering Anthony’s shins and feet: nothing could be less revealing of the
conguration of the underlying limbs; rather, the folds make two large quasi-
physiognomic or perhaps more accurately quasi-mask-like constructions, the
dark angled folds behind the knees seeming almost like eyes and the large
upside-down V-shaped folds with their strange ‘inner’ workings (at least
in the nearer of the two, the one covering the saint’s right shin) almost
like open mouths, though not quite human: the entire motif has something
86 painting with dem ons

animal- or gargoyle-like about it. I realize that this may be asking a lot of the
reader at this early stage in my argument, and in fact there is in SS Paul and
Anthony another, less conjectural grotesque face that I will hold off noting
until later, which need not discourage the reader from searching it out before
moving on.
I shall have much more to say about faces in Savoldo’s drapery as I pro-
ceed, but at this juncture it will be useful to consider another early painting
marked by what I have called internal mimesis, if only to clinch the point
that this was basic to Savoldo’s operations at this stage in his career: the
Lamentation over the Dead Christ (1515–20; illus. 46).3 A full consideration of
the Lamentation would have to take into account its relation to various cop-
ies, to the unresolved question of its initial destination, and to the likelihood
that at an early moment it was cut down at the right. But for present pur-
poses what should be noted is, rst, the conspicuous matching of the trickles
46 Savoldo, of blood from the wound in Christ’s side with the ngers of his dead hand
Lamentation over the Dead
Christ, 1515–20, oil on supported near the wound by Nicodemus and, second, the equally striking
poplar. doubling of Christ’s left hand with its wounded palm with Mary’s left hand
 faces 87

gesturing slightly above it and to the right. The latter doubling is noted in
passing by Gilbert (G, p. 88), though it is worth spelling out that there is no
question of empathic projection being inspired by any of these hands. Less
obviously there is also the close doubling of Mary’s weeping face with the
Magdalene’s partly obscured face immediately to her right (our left), as well
as, in a minor way, the semblance between the handkerchief-like cloth that
Mary has raised to her face with the piece of cloth, no doubt belonging to a
larger length of fabric, around Nicodemus’s left hand as it rests on Christ’s
left shoulder. In this connection, too, one might mention not the parallelism
but the contrast between Nicodemus’s head and features and those of Christ.
The two heads are in immediate proximity to each other, the rst alive
and warm in colour (half illuminated, half in shadow), the second dead and
discoloured (albeit illuminated, a brilliant stroke), the point being that so
emphatic a contrast, while not mimetic in the sense I have been tracking,
amounts nevertheless to a strong relationship between pictorial elements
(much the same might be said about the contrast between Nicodemus’s and
Christ’s right hands). The larger point would therefore seem to be that
for Savoldo at this early stage, at least in the Elijah, SS Paul and Anthony
and Lamentation, a painting in progress was treated by him as a relational
eld, the relations in question being, as I have said, internal to the work.
Soon, however, as we have seen, the relational axis will shift through ninety
degrees to engage the viewer, the rst such viewer being in every case the
painter, which is why the self-portrait surfaces in his art as basic to his
operations starting as early as the large Treviso altarpiece of 1521 (see
illus. 76), with the gure of St Liberale in armour gazing directly out of the
picture at the extreme right, his body mostly hidden except for his head (as
Gilbert, again, remarks).4

A l l th i s i s i m po rtan t, I think, but there is another aspect of Savoldo’s

early career that has so far gone unmentioned here except in the most gen-
eral terms: his interest in Northern art, which in fact marks his entire oeuvre,
and in particular in the sort of Northern art associated with Hieronymus
Bosch and artists like him (but also others including Albrecht Dürer, Lucas
Cranach, Martin Schongauer, Joachim Patinir and Hugo van der Goes).5 Two
works by Savoldo, both Temptations of St Anthony, are crucial here: the rst,
and almost certainly the earlier one, in the Timken Museum of Art in San
Diego (1515–20; illus. 47), the second in the Pushkin Museum in Moscow
88 painting with dem ons

47 Savoldo, Temptation (c. 1520; illus. 48).6 I will conne my remarks mainly to the rst, having seen
of St Anthony, 1515–20, the second only in reproduction.
oil on panel.
In the San Diego panel the composition is divided in two, with a large,
brownish, geologically improbable rocky mass sweeping upwards in the
middle distance; towards the right we see a nightmare landscape of ery
skies, bizarre structures, Boschian monsters and ruddy-skinned, naked,
eeing humans (some of whom have been caught and are being devoured by
the monsters), while to the left are daylight and the appearance of normal-
ity – elds, woods, a road, buildings, a few strange dark rock formations,
and in the distance blue mountains and cloudy skies. In the left foreground
the painting’s protagonist, the bearded saint, in a skullcap and a dark robe
overlaid by a twisting white tunic, ees mid-stride towards the left, looking
back over his shoulder at the hellish world behind him while raising his not
quite clasped hands in prayer. In an obvious sense the central rocky mass,
though more spectacular, is similar to the smaller mass behind Elijah, even
to the grassy stubble and the bits of twig-like growth at the top of both;
more to the point, the entire rocky structure has a bodily feel, plus there is
towards the bottom right, just above the gures and creatures exiting from
 faces 89

a cave-like space, a dark indentation in the rock that has the distinct sense 48 Savoldo, Temptation
of an eye, which would make the cave a kind of mouth and that entire por- of St Anthony, c. 1520, oil
on panel.
tion of the rocky mass a grotesque quasi-face. In the same vein, once alerted
to the possibility that physiognomies of one sort or another are in play, the
two lighter-coloured rocks immediately above the cave and to its left reveal
themselves to be viewable as a head in prole with a large forehead confront-
ing at point-blank range a smaller animal-like creature with feline ears and
a kind of tail. There also seems to be a rudimentary hand or paw with claws
beneath the two heads. Nor is this all: if one then raises one’s gaze to just
above the dark space above the creatures (or creature-like forms) I have just
described (illus. 49), one becomes aware of another face in prole on the front
of the ‘roof ’ of the space (seemingly bearded, seemingly looking down to the
left), and then inside the dark space and to the left another cartoonish face
with large round eyes and a horizontal slit-like mouth. Also (once started it
is hard to stop) it is tempting to discern still another cartoonish face with
round dark eyes to the right of the bearded face, and just possibly a head
in prole at the left of the rocky mass, silhouetted against the distant land-
scape. (So far as I know, none of this has been noted in the literature on the
 faces 91

painting.) Finally, might there not be something quasi-mask-like – two eye

sockets, a nose, a gaping mouth, all facing upwards – in the blown-back folds
in the saint’s white tunic where it crosses his upper left thigh, as well as not
quite a face but nevertheless a snoutish (mouse-like?) physiognomy with two
small eyes in the swirling convoluted end of the white tunic, just to the left of
the tiny creature reading a book? In a different vein, the large, dark rock for-
mation to the right of the central mass, proled against the hell-like ames in
the distance, has something monstrous about it, especially the suggestion of
an open maw full of menacing teeth. There is also, less tendentiously, a clear
resemblance between the landscape to the left in the St Anthony and the more
modest one in the Elijah, but what in addition I nd striking is the afnity
between the charming, or at least in no sense alarming, stone bridge-like
formation in the far middle distance in the Elijah and the sinister-seeming
dark rounded semi-arch-like form to the immediate right of and connecting
with the stony mass in the Anthony.
A quick glance at the Moscow version of the subject nds the same
two-part division with a central rocky mass that seems even more semi-
anthropomorphic (or semi-monstrous) than in the version in San Diego.
Here, however, the white-bearded saint in a voluminous red robe and cowl
has fallen backwards in terror as hideous animal-like beings crowd in on
him, tugging at his long scarf and the hem of his robe. To the right, presum-
ably seen by the saint, an almost naked man in a strange loincloth enters
bearing a naked gure on his back, the gure plainly human except for a
horrid bird-skull-like head. This pair has always been seen as based upon, in
that sense referring to, a comparable motif in Raphael’s Fire in the Borgo in
the Stanza dell’Incendio of the papal apartments.7 Knowing the painting only
in small-scale reproduction, my impression that it comprises numerous
physiognomic elements of the sort just discussed remains merely that, an
impression, albeit a strong one.
More than thirty years ago a young scholar in a short article in the Art
Bulletin pointed to certain specic sources in Northern art for the two
paintings: a print by Lucas Cranach, itself derived from one by Martin
Schongauer (1506; illus. 50), and Hieronymus Bosch’s Last Judgement trip-
tych today in the Groeningemuseum, Bruges (1486; illus. 51).8 Savoldo’s
citations from the latter have been taken to suggest that it might then have
been in Italy, but in any case recent scholarship, culminating in the land- 49 (facing page)
Savoldo, Temptation of
mark 2017 exhibition Jheronimus Bosch e Venezia in the Palazzo Ducale, has St Anthony (illus. 47),
drawn attention not only to the circulation of Northern artists and works detail.
92 painting with dem ons

50 Lucas Cranach the in early sixteenth-century Venice but also in particular to the presence of
Elder, Temptation of several important paintings by Bosch in the collection of Cardinal Domenico
St Anthony, 1506, wood-
cut in black on cream Grimani between roughly 1516 and 1523, where Savoldo would surely have
laid paper. viewed them.9 For the present study, however, the important point is not the
question of inuence in a strictly art-historical sense so much as the fact
that the impact of a Northern, in particular a Boschian, vision of a mon-
strous, demonic and oneiric world on Savoldo’s pictorial imagination appears
unmistakable, and not only in the St Anthony panels.

Wh at i h av e i n m in d is likely to seem surprising even to readers famil-

iar with Savoldo’s art – at any rate, I have never seen the topic I am about
to raise addressed in the secondary literature. In effect I have already raised
 faces 93

it, but now I mean to pursue it. Very simply, it concerns his treatment of 51 Hieronymus Bosch,
drapery, in particular his altogether personal, indeed highly idiosyncratic Last Judgement, 1486,
oil on oak panel, triptych.
tendency to proliferate folds, bulges and indentations in a manner that often
bears little or no relation to the underlying form of the body, but instead
tends towards a kind of dynamic autonomy, one might even say anarchy,
unlike anything else in Italian art with which I am familiar.10 A painting
already discussed, the Penitent St Jerome in London (see illus. 33), is a case in
point, the entire expanse of the saint’s red-purple garment below the waist
being a zone of arbitrary-seeming folds, bulges, indentations, pockets, high-
lights and shadows with only the most minimal relation to the limbs that
presumably it covers and the ground on which its lower inches rest; in fact
94 painting with dem ons

the bottom framing edge cuts through the drapery before it comes to an end.
As has been mentioned, Longhi in an early article found in Savoldo’s drapery
a precedent for Caravaggio’s (Longhi had in mind Savoldo’s Shepherd with
a Flute in the Getty, illus. 36), but in fact there is nothing in Caravaggio’s
oeuvre even roughly comparable to what we nd in the St Jerome. By this I
mean that, while it is perfectly true that drapery in Caravaggio is often con-
spicuous and ‘excessive’, he never appears to proceed, as does Savoldo in the
St Jerome and elsewhere, with a sense of near-total improvisational free-
dom, as if precisely in that regard, in the treatment of that seemingly par-
ergonal element (in Derridean lingo), his art could all but slip conventional
constraints in the interest of what might certainly be described as expres-
siveness of a sort, but of what sort?11
In the rst place, an expressiveness keyed to the implied state of mind of
Jerome himself, as he leans forwards gazing intently at the smallish gure
of Christ on the crucix, his right arm and hand in the act of preparing
to strike his breast with a stone, his left hand making the convulsive ges-
ture as if clutching or grasping discussed in Chapter Two. Note, by the way,
how the vertical division in the rocky mass to the right of Jerome may be
seen as miming Christ’s legs, or for that matter the facing-page division of
the saint’s Bible. Then there is the way in which the darkish downward-
turning indentation in the drapery to the right of, and slightly below, his
right hand clutching the stone may be seen as subtly emphasizing the action
of the saint’s right arm, an action that at once is perfectly traditional (Jerome
in the act of beating his breast with the stone) and wholly original, with
the arm and hand neither drawn further back and so silhouetted against
the sky or landscape, nor having just delivered the blow and so in contact
with the saint’s breast, but rather viewed as if in mid-trajectory, fore-
shortened against the saint’s body, which is to say against the red-purple
drapery itself. The risk of a decision to frame the arm and hand in this man-
ner is that they might thereby be made less dramatically visible than would
otherwise be the case, indeed that they might become momentarily lost
against their immediate background, despite the emphatic shadowing of the
arm’s forward contour (and the highlighting on the drapery immediately
beneath it). But by partly miming or matching the saint’s action, in par-
ticular by underscoring the position of his hand and wrist, the downward-
turning indentation helps ensure that the viewer never loses sight of the
main action, which however is then complemented by the differently focal-
ized, also sharply illuminated, grasping gesture of the saint’s left hand, with
 faces 95

its invitation to the viewer to respond to it empathically, as discussed in

Chapter Two.
Then there is this, the direction in which I have been moving all along,
not without a certain anxiety. Towards the very bottom of the same length
of drapery that contains the downward-curving, dark inward fold that I have
just discussed, one nds a cluster of indentations, together with shadows,
highlights and bulges, that I cannot help seeing as a face or mask: two eyes
in darkish sockets, a mouth, a possible nose, a jaw and chin (illus. 52). The
eyes are set wide apart, the whole face somewhat squat, the mouth open and
slightly out of alignment, with an unreadable but not particularly friendly
expression. I have to assume that the reader sees what I am seeing, now
that it has been pointed out. The question, unevadable, is what exactly to
make of it. For a start, was it, the face or mask, consciously intended as such
by Savoldo? My rst stab at an answer is that I doubt it, but is that out of
the question? Remember the heads and faces in the rocks at the bottom left
of the Elijah, clearly intended as such, in my view. This must also be true
of the heads and faces in the central rocky mass in the San Diego Temptation
of St Anthony. In the light of the face or mask in the drapery in the Penitent
St Jerome, what exactly should one make of the strange triangular structures
of folds in the lower drapery of Anthony in SS Paul and Anthony, which I
52 Savoldo, Penitent
have already suggested have something face- or mask-like about them? In St Jerome (illus. 33),
fact, looking again at the Elijah, might there not be the hint of a raven-like detail.
96 painting with dem ons

head with a strong beak in the folds immediately above his right foot? And a
related question: assuming for the moment that Savoldo did not specically
intend the face or mask in the Penitent St Jerome as such, once it was in
place did he recognize that it was there, that it had materialized in some way
or other, and did he then in effect knowingly accept it as belonging to the
structure and meaning of his painting? Or did it simply escape his notice?
For the moment let the question remain open. And if we now let our atten-
tion drift to the rather dense concatenation of folds and bulges at the lower
left of Jerome’s robe, to the left of the face or mask I have been discussing, is
there not a sense, now we have become just a bit sensitized to the possibility
– which is to say once we have begun to mobilize a certain empathic projec-
tion in this context also – in which that concatenation too might be seen as
suggesting a somewhat collapsed face or mask of sorts, with deep eye sock-
ets, a vertical divide between them, no obvious nose but a perfectly plausible
mouth slanting downwards to the right at the very bottom of the picture?
Needless to say, it would not surprise me to nd that even a reader sympa-
thetic to my general approach to Savoldo might refuse to go along with the
latter set of proposals. But I would nevertheless hope that he or she would be
struck by a hint of something like purposiveness – but to what end? – in the
very density of the collapsing cascade of folds and bulges in the region I have
described as implying the possibility of a second mask-like form.
Another painting makes relevant viewing in the light of these con-
siderations, St Matthew and the Angel (early 1530s; illus. 53),12 by univer-
sal agreement one of the works singled out by Vasari when he wrote of
having seen four ‘nocturnes with res, very beautiful’ in the mint of Milan.13
(As mentioned in the Introduction, Savoldo evidently worked for the Duke
of Milan in the 1530s.) For Longhi, as Andrea Bayer remarks in the Painters
of Reality catalogue, the St Matthew was the picture by Savoldo that more
than any other looked forward to Caravaggio: ‘To [Longhi’s] eye, every
aspect of the work conrmed his belief: gure type, illumination, mood.’14
And in fact the St Matthew is one of Savoldo’s most inspired creations: the
Evangelist, still relatively young, with a dark beard and powerful build
possibly reecting Savoldo’s own, and wearing a rose-red blouse with a
simple neck and owing sleeves, sits at a table writing with a quill pen
on a single sheet of paper (or parchment?). This would have been under-
stood by Savoldo’s contemporaries as anachronistic, like the open Bible
in the Penitent St Jerome, the idea evidently being to make the images in
question as directly accessible to contemporary viewers as possible, which
 faces 97

was standard procedure in sixteenth-century Venetian depictions of the 53 Savoldo, St Matthew

saint. In a characteristically Savoldesque invention, though, the table surface and the Angel, early 1530s,
oil on canvas.
runs across the bottom of the canvas. Barely above the bottom framing edge,
indeed touching that edge, a small oil lamp provides an unlikely degree of
illumination. Just beyond it the sheet of paper is being written upon by Mat-
thew, his right hand wielding a pen and his left hand gripping an inkwell
(anchoring the composition at the bottom, that is, making it likely that the
viewer begins his or her engagement with the picture precisely there). At
the moment depicted, Matthew appears to have paused in his writing to
receive inspiration from a Leonardesque angel with a dark blue-purple robe
and darkish wings. In greater detail, the writer, his face partly in shadow,
raises his handsome head and turns it towards the angel half-hovering
further back in space, whose face we see mainly in prole but which has
98 painting with dem ons

been turned and tilted feelingly towards Matthew. The impression conveyed
is less of mutual eye contact than of intense rapport. In fact we have the
sense of viewing the angel partly from the rear, which is to say that there is
a certain afnity between this composition and that of the Death of St Peter
Martyr, for all the disparity between their respective subjects. The quality
of communication between angel and Evangelist is exquisite, a triumph of
nuance, tenderness, suggestion. At the same time, the subtle discrepancy
in size between the two – Matthew rmly planted at his table, the angel
almost dreamlike in its relative indenition – underscores the difference in
existential registers, as in the Death of St Peter Martyr, between everyday
reality and a visitation of the divine, as does the intimation of movement
in the treatment of the winged visitor. It is not to take anything away from
Caravaggio’s profound imaginativeness, but his own versions of the subject
are distinctly less inspired. In the background to the right and left are scenes
that have been associated with Matthew’s life as narrated in the Golden
Legend; both scenes are themselves minor triumphs of nocturnal illumina-
tion, the one at the right featuring a re in a hearth, the one at the left a
moon in a night sky.15
In a compositional stroke that by now should be familiar, Savoldo situates
the Evangelist’s hands in extremely close proximity to the picture plane. As
the painting stands, however, the viewer is not thereby invited to respond
empathically, either because Savoldo on this occasion deliberately toned down
his usual masterly depiction of hands so as to make these unavailable for such
a purpose, or because – far more likely – the hands belong to a portion of the
canvas that suffered badly at some time in the past. In any case, the viewer is
placed directly before the scene, so near that one nds oneself trying to read
the writing on the sheet of paper (all one sees is a few rows of marks). What
turns out to be emphasized in the St Matthew is, rst, the marvellously imag-
ined lyrical communion between angel and Evangelist, and second, a very
different matter, the extraordinary presentation of the latter’s owing blouse,
red with dazzling white highlights, radiant beyond all rational justication
(the tiny oil lamp could not remotely have produced such an outcome), a spec-
tacular, all but phosphorescent expanse of folds, highlights, valleys, shadows
and the like with not even an approximate equivalent that I can think of in
the art of Savoldo’s contemporaries (there is nothing like it in Lotto’s work,
for example).
At this point I want simply to urge the reader to consider the blouse in as
receptive or empathic a state of mind as possible – in other words, not simply
 faces 99

taking the overall cascade of folds for granted either representationally, as

what such a blouse composed of the appropriate material would actually look
like (assuming greater illumination than that provided by the little lamp),
or merely stylistically, as characteristic of Savoldo’s pictorial handwriting,
which of course it is. The alternative I want to propose, here as elsewhere
in his oeuvre, is to ‘actively’ or, say, projectively view the folds and bulges as
continually hinting at one or another quasi-physiognomic visual metaphor,
even as the precise terms of the latter mostly escape clear denition. But
not, I think, entirely: for example, about midway down the saint’s left sleeve
there is a bunching of folds that, after multiple viewings (and much internal
debate), suggests to me a face or a mask – at any rate, I propose that the
bunching may be seen as evoking two eyes slanting outward left and right,
and below them a squat or attened nose, and below that a long downward-
curving slit of a mouth, the face or mask turned slightly to the saint’s left.
Of course I recognize that the suggestion could scarcely be more tenuous,
and that if the reader nds it unpersuasive there is no effective means by
which to counter his or her disbelief. Or, to shift attention to the saint’s right
hand holding its pen, and then to the improbably ample ‘cuff ’ of his sleeve,
in particular to the bit of the ‘cuff ’ immediately above the hand (illus. 54), is
it out of the question that one discern here a kind of semi-monstrous mask-
54 Savoldo, St Matthew
like or gargoyle-like face, two dash-like eyes and a slanting nose with the and the Angel (illus. 53),
writer’s hand issuing from an open mouth? At any rate, such a reading quite detail.
100 painting with dem ons

unexpectedly struck me with considerable force when I encountered the

St Matthew, with which I had long been familiar, in a small exhibition at the
Palazzo Barberini in 2017.16 Here too, however, the visual evidence for such
a reading is hardly dispositive. And even if it were provisionally granted that
one or another face or mask of the sort I have been evoking might be taken to
be ‘there’, the question of its signicance – of its meaning in and for Savoldo’s
art – remains to be determined.

I s h a l l late r have occasion to direct attention to faces in other works by

Savoldo as well as to pursue the question of their signicance, but one cluster
of remarkable canvases bears closely on the topic: the Magdalenes. Even more
than Matthew’s blouse in St Matthew and the Angel, the Magdalene’s shawl
in the three versions of the subject I have seen – in the National Gallery, Lon-
don, the Gemäldegalerie in Berlin and the Getty Museum in Los Angeles,
with the shawl treated differently in each of the versions as to colour and the
pattern of folds and highlights – provides an extraordinary eld for pictorial
enlivenment and what I have been calling empathic projection. In each of the
three versions such projection discovers the strong possibility of faces in the
folds, though especially as regards the London and Berlin canvases the very
notion of possibility may be felt to err on the side of tentativeness. To begin
with the London painting (see illus. 39) and its radiant silver shawl, the
most sheerly ravishing of the three canvases, there are two candidates. The
rst and less immediately persuasive may be found high on the saint’s left
shoulder, taking the narrow highlighted fold running at a slight angle from
right to left as a long nose, with two eyes or eyebrows aring off from it at
the right-hand end and an outsize open mouth under it, seen as if in prole,
making the entire head or mask into a kind of mocking caricature. As for the
second candidate, just below the mouth of the rst in the shadowy part of the
shawl, in the vicinity of the saint’s elbow, one nds two round depressions
that offer themselves as eyes (or eye sockets), and then a perfectly plausible
ski-slope-shaped nose pointing towards the left, and beneath that an equally
plausible small, partly open mouth, as if half-smiling, or in the act of speak-
ing. Elsewhere the folds in the shawl resist interpretation even as they draw
the eye, making it linger over one or another stunningly beautiful passage,
indeed teasing it here and there with intimations of further meaningfulness,
in particular – it takes a while to come to terms with this – that of a small
grotesque head and face with an impossibly long, downward-plunging nose
 faces 101

in the drapery covering the saint’s right hand. (See also the various small
indentations in the portion of the shawl covering the saint’s forehead, as well
as the round, small, dare I say moon-face-like indentation immediately to
their right.)
In the Getty Magdalene (see illus. 41) the conspicuously seamed orange-
brown drapery offers one strong candidate for a face of sorts, the pike-like
creature gliding downwards from right to left with open jaws just below
where the Magdalene’s left elbow would be (its upper contour being the
seam itself). There are two weaker ones, the pattern of folds just above the
pike’s head (a nose angled to the left, an eye socket-like depression, a mini-
mal mouth, none of these as denite as the features of the not dissimilar face
in the London version), and the light vs dark, left-facing mask-like form in
prole just to the right of the Magdalene’s ointment jar in the lower left cor-
ner of the painting, its nose a hook-like stub trending up, its eye in shadow,
its mouth open.
But perhaps the Berlin Magdalene, with its colouristically hard to spec-
ify shawl, takes the prize. In the rst place the shawl bodies forth a much
more dramatic congeries of folds and depressions than those in the other
versions; and in the second the suggestion of a monstrous grinning Bosch-
like face or mask in the region of the saint’s upper left arm, in effect between
shoulder and elbow, seems to me – again, dare I say it? – all but unmistakable
(illus. 55). Specically, I have in mind the face or mask comprising a right eye
or eye socket, a large downward-sweeping, hooked and highlighted, beak-
like nose (perhaps comparable with the bird-skull-like head of the creature
being carried on his back by an almost naked man at the right of the Moscow
Temptation of St Anthony), and directly below the nose a hideously grinning
mouth with shark-like teeth. Beneath the face or mask is a sequence of deeper
serpentine indentations that, while not quite offering themselves to physiog-
nomic reading, nevertheless disturbingly engage one’s eye and mind (one’s
bodily imagination), as if some further empathic metaphorics were poten-
tially in play.
I say ‘Bosch-like face or mask’, using an adjective that could also be applied
to the ‘pike’ at the Getty, but in fact the possible sources of such imagery
far exceed the work of any single artist. For example, there seems to me an
intriguing resemblance between the face or mask in the Berlin picture and
the monster in Giulio Campagnola’s engraving The Astrologer (1509; illus.
56), which Savoldo surely knew, although the monster does not have savage
teeth.17 It does not follow, however, that Savoldo would have intended the
 faces 103

connection, and in any case the meaning of such imagery in paintings of the 56 Giulio Campagnola,
Magdalene, which is to say in works with a Catholic subject but also ones The Astrologer, 1509,
engraving.
that feature a beautiful and alluring woman (and, if Pardo is correct, imply
the near presence of Christ), remains an open question. Necessarily, there
will be more to say about Savoldo’s phantasmic faces and their possible
signicance before this book is done.

55 (facing page)
Savoldo, Mary Magdalene
(illus. 40), detail.
 fou r • M AG I C

E a r l i e r, towa r d s t h e e n d of the Introduction, I offered as prelimi-

nary evidence of how much has not been adequately seen in and about
Savoldo’s art, much less intellectually taken into account, the observa-
tion that the subject of his universally lauded painting Tobias and the Angel
(mid-1520s; see illus. 2), is not, as has invariably been held, the angel Raphael
pointing to the sh emerging from the river at the lower left corner of the
canvas.1 Instead the angel is drawing or summoning the sh out of the water
with his extended arm and hand, a ‘magnetic’ or quasi-‘magical’ operation
for which there is no precedent either in the book of Tobit or, to the best
of my knowledge, in previous painting. In fact commentators have often
remarked on the originality of Savoldo’s interpretation of the subject. Gil-
bert, for example, nds in the Borghese picture ‘a complete transformation
of traditional iconography, in a more obvious way than the Matthew. The
Tobias and the Angel pictures of the fteenth and sixteenth centuries show
the gures walking along. Small wonder that their stationary, meditative
monumentality here has evoked the name of Giorgione’ (G, p. 376). Again,
the prevalence of stylistic categories, in this case ‘Giorgionesque’, tends to
displace attention away from the deeper stakes of Savoldo’s enterprise.
Gilbert also describes Tobias and the angel Raphael as ‘pausing to con-
template each other’ (G, p. 370), which I don’t think is accurate: the angel’s
gaze seems directed towards the sh (as one would expect, given the larger
meaning of his action), while Tobias’s is focused on the angel’s face. In fact
there is something marvellously nuanced in Tobias’s bodily position – a hint
57 (facing page)
Savoldo, Tobias and the of tension, even recoil in his head and upper body as he responds with dawn-
Angel (illus. 2), detail. ing comprehension to the angel’s gesture and perhaps also to his words. Note
106 painting with dem ons

58 Savoldo, Tobias and in this connection the subtle ribbon of light just beyond Tobias’s shadowed
the Angel (illus. 2), detail. prole, which in effect prevents the latter from being lost to view against the
more distant trees. That is not all: Tobias’s right arm and hand might strike
one at rst as pointing towards the sh, but on a closer look it feels more
accurate to say that he too is summoning the sh with his right forenger,
not deliberately but as if unknowingly, mimetically, under the inuence of
the angel’s ‘magic’ (illus. 58). If this might seem too strong, I am convinced
that it simply describes what is going on. As for the angel’s right hand, once
again it occupies almost the exact centre of the composition, as if offered to
the viewer for empathic projection, at any rate up to a point. But perhaps it
is the painting as a whole, or at least the angel’s action in its entirety, that
demands to be apprehended in empathic terms, virtually drawing the viewer
into the scene, subjecting him or her as well as Tobias and the sh to the
angel’s ‘magic’, lest the meaning of that action be missed – which is exactly
 magic 107

what has happened ever since the painting rst became the object of art-
historical attention starting in 1910, when it emerged from long obscurity.2
The failure to grasp the meaning of Tobias and the Angel is all the more
striking in that it is one of the most admired canvases (if not the most
admired) in Savoldo’s oeuvre. Bayer, however, does allude to the painting’s
‘almost magical mood’ (PR, p. 137). And Gaetano Panazza, who nds the
Tobias ‘stupendous’, sees the two gures as closed by ‘enchantment’ (incan-
tesimo) at the magic hour of dusk.3 Scholarly consensus tends to place the
Tobias in the mid-1520s, largely on the strength of various formal resem-
blances to the Adoration of the Shepherds in the Galleria Sabauda, Turin, to
which we shall turn shortly. There is the further possibility, strongly advo-
cated by Gilbert, that both the St Matthew and the Angel and the Tobias
were on display in the Milan Zecca or mint, where Vasari would have seen
them (G, pp. 434–4). Milan, of course, was where Leonardo’s inuence was
strongest, and the angel, like the one in the St Matthew, could not be more
Leonardesque. The landscape setting, on the other hand, as well as the
ickering lighting and the overall poetic mood of the scene tend to be com-
pared to Giorgione. The drapery on both gures, however, is purest Savoldo
with gleaming highlighted folds, deeply shadowed pockets (especially on
Tobias’s tunic, including an inverted heart-shaped one over his heart) and
passages that only barely stop short of suggesting partial masks or faces (for
example, at the angel’s waist near his left hand, or in the further length of
blue-grey drapery that falls between his thighs, or just below his right arm-
pit). The animistic-seeming drapery folds around the angel’s feet are also to
the point. Characteristic, too, are the large rock with a deep partial cleft on
which the angel rests his left foot, itself obscurely bodily (perhaps an auto-
mimetic ‘monstrous’ counterpart to the surfacing sh), and the sleeping dog
to its right, less ‘a particularly naturalistic touch’, as has been suggested,
than a participant in the dreamlike somatic network of the image as a whole.
Notice, too, the similar divided rock to the left of the angel’s right side; the
correspondence between the two rocks, once one notices it, is almost dis-
concerting. All this is to say nothing of the angel’s wings, which on the one
hand are widespread as if in ight and on the other anchor the angel within
the landscape setting to marvellous effect; or of the way in which the back-
ground foliage between the angel and Tobias has somewhat the force of a
distinct living presence in its own right; or, nally, of the smallish chinks in
the foliage through which we glimpse the sky at dusk, like so many bursts
of light.
108 painting with dem ons

59 Caravaggio, Calling But to return to the ‘magnetic’ gesture of the angel summoning the sh.
of St Matthew, 1599– I rst put forward this reading of Savoldo’s canvas by way of suggesting
1600, oil on canvas.
a source for what I took to be the analogous action of the gure of Christ
in Caravaggio’s Calling of St Matthew (1599–1600; illus. 59), one of his two
career-dening realist masterpieces in S. Luigi dei Francesi (the other, of
course, being the Martyrdom of St Matthew, briey discussed in Chapter
One).4 Specically, I suggested, rst, that the bearded gure seated at the
table and looking towards Christ and Peter while gesturing towards himself
(as if to say, ‘Do you mean me?’) is in fact Matthew; this by way of counter-
ing the recent proposal that Matthew is the younger man at the extreme left
who appears entirely occupied by the coins before him. Second, that Christ’s
‘pointing’ gesture with his right arm and hand is not one of pointing at all
but rather one of ‘magnetic’ summoning, like the angel’s in the Savoldo; this
is why it appears surprisingly unvigorous, a characteristic that has puzzled
art historians who have taken it to be based on the gesture of Michelangelo’s
 magic 109

God creating Adam. (In fact it came to strike certain scholars as more like
Adam’s hand than God’s, an observation that was then cashed iconologically
by the notion that Christ was ‘the new Adam’. As I remark in The Moment of
Caravaggio, I nd this unconvincing, art history’s way of saving the appear-
ances. There is a pointing gesture, but it is Peter’s, singling out the target.)
And third, I suggested that Caravaggio built on the ‘magnetic’ theme to
represent Matthew responding to Christ’s summons in three essentially
simultaneous moments: a moment of sheer priorness, expressed by Mat-
thew’s right hand still dealing with the coins on the table (and, signicantly,
mirroring the immersed youth’s right hand); one of initial uncertainty,
signied by Matthew’s gaze towards Christ and Peter and his questioning
gesture (‘Do you mean me?’, or simply ‘Me?’); and a moment of instant obe-
dience, as indicated by his muscular right leg in tights already beginning
to move, lifting his foot from the ground and taking the rst fateful step
towards martyrdom. Here, too, I take this reading to be incontrovertible.
And what is further suggestive is that, assuming the Tobias and the Angel
to have been on view in the Milan mint, it requires no mental gymnas-
tics to imagine the young Caravaggio standing before it, during his years
of apprenticeship to the Milanese painter Simone Peterzano, starting in
1584.5 So once again we are in a position to endorse the view of Longhi and
others that Savoldo provided an important precedent for Caravaggio, with
the proviso that in this case the basis for such a claim is not essentially a
matter of style (naturalism, lighting, treatment of drapery) but rather of the
two paintings’ imaginative structures.

Wi th in th e larg e r c o n te xt of the argument of this book, it matters

that the instrument of the exertion of ‘magnetic’ inuence in the Tobias is
the angel’s hand, a key Savoldesque motif, as has been seen (also his gaze,
but the hand is primary). Not that there is anything quite like the summon-
ing of the sh in any other work by Savoldo. But there is another motif that,
as I understand it, bears a denite relation to the summoning, namely the
actions of certain male gures relative to the infant Christ. To take the most
obvious example, in the Virgin Adoring the Child with Two Donors at Hampton
Court (c. 1523–8; illus. 60) we are shown the Virgin and the female donor at
the right gazing down at the naked infant Christ, who lies on a pillow (more
precisely, two pillows, a large one for his body, a small one for his head) while
raising his left leg into the air as if kicking it and his right hand in a gesture
110 painting with dem ons

60 Savoldo, Virgin that conveys a sense of incipient blessing.6 His left hand grasps a fold in
Adoring the Child with the Virgin’s robe. The Virgin and the female donor clasp their hands before
Two Donors, c. 1523–8,
oil on canvas. them, in the Virgin’s case with just the ngers touching (Savoldo’s norm for
such a gesture), in the donor’s, apparently, the entire hands. But what gives
the picture its surprising force is the action of the male donor at the left, who
gazes directly at the viewer as he delicately raises a darkish coverlet (brown
or dark red?) so as to expose the Christ Child to the Virgin and female donor
as well as to the viewer. He is a quietly impressive gure, with a thick head
of brown hair and a strong, well-trimmed beard, wearing a black robe with
a rich fur collar; one’s best guess is that he is in his forties; and his features
are strongly modelled by the fall of light from the upper left, casting the
left side of his face in shadow and giving him a density of psychological
presence that the other gures, much more evenly lit, obviously lack. His
mouth is closed, his gaze intent, and the double action of his hands – only
his forengers and thumbs actually hold the coverlet – could not be more
 magic 111

measured and exact. The setting is outdoors, with a rocky mass behind the 61 Savoldo, Adoration
Virgin, some earthworks and what may be a ruined church behind the female of the Christ Child with
SS Jerome and Francis,
donor, and a curving ruin of some kind and a glimpse of distant landscape mid-1520s, oil on canvas.
behind the man.
There will be more to say about this canvas, but rst it is necessary
to introduce another, the Adoration of the Christ Child with SS Jerome and
Francis, in the Galleria Sabauda in Turin (mid-1520s; illus. 61).7 The close
similarity between the two pictures is evident: they are roughly the same
size; even the settings are similar, though in the Turin Adoration in the place
of the female donor St Francis stands with outspread arms while gazing
at the Christ Child, and instead of the male donor one nds the bearded
St Jerome, distinctly older than the donor, here too gazing directly at the
viewer while lifting the coverlet with the same almost archly delicate ges-
ture as in the Hampton Court painting. In both cases, the action of a male
personage exposing the Christ Child has been found extremely puzzling by
112 painting with dem ons

scholars, though Gilbert suggests that it is perhaps understandable in the

case of a saint ‘but almost shocking in a donor as a casual token of familiarity
with the holy gures’ (G, p. 85). He continues:

It goes against the contemporary ethos; thus, in the exactly contempo-

rary Pesaro Madonna of Titian (nished 1526) the donors kneel and pray
as they would have in a Medieval image, and the same may be seen in
numerous examples of the same group by Palma Vecchio. The inexplicable
device in Savoldo [that is, in the Hampton Court picture] becomes under-
standable only if we see it as a repetition of the motive established by the
Saint in the earlier version.

In fact, the argument for the chronological priority of the Turin painting
rests largely on the observation that in it the coverlet is being lifted by a
saint rather than a mere donor, which is to say that there is no strong ‘exter-
nal’ or even stylistic evidence for the earlier dating. As is said apropos of the
Hampton Court picture in the standard catalogue of Italian pictures in the
Royal Collection,

It is now generally agreed that the Turin painting came rst, particularly
because the gesture of lifting the cloth to reveal the Christ Child seems
more appropriate for a saint . . . [I]t is usually the Virgin who lifts the
Christ Child’s bedcover in this way to reveal him to his worshippers . . .
[John] Shearman even suggests that the unknown donor here might have
been called Giuseppe (Joseph) and the action chosen to allude to Joseph of
Arimathea who helped at Christ’s entombment.8

In short, the chronological case is weak: the Turin picture comes rst only
if one agrees that the ‘inexplicable’ action of the male donor in the Hampton
Court Adoration can be rationalized only in those terms.9 As for Shearman’s
notion that the male donor ultimately stands in for Joseph of Arimathea
and therefore might well be named Giuseppe, it is precisely the sort of art-
historical speculation that fails to recognize itself as such, so trusting is its
faith in iconological arguments, however far-fetched. Which leaves us where?
Here is a proposal: if my readings of paintings by Savoldo in this book have
indicated anything, it is that he was an artist of remarkable originality – even
more remarkable, in fact, than has been recognized by previous commenta-
tors. As we have seen, from the very rst, or at least starting with the Elijah
and SS Paul and Anthony, one major operator or vehicle of pictorial meaning
in his art has been hands, culminating earlier in this chapter in the ‘magnetic’
 magic 113

summoning of the sh by the angel Raphael (eventually Raphael will instruct
Tobias to anoint his father’s eyes with the gall of the sh, ‘magically’ curing
his father’s blindness). My proposal is that something of the sort is in evi-
dence in both the Hampton Court and Turin paintings, regardless of which
came rst: that is, the crucial action in both paintings is the exposure of the
Child, less because of the iconographic signicance of such exposure than
because of the precise, delicate and measured gesture of the hands (the male
donor’s or St Jerome’s) in raising the coverlet. In this connection it is also
crucial that the gures in question x their gazes on the viewer: the coverlet
is being raised with the viewer in mind (with the viewer in view, so to speak),
though in the Turin painting we are also given the open-armed response to
the sight of the Christ Child of St Francis (his hands, signicantly, bearing
the stigmata), which in effect underscores the specicity of the treatment
of hands in the three gures of Jerome, the Virgin and Francis. Inciden-
tally, looking at St Francis’s outspread arms and hands with their sugges-
tion of suddenness and surprise, it is impossible not to think of the disciple
inging his arms wide apart at the right of Caravaggio’s Supper at Emmaus
(1596–8; London, National Gallery). Going a bit further, one might say that
in both Adorations there is a hint of something like ‘magic’ in the lifting of the
coverlet, as if what is being exposed to the viewer is something wholly
unexpected, something positively unnatural – and is that not in fact the case?
As if the donor in the one and St Jerome in the other were ‘magicians’ (that
is, Magi) of a sort, their gazes expressing nothing less than satisfaction in the
successful revelation to the viewer – to the paintings’ intended audience – of
the miraculous Child.
A comparison with a supercially similar work by Lotto, his Holy Fam-
ily with St Catherine of Alexandria (1533; illus. 62) in the Accademia di belle
arti G. Carrara, Bergamo, will help underscore the specialness of Savoldo’s
treatment of the revelation or disclosure theme. In Lotto’s canvas, a work of
extraordinary poetry and beauty, the sleeping Child is revealed to Catherine
by Joseph, who lifts the coverlet with his right hand and gestures towards
him with his left. Scholars have been struck by Joseph’s ‘quite exceptional’
action, but it seems clear, at least to me, that the effect of his lifting of the cov-
erlet is altogether different, by which I mean much less compelling, than the
seemingly similar actions in the Hampton Court and Turin paintings.10 Put
another way, Lotto’s treatment of hands is singled out for praise by Berenson
in his early book on the painter: ‘Far from treating the hand as a mere append-
age, he makes it as expressive, as eloquent, as the face itself, and in some of
114 painting with dem ons

62 Lorenzo Lotto, Holy his pictures . . . the hands form a more vital element in the composition than
Family with St Catherine even in Leonardo’s Last Supper.’11 The Bergamo Holy Family is a case in
of Alexandria, 1533, oil
on canvas. point: the faces are relatively inexpressive but the hands are full of feeling;
note especially the Virgin’s right hand with its improbably slender, illumi-
nated ngers gently raised as if in protective reaction to Catherine’s rapt,
forward-leaning response to the sight of the Child and Joseph’s one-handed
lifting of the coverlet while gesturing quietly – ‘Here he is’ – with his other
hand. More broadly, a highly motivated concern for the treatment of hands
characterizes Lotto’s art as it does Savoldo’s. But their respective attitudes
towards hands are fundamentally distinct, Lotto being chiey concerned, as
Berenson recognized, and as is vividly the case in the Bergamo canvas, both
with the evocation of expressive nuance and with the compositional inter-
play between the superbly rendered hands across the lateral expanse of the
painting. In contrast, Savoldo’s hands, put as simply as possible, play a vital,
indeed all-important role in the articulation of certain highly specic, often
empathic projection-soliciting, structures of beholding, which is why it is a fun-
damental fact about the Hampton Court and Turin pictures that the male
donor in one and St Jerome in the other x the viewer with their respective
gazes. Were this not the case – were their gazes directed inside the paint-
ing, as in the Lotto – what I have described as the quasi-‘magical’ viewer-
 magic 115

transxing force of both canvases would be sharply diminished. Put slightly

differently, the unusual actions of the male donor and St Jerome, along with
their equally unconventional gazes, are saturated with thought in a manner
unique to Savoldo. Is it not plain that the viewer is invited to participate in
that thought, to the extent that he or she can?
One other point worth making before moving on is to do with the role
played by the Christ Child in the structure I have been trying to elucidate.
Until now I have stressed the importance of the donor’s and St Jerome’s
actions of delicately lifting the coverlet and xing the viewer with their
respective gazes. Such an account implicitly treats the Christ Child as essen-
tially the object of those acts of revelation, but in fact the Child in both can-
vases is depicted as unusually active (comparison with the Lotto is again
relevant), so much so that it is tempting to see his actions (kicking, grasping,
indeed blessing) as the inaugural actions in the painting and everything else,
including the revelation of his presence, as a secondary effect – the con-
sequence of his ‘magic’ or ‘inuence’. It was a brilliant stroke to make the
Child’s most conspicuous action the kicking of his upraised leg, in counter-
point to the surrounding play of hands. Again, a more sophisticated artist
than Savoldo is hard to imagine.
In fact there is another dimension to all this, as is persuasively argued in a
brilliant essay by Stephen Campbell and then developed in greater detail in
his recent book The Endless Periphery: Towards a Geopolitics of Art in Lorenzo
Lotto’s Italy. Briey, Campbell argues that in Brescia between 1520 and 1550,
in response to a growing sense of religious crisis within the Catholic Church,
certain painters, Savoldo among them, came to associate not only the Christ
Child (and more broadly the body of Christ) but also naturalistic represen-
tation itself with the miracle of the Eucharist, the latter being conceived
‘as a supreme mode of representation with a unique purchase on the real,
the sign that is consubstantial with what it represents’.12 The other major
painters he treats in this connection are Romanino and Moretto. His term
for the pictorial mode in question is ‘sacred naturalism’ (also ‘Eucharistic
naturalism’), which he sees as ‘seek[ing] to occupy a kind of second rung . . .
between the “ultra-true” Eucharist on the one side and the ctive or poetic
mode on the other’ (p. 209). Such a summary, however, does less than justice
to the richness of Campbell’s evidence, the subtlety of his visual analyses,
the cumulative force of his arguments. As he emphasizes, the nature of the
Eucharist, which is to say the question of the ‘real presence of Christ’, was
a bitterly divisive topic in the disputes between Luther and other Northern
116 painting with dem ons

theologians and the Catholic Church, just as the desecration of the Host
was a ‘demonic’ threat of which Savoldo’s contemporaries were acutely con-
scious.13 This suggests that the ‘magic’ I associate with the Hampton Court
and Galleria Sabauda canvases, with their intense close-range focus on the
infant Christ, may be thought of as Eucharistic in nature, a view I take to be
perfectly compatible with my emphasis on the actions and gazes of the male
donor and St Francis.

A pa i n ti n g re lat e d to both the Hampton Court and Turin pictures

but scenographically more ambitious than either, though close in size to
both, is the Adoration of the Shepherds also in the Galleria Sabauda, Turin
(c. 1522–3; illus. 63).14 In an expansive dusk-lit outdoor setting – as usual
with a central dark mass, this one seemingly a brickwork ruin – three shep-
herds make their way from the left towards the kneeling Virgin and the
Christ Child, who lies partly wrapped in swaddling clothes on a blanket on
the ground and with a pillow under his head, while gesturing towards his
mother. Freedberg beautifully describes the gures as ‘mov[ing] with slow
arrested cadence in a crepuscular landscape, shining in it’ (PI, p. 227). Each
of the shepherds, wearing what one takes to be contemporary dress, is clearly
characterized: the leftmost, by far the most youthful, is shown walking, his
arms crossed over his chest and holding a long staff; the second, bearded
but by no means aged, also with a staff, kneeling on one knee, clasps his
hands (ngers touching) while gazing down at the Child; and the third, also
bearded, probably the oldest and certainly the most strongly individuated of
the three, kneels on both knees (not that we see them; they are covered by
his owing robe) while hovering almost directly above the Child. Gilbert
assumes the last of these is Joseph, but that seems doubtful; in any case, one
feels he belongs rhythmically to the procession of visitors.15 More precisely,
he has come to ground just to the left of the Child but his hands, conspicu-
ously large and open with ngers separated in a gesture (of what exactly?),
do the hovering to mysterious effect. As to the gesture: at rst glance one is
tempted to take it as a sign of marvelling, which would be consistent with
numerous precedents of images of Joseph, but on looking closer – and, let
me acknowledge, with a focused empathic effort to feel one’s way into the
hands’ actions, including the cocking of the right wrist – the meaning of the
gesture seems otherwise, at once a quasi-‘magical’ evocation of the Child,
as if the kneeling adorer were conjuring his presence, and a ‘magnetic’ – in
 magic 117

the sense of unconscious, automatistic – response to the sight and perhaps

also the action of the Child, which is to say that once again the implied ow
of ‘magic’, or what I have called ‘inuence’, runs in both directions, towards
and from the Child, the brightly illuminated focus of the composition. All
this is apart from the two-way ow of affection between the Child and the
Virgin, another ‘magical’ relationship, albeit a traditional one. Savoldo,
however, makes one feel it as ‘magical’, which is not quite traditional.16
Also fascinating are the shepherds’ costumes, which appear to have noth-
ing to do with shepherd life; on the contrary, the impression conveyed is
of fashionable materials, and of a stylish interplay of colours and textures.
The contrast between the pale violet of the youngest shepherd’s jerkin and
his knee-length yellow-gold trousers is particularly striking. His bare feet
63 Savoldo, Adoration
hardly establish his shepherd credentials, nor, balletically posed as they of the Shepherds,
appear, are they intended to. The entire painting has somewhat the sense c. 1522–3, oil on panel.
118 painting with dem ons

of a contemporary quasi-theatrical re-enactment before a select audience.

And then there is the animistic-seeming play of folds, particularly evident in
the loosely cut blue jodhpurs of the second shepherd (is that a sort of snout
over his right knee?), the yellow-gold drapery beneath the kneeling shep-
herd’s hands (in which I think I see an eye with eyelids, a sort of nose and a
mouth), and the groundcloth on which the Child has been placed, as well as
the smallish bundle of something unidentiable near the Virgin in the lower
right corner of the canvas. Resting against that bundle is a small metal can-
teen, another contemporary item; Savoldo appears to be making a denite
point. And is there just possibly the suggestion of a rosary with large dark
beads on the ground beneath the male gure with open hands? It is impos-
sible to be certain. Finally, it is worth stressing that with the covering of the
gesturing man’s knees, his posture takes on a somewhat ‘magical’ air, as if he
himself were in fact hovering rather than kneeling – no doubt a momentary
impression, but a suggestive one.

O n e m o re a d o rat io n - ty pe painting is relevant here: the Adoration of

the Shepherds in the National Gallery of Art in Washington, DC (c. 1530–35;
illus. 64).17 As in the Hampton Court and the rst of the Turin pictures, the
gures are half-length (in the latter they are slightly more), but the setting is
nocturnal and the scene as a whole feels strangely compressed, concentrated.
The focus of the scene, even more than in the other Adorations, is the myste-
riously illuminated naked Christ Child in the immediate foreground – indeed
the Child, shown lifting a transparent veil with his left hand as if discreetly
exposing himself, seems a source of light rather than its recipient (there is no
other source of light on view) – while the shepherds, assuming that is who
they are, are only somewhat particularized. At the left we see two bearded
gures who appear almost to double one another, and at the right a beardless
youth who somewhat oddly gazes upwards, as if towards the Virgin. Three of
the four gures – the second bearded man, the Virgin and the youth – clasp
their hands in admiration of the Child (Mary, as is customary with Savoldo,
with only ngertips touching), but it is the action of the shadowy bearded
man at the extreme left that catches one’s attention. As in the Turin Adora-
tion, his two dramatically poised hands hover feelingly above the Child, and
even more strongly than in that painting his gesture, consistent with his
expression, appears one less of surprise or wonderment than of practised
conjuration, as if the painting were the locus, the arena, of two equally pow-
 magic 119

erful feats of ‘magic’: one Eucharistic, emanating from the Child, the other of 64 Savoldo, Adoration of
an uncertain nature, though probably it too is best thought of as Eucharis- the Shepherds, c. 1530–35,
oil on panel.
tic, reective of, in that sense channelling, the divine mystery the painting
emanates. Note, by the way, the Savoldesque motif of a distant burst of
light in the dark sky at the top right, within which is depicted the angel of
the Annunciation in ight, as well as the tiny gures on the ground down
below. This is undoubtedly a Christian motif, but also, in this wholly unusual
work, painted in an age and place where witches were a constant concern,
one perhaps not devoid of a tinge of uncanniness; there will be more on
this later.18

A f i nal work in a different vein will bring this chapter to a close: the
impressive Portrait of a Woman as St Margaret in the Pinacoteca Capitolina,
Rome (c. 1525–30; illus. 65).19 The presentation of a specic individual in the
guise of a saint was a previously established convention, one, as also in the
120 painting with dem ons

65 Savoldo, Portrait of Portrait of a Young Man as St George, that Savoldo found congenial. What
a Woman as St Margaret, establishes the reference to Margaret is the unlikely-looking ‘dragon’ at the
c. 1525–30, oil on canvas.
lower left, which at rst appears to be attached to the sitter’s waist by a
silver(?) chain, one of the fourth-century saint’s apocryphal feats having been
the repulsing (in one account the destruction) of a dragon that had come to
devour her while in prison. In an obvious sense, the ‘dragon’ in this painting
is meant to appear ridiculous, in effect a not particularly attractive pet. But
there is no mistaking the originality of the composition, not only because it
is wider than it is high, unusual for a portrait at that time, but also because
of the sharp division between the two halves of the picture, with the large
open vista on to a Northern-seeming landscape at the right setting off the
elaborately dressed and coiffed sitter (plus ‘dragon’) at the left.
As for the woman herself, whose identity remains unknown, for all her
rich, meticulously depicted costume with pearls and embroidered owers
 magic 121

(marguerites) bearing symbolic asso-

ciations, she communicates a palpable
strength of purpose as she sits erect
with her left hand keeping her place in a
small book, doubtless a breviary, and her
right poised backhandedly against her
right hip, all the while directing her seri-
ous, thoughtful, almost frowning gaze
towards the viewer. She indeed almost
appears troubled: what is the viewer to
make of the not quite vertical crease in
the woman’s brow, an expressive marker
of considerable force? Gilbert singles out
for praise Savoldo’s treatment of her left
hand holding the breviary, and it is easy
to see why (G, p. 97). But what I nd most
suggestive within the framework of the
present chapter is the detailed rendering
not so much of her right hand (we are
given only a glimpse of several ngers)
but of the left-hand glove that she holds
in that hand, and in particular of the way
in which it may be seen as directed ‘mag-
ically’ against the ‘dragon’, as if casting a
spell that keeps the latter subjugated and harmless (illus. 66). (On close look- 66 Savoldo, Portrait of
ing, too, it emerges that the chain-like leash is also being held in the lady’s a Woman as St Margaret
(illus. 65), detail.
right hand, which appears to be wearing its glove. What is notable, in any
case, is the amount of detailed care Savoldo has lavished on this seemingly
marginal precinct of the painting.) This would also explain the half-comic,
half-pathetic upward and backward gaze of the ‘dragon’, afraid or unable to
tear his eyes away from the controlling glove. In other words, in the Por-
trait of a Woman as St Margaret, too, a certain perhaps not entirely orthodox
‘magic’ is conspicuously in play, and a further implication of the woman’s
steady gaze might well be that she is perfectly aware of that fact.
 five • T H E B R E S C I A ADORATION OF
THE SHEPHERDS

L ate i n h i s c ar e e r , around 1540, Savoldo conceived and painted a

work of considerable originality for the Bargnani chapel in San Barnaba
in Brescia (c. 1540; illus. 68).1 It hangs today in the Pinacoteca Tosio
Martinengo in that city and, so far as we know, was his last compositional
invention. Once again it is an Adoration of the Shepherds, and like two of his
four previous depictions of that subject the Brescia painting is on panel, not
canvas; it is also signicantly larger than the others, 192 cm high by 178 cm
wide (the format, essentially square, is all but unique in his production). It
should also be said that the painting’s execution is not that of Savoldo at his
nest. Most importantly, however, there is no precedent in his oeuvre for
the basic compositional idea, which presents in the foreground the kneeling
Joseph and Mary gazing down at the naked Christ Child, who kicks his left
leg and makes the same gesture with his right hand as in the Hampton Court
and Turin canvases, while three shepherds, separated from the Holy Family
by a ruined but still functional structure in stone and wood, look on. The
shadowy presence of animals, a cow and a mule, behind the Virgin indicates
that the structure is a stable; beyond the animals two of the shepherds, a
man with a dark beard and a young man, whose face only we see, appear at a
window; while to the left, behind and beyond Joseph, a third shepherd with
67 (facing page) a dark beard, green jacket and wide-brimmed hat leans on a stone sill. The
Savoldo, Adoration of the hour is nocturnal, with clouds in the sky; possibly we are to understand the
Shepherds (illus. 68),
detail showing faces or moment as shortly after dusk. But whatever the hour, the mood of the paint-
masks in drapery. ing is difcult to specify. Joseph and Mary seem unusually inexpressive and
 the bres cia adoration of the shepherds 125

both appear as if withdrawn into private states of mind, Joseph with lowered
eyes and arms crossed on his breast, Mary, distinctly unbeautiful, with hands
pressed together as if in prayer (conventional poses, in other words). The
shepherds too seem merely to look on, not to respond to the scene in any
obvious way. In fact only two of the shepherds, the older men, can plausibly
be said to be taking in the Holy Family; at the window beyond the Virgin,
the youth or boy at the lower right looks up towards his right (our left), not,
it would appear, at the dark-haired man at his side but at something else –
perhaps the unseen moon (this is suggested by a splash of light on his fore-
head as well as by the illumination from above of the bearded man’s blue
blouse or jacket). The upwards gaze of the young man matches that of the
similar gure in the Washington Adoration, also a late work. In any case the
physical distancing of the shepherds, positioned outside the stable, gives them
the character of witnesses, not participants, as is normal in Adorations and
is manifestly the case in Savoldo’s other paintings of the theme. Impassive
witnesses, at that: how else are we to understand the action of the shepherd
at the left, who seems to partly support his head with his right hand while
simultaneously pushing his hat back on his head, as if quietly puzzling over
the scene before him? The other bearded shepherd seems almost frowning
with concentration, but with his face in shadow it is hard to make this out.
Besides which, concentration or indeed melancholy, another possible char-
acterization of his state of mind, is far from a standard response to the sight
of the Christ Child. Finally, beyond the peasant on the left we see, rst, a
zig-zagging brick or stone-slab wall partly illuminated from above; a hillside
with ruins and trees; and in a distant clearing a trio of tiny gures apparently
reacting to the sight of the angel of the Annunciation to the Shepherds at the
centre of a burst of light in the sky – the most spectacular instance of this
device in Savoldo’s art.
Other features of the scene relate to effects and motifs with which we are
by now familiar: the (Eucharistic) Christ Child seems almost as palpable a
source of radiance as in the Washington Adoration; more broadly, the treat-
ment of nocturnal light in the Brescia panel has always been singled out
for praise; the large stone blocks to the right of the Virgin may be imag-
ined as crudely miming the prole of the cow; and a grotesque face or mask
with a large grimacing mouth is formed by the richly black folds of Joseph’s
right sleeve, just below the shoulder. (Assuming one grants the possibility 68 (facing page)
Savoldo, Adoration of the
that such faces or masks are to be found in Savoldo’s art, this one almost Shepherds, c. 1540, oil on
leaps into view; see illus. 67.) Note, too, how the eroded wood above and panel.
126 painting with dem ons

beyond the Christ Child’s kicking leg

and blessing hand may be seen as having
been ‘magically’ removed by the Child’s
action of kicking, as Campbell more or
less suggests.2 And does not the angel of
the Annunciation in the burst of light in
the sky to the left have something dis-
concertingly witchlike or Boschian about
it (those long, raptor-like wings, so
unlike the angel wings in other Renais-
sance paintings, including the Tobias),
especially when seen in relation to the
small gures gesturing towards it on
the hillside below? In addition, the very
idea of placing the Holy Family in such a
setting has rightly been seen as indebted
to Northern precedents. But the differ-
ences between the Brescia panel and
Savoldo’s other Adorations are in the end
more striking: rst, the affective neutral-
ity, verging on nullity, of all the gures
except the Christ Child, including Joseph
and the Virgin (one would never know
69 Unknown artist, on the basis of this painting what a master of facial expression Savoldo was,
Adoration of the Child, or perhaps I should say had been); second, the absence of any signicant
c. 1440, fresco, detached
and moved to its present play of hands, the decisive operator of pictorial relationships throughout
location in the Brescian Savoldo’s oeuvre (the deployment of the hands of the leftmost shepherd
church of Santa Maria plainly shows the painter’s desire to normalize them, make them merely
delle Grazie in 1539.
ordinary); and third, the basic architecture of the composition, which gives
pride of place to the ruined stable with its thick stone wall viewed end on,
ramshackle wooden roof, and dark inner space punctuated by a simple win-
dow, as well as, this being the point of the entire structure, to the theme of
the separation of onlookers from the central scene.3 A local term of compari-
son, illustrated and briey discussed by Campbell, is a miracle-working (in
that sense ‘magical’) fresco of the Nativity (c. 1440; illus. 69) that was moved
to the Brescian church of Santa Maria delle Grazie in 1539.4 But the archi-
tecture in the fresco is nominal, no shepherds are present, and there is not a
hint of separation between the members of the Holy Family, two angels with
 the bres cia adoration of the shepherds 127

clasped hands, and a curious cow. So the resem-

blance between the two works, albeit undeni-
able, is limited. Another term of comparison,
compositionally and indeed thematically less
close but surely on Savoldo’s mind, is Moretto’s
monumental Nativity with the Shepherds, St
Jerome and a Hieronymite Donor (c. 1530–35;
illus. 70), known as the Delle Grazie Altarpiece
and today in the Pinacoteca Tosio Martinengo.
The question is what all this comes to, or, to
put it slightly differently, what was Savoldo’s
intention, his project, in proceeding in this
way? Throughout this book, starting with my
reading of the Death of St Peter Martyr in Chap-
ter One, I have drawn attention to Savoldo’s
determination to build into his paintings one
or another specic relational structure with
respect to the viewer, characteristically imply-
ing extreme proximity between the two and
often, as in the Death of St Peter Martyr, involv-
ing a quite particular emphasis on the depic-
tion of hands. I suggest that what we nd in the
Brescia Adoration, looked at from this perspec-
tive, is nothing less than an effort to construct
a new mode of pictorial organization (a new dispositif ) that in effect identies 70 Moretto da Brescia,
the viewer before the painting as standing in a comparable relationship to the Nativity with the
Shepherds, St Jerome and
scene as the witnessing shepherds, by which I mean that the viewer’s con- a Hieronymite Donor,
sciousness of being held at a distance from the Virgin, Joseph and the Christ c. 1530–35, tempera on
Child is mirrored or, perhaps better, modelled within the painting by the canvas.

shepherds’ conspicuous separation from the Holy Family. And not only that:
there is also a sense in which the impassiveness of the shepherds, together
with the neutralizing of their hands by virtue of the latter’s engagement in
strictly conventional gestures, is designed to force the issue of the viewer’s
personal response to the sight of the Holy Family, as if all the emotion that
may or may not turn out to be in play is to be understood by the viewer as
coming from, originating in, his or her own response to the Christ Child, a
response inevitably attenuated, not to say disabled, by the circumstances I
have just described.
128 painting with dem ons

To take one further, frankly speculative step: it is as if the organization

of the painting, with its unusual square format and its foregrounding of the
partly ruined stable, aspires and in a sense gestures towards a realization in
three dimensions, an actual architecture, that would in effect embrace the
viewer while at the same time holding him or her at a distance from the prin-
cipal gures – even while the literal embracing of the viewer would be an
impossibility within the limits of the art of painting. But the very framing of
that impossibility may be understood as acknowledging that Savoldo’s until
then fundamental commitment to a far more direct, intimate and involving
mode of entanglement of painting and viewer (in the rst instance, of paint-
ing and painter) was no longer available to him. Or, to put it more strongly,
at this late stage in his personal evolution painting, as he had hitherto prac-
tised it, was leaving him or at least becoming distant from him, which is
also to say from his still active and productive but no longer fully efcacious
hands.5 (More on the notion of entanglement will be found towards the close
of Part Two.) In short, the Brescia Adoration marks a valedictory and all but
nal phase, pictorially and ‘ontologically’, in what I hope can now be seen as
among the most singular and artistically compelling careers in all sixteenth-
century painting.
I describe this as ‘all but nal’ because, some time after completing the
Brescia Adoration, Savoldo painted two variants of the composition, one for
the church of San Giobbe in Venice (also dated c. 1540; illus. 71) and the sec-
ond for Santa Maria la Nova in Terlizzi.6 Without question the San Giobbe
painting is much less pictorially authoritative than the original: the hour is
no longer nocturnal but indeterminate; the architecture of the stable is much
less powerful; the Holy Family is pitched nearer the viewer (the sense of
blockage, of distancing, is largely done away with) without, however, yield-
ing a sense of genuine proximity; the shepherd at the left of the Brescia pic-
ture has been replaced by a dark-bearded man, doubtless a donor, with hands
clasped in prayer, who seems to look on almost furtively (though if one fol-
lows his implied sightline one cannot be certain that he can see much of
anything); and the Christ Child no longer kicks his leg but rather seems to
unveil himself, at least to the extent of holding up a bit of coverlet with his
left hand (probably the strongest detail in the work). The painting as a whole
is slightly smaller and considerably narrower than the Brescia prototype,
while the addition of the rounded top of a traditional pala further diminishes
its force. The execution too is weaker, hastier, than in the original. Not a
happy sequel, in other words. I have not seen the Terlizzi picture, but in
 the bres cia adoration of the shepherds 129

reproduction its condition appears very poor; it also dispenses with the left- 71 Savoldo, Adoration of
most gure found in the other versions. the Shepherds, c. 1540, oil
on canvas.
One last observation about the San Giobbe picture is that it lacks the face
in Joseph’s sleeve in the Brescia original. Does this suggest that Savoldo was
unaware of the face in the rst place? Again, it is impossible to say.
part t wo
 Wi t h t h e p re v i o u s ve chapters in play, I now want to pursue a range
of further reections under various headings, all with a view to bringing
the special character of Savoldo’s art into overall focus. Necessarily, there
will be some repetition of observations and suggestions already put for-
ward. But my aim in the remarks that follow will be to relate those observa-
tions and suggestions to each other with new clarity, as well as to introduce
and develop relevant contextual material that has not yet been part of my
account. By so doing I hope to bring out unexamined depths and reaches
of Savoldo’s unique achievement, and beyond that to provide some portion
of a new understanding of what was artistically possible in the imaginative
universe of early sixteenth-century Venice and its environs.

b r e at h i n g t h e s a m e a i r
Going on from the previous chapters, as well as from my observations about
the structure of the late Brescia Adoration, I now want to expand on the claim
that Savoldo’s determined and resourceful close-range engagement with the
viewer has no equivalent in the work of any other sixteenth-century Italian
72 (previous page) painter. Gilbert does not pursue the point, but he does remark that it is as if
Savoldo, Adoration of
Savoldo’s gures and the viewer breathe the same air (G, p. 7).
the Christ Child with
SS Jerome and Francis So, for example, in the early Elijah Fed by Ravens and SS Paul and Anthony
(illus. 61), detail. (illus. 43 and 44) the viewer is invited, indeed positively encouraged, to
approach the canvas very closely in order to register certain details of the
73 (facing page)
Savoldo, Death of St Peter depiction, such as the fur on the inside of Elijah’s cloak or, a subtler feature,
Martyr (illus. 8), detail. the brown hatchwork on Paul’s palm-leaf robe, which cannot be perceived
134 painting with dem ons

74 Savoldo, SS Paul at a distance of more than a few feet (illus. 74). (Not to mention the faces
and Anthony (illus. 44), in the rocks in the left foreground of the Elijah, about which there will be
detail.
more to say.) Then there is the conspicuous deployment of hands in paint-
ings such as the Portrait of a Man in Armour in the Louvre, the Prophet or
Apostle in Vienna, the Young Man (John the Evangelist?) in the Borghese, the
Evangelist (?) in a private collection in Milan, and the so-called Self-portrait
as St Jerome in a private collection in Bergamo. Not surprisingly, perhaps,
the hands in question occupy the bottom immediate foregrounds of those
paintings, which is to say the portion of the representational eld nearest the
viewer. What is surprising, however, or at least far from usual, is that in cer-
tain pictures, such as the Man in Armour and the Milan Evangelist (?), hands
extend or thrust towards the viewer, and that more broadly – not only in
those canvases – the hands repeatedly call for the viewer’s imaginative par-
ticipation in their implied activities (grasping, clutching, ultimately, I have
suggested, the acts of holding a palette and wielding a paintbrush), a feat of
empathic projection that can be conducted only at extremely close range, as
 breathing the same air 135

if involving virtual merger. In the case of the

Portrait of a Man in Armour there is also the
important matter of the reection in the mir-
ror on the (not at all distant) rear wall, which
requires close and intensive looking if it is to
be properly construed. Note too the breast-
plate or gorget in the lower right foreground,
cut off by the bottom framing edge and to be
imagined extending into the viewer’s space.
Moreover, if I am right, the entire image is
to be understood as the reection in a mirror
‘this’ side of the picture surface.
Continuing with the theme of reections,
there is also the right hand of the Gentleman
as St George in the National Gallery of Art in
Washington, DC, which is reected reversed,
that is, as a left hand, in the man’s armour
breastplate; partly obscured by the sitter’s
right hand, the reection requires a distinct
close-range act of attention to be noticed and
is easily missed. Once seen, however, it raises
the question of the viewer’s relation to that
virtual left hand, which again invites (or at least suggests) a kind of pro- 75 Savoldo, Portrait of
jective ‘identication’. Another observation, not made earlier, is that might a Gentleman as St George
(illus. 38), detail.
there be a sort of one-eyed, upward-gazing, tragically grimacing face-like
mask in the phosphorescent green drapery at the bottom of the canvas near
the sitter’s left hand? For that matter, what of the pair of mask-like (frog
mask-like) forms in the drapery above the man’s left hand and wrist, espe-
cially the uppermost of the two? Finally, what is one to make of the strange,
narrow, tube-like form improbably assumed by a length of drapery (is this
right?) just in front of the staff resting against the sitter’s shoulder (illus.
75)? Is it possible, for example, that there is a face of sorts – an eye, a long
slotted nose, a rudimentary mouth – in the swollen area towards its top?
This can be made out only at very close range, virtually ignoring the rest of
the painting. All this is in contrast to the conventional, in no way unusual,
appearance of the sitter himself. In any case, one is dealing with a paint-
ing that on rst encounter appears unremarkable but on close examination
turns out to be charged with puzzling provocations. The green drapery at
136 painting with dem ons

the lower left with its central ‘spine’ and strange indentations and highlights
seems almost abstractly such, though there too one notes a possible pike-
jawed prole of sorts facing downwards towards the left.
In a different vein, there is the lamp, page of writing and inkwell, and
also the writer’s curiously uncompelling hands, in St Matthew and the Angel,
which together stake out the bottom few inches of the painting as a zone
demanding particular attention.1 As well as, if my earlier suggestion is found
persuasive, the gargoyle-like face or mask in the ‘cuff ’ immediately above the
saint’s right hand. Coming to terms with the St Matthew involves several
different acts or phases of seeing: a rst (let us say) directed close-range and
down towards the hands and their activities (and the inkwell, lamp and sheet
of paper with ‘writing’ on it); a second registering the exquisitely imagined
communication between the angel and the Evangelist (implying a greater
distance but not by much); and then perhaps a third and fourth taking in the
smaller scenes further back in space to the left and right (approaching the
canvas again). There is also conceivably a oating eye (or ‘eye’) gazing as
if placidly at the viewer in the angel’s drapery, just to the side of his elbow
(this may not be apparent in the illustration). As remarked earlier, the entire
scene is permeated by a sense of psychic, not to say quasi-physical intimacy
extending, if one allows it, to the relationship between painting and viewer.
Nothing quite so deliberately scripted is found in Tobias and the Angel, but
it speaks volumes as to Savoldo’s intent in that deeply poetic production that
the surfacing sh – the crucial focus of the composition, once one comes to
understand the latter correctly, or rather one of the two crucial foci, along
with Raphael’s right hand and arm – is found at the very bottom of the
canvas (and in a corner), where it is easily missed by the viewer, at least at
rst; this may be one reason why the picture’s true subject, the ‘magnetic’
extraction of the sh from the water, has never been recognized. (‘Never’
is almost certainly too strong: my proposal in The Moment of Caravaggio is
that Caravaggio understood exactly what is going on and adapted the idea
in his Calling of St Matthew. Possibly other early viewers got the point as
well.) And it is in that same precinct of the canvas that one nds the subsid-
iary ‘mimetic’/‘magnetic’ detail of Tobias’s beckoning forenger. Note, by
the way, with what nesse Savoldo has depicted the ripples caused by the
sh’s surfacing, just in case a viewer might be inclined to take that surfac-
ing for granted. Signicantly, too, because suggesting how deeply Savoldo
rethought this traditional subject, another standard ‘character’ in the story,
Tobias’s dog, usually portrayed accompanying the walking gures, is shown
 breathing the same air 137

asleep in the lower right-hand corner of the canvas. This calls for a further,
separate act of seeing, as if the dream-likeness of the scene were anchored
precisely there.
Also requiring close attention is the byplay in the lower left corner of
the Portrait of a Woman as St Margaret between the hapless ‘dragon’ on its
chain and the lady’s left glove (a surrogate hand), another brilliant stroke
that seems to have escaped recognition over the centuries. And of course
there is the altogether stunning tour de force that the present book began
by unpacking: the implication in the Death of St Peter Martyr of a descending
angel bearing the palm of martyrdom to the astounded saint just ‘this’ side of
the picture surface and above the painting’s upper framing edge, along with
the (to me) irresistible solicitations to empathic projection proffered by the
saint’s gesturing hands.
Finally, capping everything as regards the issue of nearness, there is
the astonishing implied scenario of the sublime Magdalenes, which posits the
viewer (in the rst place the painter) standing in intimate proximity to
the beautiful saint with the risen Christ, the presumed source of the radiance
ooding the saint’s shawl, at his or her side – a pictorial and ‘ontological’
invention of almost inconceivable daring and genius. I shall return to the
theme of nearness shortly.

s av o l d o a n d t h e s e l f - p o r t r a i t
To Savoldo’s most devoted modern scholar, Creighton Gilbert, is owed the
recognition that among the painter’s surviving paintings (roughly fty) are
at least four self-portraits, or works including a self-portrait.2 First is the
Prophet or Apostle in Vienna, which, as Gilbert notes, bears Savoldo’s sig-
nature on the scroll in the sitter’s left hand (G, pp. 426–7; see illus. 28).3
We may therefore take this work as providing certain evidence of Savoldo’s
physiognomy.
Second is the gure of St Liberale at the right of the large altarpiece in
Treviso, Madonna and Child Enthroned among SS Nicholas, Dominic, Thomas,
Jerome and Liberale (1521; illus. 76), the earliest and most ambitious of
Savoldo’s four altarpieces.4 In fact the altarpiece was begun by another
painter, Marco Pensaben, who worked on it between April 1520 and May
1521, after which Savoldo replaced him. Scholars have disagreed as to the
extent of Savoldo’s participation in the nal work, but it appears to have
been signicant. Gilbert notes that the St Liberale gure closely resembles
 s avoldo and the self-portrait 139

the Vienna Prophet or Apostle, which is plainly the case. The saint also, in
Gilbert’s view, ‘occupies the exact place and pose which is [sic] correct for a
self-portrait. He is at the extreme right, in prole, his head turned towards
us, and unlike all the other gures in being entirely hidden except for his
head. In every one of these respects he is identical, for instance, with the
most famous of all self-portraits in the decade preceding, Raphael’s in the
School of Athens’ (G, p. 428).
Third is the painting in the Louvre long regarded as a portrait of the con-
dottiere Gaston de Foix and today called Portrait of a Man in Armour, a work
analysed both in The Moment of Caravaggio and in Chapter Two above (see
illus. 20). In my view Gilbert is right in taking this to be a self-portrait, and
I will forgo summarizing his arguments; sufce it to say that his emphasis
on the presence of mirrors and reecting armour is to the point, as well as
his suggestion that the subject of the painting facially resembles the previ-
ous two images of Savoldo. What Gilbert leaves out, of course, is the entire
dynamic of mirror-reversal and the act of painting implicit in the treatment
of the sitter’s arms and hands and their reections.
Fourth is the painting in a private collection in Bergamo that Francesco
Frangi catalogues as a self-portrait as St Jerome (see illus. 30). Frangi him-
self expresses reservations as to the identication of the sitter,5 but I am
persuaded by Gilbert’s insistence that the latter resembles the persons in
the other self-portraits, quite apart from the fact, which by itself is hardly
dispositive, that the name Gerolamo (Jerome in Italian) is that of the painter.6
As is spelled out in Chapter Two, the knockdown evidence for me, quite
apart from matters of physiognomy, as regards the Vienna and Bergamo can-
vases again concerns the treatment of the respective sitters’ hands, which I
understand in each case as referring at a remove to the act of painting: in
the Self-portrait in the Costume of St Jerome, to one involving the right-angle
dispositif that will later come into play in Annibale’s Self-portrait with Figures
and Caravaggio’s Boy Bitten by a Lizard. All this is to say nothing of other
paintings such as the Death of St Peter Martyr and Penitent St Jerome, which
are not self-portraits but which nevertheless share with the self-portraits a
76 (facing page)
common impulse by way of the depiction of hands. Savoldo, Madonna and
Gilbert also cites the charcoal drawing in Windsor (c. 1515–20?; illus. 77), Child Enthroned among
SS Nicholas, Dominic,
persuasively in my view.7 But nothing crucial hangs on this.
Thomas, Jerome and
As Gilbert remarks, Savoldo’s preoccupation with the self-portrait was Liberale, 1521, oil on
highly unusual. In his words: panel.
140 painting with dem ons

77 Savoldo (attributed), It seems paradoxical that an artist simultaneously was so concerned

The Head of a Man, to paint himself over and over, in perhaps the largest proportion of his
c. 1515–20?, black and
white chalks on blue total oeuvre of any artist [competitors for the distinction would include
paper. Sofonisba Anguissola, Caravaggio, Artemisia Gentileschi, Rembrandt,
Courbet, Van Gogh and Corinth], and yet always concealed it [by which
I take Gilbert to mean that Savoldo never portrayed himself unmistakably
in the act of painting, nor even holding brush and palette] . . . The mat-
erial here for a psychological interpretation is obviously tempting.8
 s avoldo and the self-portrait 141

Not tempting to me, though. For one thing, the notion of concealment
is not quite right. In fact Savoldo’s treatment of hands has been viewed as
one of the signatures of his art by a succession of art historians going back
to Longhi in 1917; the failure of the same commentators to understand the
action of those hands in the terms developed in Chapter Two probably says
more about the limitations of traditional stylistic categories than about any
desire on Savoldo’s part to obfuscate what he was doing. As to Savoldo’s
motivation in thematizing the act of painting in this way, here too I would
minimize psychological considerations in favour of what I have been calling
‘ontological’ ones. In that perspective it makes perfect sense that an artist
who found himself driven to emphasize the seeming nearness to the viewer
(in the rst instance the painter) of a painting’s contents, and by implication
of the painting itself, would be led to become aware to a particularly acute
degree of his own role in bringing the painting about. And that one offshoot
of that awareness might plausibly be a tendency to portray himself in the
act of doing just that, at least in the necessarily somewhat disguised (but
by no means ‘concealing’) form that I have tried to evoke, just as another
would be the calling into question of the impermeability of the picture
surface as an ‘ontological’ barrier, as in the self-portraits, but also as in the
Death of St Peter Martyr, the Penitent St Jerome, the Magdalenes and other
non-self-portraits in which an effect of nearness is carried to a hyper-
bolic extreme. This is to suggest that throughout Savoldo’s oeuvre – not
in every work but in a sufcient number to make this one of the dening
characteristics of his vision – the painting itself, the painting as artefact,
emerges as a kind of transactional eld belonging at once to both the virtual,
depicted world situated ‘beyond’ the picture surface and the actual, physical
world grounded ‘this’ side of that surface, in the latter of which, in intimate
proximity to the canvas, the painter, paintbrush in his right hand and
(traditionally) palette in his left, actively pursued his enterprise. A further
implication of all this is that such paintings eschew the essentially lateral
mode of organization that was common in Venetian painting in his time.
Exactly why Savoldo proceeded in this way, why he imagined the picto-
rial arena in these highly charged, partly somaticized, intensely relational
terms, is impossible to specify, but again it seems unlikely that personal psy-
chology holds the key to what he repeatedly did. Perhaps the best one can
say at this point in our analysis is that he held a vision of painting unlike
that of any Italian contemporary, one that looked forwards to Caravaggio’s
epochal achievement even as it stopped short of the full-blown thematization
142 painting with dem ons

of embodiment that I have suggested marked Caravaggio’s canvases of the

1590s and early 1600s, as well as those of successor artists such as Ribera,
Manfredi, Valentin, Régnier and Cecco del Caravaggio. More on this too
before we are done.

fac e s , m a s k s , n o rt h e r n a rt
With considerations such as these we put Caravaggio aside for the time
being; more to the point, we acknowledge that Savoldo belonged to the rst
half of the sixteenth century, when many of the oppositions that character-
ize the late sixteenth and early seventeenth – crucially between religion and
science, or for that matter between magic and science – were not yet in place.
Nor, during the decades of Savoldo’s activity in North Italy, had the religious
crisis within the Catholic Church that already deeply troubled his environ-
ment developed to the point that we associate with the later period, with
a militant Protestantism, itself far from unitary, centred in Germany and
Switzerland, faced off against an equally militant Catholic Church, with the
Jesuits at the point of the spear. A decisive event, the Council of Trent, began
in 1545 and did not conclude until 1563.9
Another feature from the 1510s to the 1540s, as mentioned earlier, was the
inuence on painting in and around Venice of Northern art, much (though
by no means all) of it conveyed through prints, an inuence that in the case
of Savoldo resulted in two works that form a distinct unit in his oeuvre,
the St Anthony panels in San Diego and Moscow, though it remains an
open question whether they belong to the same or different moments in
time. I have suggested that this is also to be detected in the proliferation,
as I see it, of grotesque or monstrous faces, semi-faces and masks in rocks
and drapery folds and creases throughout his art. Key gures in this con-
nection are Bosch, Dürer, Cranach, Schongauer and Patinir, both as regards
specic imagery, the St Anthony panels being obvious cases in point, and as
regards Savoldo’s treatment of drapery generally, which has no parallel else-
where in Italian art. A striking afnity between Savoldo and Dürer is sug-
gested by a comparison between the various faces I have claimed to discern
in Savoldo’s paintings and the marvellous early Dürer pen-and-ink drawing
Six Pillows (c. 1493; illus. 78), which as it happens is found on the verso of
a Self-portrait at Age Twenty-two (c. 1493), one of a series of remarkable self-
portrait drawings and paintings by the Northern master. (A predilection for
self-representation is something else Savoldo and Dürer have in common.)
 faces, masks, northern art 143

Both drawings are discussed by Joseph Koerner in his ambitious study 78 Albrecht Dürer, Six
The Moment of Self-portraiture in German Renaissance Art, where the verso is Pillows, c. 1493, pen and
brown ink.
seen, reasonably enough, as depicting a single pillow that has been punched
and twisted into six different congurations.10 To that extent it is a study
in realistic description. But there is another aspect to the sheet. Koerner
continues:

For as commentators have long noted, the pillows themselves contain hid-
den faces in their folds and indentations. For example, the curious spi-
ral on the left edge of the pillow at center left acts as a possible eye for
faces discoverable in the lower left corner and, upside down, at the upper
left. Similarly, faces appear at the far edge of the pillow just to the right;
the curved horizontal fold just below center reads both as the mouth of
one face whose nose is the pillow’s upper corner, and as the nose and eye
of a smaller face in the lower corner. Once set in motion, this game of
144 painting with dem ons

‘seeing as’ can be played indenitely, transforming corners into noses,

chins, or satyrs’ horns, and creases into mouths and brows, until each
pillow is animated by a number of hypothetical masks frowning, laughing,
fretting, and speaking. None of these faces, it is true, can be posited with
certainty as really being there, nor do they remain stable for us. We lose
one possible physiognomy once we discover another that overlaps with
it. Yet even the skeptic will admit that the pillows, arranged in pairs and
gesturing dynamically, have been rendered animate, if not anthropomor-
phic, by the artist’s pen. These drawings exercise the visual fantasy not
only of the artist inventing these grotesque physiognomies, but also of the
viewer discovering faces that others might miss (pp. 27–8).

A strange statement in some ways, as if Koerner wants both to endorse the

presence of the faces and at the same time not disown them exactly, but
rather emphasize the role of the viewer in construing the pillows in those
terms, which is fair enough so long as the two formulations are balanced
against one another. (So long as ‘once set in motion’ is understood as mean-
ing ‘once having started responding to the drawing’s cues and solicita-
tions’ rather than ‘once the viewer starts playing the game of “seeing as” ’,
as if doing so were an entirely subjective choice on the viewer’s part. I take
Koerner and myself to be in agreement on this point.) To shift the discussion
to Savoldo, it is important to acknowledge the extent to which the grotesque
faces and masks that I have pointed out, sometimes hesitantly or tentatively,
fall short of declaring themselves unmistakably, unequivocally, which doubt-
less is largely why until now they have been overlooked. But it is equally if
not more important to register the fact that the treatment of folds and such
in clothes and drapery in Savoldo’s paintings repeatedly holds out the pos-
sibility that something very far out of the ordinary with respect to animistic
physiognomic imagery is in play, and that if we fail to take that aboard we
are ignoring one of the most intriguing and perplexing features of his art.
Simply put, nothing remotely similar can be found in the production of any
other Italian painter, though in the work of various artists – Giotto, Andrea
Mantegna, Piero di Cosimo, Lorenzo Lotto, Bernardo Zenale and, I would
add, Giovanni Bellini – isolated instances of faces or bodies in clouds, trees
or rocks can be made out.11 While anthropomorphic imagery is notoriously
present in the work of more than a few Northerners, including Dürer, Bosch
and Herri met de Bles, the forms it takes differ fundamentally from what I
have claimed to nd throughout Savoldo’s oeuvre.12
 faces, masks, northern art 145

Of course, not all the faces and masks I have pointed to are equally strik-
ing. To consider several examples, in the rst place there are the heads and
faces in the rocks in the left foreground of the Elijah. I take them to be
incontrovertible and in that sense to provide a kind of evidentiary ‘anchor’
to the rest of my proposals. But what of the possible raven’s head in the
drapery folds immediately above Elijah’s right foot? At rst glance that may
seem unlikely, but is it not also consistent with the general play of auto-
‘mimetic’ relationships in this and other paintings, for example between the
raven on its branch and Elijah’s head supported by his arm and hand? Or,
in the same canvas, looking patiently, may one not make out a frontal face of
sorts in the large rocky form alongside Elijah’s left shoulder, one nowhere
near as striking as the faces at the bottom left but not simply to be ruled out
on that account? Then there is the large monstrous face with an open cave
mouth in the central rocky mass in the San Diego Temptation of St Anthony,
another plainly intentional ‘anchor’ as well as the clearest possible indication
of a connection with Northern art, Bosch in particular. (I also see the other
heads and faces in the central rocky mass as incontrovertible, but am willing
to grant that others might not. Or not at rst; in the end, they can scarcely
be doubted.) In a somewhat different register, in the Penitent St Jerome what
is the status of the possible face or mask (two faces or masks, actually) near
the bottom of the painting in the vicinity of the saint’s covered right knee,
again in the context of other ‘mimetic’ relationships elsewhere in the canvas?
Or in St Matthew and the Angel, the face or mask with attened nose in the
folds of the left sleeve or the gargoyle-like face with wide-open mouth in the
glowing red right cuff of the writing evangelist? Or in the late Adoration of
the Shepherds in Brescia, of the grimacing face or mask in the indentations
of Joseph’s right sleeve? Or in the London, Berlin and Getty Magdalenes of
the various faces in the radiant shawls among the latter’s deep folds and
gleaming highlights (see illus. 55)? Or in the Death of St Peter Martyr, of
the grimacing face-like folds in the upper sleeve of the impassive murderer
preparing his backhanded death-blow (a stretch, I admit, but in the light of
more perspicuous faces elsewhere in Savoldo’s oeuvre, perhaps not out
of the question; see illus. 73)?13 Or to look again at a painting we have not
yet considered in this regard, what of the dark folds towards the bottom of
the shepherd’s blue tunic in the Getty Shepherd with a Flute (see illus. 36),
which suggest an exaggeratedly long downward-turning mouth, a squashed
nose and a deep eye socket, all directly below the shepherd’s right elbow? (It
was the Shepherd with a Flute that Longhi saw as anticipating Caravaggio’s
 faces, masks, northern art 147

handling of drapery.) Finally, turning to a work that has so far gone unmen- 80 Savoldo,
tioned, what of the possible face- or mask-like intimations of the yellow and Transfiguration (illus.79),
detail.
red drapery in the region of the disciples’ feet in the Transfiguration in the
Ufzi (1530s; illus. 79), the yellow folds suggesting a masked visage with
deep eye sockets, the red ones hinting at a small animal’s head with a short
snout and large ears (illus. 80)?
Ultimately, I am tempted to let my entire argument rest on a single crux,
the large grinning beak-nosed face or mask in the Berlin Magdalene (see
illus. 55), which I earlier compared both with the bird-like skull in the Mos-
cow Temptation of St Anthony and with the Bosch-like creature (an unlikely
dragon?) in Giulio Campagnola’s engraving The Astrologer. To be as plain
as possible: I think the grinning face or mask is there, despite the fact that
until now it has gone unrecognized. A further, crucial question is whether
or not Savoldo was aware of its presence, and if he was, whether he can
be imagined as having put it there deliberately. On the one hand, the lat-
ter possibility seems to go too far. On the other hand, assuming the head
came about partly or wholly automatistically, in the course of developing
the ostensibly merely decorative and enlivening bulges and creases in the
radiant fabric of the shawl, is it plausible to imagine that the painter
79 (facing page)
remained oblivious to what he had done, in spite of the singular strength Savoldo, Transfiguration,
of the hideous image once it is given its due, and bearing in mind the facial 1530s, oil on panel.
148 painting with dem ons

and mask-like ‘anchors’ that I have cited in other of his works? I mean this
as other than a rhetorical question.
In this connection, some remarks by Savoldo’s former student, Paolo Pino,
in his 1548 Dialogo di pittura, make intriguing reading. At a crucial juncture
Pino has one of his dialogists, Fabio, expatiate on the sweetness of the art
of painting:

This is evident, for nature imitates herself, and by nature do all articers
have the things she makes [a Leonardesque claim, needless to say]. Nature
demonstrates this often by painting on her own, in marble and tree trunks,
a diversity of gurable forms, and elsewhere in smoke and cloud is she
similarly engaged; and nature does this with that same delight one experi-
ences in seeing one’s efgy in the mirror. (D, p. 330)

Earlier Pino states that ‘Our art creates the same effect as the mirror, which
receives within itself whatever form (minus the movement) is present before
it’ (D, p. 305).14
Elsewhere in Pino’s Dialogo, as discussed in Chapter Two, is his inuential
account of Giorgione’s legendary invention of a gure of St George
multiply reected so as to provide a fully in-the-round depiction of a single
gure on the at surface of a painting, thereby equalling sculpture’s abil-
ity to render three-dimensional form (D, pp. 367–8), an account obviously
related to Savoldo’s Portrait of a Man in Armour (see illus. 20), a self-portrait
of the painter in the act of painting, as I have said.
Returning to Pino’s remarks about nature taking pleasure in painting her-
self, it is striking that he understands this as a practice of self-portraiture
(also self-relishing) on nature’s part via the metaphor of the mirror, which
raises the question as to whether Pino had Savoldo in mind when he wrote
these sentences. The answer would seem to be yes, and at any rate he soon
singles out ‘Messer Gerolamo of Brescia [who] was most learned in [the
painting of vistas]. I once saw by his hand certain sunrises with solar
reections, certain nocturnes with a thousand most ingenious and rare par-
ticularities, all of which seemed truer images of actuality than the Flemish
ones’ (D, p. 372). To this he adds: ‘This specialty is very natural to the painter
and a source of pleasure to himself and to others; and that method used by
the Germans, of copying landscapes [paesi] in a mirror, is very much to the
point [here]’ (D, p. 373).
Not that it is entirely clear what Pino is getting at. As Mary Pardo, Pino’s
translator, writes in her invaluable commentary on the Dialogo:
 faces, masks, northern art 149

Whatever its source, Fabio’s advice to follow the ‘German’ manner of

depicting landscapes in a mirror – it is hard to tell whether he means
using the mirror as a pictorial aid, or making Eyckian mirrored vistas –
recalls his earlier words on nature ‘depicting’ herself in clouds and rocks
with the same delight with which we behold ourselves in a mirror; Fabio
insists that the skill in the making of landscapes is most ‘natural’ to the
painter, and as much a source of personal delight as it is to the ‘mother’
whose works he imitates. In fact, the passage reads like a deliberate par-
tial commentary on the earlier Natura pictrix statement. It would seem
that ‘paesi ’ here has the double sense of ‘nature’s image’ and ‘nature’s
inventione ’, in to which both nature and the painter ‘project’ themselves.
(P, p. 277)

One more reference to Savoldo is worth noting. The topic is the difculty
of depicting three-dimensional form on a at surface and the failure of per-
sons not versed in painting to grasp the complexities of foreshortening and
the like. Thus Fabio:

we cannot make every gure distinctly visible in its entirety, and this
occurs because of the promptness of its gestures, as [is apparent] in fore-
shortened forms, where certain parts evade the sight; we [painters] com-
prehend these with difculty, and they cannot be understood by anybody
without [a knowledge of] art. And so it will happen that an excellent
painter will make a gure resembling the living model, in a pose of such
difculty, and that it will be not only not understood, but censured by
whomever does not know the scope of our art. And so the man will divest
himself of honor with the very toils which he invests in its acquisition.

To which the other dialogist, Lauro, responds:

You speak the truth. Be there a master as learned as it is possible to be

about the art, his works will still keep him shackled between the hope
of praise and the fear of censure; and sometimes the ignorant are lled
with such a bad impression that if a gure by the painter, or a single hand
displeases them, they take such a dislike to him that they never again
nd pleasure in his works. See how Messer Gerolamo of Brescia, Paolo
Pino’s master and a man most rare in our art, excellent in the imitation
of all things, spent his life on few works and with scarce renown. How-
ever, it is true that at one time he was salaried by the late Duke of Milan.
(D, p. 304)
150 painting with dem ons

‘A pose of such difculty’ (perhaps that of the man in armour in the paint-
ing in the Louvre) and ‘a single hand displeases them’: it is impossible not
to think not only of the Louvre canvas but also of the other self-portraits,
with their prominent hands as if vigorously gripping or wielding imagi-
nary implements. This gives added force to Pino’s metaphorics of mirror-
ing, which his remarks allow us to extend beyond the self-portraits proper
to Savoldo’s oeuvre generally, for example to the electrifying painting dis-
cussed in Chapter One, the Death of St Peter Martyr, and in particular to
the nature of its solicitations towards empathic projection as conveyed by
the saint’s left and right hands, which I suggested are chiey to be seen as
mirroring the viewer’s (originally the painter’s) right and left hands as he or
she stands before the canvas (having approached it closely, as already said).
To repeat a question posed earlier: where could a painting such as the St
Peter Martyr have been hung in Savoldo’s time so that the proximity-effects
just mentioned could come into play? It is as if a painting like that nds
its ideal place in a modern museum. As we saw in Chapter Two, a similar
mirror-reversal effect with respect to the sitter’s hands is to be found in the
Self-portrait as St Jerome and the Portrait of a Man in Armour. The Prophet or
Apostle is more ambiguous, at once evoking a ‘normalized’ self-portrait in
which the brush would be held in the sitter’s right hand and keeping open
the possibility of a mirror-reversed relationship by virtue of which the sit-
ter’s right hand stimulates empathic projection on the part of the viewer’s
left hand.
All of which leads to the further suggestion that Pino’s mirror meta-
phor, far from being merely a literary ourish, captures something fun-
damental about Savoldo’s art: that his paintings often evoke the thought
that they are, to an uncanny degree, mirrors (perhaps I should say ‘magic’
mirrors), in which both world and artist paint themselves, mirror-reversal
with respect to hands being one indication of that state of affairs. Another
such indication, of course, are the gleaming, light-struck surfaces of the
intensely coloured drapery in many of his works, culminating in the Magda-
lene’s highly reective, dazzlingly radiant shawls in the London, Berlin and
Los Angeles canvases. In her important essay on the Magdalenes, Pardo
describes them as ‘constituting the image as a virtual mirror’, which seems
perfectly apt.15
As regards the faces and masks, Pino’s appeal to the idea of nature’s self-
depiction by no means settles the questions I have raised, but the cited pas-
sages present Savoldo’s art in a context of ideas, concerns and metaphors
 ‘magic’, ‘influence’, demons 151

that goes at least some of the way towards accounting for the production
of imagery whose intentional structure remains suggestively in question.
We are about to take a step farther, raising the stakes of the discussion as
we go.

‘magic’, ‘influence’, demons

We touch here on another crux that belongs to the sixteenth century, or
more precisely to the late fteenth century and the sixteenth, with roots
going back much earlier. Simply put, Savoldo lived at a time and in a cul-
ture when and where belief in astrological thinking was rife, and various
notions of magical ‘inuence’ and, especially pertinent in the present con-
text, both good and bad demons were seriously entertained by a diverse
array of unorthodox gures, both Italian and foreign, with Marsilio Ficino,
Cornelius Agrippa, Trithemius of Würzburg, Girolamo Cardano, Gianfran-
cesco Pico della Mirandola and Giordano Bruno prominent among them. (At
this point I am discarding the quotation marks around ‘magic’ and ‘magical’.)
As D. P. Walker and others have emphasized, one major source or Ur-source
of magical ideas was the Catholic Mass,

with its music, words of consecration, incense, lights, wine and supreme
magical effect – transubstantiation. This, I would suggest, is a fundamen-
tal inuence on all mediaeval and Renaissance magic, and a fundamen-
tal reason for the Church’s condemnation of all magical practices. The
Church has her own [Eucharistic] magic; there is no room for any other.
The effort to make a sharp distinction between Christian rites and any
kind of secular magic is . . . apparent in many 16th-century discussions of
such subjects.16

In particular, according to Ioan P. Couliano,

any form of magic invoking demons was held to be suspect and was per-
secuted. This is why Marsilio Ficino, who had to endure the attacks of the
Church for his treatise De via coelitus comparanda [On Obtaining Life from
the Heavens] – which the pope nally judged to be inoffensive – did not
know what precautions to take to demonstrate that the ‘natural’ magic
he practiced was not demonic. Probably he was right only in the sense
that the magician was able to restrict his own processes, but that did not
prevent demonomagic, in certain if not all cases, from being a form of
spiritual magic.17
152 painting with dem ons

For another key gure, Agrippa, even more open to the thought of demons
than Ficino, the universe was ‘permeat[ed] by spiritual beings, both good
and evil. Not only the heavens, stars, and elements but also the whole world
itself have rational souls. Man can by careful preparation and correct ritual
summon these spirits and souls and induce them to do his bidding.’18 The
personage who could contrive this was the Magus.
It is not my aim to associate Savoldo with one or another school of magi-
cal thinking in Italian culture of the late fteenth century and the sixteenth;
nor are these pages the place for a consideration of the role of demons in the
contemporaneous European imaginaire, a vast and fascinating topic.19 But
it does seem plausible to think of the often snarling or glowering or other-
wise distorted faces and masks in Savoldo’s drapery as essentially demonic,
and to associate their occurrence with the idea of ‘inuence’, understood in
the rst place as based on Savoldo’s familiarity with Northern paintings
and prints. (The modern art-historical notion of inuence, such as I have
been appealing to until now, would be a ‘scientic’ reduction of the earlier,
richer conception.20) Furthermore, as emerged in Chapter Four, I nd an
intimation of magical proceedings in a number of Savoldo’s paintings. This
is most obvious, in the Tobias and the Angel, in the ‘magnetic’ gesture of the
angel Raphael (also, by implication, in the Tobias narrative, which includes
the idea of burning the sh’s liver and heart to drive away a demon!); in the
Hampton Court and Turin Adorations, in the ‘conjuring’ actions of the male
donor in the rst and St Jerome in the second; in the Turin and Washington
Adorations of the Shepherds, in the far from ordinary hovering hand gestures
of the Magus-like bearded men; and in the Portrait of a Woman as St Margaret,
in the implied ‘dragon’-suppressing action of the lady’s right glove, perhaps
the most outlandish of all the magical operations in Savoldo’s art. Is there
not also a sense in which the intense close-range concentration on the Christ
Child of the main bearded gure in the Washington Adoration seems akin to
what one imagines would have been that of a person indulging in the six-
teenth- and seventeenth-century practice of scrying, that is, gazing intently
into an instrument like a crystal ball? In a different, manifestly ‘Eucharistic’
register, there is also the incandescent presence of the Christ Child in the
Adorations, a powerful, or should one say all-powerful, source of ‘inuence’
in its own right – which, however, is not to deny the witch-like aura of the
unconventional angels of the Annunciation in at least the Washington and
Brescia paintings.
 ‘magic’, ‘influence’, demons 153

In my discussion of the St Margaret I imply that the magical power I asso-

ciate with her glove is perhaps somewhat unorthodox, but of course there is
no reason not to regard it as belonging to an essentially Catholic imagina-
tive universe, as would have been the professed belief of Ficino, Agrippa and
the others, whatever the various ambiguities of their writings. Similarly, as
already noted, the actions of the hands as well as the facial expressions of
the male gures in the Hampton Court and Turin Adorations, albeit icono-
graphically unusual (uncovering the Christ Child to the puzzlement of art
historians) and charged with a special intensity (keyed to the male gures’
seemingly pointed gazes directly at the viewer), are perfectly intelligible in
Catholic terms, especially if we understand them, as doubtless we should,
as channelling the Christ Child’s divine presence. The same holds for what
I have just characterized, no doubt tendentiously, as the ‘witch-like’ angels
of the Annunciation in their nocturnal heaven-rending bursts of light in
the Washington and Brescia Adorations, another Savoldesque motif with no
equivalent in the art of his Italian contemporaries.
As for the grotesque or monstrous faces and masks, our particular
concern in these pages, my suggestion is that these may be understood as
evoking what might (almost? not quite?) oxymoronically be described as a
Catholic world of demonic forces, such as was typically depicted in scenes
of the Temptation of St Anthony (Bosch’s great triptych in Lisbon being
the ne plus ultra of this), which is to say that, to take a particularly striking
example, the presence of a frightening visage in the drapery of the Berlin
Magdalene (an apparition that I have compared with the creature depicted
in Campagnola’s The Astrologer) is plausibly to be viewed as offering
special insight into the demon-ridden nature even of scenes of purest
sainthood. The Magdalenes as a group all but spell this out, as do the expres-
sively urgent Penitent St Jerome and the lyrically intense St Matthew and
the Angel.
But would the Italian Church in Savoldo’s time possibly have been accept-
ing of such a vision? It seems extremely unlikely: the uncompromising, life-
threatening hostility of the religious authorities to any hint of demonism in
art or thought, which would without question have included the physiog-
nomies I have claimed to nd in Savoldo’s work, is the strongest evidence
imaginable that they went unremarked at the time, most likely by the art-
ist himself. To cite Stuart Clark in Thinking with Demons, ‘an unbridgeable
gulf between [what they saw as] religion and magic came to dominate the
154 painting with dem ons

sensibilities of churchmen and their evangelical efforts.’21 With regard to

real-world consequences, in the Val Camonica in the Venice-ruled diocese
of Brescia around sixty women and twenty men in the years just before
1520 were burned as witches or ‘perdious and malecent persons’.22 Indeed
Savoldo in the course of his career appears to have been in signicant con-
tact with the Dominican Order, which, as was earlier remarked apropos of
his Death of St Peter Martyr, was at the forefront of efforts to stamp out her-
esy wherever it was suspected to have arisen.23 Savoldo’s St Anthony panels
would have been acceptable, of course, licensed by the particular subject of
a saint famously understood to have been tormented by demons and by its
Northern origins. Apropos the latter, Paul Vandenbroeck, a leading Bosch
scholar, remarks about Bosch: ‘For the same spirit permeates his entire
oeuvre: the cosmos teems with demons, unbridled sexuality ows through
nature’s very veins, humanity is composed of fools and sinners, and the saints
– even Christ himself – are constantly beleaguered and tormented.’24 Need-
less to say, so extreme a formulation cannot be applied broadly to Savoldo’s
art. Nevertheless, the notion of a Catholic universe not just teeming with but
interpenetrated by demonic agents, and therefore subject to the operations
both of the ‘good’ magic of the Mass and of demonic magic (unequivocally
‘bad’), perhaps bears more closely on the works we have been considering
than is comfortable to think. Put slightly differently, in certain paintings by
Savoldo the Eucharistic and the demonic – the sacred and the monstrous
– inhabit the same representational eld: as if the painter himself were the
locus, a nodal point, of their interpenetration, imaginatively speaking.
One further face or mask, notice of which I have been holding back, bears
on this issue as well as on the question of Savoldo’s awareness of what he
was doing. In SS Paul and Anthony (illus. 44), discussed in detail in Chapter
Three, there is, to my mind unmistakably, in the brownish folds of Anthony’s
cloak in the lower left-hand corner of the canvas a rather large, grotesque
face or mask more or less aligned with the saint’s body (that is, facing right,
in our terms), with two deeply shadowed, outwardly slanting eyes, a rudi-
mentary nose and a large, partly open mouth, the entire ensemble conveying
a sense of hard-to-read emotion, perhaps melancholy (illus. 81). What makes
this face or mask particularly signicant is that it is located exactly where
in the immediately previous Elijah there appears the cluster of undeniable
heads and faces to which I called attention at the start of Chapter Three.
This raises the possibility if not the likelihood that Savoldo in the SS Paul
and Anthony was indeed conscious of what he was doing. If that is the case,
 ‘magic’, ‘influence’, demons 155

it perhaps works against the thought that he was oblivious to the greater
part of the masks and faces in drapery that I have claimed to detect. (And
what, it might be asked, of the two faces or masks towards the bottom of the
saint’s robe in the Penitent St Jerome, that is, in a similar position relative to
the gure?)
And yet it seems to go much too far to imagine Savoldo throughout his
career knowingly suggesting one face here, another there, except in the Eli-
jah, SS Paul and Anthony and Temptation of St Anthony pictures – and perhaps
in the Penitent St Jerome. Rather, I think of the appearance – the embedding,
surfacing, condensing, materializing – of such faces and masks in the folds,
creases, bulges and pockets of drapery in the other works considered in these
81 Savoldo, SS Paul
pages as much more likely the outcome of a technique-based improvisatory and Anthony (illus. 44),
procedure, according to which Savoldo’s markedly original, largely automa- detail.
tistic treatment of the typically radiant,
colouristically ‘surcharged’ (Freedberg’s
epithet) fabric of his personages’ cloth-
ing turns out to give visible expression
to demonic presences and ‘inuences’ that
otherwise would remain mostly unde-
tectable. As for whether or not Savoldo,
except in the obvious cases mentioned
above, became aware of what in effect he
was doing or had done, I trust it is clear
how uncondent I am as to where in the
end to come down. Just conceivably, in the
spirit of Bosch, he understood the demons
and monsters as belabouring saints and
others but in the end overcome by them
(and as meant to be seen in those terms if
they were noticed as such by contempo-
rary viewers). But is it at all likely that an
Italian painter of Catholic subjects, much
less one linked with the Dominicans, in
the far from serene religious climate of
sixteenth-century Venice and Brescia,
would have dared deliberately to place one
or more blatantly demonic faces in imme-
diate juxtaposition to the Magdalene’s
156 painting with dem ons

own? Or to depict St Matthew’s writing hand emerging from the mouth of

a demon? Or to position another grimacing face on the kneeling Joseph’s
sleeve, in close proximity to the Virgin and the Christ Child (see illus.
67)? Regardless of the fact that these and other faces and masks have gone
unrecognized until now.25
A nal turn of the screw would be to emphasize the instrumental role
in such a procedure of the activity of Savoldo’s hands, which in this regard
are to be imagined working as if on their own, disengaged from other than
the most improvisational cast of mind – which is to say as if responding to a
play of ‘inuence’ that could be made manifest, given palpable form, only in
this way.

w h o , t h e n , wa s s av o l d o ?
By now it need hardly be stated that this is not an easy question to answer,
even if one puts to one side, as for the moment I mainly shall, the matter
of faces, demons and magic. As Gilbert succinctly remarks, ‘Few Italian
painters of his time, of any interest at all, are so slightly known as persons’
(G, p. 49), and ‘He was one of the last artists to be raised to the rank of the
major High Renaissance masters.’26 As was noted in the Introduction, we
have only a handful of dates with the aid of which to construct a chronology,
with the result that a great deal of uncertainty prevails when it comes to
establishing an artistic trajectory. What we think we know is this: Savoldo
was very likely born in Brescia, probably around 1480; his training, Gilbert
proposes, was in the style of Alvise Vivarini and Cima da Conegliano; he
seems to have been in Parma in 1506 and Florence in 1508, which perhaps
indicates a wide experience of early sixteenth-century central Italian paint-
ing; by 1521 he was living in Venice, and it is possible that he settled there
years before that. At some point he established a studio, which we know of
only thanks to a reference in his former student Paolo Pino’s 1548 Dialogo di
pittura. Beyond that, the studio is a blank.
Also according to Pino, who clearly admired him, Savoldo’s oeuvre was
small and he enjoyed only modest success except, it seems, for a period in the
employ of Francesco Maria Sforza, the last Duke of Milan. This was likely
during the years 1531–5, and resulted in four paintings in the Milan mint,
described by Vasari as night scenes and admired by him; among those, it is
widely agreed, were two of Savoldo’s supreme achievements, St Matthew and
the Angel and Tobias and the Angel (not a night scene but the same dimensions
 who, then, was savoldo? 157

as the Matthew). Savoldo appears still to have been alive when Pino wrote his
Dialogo, but by then he may well have ceased painting. The date of his death
is unknown.
And in fact his oeuvre is unusual, to say the least. Not only is it small,
it is also strangely unbalanced, containing only a half-dozen altarpieces,
fewer, Gilbert says, than by any other signicant Italian painter of the time
(G, p. 137). Moreover, the altarpieces themselves tend to be rather bare bones
in comparison with the magnicent productions in that genre by the likes
of Giovanni Bellini, Titian, Lotto, Veronese and Tintoretto or, in Brescia, by
the prodigious, younger Moretto. Savoldo’s corpus also includes relatively
few portraits, and the ones that survive, with the exception of the Portrait
of a Young Man (St John the Baptist?) and the Portrait of a Young Man with
Flute, are by no means among his most distinguished productions; even in
those technically superb canvases the sitters’ inner lives, what Berenson
would have called their psychology, are left untouched, unexplored. Finally,
as already mentioned, there is a disproportionately large number of self-
portraits. All in all, a highly idiosyncratic body of work.
Like so many sixteenth-century masters, Savoldo was rediscovered in the
second half of the nineteenth century, but his modern recuperation owes
much to Roberto Longhi, on the strength of a few paragraphs in early essays.
(The enthusiasm of Adolfo and Lionello Venturi also contributed to this.) In
particular Longhi advanced the notion of a distinct Brescian school or tradi-
tion going back to Vincenzo Foppa, along with that of a Lombard tradition
of naturalism and concern with effects of light. (Gilbert demurs, emphasiz-
ing Venetian painting, including that of Titian, and during Savoldo’s time in
Milan, assuming he worked there in the rst half of the 1530s, the palpable
inuence of Leonardo on both the St Matthew and the Tobias. Giorgione is
also frequently mentioned, though Gilbert doubts his relevance.) Another
feature of Longhi’s appreciation of Savoldo was the important claim that
he was a major predecessor of Caravaggio, an idea that has proven greatly
inuential in the subsequent literature. Perhaps most important, as we
have seen, Longhi brilliantly pioneered the recognition of the originality of
Savoldo’s treatment of hands, one of the most distinctive features of his art
and a major preoccupation of this book.
My own impulse, intermittently expressed in these pages, has been to
see Savoldo alongside and in more or less continual juxtaposition to Lotto,
another Northern Italian born around 1480 whose art has rightly been
viewed as marked by a certain quattrocento archaism, who remained open
158 painting with dem ons

to the inuence of Northern art, and whose career also seems to have been
partly shaped by a desire to avoid direct competition with Titian, hence
his projects in and travels to provincial towns. To that extent, both Lotto
and Savoldo stand apart from the central line of pictorial development in
Venice; Gilbert refers in passing to their stylistic ‘cousinship’ (G, p. 308).
But at that point the differences between them become salient: for one thing,
we know vastly more about Lotto, his commissions, his travels, his stays
in various towns in North Italy and the Marches (most importantly, per-
haps, in Bergamo in 1513–25), even, modern scholarship has shown, his
religious commitments, which suggest an openness to reform movements
within Catholicism.27 For another, Lotto was a vastly more productive artist,
with more than 250 surviving works in his oeuvre. And the difference is not
only quantitative: there is no equivalent in Savoldo’s corpus to Lotto’s 38
magnicent altarpieces, the best of them works of nonpareil beauty and
imaginative richness, such as the Martinengo, Santo Spirito and San
Bernardino altarpieces in Bergamo or the stupendous Crucifixion in Santa
Maria della Pietà in Telusiano, Monte San Giusto. Or, more broadly,
for the cornucopia of Lotto’s pictorial gifts, his marvellous colourism, his
compositional inventiveness, his particular brand of lyricism (amounting
at times to a ‘mysticism of the affections’, in Longhi’s formulation28), and
his stylistic range, from works of exquisite renement to the ‘popular’
imagery of the St Lucy altarpiece (Iesi, Pinacoteca Civica) and the frescoes
in the Suardi oratory at Trescore Balneario. Finally, nothing in Savoldo
remotely approaches the authority of Lotto’s portraits, as compelling
as those of any other sixteenth-century painter. In short Lotto, for all his
idiosyncratic tendencies, emerges as a larger gure in every way, fully
worthy of Berenson’s attention in his path-breaking monograph of 1895.29
And yet . . .
And yet, nothing in all this takes away from a sense of Savoldo’s remark-
able distinction – of his particular genius, I am inclined to say. By this
I mean, to begin with, that Savoldo’s handful of truly great paintings – the
Tobias, the St Matthew, the St Jerome, the Magdalenes, the Portrait of a Man in
Armour, the Prophet or Apostle, the Adoration of the Shepherds with full-length
gures in Turin, the Portrait of a Young Man with Flute, the Shepherd with a
Flute and, in my view, the St Peter Martyr – are utterly unique in their age. In
each Savoldo radically reconceived traditional subject matter; from each radi-
ates a sense of imaginative, technical and rhetorical (I would also say ‘onto-
logical’) singularity with no parallel in the art of his contemporaries. This
 who, then, was savoldo? 159

is apart from the question of hitherto unremarked faces, masks, monsters,

demons; but perhaps it was his singularity that made him, uniquely among
Italian painters of his age, open to the suggestion of demonic ‘inuence’ that
I have found myself tracking in this book. And although his generic portraits
are undistinguished, his mastery of facial expression in paintings like the
Death of St Peter Martyr and St Matthew and the Angel is beyond compare; the
half-dozen charcoal drawings of individuals are also magnicent.
The mystery, we might say, is why there are so few paintings such as
those I have just cited, even assuming that a number of works have been
lost, including the two other canvases that Vasari reports having seen in
the Milan mint. And indeed Savoldo’s singularity has been registered by his
admirers. Vasari, as noted earlier, calls him ‘fanciful and artful’. For Longhi
in 1917, Savoldo was the most civilizzato and cittadino – the least provincial
– of all the Brescians.30 Gilbert sees in his half-hidden signatures a sign that
he was reluctant to assert himself and that his obscurity in his lifetime may
have owed as much to this psychological trait as to a lack of appreciation
(G, p. 49). (Glossing Pino, he also refers to ‘the hierarchical lordliness of
gure and hand that was apart from Venetian taste’ [G, p. 394].) Alessandro
Ballarin nds his intellect sensitive to Northern art, to every extreme of
‘escape into the realm of dreamlike imagination’ (‘evasione nel regno della
fantasia onirica’), which, in the context of the present book, is a suggestive
remark, to say the least.31 Gaetano Panazza, who notes the smallness of his
oeuvre, writes that by the end Savoldo was solitary, apart, bitter, in the grip
of deep sadness – an interesting view, but based on what, exactly? Savoldo’s
work, he goes on to say, reveals a severity of temperament, a felt religiosity
(like Moretto’s), but also great renement and control; one can only agree.32
Gilbert again, as was remarked earlier, discerns in Savoldo’s gures a capac-
ity for profound emotion; they invite us to participate in their lives via an
interior afnity like the characters of a great novelist. That is not an anal-
ogy that would have occurred to me; one has the sense that Gilbert’s deep
admiration for Savoldo continually escapes adequate characterization. I say
this without wishing to detract from a recognition of Gilbert’s fundamental
contribution to Savoldo studies.33
By all odds the most stimulating attempt to capture Savoldo’s special-
ness is by Mary Pardo, who in her indispensable and wide-ranging essay
on the Magdalenes emphasizes the importance of pictorial artice and the
aesthetic dimension in his art. In her account, the Magdalene (she focuses
on the London version) ‘engages our self-conscious awareness of pictorial
160 painting with dem ons

artice’, noting that ‘by 1500 chiaroscuro, contraposition and foreshorten-

ing . . . functioned as demonstration subjects in their own right’. She contin-
ues: ‘If I am correct about the importance of the aesthetic to the devotional
experience of Savoldo’s painting, then it is possible to say that the Magdalene
afrms its own critical dimension – which is that of artistic problem-solving
– even as it opens the way for higher contemplative activity’ (p. 84; the latter
presumably centred on its Christian content, a surprising ordering of its pri-
orities). In this connection she cites Pino, who lamented that while Savoldo
was ‘raro nell’arte’, he never achieved popularity and ‘spent his life on
few works’. ‘Even so gorgeous a painting as the Magdalene ’, she suggests,
‘seems to propose a restricted audience of fellow-painters and sophisticated
patrons willing to savor its mirror play of sense and surface’ (p. 90).34 Pardo
goes on:

It seems likely that, in calling Savoldo capriccioso e sofistico, Vasari was tar-
geting the Brescian master’s ironic reserve, his pleasure in putting grand
artistry into modestly scaled, isolated images. The Magdalene is especially
suited to the display of artice: since it does not pretend to ‘contain’ truth,
only to reect it [that is, to reect the light of Christ standing before
it], its ostensible content is wholly exterior to it. Yet the resultant ‘emp-
tiness’ is also a kind of limitless potentiality (since it holds the viewer
in thrall), and guarantees the painter’s essential autonomy in spinning
out his ction. The gleaming shawl ‘re-represents’ the pigmented and
brush-imprinted canvas surface in terms of illusion, and thus invites us to
contemplate on its own terms that other content, the artistic process itself.
It is an extraordinary sleight-of-hand (pp. 90–91).

This is dazzling, as bets its topic, but my sense is that it perhaps portrays
Savoldo as somewhat more modern, not to say proto-modernist, than he
really was (who is Pardo’s ‘us’?). In her account, at any rate in the London
Magdalene, he emerges as a ‘pure’ painter, working for a small, cultivated
audience of cognoscenti, setting for himself and then brilliantly solving pic-
torial problems that, however fascinating and ingenious, were inevitably of
limited appeal. Also, her account implies, such problems were ahead of their
time, the entire issue of ‘guarantee[ing] the painter’s autonomy as he spins
out his ction’ seeming not quite to match either his situation or his proj-
ect.35 (As this makes clear, the faces and masks escaped her notice.) But her
appreciation of Savoldo’s profound originality, astounding sophistication and
sheer intellectual renement feels exactly right.
 savoldo and caravaggio 161

s av o l d o a n d c a r ava g g i o : t h e i n e s c a pa b l e r e l at i o n
With the following reections I aim to draw this book to a close. In the
Introduction I used the term ‘independent’ paintings to refer to the bulk
of Savoldo’s work, remarking that I considered this a place-marker for a
further discussion of the character of his art. I also said that I used the term
by way of avoiding the designation ‘devotional’ paintings, which during the
past decades has become more or less standard for late fteenth- and early
sixteenth-century Italian, especially Venetian, works that are not altarpieces,
portraits or narratives. The Venetian master of such ‘devotional’ painting
is of course Giovanni Bellini, in his (and his workshop’s) many images of
the Virgin and Child, sometimes alone, sometimes anked by two or more
saints.36 Typically such paintings are modest in scale, smaller than the Savol-
dos we have considered, as bets the fact that they were intended as aids to
private devotion, both by presenting the viewer with an image to be con-
templated in a devout frame of mind and also, when two or more saints are
present, by offering in the persons of those gures images, or say models,
of appropriate religious response. A particularly affecting example of such
a painting is Bellini’s Madonna and Child with SS Catherine of Alexandria and
Mary Magdalene (c. 1490; illus. 82), the gures depicted half-length against 82 Giovanni Bellini,
a dark background. The Christ Child, naked, is seated on a small pillow, The Madonna and Child
apparently on Mary’s lap; her right hand delicately embraces his body and with SS Catherine of
Alexandria and Mary
her left hand gives support to his left hand in a marvellously subtle gesture Magdalene, c. 1490, oil
of mutual touching. Her expression with its averted gaze, which drifts off to on panel.
162 painting with dem ons

83 Giovanni Bellini, her right, suggests pensiveness (a traditional notion: Mary has foreknowl-
St Jerome in the Desert, edge of what is to come); the Child looks up, towards what we cannot tell;
1480, oil on panel.
St Catherine of Alexandria, a frequent participant in such scenes, gazes some-
what sadly down at the child with her hands lightly clasped in prayer; while
Mary Magdalene, with loosened hair, her facial expression not happy but
otherwise unreadable, has her hands crossed on her breast. It is as though
the entire composition in its utter repose and relative inexpressiveness,
though the general tonality is quietly sad, has been devised to leave room for,
in that sense to elicit, the viewer’s inward act of devotion, which I imagine as
comparably stilled, reective, grave. Or, to take another ‘devotional’ subject
popular in Venice at the time, Bellini’s depictions of St Jerome in the wilder-
ness conveyed to urban collectors a vision of ‘an alternative, solitary life to
be lived in nature’ (Hans Belting), as in the work Belting calls the painter’s
 savoldo and caravaggio 163

masterpiece in this vein, the St Jerome in the Ufzi (1480; illus. 83). These
too are outwardly inexpressive scenes of quiet reection.37 Other Venetian
painters found in the subject a devotional theme par excellence, Lotto’s St
Jeromes in Paris and Rome (1506 and 1508) being two much-admired varia-
tions on the Bellinesque ideal.
My aim in adducing such works at this juncture is to drive home the point
that nothing of the sort can be found in Savoldo, who, simply put, is not in
the least a ‘devotional’ painter. It is no accident, for example, that his Peni-
tent St Jerome is not a small-scale, distanced image inviting quiet contempla-
tion but rather an altogether original close-range depiction of intense con-
centration and violent physical self-punishing (on a literal level; what else
might be involved in the actions of Jerome’s two hands is a topic broached in
Chapter Two). The St Jerome is also considerably larger than the ‘devotional’
norm, and its relation to the viewer, as I suggest in Chapter Two, is a matter
not simply of inducing subjective feeling but rather of inviting what I have
been calling empathic projection, that is, a quasi-active, quasi-bodily sense of
identication with the saint’s actions and gestures, including, crucially, the
focalized clutching gesture of his left arm and hand. Nor for that matter are
the Tobias and the Angel, St Matthew and the Angel and the Death of St Peter
Martyr plausibly to be understood as ‘devotional’ in intent; in contrast (but
not really) the Hampton Court and Turin Adorations at rst might appear to
t the ‘devotional’ mould, but very quickly the ‘conjuring’ (or ‘disclosing’)
actions of the male donor in the rst and St Jerome in the second, along
with their penetrating gazes directly at the viewer, shift both works into
another, far more challenging register.38 A similar dynamic marks the Wash-
ington Adoration of the Shepherds, in which the unconventional hand gesture
and Magus-like expression of the bearded man refuse to allow the viewer an
unconsidered response.
One other cluster of works that seem ‘devotional’ on rst view, not men-
tioned until now, comprises several canvases of the Rest on the Flight into
Egypt, the most impressive of which may well be the version in a private
collection in Milan with a highly detailed and topographically accurate vista
of the Riva degli Schiavoni in Venice in the right-hand portion of the canvas
(1525–7; illus. 84). However, the composition with its multiple centres of
interest – the seated Virgin and Child, Joseph in the middle distance seeing
to their donkey, the small group of a red-cloaked and hatted man on horse-
back, a dog and a page or soldier, the tower at the extreme right, as well as
the numerous workmen in its vicinity hauling and assembling lengths of
164 painting with dem ons

84 Savoldo, Rest on the wood (to what end?), the Riva degli Schiavoni itself and St Mark’s Basin with
Flight into Egypt, 1525–7, its many vessels, and nally the dark entranceway behind the Virgin with its
oil on canvas.
round window within which is silhouetted the upper body of a person (a bust
or statue?) – is far too complex and dispersive to allow a merely ‘devotional’
response on the part of the viewer.39
More broadly, throughout this book I have been stressing the uniqueness
of Savoldo’s paintings in their time, the designation ‘independent’ serving
both to acknowledge that fact and to leave open the question of how they
might most accurately be characterized in general terms. The relative lack
of distinction of Savoldo’s few altarpieces makes the same point differently:
nothing could be less Savoldesque than the structurally distanced and formal
mode of address to a gathering of believers that the classic late fteenth- or
sixteenth-century altarpiece entailed.
And here is my further point, which brings me back to Caravaggio. In
The Moment of Caravaggio and then in its sequel, After Caravaggio (2016),
 savoldo and caravaggio 165

I emphasize the invention of what I call the full-blown gallery picture by

the master and his successors. By this I mean that in Caravaggio’s work
of the 1590s and early 1600s, and then in that of successor artists such as
Bartolomeo Manfredi, Valentin de Boulogne and Nicolas Régnier, there
came to the fore a new kind of painting aimed at the patronage of an elite
group of ambitious and cultivated collectors – a kind of painting

not small but not outstandingly large, often religious in subject matter but
not necessarily devotional in intent, framed and portable rather than xed
permanently in place, open to compositional and interpretive innovation,
and typically, as in Caravaggio’s case, executed with a care that signaled
the painter’s alertness both to the cultivated tastes of his elite viewer-
ship and to the circumstances of display that would make possible unusu-
ally close scrutiny of the nished artifact . . . Such works are sometimes
called ‘gallery pictures’ and are best thought of as a version of – perhaps
more accurately, an immediate forerunner of – the autonomous and essen-
tially portable or, as is sometimes said, ‘homeless’ easel picture that would
presently emerge as the dominant pictorial form of the modern era. (MC,
pp. 83–4)

The notion of autonomy is crucial here, the idea that such a painting was
meant to be entirely sufcient to itself, independent of its surroundings, in
effect cut off or severed from them (as it were by its frame). At the same
time, the ideal of autonomy as I develop it in those books goes along with
a pursuit of what I characterize as pictorial density, keyed in the rst place
to the bodies of both painter and viewer, which is also to say to the ques-
tion of bodily orientation: the painter and viewer facing into the painting
(hence the proliferation of gures seen from the rear, especially in the work
of the Caravaggisti), while the painting itself, as a worked artefact, faces
outwards, in effect addressing both painter and viewer (hence the prolif-
eration not simply of facing gures in Caravaggio and his successors but
also of an unprecedentedly acute practice of interpellation, to use a nakedly
anachronistic term). (In many multi-gure works by the Caravaggesque
painters the theme of pictorial density is driven home by the placement in
the middle of the composition of a solid, block-like table, often of marble,
almost always skewed relative to the picture plane so as to be as obtrusive as
possible.40) The issue of empathic projection also is in play, not quite accord-
ing to the close-range, essentially ‘mimetic’ dynamic I have attributed to
Savoldo but by way of a new pictorial poetics of absorption, which I contend
166 painting with dem ons

rst enters European painting in force, as a major resource for the art, in
such paintings by Caravaggio as the Penitent Magdalene in the Galleria Doria
Pamphilj, the Death of the Virgin in the Louvre, and the St Jerome Writing in
the Galleria Borghese. Roughly, I see these and other expressively ‘minimal’
paintings as eliciting the viewer’s spontaneous and unreective conviction as
to the self-contained inner life of the paintings’ protagonists.
All this, I want to argue, is both somewhat like but also, equally important,
crucially unlike what we nd in Savoldo: like above all in that at the very
core of Caravaggio’s enterprise, as is true of Savoldo’s, is an intense one-to-
one and in important respects bodily relation between painting and painter
(the relation between painting and viewer being as it were subsequent to
that), which in both their oeuvres leads to a practice of self-portrayal that is
anything but common in their respective pictorial universes. (The Moment
of Caravaggio begins with a detailed consideration of Caravaggio’s early Boy
Bitten by a Lizard, which I interpret as a disguised and displaced self-portrait
of a particular kind; I go on to argue that Savoldo’s Portrait of a Man in
Armour is also a self-portrait, each of the two works involving the mirror-
reversal of right and left.) And unlike in that Caravaggio, as I understand
him, soon came in works such as the Martyrdom of St Matthew in San Luigi
dei Francesi, Rome, and the tremendous David with the Head of Goliath in
the Borghese to pursue a project involving his ultimate expulsion from his
paintings. In the Martyrdom this was in the person of the eeing bravo at
the left looking back over his shoulder with an expression Mina Gregori
acutely describes as ‘overwhelmed with sadness’,41 in the David through an
act of decapitation, that is, of violently separating the work in question from
himself, which I take to be a further manifestation, a hyperbolization, of the
desire to sever the full-blown gallery picture from its immediate context,
thus securing its aesthetic autonomy in the most declarative terms possible.
(Both the bravo and Goliath are self-portraits, I need hardly add.) In sum,
I regard the unprecedented proliferation of scenes of decapitation in the art
of Caravaggio and his successors in this light, as a virtual acting out of the
gallery picture ideal.42
Here we come to a basic difference, in some respects the basic difference
(bracketing once again the matter of faces and demons), between Cara-
vaggio and Savoldo, in whose paintings nothing remotely like a pursuit of
separation is to be found. By this I refer not only to the forthright, in no
sense ‘removed’ or ‘severed’ presence of the sitter in Savoldo’s self-portraits
– even his marginal self-portrayal as St Liberale in the altarpiece in San
 savoldo and caravaggio 167

Nicolò, Treviso, has something quietly assertive about it – but also, return-
ing one last time to a central motif of this book, to the painter’s hands, into
the depiction of which, starting with my reading of the Death of St Peter
Martyr, I have tried to show the viewer is invited empathically to project
time and again. (That the invitation has been declined even by Savoldo’s
most prescient and devoted commentators speaks to the norms of the art-
historical discipline rather than to any merely personal failure.) Note, by the
way, that the viewer’s relation to the depicted hands in Savoldo’s paintings
is sometimes one of mirroring, sometimes one of congruence, sometimes
possibly both, which I see as contrasting signicantly with the polarity
between facing and facing away in Caravaggio and his successors discussed
in Chapter Two – so much more rigorous, more nearly systematic, than what
we nd in Savoldo. (Facing versus facing away is not a Savoldesque trope.)
In any case, with a few exceptions, the early Boy Bitten by a Lizard being
one, nothing comparable to such an intense thematization of hands is to be
found in Caravaggio, which is not at all to say that hands play an unim-
portant role in his art. They are certainly important, but they are not the
decisive factors – the foregrounded, often central, dramatically active pres-
ences – that they are in Savoldo’s. (Also, in a different sense, as often in the
tavern paintings of such as Valentin, Manfredi and Régnier.43 And that they
will be again, more than three centuries later, in the early self-portraits and
key Realist canvases of the great French painter Gustave Courbet: but that
is another story, or, perhaps more accurately, another chapter of a single, still
incompletely understood dialectic.44)
Even more strongly, the presence – also the actions, the thrusting and
grasping gestures – of hands in Savoldo’s pictures testify to the fact that he
never achieved, or never sought to achieve, the separation and distancing
from his paintings that I am claiming was basic to Caravaggio’s enterprise.
Put slightly differently, there is in Savoldo no equivalent to the pursuit of
autonomy at all costs that marks Caravaggio’s project and the emergence
of the full-blown gallery picture in the 1590s and after. (Savoldo’s are not
gallery pictures in that sense, despite having broken with the ‘devotional’
convention. By the same token, no subject matter could be more foreign
to his project than decapitation.45) Rather, Savoldo remains productively,
if at times also somewhat confusingly, entangled with his paintings and, I
have suggested, the ideal empathic viewer becomes entangled with that
entanglement. This at once largely accounts for the interpretive challenge
Savoldo’s paintings have posed for even his most appreciative modern
168 painting with dem ons

85 Savoldo, Adoration commentators, and perhaps also helps explain what seems to have been, in
of the Shepherds (illus. 63), their own time and country, their somewhat limited appeal, as is suggested by
detail.
Pino’s remarks cited earlier about viewers being put off by ‘a single hand’.46
 savoldo and caravaggio 169

It follows that the afnity between Savoldo and Caravaggio is at once

partial and profound, the latter to the extent that it is only in the retrospec-
tive light of Caravaggio’s momentous achievement and the emergence in the
1590s and after of the full-blown gallery picture that Savoldo’s supremely
intelligent, mysteriously ‘inuenced’ and, at their best, almost literally spell-
binding paintings can be brought into art-historical focus. This in turn
means that Longhi’s pioneering insight was on target, even if the grounds of
that insight stand in need of the radical reimagining that the present study
has sought to provide.47
One last thought. What I have just called Savoldo’s entanglement with
his art – the fact that he seems never to have achieved or indeed sought to
achieve separation and distance from it (not even in the late Adoration in
Brescia, where a tendency towards both is in play) – may also help explain
his not having been aware of the greater number of the faces or masks I have
claimed to discover in his paintings, assuming that such awareness escaped
him (the grotesque face in Joseph’s sleeve in the Adoration being a nal case
in point). But this is speculation.
A F T E RWO R D

A b r i e f ac c o u n t o f h ow t h i s b o o k came about is perhaps in order.

As readers of The Moment of Caravaggio will be aware, early on in that book
I discuss in some detail Savoldo’s Portrait of a Man in Armour in the Louvre
(long thought to be a portrait of the French military commander Gaston
de Foix), which the late Creighton Gilbert was the rst to suggest should
be recognized as a self-portrait. I found myself in agreement with Gilbert,
and went on to discuss the Man in Armour ’s fascinating play of hand and
arm gestures and mirror-reections, which I understood as representing in
displaced form the actions of Savoldo’s left and right hands in the course of
bringing the painting into being. (I rehearse that reading in Chapter Two of
this book.)
Subsequently I realized that there was much more to say about the treat-
ment of hands in Savoldo’s art, and sat down to write an essay that would
spell that out, with the working title ‘Savoldo’s Hands’. But it quickly became
clear that a short book, not an essay, was required in order to do justice to
the topic. And then, after several months of close looking, research and writ-
ing, on a visit to the National Gallery of Art in Washington, DC, I noticed
that there were unmistakable heads and faces in the rocks at the bottom
left of the early Elijah Fed by Ravens. This was surprising, but even more so
was the fact that no one had ever seen (or at least had ever commented on)
the heads before. And in the months that followed, mainly during 2017–18,
I discovered or came to believe that I discovered a signicant number of
other heads and faces, mainly though not exclusively in the folds, bulges,
creases and depressions of the drapery of various personages in Savoldo’s
pictures.
 afterword 171

Working out what this meant for the understanding of his art was a
tremendous challenge and became the other central theme of the present
book. Eventually I came to be persuaded that for the most part Savoldo was
less than fully aware of the presence of such heads, faces and mask-like forms
in the bulk of his work (the early Elijah, SS Paul and Anthony and the Tempta-
tion of St Anthony paintings in San Diego and Moscow being exceptions to
this), and also that the heads and faces were essentially demonic, which is to
say that they belonged to a sixteenth-century frame of mind that imagined
the universe as saturated with demonic presences. That such a frame of mind, 86 Savoldo, Adoration
of the Shepherds (illus.
which at least to some degree I take to have been Savoldo’s, could also be
68), detail showing the
rigorously Catholic and indeed orientated to the reality of the Eucharist is appearance of the angel
also a basic claim of this book. of the Annunciation.
REFERENCES

a b b r e v i at i o n s
D Paolo Pino, Dialogo di pittura, trans. in Mary Pardo, ‘Paolo Pino’s Dialogo
di Pittura, a Translation with Commentary’, PhD diss., University of
Pittsburgh, 1984
F Francesco Frangi, Savoldo: Catalogo completo (Florence, 1992)
G Creighton Gilbert, ‘The Works of Girolamo Savoldo’, PhD diss., New
York University, 1955. Subsequently published as Creighton E.
Gilbert, The Works of Girolamo Savoldo: The 1955 Dissertation, with a
Review of Research, 1955–1985 (New York and London, 1986). In the
latter publication pp. 71–158 of the original dissertation are omitted,
replaced by a fresh discussion of the post-1955 scholarly literature and
questions of chronology on pp. 524–69. All references in the present
book are to the original dissertation.
MC Michael Fried, The Moment of Caravaggio, The A. W. Mellon Lectures
in the Fine Arts (Princeton, NJ, and Oxford, 2010)
P Mary Pardo, Commentary on Pino’s Dialogo di pittura, in Pardo, ‘Paolo
Pino’s Dialogo di Pittura, a Translation with Commentary’, PhD diss.,
University of Pittsburgh, 1984
PI S. J. Freedberg, Painting in Italy, 1500–1600, Pelican History of Art
(Harmondsworth and Baltimore, MD, 1971)
PR Painters of Reality: The Legacy of Leonardo and Caravaggio in Lombardy,
ed. Andrea Bayer, exh. cat., Cremona, Museo Civico ‘Ala Ponzone’; New
York, Metropolitan Museum of Art (New Haven, CT, and London,
2004)
Savoldo 1990 Giovanni Gerolamo Savoldo tra Foppa, Giorgione e Caravaggio, ed. Bruno
Passamani, exh. cat., Brescia, Monastero di San Salvatore-Santa Giulia;
Frankfurt, Schirn (Milan, 1990)
T Tiziano e la pittura del cinquecento tra Venezia e Brescia, ed. Francesco
Frangi, exh. cat., Brescia, Museo di Santa Giulia (Milan, 2018)
 references to pages 10–34 173

introduction quali tuttavia è bandita ogni traccia di accademismo,

che non di rado gli eroi artistici portano con sé, ma che
1 Aretino’s letter is quoted in the original in G, pp.
piuttosto sono interpretati con una sensibilità poetica
57–8. See G for a summary of basic facts about Savoldo’s
e una capacità di approfondimento emotivo che ci invi-
life and career, as well as for a useful survey of Savoldo
tato a participare all loro vita . . . Sono in denitiva
scholarship up to 1955.
personaggi legati tra loro da profonde afnità interi-
2 D, p. 304. Further page references will be in paren-
ori, come quelle che talvolta si incontrano in un grande
theses in the text.
romanziere, ma tuttavia non possiamo dire che in essi
3 Giorgio Vasari, The Lives of the Painters, Sculptors
vi siano afnità di rapporti letterari: i loro caratteri
and Architects, 4 vols, trans. William Gaunt (London and
sono deniti unicamente attraverso i mezzi visive,
New York, 1963), III, p. 321. Vasari’s remarks on Savoldo
in parte dei gesti, dalle espressioni, dei costume, ma
read in their entirety: ‘Many works of Giangirolamo
anche alla forma nel senso astratto della parola.
Bresciano may be seen in Milan and Venice, and the mint
contains four very ne representations of night and res. 13 Jaś Elsner, ‘Style’, in Robert S. Nelson and Richard
In the house of Tommaso di Empoli there is a very ne Shiff, eds, Critical Terms for Art History, 2nd edn (Chicago,
Nativity at night, and some similar fancies, in which he IL, 2003), p. 102.
excelled. But as he only did such things and no large 14 One recent exception is the pages on Savoldo in
works, I can say no more of him than that he was imagi- Stephen J. Campbell’s important The Endless Periphery:
native and fantastic [capriccioso e sofistico], and his works Toward a Geopolitics of Art in Lorenzo Lotto’s Italy (Chicago,
deserve much praise.’ Mary Pardo translates ‘capric- IL, and London, 2019). These will be cited again and dis-
cio e sostico’ as fanciful and artful in ‘The Subject of cussed later in this book.
Savoldo’s Magdalene ’, Art Bulletin, LXXI/1 (1989), p. 69. 15 I make these observations in MC, pp. 200–201.
4 See the catalogue entry by Andrea Bayer on the
Crucifixion in PR, p. 136.
5 See Antonio Boschetto, Giovan Gerolamo Savoldo
pa r t o n e
(Milan, 1963).
6 Savoldo 1990. o n e • d e at h o f s t p e t e r m a r t y r
7 Keith Christiansen, ‘Gerolamo Savoldo’, in The
1 The painting was rst shown at Hall & Knight
Age of Caravaggio, exh. cat., Naples, Museo Nazionale di
Ltd (London and New York) in 2001, accompanied by
Capodimonte; New York, Metropolitan Museum of Art
an exhibition catalogue titled MMI, including an essay by
(New York, 1985), pp. 79–85.
Mina Gregori, ‘Giovanni Girolamo Savoldo, The Death of
8 Andrea Bayer, cat. entries in PR, pp. 136–45.
St. Peter Martyr ’, pp. 70–77. Further page references to
9 T.
her essay will be in parentheses in the text.
10 For more on Lombardy, see Bayer, ‘Dening Natu-
2 On Titian’s Death of St Peter Martyr, see also the
ralism in Lombard Painting’, in PR, pp. 4–5.
interesting discussion in Una Roman D’Elia, The Poetics
11 Roberto Longhi, ‘Quesiti caravaggeschi, II: I prec-
of Titian’s Religious Paintings (Cambridge, 2005), chap. 4,
edenti’, Pinacotheca, 5–6 (1929), pp. 258–320; repr. in ‘Me
‘Christian Tragedy’.
pinxit’ e quesiti caravaggeschi, Edizione delle opere com-
3 Creighton E. Gilbert, ‘Savoldo’s Death of Peter
plete di Roberto Longhi, vol. IV: 1929–34 (Florence, 1968),
Martyr’, in Venezia, le Marche e la civiltà adriatica: per
pp. 97–143.
festeggiare i 90 anni di Pietro Zampetti, Arte documento
12 Creighton Gilbert, ‘Savoldo Cortese’, in Savoldo
17–19 (2003), p. 291. Further page references will be in
1990, p. 38 (my translation). The original Italian reads:
parentheses in the text.
Nasce così una galleria di personaggi imponenti nei 4 Keith Christiansen, ‘Caravaggio and “L’esempio
quali sembrano connaturate la stabilità morale e aristo- davanti del natural”’, Art Bulletin, LXVIII/3 (1986), p. 422.
crata e l’inclinazione all’intimità dei pensieri. Si tratta 5 Keith Christiansen, ‘Giovanni Gerolamo Savoldo’, in
di personaggi solitari, qualche volta anche eroici, nei The Age of Caravaggio, exh. cat., New York, Metropolitan
174 references to pages 34–48

Museum of Art; Naples, Museo Nazionale di Capodi- Campbell writes: ‘We might describe Lotto’s subject’s
monte (1985), p. 79. postures as labored. It is as if the body is straining
6 See Mina Gregori, ‘I temi della luce articiale en against its own opacity in order to make itself articu-
Savoldo e le radici lombarde di Caravaggio’, in Savoldo late and intelligible. The soul is not a light that shines
1990, p. 91. forth from within, as the character “Pietro Bembo” con-
7 Stanley Cavell, The Claim of Reason: Wittgenstein, ceives it at the conclusion of Baldassare Castiglione’s Il
Skepticism, Morality, and Tragedy (New York and Oxford, Cortegiano. Interiority is, as it were, produced and main-
1979), esp. pp. 440–41. I make use of the concept apro- tained through the application of the subject’s entire
pos of the invention/discovery of absorption as a major physical being . . . Lotto clearly wants to signal the
resource for painting in MC, pp. 105–6. disparity between Bembo’s identication of grace and
8 S. J. Freedberg, Circa 1600: A Revolution of Style in virtue, and the lived experience of embodied individuals
Italian Painting (Cambridge, MA, 1983), pp. 65–6. in their everyday world’ (pp. 149–50). To this I would add
9 Personal communication (2007). This was said apro- that Lotto’s sitters’ characteristically charged address to
pos of my discussion of the ‘invention of absorption’ in the viewer solicits the latter’s awareness of some such dis-
MC, which Cavell had read in draft form. parity, as well as a sense of the ‘inquisitional’ character of
10 Or in Hegelian language, from the Lectures on Fine the world in which they (and Lotto) increasingly found
Art: ‘This being at one with itself in its other is the really themselves. At the limit, it is as if certain of Lotto’s por-
beautiful subject matter of romantic art [by which Hegel traits seek to induce a comparable bodily and psychologi-
means Christian art], its Ideal which has essentially for cal response in the viewer.
its form and appearance the inner life and subjectivity, 12 Campbell’s The Endless Periphery brings new in-
mind and feeling. Therefore the romantic Ideal expresses sight to bear on these matters; see esp. chaps 2 and 4.
a relation to another spiritual being which is so bound up I shall have more to say about Campbell’s ndings later
with depth of feeling that only in this other does the soul in this book.
achieve this intimacy with itself ’ (G.W.F. Hegel, Aesthetics:
Lectures on Fine Art, trans. T. M. Knox [Oxford, 1975],
I, p. 533).
t wo • hands
This reciprocity, if I may call it that, the thought ‘that
only in this other does the soul achieve this intimacy with 1 MC, pp. 7–12.
itself ’, may appear to critique the idea of empathic projec- 2 The change from normalization to mirror-reversal
tion, and perhaps it does. (As if the italicized term can around 1860 was rst remarked by Zirka Zaremba Fil-
be taken as implying too one-way an operation.) But my ipczak, Picturing Art in Antwerp, 1550–1700 (Princeton,
use of Cavell’s term has always for me implied the notion NJ, 1987), p. 202. In MC I illustrate and discuss Henri
that the projection in question is elicited by the painting, Matisse’s Self-portrait of 1918 as exemplifying the same
in one way or another. The question, in any case, is to right-angle mirror-reversed structure I claim to detect
what extent Hegel’s notion of romantic art remains in in the Boy Bitten by a Lizard (pp. 9–11). See also my essay
force in ‘our’ encounter with the Death of St Peter Martyr, ‘David/Marat: The Self-portrait of 1794’, in Another Light:
or indeed with other works by Savoldo to be treated in Jacques-Louis David to Thomas Demand (New Haven, CT,
this book. My thanks to Robert Pippin for discussing the and London, 2014), pp. 40–51.
above passage with me. See also Robert Pippin, ‘Hegel on 3 F, cat. 12.
Painting’, in The Art of Hegel’s Aesthetics: Hegelian Philo- 4 G, pp. 129–30, 428–31.
sophy and the Perspectives of Art History, ed. Paul A. Kott- 5 Mary Pardo makes the connection between the ‘Gas-
man and Michael Squire (Paderborn, 2017), pp. 211–37. ton de Foix ’ and the Giorgione, adding that we may never
11 See in this connection the superb discussion of know whether Giorgione actually painted such a picture
Lotto’s portraits by Stephen J. Campbell in The End- (P, p. 266).
less Periphery: Toward a Geopolitics of Art in Lorenzo 6 See in this connection Rona Goffen, Renaissance
Lotto’s Italy (Chicago, IL, and London, 2019), pp. 149–51. Rivals: Michelangelo, Leonardo, Raphael, Titian (New
 references to pages 49–85 175

Haven, CT, and London, 2002), pp. 31–67. My thanks to 15 F, cat. 8.

Stephen Campbell for the reference. 16 F, cat. 24.
7 Interestingly, Gilbert in passing associates the ‘Gas- 17 F, cat. 22.
ton de Foix ’ with Caravaggio’s Boy Bitten by a Lizard, but 18 See Hans Belting, ‘Poetry and Painting: Saint
says nothing more about the relation of the one to the Jerome in the Wilderness’, in Giovanni Bellini: Landscapes
other (G, p. 401). of Faith in Renaissance Venice, ed. Davide Gasparotto, exh.
8 See in this connection Michael Fried, After Caravag- cat., Los Angeles, CA, J. Paul Getty Museum (2017), pp.
gio (New Haven, CT, and London, 2016); and idem, Cour- 25–35; and (more broadly) Eugene F. Rice, Jr, Saint Jerome
bet’s Realism (Chicago, IL, and London, 1990). in the Renaissance (Baltimore, MD, 1985).
9 See for example G, pp. 346–7 and passim. 19 F, cat. 13.
10 F, cat. 8. 20 Keith Christiansen, The Age of Caravaggio, exh. cat.,
11 F, cat. 24. New York, Metropolitan Museum of Art; Naples, Museo
12 F, cat. 19. Nazionale di Capodimonte (New York, 1985), cat. 11, pp.
13 Creighton Gilbert, ‘Newly Discovered Paintings 79–81. See also the essays by Gilbert, Frangi and Maria
by Savoldo in Relation to their Patronage’, Arte Lombarda, Teresa Rosa-Barezzani in Giovanni Gerolamo Savoldo,
n.s., no. 96/97 (1–2) (1991), p. 41. Gilbert cites an essay Ritratto di Gentiluomo con Flauto (Brescia, 1994).
by Veca of 1981. 21 Bayer in PR, p. 139.
14 Roberto Longhi, ‘Cose Bresciane del Cinquecento’, 22 Cf. Gilbert’s extended and largely bafed discus-
in Scritti giovanili, 1912–1922, 2 vols, Edizione delle sion of the ‘strange gesture’ of the shepherd’s hands in
opere complete di Roberto Longhi (Florence, 1961), I, pp. G, pp. 342–3.
339–40 (my translation). The paragraphs read in the 23 F, cat. 32.
original: 24 F, cats 28, 31, 29. I owe to Francesco Frangi the
information that a fourth Magdalene is at present in the
Il Savoldo si respira alla prima; se non restasse che la
Contini Bonacossi collection in the Galleria degli Ufzi
mano, essa basterebbe per riconoscerlo. Eppoi v’è la
in Florence. Not having seen it, I shall not discuss it in
pasta fusa e compatta della casacca rossa studiata nelle
this book. See F, cat. 30.
solite tortuose chiavi di pieghe ove Savoldo si compìa
25 Mary Pardo, ‘The Subject of Savoldo’s Magdalene’,
ceva di risolvere certi suoi quesiti di forme; e v’è anche
Art Bulletin, LXXI/1 (March 1989), pp. 67–91. Further
– ed è peccato che sia caduto fuor della fotograa – un
page references will be in parentheses in the text. Pardo
bolla di luce che vaga sul margine del libro e ridona lib-
briey reviews several opinions of the relative dating of
ertà e vita imprevista a tutta la composizione dei lumi.
the four paintings, none of which bears a date (p. 69, n.3).
Ma bisogna ritornare a quella mano che afferra
26 On Titian’s Noli me tangere, see the brief discus-
l’occhio in una complicatezza formale da risolversi
sion in Una Roman D’Elia, The Poetics of Titian’s Religious
como il nodo di Salomone; credere ch’esse sia un gioco
Paintings (Cambridge, 2005), pp. 20–22.
prospettico sarebbe alquanto superciale; poichè è
scelta troppo ossuta e carnosa ad un tempo per potersi
installare entro pochi piani, entro un prismo prospet-
tico. In verità, esse nasconde e rivela un più semplice three • faces
e profondo anagramma pittorico; è un momento, è
1 F, cats 3, 4.
un’impressione di mano ssata con perspicacia mirabile
2 Roberto Longhi, ‘Due dipinti inediti di Giovan
da una specula appostata nel punto più strano; è uno
Gerolamo Savoldo’, in Saggi e richerche, 1925–1928, 2
scorcio ‘corsivo’ che sebbene attuato in pasta ancora
vols, Edizione delle opere complete di Roberto Longhi
‘quattrocentesca’ prepara le rapide mani di Caravaggio,
(Florence, 1967), I, p. 153. Andrea Bayer, in her cata-
et forse di Degas.
logue entry in PR on Savoldo’s Shepherd with a Flute in
My thanks to Walter Stephens for his assistance in trans- the Getty, notes that ‘Longhi said of this painting that
lating Longhi’s difcult Italian. the drapery, modeled with a directed fall of light and
176 references to pages 86–98

deep shadows, seemed “like a fragment of a Caravaggio”’ 8 M. A. Jacobsen, ‘Savoldo and Northern Art’, Art
(p. 139). Longhi’s Italian reads: ‘Il massimo di lume e il Bulletin, LVI/4 (1974), pp. 530–34.
massimo d’ombra sono qui congiunti direttamente per 9 Bernard Aikema, ed., Jheronimus Bosch e Venezia, exh.
opposizione e si giunge a quella sommità quasi abba- cat., Venice, Palazzo Grassi (2017). Essays by Aikema,
gliata della casacca sul ginocchio che, presa per sé sola, Rosella Lauber, Isabella di Lenardo, Jos Koldeweij, Giulio
apparirebbe come un frammento del Caravaggio, o di un Bono and Maria Chiara Maida.
Velázquez del 1620.’ 10 This is perhaps as good a place as any to cite a
3 F, cat. 1. recent book that bears indirectly on Savoldo’s treatment
4 In his dissertation Gilbert identies the gure as of drapery: Paul Hills, Veiled Presence: Body and Drapery
St George (G, p. 428). The correction to St Liberale is from Giotto to Titian (London, 2018). Hills never men-
made in Gilbert, ‘Discovered Paintings by Savoldo in tions Savoldo, but his Chapter Seven, ‘Lorenzo Lotto:
Relation to their Patronage’, Arte Lombarda, n.s., no. 96/ Drapery Possessed’ (pp. 149–71), is of interest.
97 (1–2) (1991), p. 41. 11 My reference, of course, is to Jacques Derrida’s
5 I mention Hugo van der Goes because of the prob- essay ‘Parergon’, in The Truth in Painting, trans. Geoffrey
able inuence on Savoldo’s treatment of hands, especially Bennington and Ian McLeod (Chicago, IL, and London,
in the early Elijah and SS Paul and Anthony, of the great 1978), pp. 15–147.
Flemish painter’s monumental Portinari Altarpiece 12 F, cat. 23.
(1475–8), which Savoldo would have seen in Florence 13 Giorgio Vasari, The Lives of the Painters, Sculptors
in Santa Maria Nuova, assuming that it is he and not and Architects, trans. William Gaunt, 4 vols (London and
another Brescian painter who is referred to in the let- New York, 1963), IV, p. 321.
ter from Michelangelo’s brother mentioned in the Intro- 14 Roberto Longhi, ‘Quesiti caravaggeschi, II: I prec-
duction. The connection has been recognized by several edenti’, Pinacotheca, 5–6 (1929), p. 259; repr. in Longhi,
authors, including Gaetano Panazza, Sybille Ebert- ‘Me pinxit’ e quesiti caravaggeschi, 1928–1934, Edizione
Schifferer and Mina Gregori in Savoldo 1990. delle opere complete di Roberto Longhi, vol. IV (Florence,
6 It should be noted that the date of the Moscow 1968), p. 98.
picture has been the object of considerable dispute. See 15 Andrea Bayer in PR: ‘The diminutive scenes in
the brief discussion in F, p. 37. The San Diego canvas is the background have been identied as episodes in
less problematic, but Gilbert in 1963 proposed a date of Matthew’s life, as known through the Golden Legend, a
1535–8, which seems at wrong; see Creighton Gilbert, thirteenth-century compilation of lives of the saints. That
ed., Major Masters of the Renaissance, exh. cat., Waltham, on the right most probably represents the saint – now
MA, Poses Institute of Fine Arts, Brandeis University elderly and seated before a re – receiving hospitality
(1963), p. 23. at the house of the eunuch of the queen of Ethiopia,
7 F, p. 36. See also Beverly Louise Brown, ‘From Hell where he had preached and exposed the chicanery of two
to Paradise: Landscape and Figure in Early Sixteenth- magicians. The scene on the left, in which four small
century Venice’, in Renaissance Venice and the North: gures, one lower to the ground, are silhouetted against
Crosscurrents in the Time of Bellini, Dürer, and Titian, ed. a towering edice seen in moonlight, is more difcult
Bernard Aikema and Beverly Louise Brown, exh. cat., to identify. One interpretation regards the gures as
Venice, Palazzo Grassi (New York, 2000), p. 427, n.57. the citizens of Ethiopia, whom Matthew is healing of
Brown suggests that the two gures are probably based the malign sorcery of the two magicians; the other sees
on Jacopo Caraglio’s engraving of the Fire in the Borgo the tower as the symbolic center of the vignette, meta-
from c. 1525, which, if true, would imply a later date phorically representing the virtuous edice the apostles
for the painting. The engraving is erroneously labelled could build through their knowledge of many tongues,
‘Aeneas rescuing his father’. Brown also thinks Savoldo’s contrasting it with the Tower of Babel. Another moment
Moscow picture probably reects a lost Giorgione ‘re during Matthew’s long sojourn in North Africa might
landscape’. also be represented. After the magicians were unable to
 references to pages 100–116 177

bring the dead son of the king of Ethiopia to life (perhaps p. 37). Savoldo’s painting, in other words, amounts to a
he is the gure on the ground), Matthew miraculously reinterpretation of this sequence of events, the sh not
did so, and in response the king and his people built a emerging from the water on its own initiative but because
great church’ (p. 138). summoned by Raphael.
16 The exhibition, called ‘Venezia Scarlatta: Lotto, 3 Gaetano Panazzi, ‘Gian Gerolamo Savoldo: quesiti
Savoldo, Cariani’ and comprising six paintings, was held risolti e problemi insoluti’, in Savoldo 1990, p. 34.
in 2017. 4 MC, pp. 195–201.
17 On Campagnola, see Antonio Corradero, ‘Giulio 5 As is suggested by Mina Gregori, ‘I temi della luce
Campagnola, un artista umanista’, Venezia Cinquecento, articiale nel Savoldo e le radici lombarde di Caravaggio’,
20 (2010), pp. 55–134, esp. p. 102, where the print is in Savoldo 1990, p. 91. Interestingly, a depiction of Christ
called ‘a work of rare and personal interpretive tension’. making a similar beckoning gesture is found in Giacomo
See also Brown, ‘From Hell to Paradise’, pp. 442–3, cat. Jaquerio’s small painting from the rst half of the 1400s
115, on The Astrologer. She calls the monster an ‘irides- of the Vocation of St Peter and the Liberation of St Peter from
cent dragon’, and sees it as owing less to Bosch than to Prison in the Palazzo Carignano, Turin.
Dürer (p. 442). She also states that the date implied by 6 F, cat. 15.
the sphere is 13 September 1509, when there was to take 7 F, cat. 16.
place a conjunction of the sun and moon, understood as a 8 Lucy Whitaker and Martin Clayton, with Aislinn
sign of impending devastation. She adds: ‘Demons were Loconte, The Art of Italy in the Royal Collection: Renaissance
thought to wait until certain phases of the moon made and Baroque (London, 2007), pp. 210–11.
it more favorable for them to infest bodies.’ More on 9 In fairness to Gilbert, he adduces various ‘stylistic’
demons in Part Two of this book. reasons for the earlier dating of the Turin canvas, but
Highly interesting remarks about Campagnola, one gets the impression that the iconological argument
including a brief discussion of The Astrologer, are also is decisive for him.
to be found in Stephen J. Campbell, ‘Naturalism and the 10 Peter Humfrey, in Lorenzo Lotto: Rediscovered
Venetian “Poesia”: Grafting, Metaphor, and Embodiment Master of the Renaissance, ed. David Alan Brown, Peter
in Giorgione, Titian and the Campagnolas’, in Subject Humfrey, Maure Lucco et al., exh. cat., Washington, DC,
as Aporia in Early Modern Art, ed. Lorenzo Pericolo and National Gallery of Art; Bergamo, Accademia di belle
Alexander Nagel (Burlington, VT, and Farnham, 2010), arti G. Carrara; Paris, Galeries nationales du Grand Pal-
pp. 113–40. ais, 1998–9 (New Haven, CT, and London, 1997), p. 182.
11 Bernard Berenson, Lorenzo Lotto: An Essay in
Constructive Art Criticism (New York and London, 1895),
p. 327.
four • magic
12 Stephen J. Campbell, ‘Renaissance Naturalism and
1 F, cat. 10. the Jewish Bible: Ferrara, Brescia, Bergamo, 1520–1540’,
2 In fact, in the book of Tobit, where the Tobias story in Judaism and Christian Art: Aesthetic Anxieties from the
is found, the capture of the sh is related as follows: Catacombs to Colonialism, ed. Herbert Kessler and David
‘Tobias set out on his journey with a dog following, and Nirenberg (Philadelphia, PA, 2011), p. 209; and Camp-
the rst stop was by the river Tigris. When he went to bell, The Endless Periphery: Towards a Geopolitics of Art
wash his feet, he saw a huge sh come up and try to eat in Lorenzo Lotto’s Italy (Chicago, IL, and London, 2019),
him. Terried by the sh, Tobias cried out and said, “It’s chap. 5, ‘Brescia and Bergamo, 1520–1550: Sacred Natu-
coming after me, sir!” And the angel said to him, “Grab ralism and the Place of the Eucharist’, pp. 181–226. Fur-
it by the gills and pull it toward you.” When he did this ther page references to Campbell, ‘Renaissance Natural-
and dragged it on to the dry land, the sh began to gasp ism’, will be in parentheses in the text.
near his feet’ (Brian Copenhaver, ed., The Book of Magic: 13 See, in particular, Walter Stephens’s indispensable
From Antiquity to the Enlightenment [London, 2015], book Demon Lovers: Witchcraft, Sex, and the Crisis of Belief
178 references to pages 116–28

(Chicago, IL, and London, 2002), esp. chap. 8, ‘“This Is f i v e • t h e b r e s c i a a d o r at i o n o f

My Body”: Witchcraft and Desecration’, pp. 207–40. The the shepherds
core of Stephens’s compelling argument is that the real-
1 F, cat. 40. See also the catalogue entries by Pier Vir-
ity of witches’ experiences, including sexual contact with
gilio Begni Redona on the Brescia Adoration and the sub-
demons, was important to Renaissance inquisitors pre-
sequent variants in San Giobbe, Venice, and the Chiesa di
cisely because it served indirectly to conrm the reality –
Santa Maria la Nova, Terlizzi, in Savoldo 1990, pp. 106–
the divine status – of the Eucharist, belief in which turns
15; and the important essay by Bruno Passamani, ‘La
out to have required constant reinforcement. There will
“Natività” della Pinacoteca Tosio Martinengo di Brescia’,
be more to say about demons in Part Two of this book.
in Giovanni Gerolamo Savoldo, Pittore Bresciano, ed. Mario
14 F, cat. 9.
Pedini (Brescia, 1984), pp. 89–98.
15 A length of yellow-gold drapery trails across his
2 See Stephen J. Campbell, ‘Renaissance Naturalism
left upper arm before falling to the ground behind him;
and the Jewish Bible: Ferrara, Brescia, Bergamo, 1520–
might this be an indication that he is Jewish, hence that
1540’, in Judaism and Christian Art: Aesthetic Anxieties
he is to be taken as Joseph after all? In this connection
from the Catacombs to Colonialism, ed. Herbert Kessler and
see, for example, Joseph Leo Koerner, The Reformation of
David Nirenberg (Philadelphia, PA, 2011), p. 305.
the Image (London, 2004), who refers to yellow as ‘the
3 Gilbert glosses the painting’s ‘astounding spatial
colour that Jews – in Christian Europe, since the thir-
arrangement’ by noting that ‘[e]ach of the four major
teenth century, and in Islam since about the eighth – were
gures has his own little space, marked off from the
forced to wear to symbolize their infamy’ (p. 371). This
others by a wall, a ledge, or a window, making a kind of
is said apropos of the gure of Judas in Lucas Cranach’s
labyrinth of two-dimensional design’ (G, p. 137).
Wittenberg Altarpiece, the artefactual focus of Koerner’s
4 Campbell, ‘Renaissance Naturalism’, pp. 303–5.
deeply learned and closely argued book.
5 Campbell’s perspective is somewhat different. After
16 Compare Gilbert on the hands, which he sees as
noting various signs of the supernatural, such as the
recalling works from disparate moments in Savoldo’s
‘negative halo produced by the decay of the wooden
career; this suggests to him that ‘Savoldo kept on hand
planks’ behind the Christ Child, he remarks (ibid., p. 305):
a le of motives’, in this case, hands. Specically, he sees
‘But supernatural effulgence is now only a supplement or
the right hand as that of ‘an elderly male saint lifted to
foil for an emphatically material and everyday manifes-
express wonderment over the Christ Child [cf. Jerome
tation of the divine: as the shepherds regard the Christ
in the Turin Adoration] and his left hand [as] the hand
child through frames and across parapets, it is impressed
of an elderly male saint in kneeling adoration with his
upon us in our parallel condition as viewers of the paint-
elbow behind it pressed down on a ledge [cf. the Peni-
ing that the divine does indeed exist in the realm of facts
tent St Jerome]’ (G, p. 149). Such a manner of proceeding
accessible to human vision. This is the case even if we
seems to me foreign to Savoldo’s aims.
perceive it through the mediation of frames and thresh-
17 F, cat. 37.
olds that mark off the domain of the sacred but do not
18 Gilbert doubts the painting’s authenticity on vari-
disrupt its continuity with the world from which we
ous grounds, including the gesture of the gure at the
regard it. I will again refer to this style of handling reli-
left: ‘The emphasis given to a genre motive, in which a
gious subjects as “sacred naturalism”; it is distinguished
shepherd plays a little game with his ngers to amuse
not simply by delity to natural appearances, but by hier-
the child, is also outside Savoldo’s ways of proceeding’
archically ordered degrees of reality within the pictorial
(G, p. 154). Needless to say, I take this reading of the
eld itself. It is as if that which we, like the shepherds,
gesture to be off-key. In the end Gilbert thinks the
perceive just beyond the frame is the most “real” of all –
National Gallery canvas ‘may be accepted as a fairly close
the incarnate body of Christ.’
copy of a lost work’ (G, p. 155).
6 F, cat. 41. See also the brief discussion of the
19 F, cat. 20.
San Giobbe painting in a brochure devoted to the
church, Lorenzo Finerchi Ghersi, Augusto Gentili and
 references to pages 136–44 179

Carlo Corsato, The Church of San Giobbe (Venice, 2007), work, specically in his Martyrdom of St Sebastian (1459;
pp. 23–7. Vienna, Kunsthistorisches Museum) and Pallas Expelling
the Vices from the Garden of Virtue (1503; Paris, Louvre).
(In the rst of these, a rocky mass in the form of a head
is also evident.) To Piero di Cosimo, to his Misfortunes
pa r t t w o
of Silenus (c. 1500; Cambridge, MA, Fogg Art Museum),
1 It is almost as if the tabletop in the immediate fore- in which Erwin Panofsky found a tree-trunk with ‘an
ground is shared with the viewer. Compare the discus- excrescence resembling the head of a deer’ (see Panofsky,
sion of a comparable structure in an early self-portrait ‘The Early History of Man in Two Cycles of Paintings
drawing by Henri Fantin-Latour in Michael Fried, by Piero di Cosimo’, in Studies in Iconology [New York,
Manet’s Modernism, or, The Face of Painting in the 1860s 1972], p. 63). To Lotto, to tree-trunks resembling the
(Chicago, IL, and London, 1996), pp. 377–8. limbs of a human body in his St Jerome in the Wilderness
2 Gilbert twice discusses the self-portraits as a group: (1509–10; Rome, Castel Sant’Angelo), as well as a pos-
rst in his dissertation (G, pp. 426–34) and subsequently sible head emerging from the rocks behind the saint in
(now adding the Self-portrait as St Jerome) in his article the Penitent St Jerome (1506) in the Louvre. To Zenale,
‘Newly Discovered Paintings by Savoldo in Relation to to a mysterious demonic face in the rock formation over
their Patronage’, Arte Lombarda, n.s., no. 96/97 (1–2) the Virgin’s head in his altarpiece (c. 1510) for the Mila-
(1991), pp. 41–4. nese church of San Francesco Grande, today in the Den-
3 Gilbert, ‘Newly Discovered Paintings’, pp. 41–2. ver Museum of Art. And to Giovanni Bellini (not previ-
4 F, cat. 7. ously remarked, as far as I can tell, though that scarcely
5 This remains the case in his catalogue entry for the seems possible), to a large partial head and face in prole
painting in T, cat. 48, p. 162. in the rock wall confronting the saint in his St Jerome
6 Gilbert, ‘Newly Discovered Paintings’, p. 41. in the Wilderness (c. 1480; Florence, Ufzi). My thanks
7 Ibid., p. 42. to Stephen Campbell for the Zenale reference, as well as
8 Ibid., p. 43. for the information that the detail is shown in isolation
9 On the Council of Trent, see John O’Malley, Trent: on the museum website (https://denverartmuseum.org).
What Happened at the Council? (Cambridge, MA, 2013). I will add that the face in the Zenale altarpiece is at least
Various dimensions of the struggle for reform within the as difcult to make out as any cited by me in connection
Catholic Church (with particular emphasis on Venice) are with Savoldo.
analysed by Elizabeth Gleason, Gasparo Contarini: Venice, Relevant studies of such imagery include Daniel
Rome, and Reform (Berkeley, CA, Los Angeles and Oxford, Arasse, ‘Lorenzo Lotto dans ses bizarreries: le peintre et
1993). Also informative are Thomas F. Mayer, Reginald l’iconographie’, in Lorenzo Lotto: atti del convegno inter-
Pole: Prince and Prophet (Cambridge, 2000); O’Malley, nazionale di studi per il V centenario della nascita, ed. Pietro
The First Jesuits (Cambridge, MA, 1995); and Adam Pat- Zampetti and Vittorio Scarbi (Treviso, 1981), pp. 365–82;
rick Robinson, The Career of Cardinal Giovanni Morone see also Arasse, Le Détail: Pour une histoire rapprochée de
(1509–1580): Between Council and Inquisition (Burlington, la peinture (Paris, 1992), pp. 244–5. Stephen J. Campbell,
VT, 2012). My thanks to John O’Malley for suggesting ‘Cloud-poiesis: Perception, Allegory, Seeing the Other’,
Gleason’s, Mayer’s and Robinson’s books to me. in Senses of Sight: Towards a Multisensory Approach of the
10 Joseph Leo Koerner, The Moment of Self-Portraiture Image. Essays in Honor of Victor I. Stoichita, ed. Henri de
in German Renaissance Art (Chicago, IL, and London, Riedmatten et al. (Rome, 2015), pp. 1–29 (mainly on
1993), pp. 27–8. Mantegna). Dario Gamboni, Potential Images: Ambigu-
11 The reference to Giotto is to the surprising recent ity and Indeterminacy in Modern Art (London, 2002), pp.
(2011) discovery of a demonic-seeming face in the clouds 27–31. James Elkins, Why Are Our Pictures Puzzles? On
in the upper portion of his fresco of the Death of St the Modern Origins of Pictorial Complexity (New York and
Francis in the Basilica of San Francesco in Assisi. To London, 1999), chap. 7, ‘Hidden Images: Cryptomorphs,
Mantegna, famously, to certain images in clouds in his Anamorphs, and Aleamorphs’, pp. 178–230. Famous
180 references to pages 144–8

texts by Leon Battista Alberti and Leonardo da Vinci are Krönig’, Aachener Kunstblätter, 51 (1971), pp. 229–30;
cited and discussed by H. W. Janson, ‘The Image Made Karl Möseneder, ‘Blickende Dinge: Anthropomorphes
by Chance in Renaissance Thought’, in De artibus opuscula bei Albrecht Dürer’, Pantheon, 45 (1986), pp. 15–23; Her-
XL: Essays in Honor of Erwin Panofsky, ed. Millard Meiss, mann Leber, Albrecht Dürers Landschaftsaquarelle: Topog-
2 vols (New York, 1961), I, pp. 254–66. In Le Détail, raphie und Genese (Hildesheim, Zürich and New York,
apropos of the anthropomorphic imagery in Lotto’s St 1988), pp. 33–9, 59, 147–69; and Felix Thürlemann,
Jerome paintings, Arasse refers to Lorenzo Giustiniani, ‘Der Balkon im Auge: Dürers Arco-Aquarell als Theo-
who ‘in his De Vita religiosa, published in Venice in 1494, rie der Mimesis’, Neue Zürcher Zeitung, no. 227 (30 Sep-
[gives the description] of diverse modes of temptation, tember–1 October 1993), pp. 67–8. Also on Dürer, see
whether the devils “offend the spirit” and lead it to forget Peter Parshall, ‘Graphic Knowledge: Albrecht Dürer and
to invoke God, or whether they “show themselves to the the Imagination’, Art Bulletin, 95 (September 2013), pp.
human spirit in a horrible manner by means of obscure 393–410. On Herri met de Bles, see Michel Weemans,
images and inhabitual aspects”’ (p. 244, my translation). Herri Met de Bles: Les Ruses du paysage au temps de Brue-
See also Campbell, ‘Naturalism and the Venetian “Poe- gel et d’Erasme (Paris, 2013). My thanks to Giovanni
sia”: Grafting, Metaphor, and Embodiment in Giorgione, Careri and Angela Mangoni for alerting me to Weeman’s
Titian and the Campagnolas’, in Subject as Aporia in Early work. Also pertinent are Joseph Leo Koerner, ‘Bosch’s
Modern Art, ed. Lorenzo Pericolo and Alexander Nagel Contingency’, in Kontingenz, ed. Henri de Riedmatten
(Burlington, VT, and Farnham, 2010), pp. 113–40. And et al. (Munich, 1998), pp. 242–83; Koerner, ‘Impossible
on metamorphic transformations generally, see Michel Objects: Bosch’s Realism’, RES: Anthropology and Aesthet-
Jeanneret, Perpetual Motion: Transforming Shapes in the ics, 46 (2004), pp. 73–97; Reindert L. Falkenburg, ‘The
Renaissance from da Vinci to Montaigne, trans. Nidra Poller Devil Is in the Detail: Ways of Seeing Joachim Patinir’s
(Baltimore, MD, and London, 2001). “World Landscapes”’, in Patinir: Essays and Critical Cata-
Two additional observations (the rst owed to Keith logue, ed. Alejandro Vergara, exh. cat., Madrid, Museo
Christiansen): one nds a deliberate, Mannerist play del Prado (Madrid, 2007), pp. 61–79; and Stephen Camp-
with heads and faces, including a large partial mask-like bell’s brief discussion of the sphinx-like rock formation
face (eyes, nose, nostrils), in the lower part of the sitter’s in the background of Cosmè Tura’s Annunciation on the
costume in Agnolo Bronzino’s Portrait of a Young Man Ferrara Cathedral organ-shutters (Cosmè Tura of Ferrara:
(1530s; New York, Metropolitan Museum of Art); and is Style, Politics and the Renaissance City, 1450–1495 [New
there not an unmistakable face or mask looking aggres- Haven, CT, and London, 1998], p. 156). Koerner further
sively, one might say deantly at the viewer (out of its develops his account of Bosch in Bosch and Bruegel: From
left eye), in the folds of the silver-grey left sleeve of the Enemy Painting to Everyday Life, The A. W. Mellon Lec-
protagonist of Jacopo da Pontormo’s Portrait of a Young tures in the Fine Arts (Princeton, NJ, and Oxford, 2016);
Man in a Red Cap (Carlo Neroni?) (c. 1530; private col- see in particular his discussion of and around Bosch’s Lis-
lection), another plainly deliberate, Mannerist tour de bon Temptation of St Anthony, pp. 155–78. Finally, both
force? This too seems to have gone unremarked by com- Italian and Northern images (including nine by Dürer)
mentators. are illustrated and discussed under the rubric of ‘double
Finally, let me cite ‘The End of the Masquerade’, the images’ in Jean-Hubert Martin and Dario Gamboni, eds,
last chapter of Charles Dempsey’s Inventing the Renais- Une image peut en cacher une autre, exh. cat., Paris, Galeries
sance Putto (Chapel Hill, NC, and London, 2001), pp. 219– nationales du Grand Palais (Paris, 2009).
31, with its beautiful closing pages on masks in Michel- 13 And an equal stretch in the same painting: note
angelo’s Medici Chapel. how the white robe of the second Dominican friar being
12 See, for example, the discussion of various works attacked in the right middle distance may (almost?) be
by Dürer, including the ink drawing of six pillows, in seen as a head in prole, with a long Jimmy Durante-like
Gamboni, Potential Images, pp. 31–3. Gamboni cites four nose pointing downward at the left.
previous texts in particular: Heinz Ladendorf, ‘Ein Fels- 14 Compare Leonardo: ‘The painter’s mind endeav-
gesicht bei Albrecht Dürer’, in ‘Festschrift für Wolfgang ours to be a mirror, for a mirror always makes itself have
 references to pages 150–52 181

the colour of the object that is reected by it. A mirror (Geneva and Lille, 1954); Stuart Clark, Thinking with
is indeed full of as many things as happen to stand oppo- Demons: The Idea of Witchcraft in Early Modern Europe
site it. A painter ought to know therefore that he can- (Oxford, 1997); Clark, Vanities of the Eye: Vision in Early
not be good at his craft unless he is a universal master Modern European Culture (Oxford, 2007), esp. chap. 2,
of creating through his art all the qualities of the forms ‘Fantasies: Seeing Without What Was Within’, pp. 39–77,
which nature produces. But this cannot be done unless and chap. 4, ‘Glamours: Demons and Virtual Worlds’, pp.
he sees them and retraces them in his mind . . . And in 123–60; Norman Cohn, Europe’s Inner Demons: The Demon-
fact, whatever is in the universe as essence, occurrence, ization of Christians in Medieval Christendom (New York,
or imagination, the painter must rst have in his mind 1975); Michael Cole, ‘The Demonic Arts and the Origin
and then in his hands. His hands must be of such excel- of the Medium’, Art Bulletin, LXXXIV/4 (December 2002),
lence that they can shape things into a well-proportioned pp. 621–40; Brian Copenhaver, ‘How to Do Magic, and
harmony by a single glance and take no more time doing Why?’, in The Cambridge Companion to Renaissance Philos-
it than it takes the things to be.’ Cited in Eugenio Garin, ophy (Cambridge, 2007), pp. 137–70 (on Ficino); Marsilio
Science and Civic Life in the Renaissance, trans. Peter Munz Ficino, The Book of Life, trans. Charles Boer (Woodstock,
(Garden City, NY, 1969), pp. 51–2. Compare the slightly CT, 1980); Eugenio Garin, Astrology in the Renaissance:
different version in Edward MacCurdy, ed., The Notebooks The Zodiac of Life, trans. Carolyn Jackson and June Allen,
of Leonardo da Vinci (New York, 1955), p. 857. rev. by the author and Clare Robertson (London, 1983),
15 Mary Pardo, ‘The Subject of Savoldo’s Magdalene’, esp. pp. 61–78 (on Ficino); Anthony Grafton, Cardano’s
Art Bulletin, LXXI/1 (March 1989), p. 83. Further page Cosmos: The Worlds and Works of a Renaissance Astrologer
references to Pardo’s article will be in parentheses in the (Cambridge, MA, and London, 1999), esp. chap. 9, ‘Rival
text. Disciplines Explored’; Frank Klaassen, The Transforma-
16 D. P. Walker, Spiritual and Demonic Magic from tions of Magic: Illicit Learned Magic in the Later Middle Ages
Ficino to Campanella (London, 1958), p. 36. and Renaissance (University Park, PA, 2013), esp. chap.
17 Ioan P. Couliano, Eros and Magic in the Renaissance, 7, ‘Medieval Ritual Magic and Renaissance Magic’, pp.
trans. Margaret Cook (Chicago, IL, and London, 1987), 187–218 (on Ficino and Agrippa); Philippe Morel, Magie,
p. 161. A book of considerable originality, with particular astres et démons dans l’art italien de la Renaissance (Paris,
emphasis on the thought of Ficino and Bruno. 2008); Ingrid Rowland, Giordano Bruno: Philosopher/Her-
18 Charles G. Nauert, Jr, Agrippa and the Crisis of etic (Chicago, IL, and London, 2008); Juanita Feros Ruys,
Renaissance Thought (Urbana, IL, 1965), p. 269. Nauert’s Demons in the Middle Ages (Kalamazoo, MI, and Bradford,
basic view owes much to Eugenio Garin, as when he 2017); Walter Stephens, Demon Lovers: Witchcraft, Sex,
writes: ‘so far was medieval magic from perishing in the and the Crisis of Belief (Chicago, IL, and London, 2002),
Renaissance, that some intellectual historians have con- esp. chap. 8, ‘“This Is my Body”: Witches and Desecra-
cluded that magic became more important, not less so, tion’, pp. 207–40. Stephens’s article ‘“In the Body”: The
in the later epoch. From being an unspoken and often Canon Episcopi, Andrea Alciati, and Gianfrancesco Pico’s
unacknowledged element in the mental world of Euro- Humanized Demons’, in Demonology and Witch-hunting
pean men, magic during the Renaissance stepped forth in Early Modern Europe, ed. Julian Goodare, Rita Volt-
into the light of day as a central element of culture; and mer and Liv Helene Willumsen (forthcoming), is also of
the Renaissance marks not a stage in the abandonment great interest, as is his translation of an important text
of the occult in favor of pure reason, but a re-emphasis discussed in that article, Gianfrancesco Pico della Miran-
of the magical world view’ (pp. 225–6). See in this con- dola’s The Witch, or On the Deceptions of Demons, to appear
nection Garin’s brilliant chapter ‘Magic and Astrology in the I Tatti Renaissance Library (in preparation).
in the Civilization of the Renaissance’, in his Science and Stephens, ‘Habeas Corpus: Demonic Bodies in Ficino,
Civic Life in the Renaissance, esp. pp. 163–5. Psellus, and Malleus Maleficarum’, in The Body in Early
19 In addition to the books by Walker, Couliano and Modern Italy, ed. Julia L. Hairston and Walter Stephens
Nauert cited above, I have found particularly instruc- (Baltimore, MD, 2010), pp. 74–91; Claudia Swan, Art, Sci-
tive the following: André Chastel, Marsilio Ficino et l’art ence and Witchcraft in Early Modern Holland: Jacques de
182 reference to page 152

Gheyn ii (1564–1629) (Cambridge and New York, 2005), See also Cardano, The Book of My Life, trans. Jean Stoner,
esp. chaps 5 and 6; Stéphane Toussaint, ‘L’ars de Marsile intro. Anthony Grafton (New York, 2002).
Ficin, entre esthétique et magie’, in L’art de la Renaissance The second, from Cole’s superb article on the demonic
entre science et magie, ed. Philippe Morel (Rome, 2006), arts, cites Francesco Cattani da Diacceto’s discourse
pp. 455–67 (a rich essay, which tellingly contrasts Pan- ‘On the Art of Magic’s Superstition’ (Discorso . . . sopra
ofskyan and Warburgian conceptions of Renaissance art, la superstizzione dell’arte magica [1567]): ‘[Demons] can
the latter, of course, open to the demonic); Christopher also form themselves into bodies and present themselves
S. Wood, ‘Countermagical Combinations by Dosso to our eyes in various aspects [specie], it being within
Dossi’, RES: Anthropology and Aesthetics, nos 49/50 their power to operate those things that one conducts to
(Spring–Autumn 2006), pp. 151–70; Frances A. Yates, an end with the local motions of inferior bodies. One of
Giordano Bruno and the Hermetic Tradition (Chicago, IL, the things demons can do is operate bodies that appear
and London, 1964) and Charles Zika, The Appearance of to be men, or some sort of animal, the likeness of this
Witchcraft: Print and Visual Culture in Sixteenth-century animal consisting in its gure and its color. The gure
Europe (London, 2007). The title of Gianfrancesco’s text is induced by means of local motion, just as painters, by
alerts one to the fact that demons were frequently said means of brushes and other instruments, color their bod-
to cause delusions in their victims, as was also remarked ies. In this manner, then, they gure and color their bod-
by contemporary writers on melancholy, another major ies, and the bodies then appear at one moment in the form
theme in the early modern period. On melancholy and of a man, in the next in the form of a woman, in the next
delusions see, in particular, Clark, ‘Fantasies’, and of in the form of an animal, or of another thing, according
course Robert Burton, The Anatomy of Melancholy (New to what the demons judge most harmful to others’ (‘The
York, 2001) and Raymond Klibansky, Erwin Panofsky Demonic Arts and the Origin of the Medium’, p. 623).
and Fritz Saxl, Saturn and Melancholy: Studies in the His- Cole goes on to argue, persuasively to my mind, that
tory of Natural Philosophy, Religion, and Art (Cambridge, for Diacetto and others the inferior bodies in question
1964). were mainly those of air. But in the present context might
Two passages from works just cited are particularly one not also imagine that they could be folds, creases,
arresting. The rst, from Grafton’s study of Cardano (a depressions, bulges and other irregularities in drapery?
famous astrologer, physician and mathematician), conveys Or indeed (coming full circle metaphorically) the imita-
one version of mid-sixteenth-century Italian thought on tion of such folds, creases, bulges and other irregularities
the topic of demons: ‘In [his book] On Variety, Cardano . by a painter’s brushwork?
. . described, at great length, the world and habits of the One further set of remarks. Although there exists no
daemons who inhabited the upper realms . . . Later in hard documentation for the events in question, in Mex-
the work, in the detailed discussion of daemons to which ico in 1531 a recent Catholic convert named Juan Diego,
Cardano devoted book 16, he described at horric length passing the hill of Tepeyac on his way to Tlatelolco, was
the ways in which they interacted with humans: “They accosted by a resplendent vision of the Virgin Mary, who
come, sometimes, when called, or produce the image of instructed him to go to his bishop and tell him to have a
one coming. Sometimes they are gentle and wise, and church built on the spot. Juan Diego tries to do so but is
predict certain future things, surrounding them with a put off; but the Virgin appears to him again and instructs
thousand ambiguities and mingling them with lies. Oth- him to gather owers and conceal them in his filma or
ers choke humans, or, if they cannot do that, drive them cloak, then return to the bishop and open the cloak in his
to despair. They make some think they are entering their presence. When Juan Diego does so the owers fall to
bodies. They kill the sons of others, not on their own, but the ground and reveal, imprinted on the cloak, a beauti-
by a certain art, in exactly the same way that men use net ful full-length picture of the Virgin. The image in ques-
and trident to kill sh on the bottom of the sea.” Cardano tion is the famous Our Lady of Guadalupe, which eventu-
advised against becoming acquainted with daemons, ally became the most sacred image in Mexican history
since, like tyrants, powerful men, and wild animals, they and is today found in the Minor Basilica of Our Lady of
made dangerous company’ (Cardano’s Cosmos, pp. 167–8). Guadalupe in Mexico City. Not surprisingly, modern
 references to pages 152–6 183

scholarship has raised a host of serious questions about already and from the start entirely relational (pp. 91–2;
this account, starting with the supposed date of the event, my translation).
but what strikes me, if one regards the date of 1531 not
as a chronological fact but nevertheless as a signicant (Ils sont donc à la fois ce qui caractérise notre individ-
feature of the story, is the coincidence with the years of ualité et ce qui nous permet de communiquer avec les
Savoldo’s activity – by which I mean that the alleged spec- autres et avec le monde. Car chacun entre nous n’est
tacular appearance of the image of the Virgin in a piece que le noeud n’un réseau, de telle sorte qu’il est vain
of clothing in the New World forms an antithetical, reli- de vouloir séparer ce qui nous appartient en propre et
giously positive counterpart to the proliferation in North ce qui nous a été transmis. De même qu’il est inutile
Italy of (until now) unseeable demonic faces and masks in de vouloir distinguer une intériorité qui serait notre
the clothing and drapery in Savoldo’s paintings. On the seule affaire et nos gestes et manière qui relèveraient
Mexican image, see, for example, Stafford Poole, Our Lady du collectif. Car nous n’avons jamais que les senti-
of Guadalupe: The Origins and Sources of a Mexican National ments et les sensations et les pensées que notre entou-
Symbol, 1531–1797 (Tucson, AZ, 1995, rev. 2017). rage nous a appris et qui se sont inscrits en nous. Nous
For a stimulating treatment of the concept of a medium ne sommes qu’une résultante de forces multiples.
(I am here referring back to Cole’s article), see Antonio L’entrecroisement de la trame et de la chaîne du tissu
Somaini, ‘Walter Benjamin’s Media Theory: The Medium socioculturel qui nous constitue est à la fois ce point
and the Apparat’, Grey Room, no. 62 (Winter 2016), pp. particulier et l’ensemble de la toile. Ce qui en nous est
6–41. le plus inaliénable est déjà et depuis toujours instance
20 My deployment of the concept of ‘inuence’ in this relationnelle.)
book has also been coloured by the arguments developed
21 Clark, Thinking with Demons, p. 458.
in François Roustang’s subtle account of Ericksonian
22 Stephen B. Dowd, Venice’s Most Loyal City: Civic
suggestion and its implications for psychoanalytic prac-
Identity in Renaissance Brescia (Cambridge, MA, 2010),
tice in the former’s Influence (Paris, 1990). Thus Roustang
p. 188. The entire chapter ‘Witches’, pp. 174–91, which
remarks apropos of all the traits, habits, assumptions and
also describes a political struggle between the Venetian
elements of seemingly personal style a man or woman
authorities and the inquisitorial branch of the Church,
acquires without being aware of it from his or her accul-
is of interest in this context. See also Stephen J. Camp-
turation, surroundings, interactions with others, and
bell, The Endless Periphery: Towards a Geopolitics of Art in
the like:
Lorenzo Lotto’s Italy (Chicago, IL, and London, 2019), in
They are at the same time that which characterizes our particular the (to me) indispensable chap. 5, ‘Brescia and
individuality and that which allows us to communicate Bergamo, 1520–50: Sacred Naturalism and the Place of
with others and with the world. Because each of us is the Eucharist’, pp. 181–226.
only the nodal point of a network, in such a way that 23 Savoldo’s ‘continual contact’ with the Dominican
it is vain to wish to separate that which truly belongs Order is noted by Gregori in her essay on the Death of St
to us and that which was transmitted to us. Just as it Peter Martyr (p. 74), with particular reference to an article
is useless to want to distinguish an interiority that by Alessandro Nova detailing those relations, ‘Brescia
would be our exclusive affair and those of our gestures and Frankfurt: Savoldo’, Burlington Magazine, CXXXII,
and manner that derive from the collective. Because no. 1047 (1990), pp. 433–4.
we always have only the feelings and sensations and 24 Paul Vandenbroeck, ‘The Axiology and Ideology
thoughts that our entourage has taught us and which of Jheronimus Bosch’, in Bosch: The 5th Centenary Exhibi-
are inscribed in us. We are only the product of mul- tion, ed. Pilar Silva Maroto, exh. cat., Madrid, Museo del
tiple forces. The intermingling of the weave and of the Prado (London, 2017), p. 92.
chain of socio-cultural tissue that constitutes us is at 25 Apropos of what he takes to be the relative dis-
the same time this particular point and the entirety of appearance of double images and the like from Italian
the canvas. That in us which is the most inalienable is painting after Mantegna, Stephen J. Campbell remarks:
184 references to pages 156–60

‘Most likely, the need to sustain the enterprise of Chris- heroes. Even when portraits are not counted, paintings of
tian art, increasingly the target of iconoclastic critiques only one or two gures play a role in his oeuvre perhaps
as the century progressed, could no longer sustain such greater than in that of any other contemporary. He makes
open avowals of the vagaries of sight’ (‘Cloud-poiesis’, them, in their emphatic airy empty worlds, subjects as
p. 19). Which, if true, makes the occurrence of faces and well as objects of monumental meditation’; ‘Several of
masks in the drapery and clothing in Savoldo’s paintings the Contexts of Savoldo’s “Dead Christ”’, Bulletin of the
all the more remarkable. Cleveland Museum of Art, LXXIX/1 (1992), p. 32. Here,
26 Creighton E. Gilbert, ‘Savoldo, Giovanni Girolamo’, too, the notion of ‘emphatic airy empty worlds’ is hard to
in The Dictionary of Art, ed. Jane Turner (London, 1996), reconcile with what I have tried to show is the intensely
XXVII, p. 892. Also: ‘Savoldo was an “independent” work- somatic internal dynamic of Savoldo’s paintings. In a
ing in Venice in the age of Titian. His failure to follow similar vein, Gilbert in his dissertation remarks in cer-
that great painter’s taste cost him fame and success; tain paintings Savoldo’s gures’ ‘lassitude, the sense that
indeed, the best known fact about him in his lifetime was their heroism has nothing to apply itself to in this perfect
that he was little known’ (Creighton Gilbert, ed., Major empty world’ (G, p. 353).
Masters of the Renaissance, exh. cat., Waltham, MA, Poses 34 Pardo adds in a note: ‘Gilbert [in his disserta-
Institute of Fine Arts, Brandeis University (1963), p. 23). tion] has a very seductive hypothesis for the Magdalene ’s
27 See Adriano Prosperi, ‘The Religious Crisis in original ownership: in 1527, Savoldo was commissioned
Early Sixteenth-century Italy’, in Lorenzo Lotto: Redis- to do a Saint Jerome (very probably the painting in Lon-
covered Master of the Renaissance, ed. David Alan Brown, don, National Gallery, no. 3092) for Giovan Paulo Aver-
Peter Humfrey and Mauro Lucco, exh. cat., Washing- oldi, a member of the prominent Brescian family . . . The
ton, DC, National Gallery of Art; Bergamo, Accademia di archives show that G. P. Averoldi . . . regularly invested
belle arti G. Carrara; Paris, Galeries nationales du Grand in vanguard Venetian painting. The London Magda-
Palais (New Haven, CT, and London, 1997–9), pp. 21–6. lene was recorded in the Averoldi Collection by 1620 .
Also Massimo Firpo, Artisti, gioiellieri, eretici: Il mondo di . .; perhaps it had been there, like the nearly contempo-
Lorenzo Lotto tra Riforma e Controriforma (Rome, 2001). rary Jerome, since the cinquecento. If it had, it was indeed
28 Cited by Mauro Lucco in the catalogue entry for painted for an audience of cognoscenti’ (p. 90, n.88).
Lotto’s Virgin and Child with SS Jerome and Nicholas of More broadly, in the brief review of the 1990 exhibi-
Tolentino (1523–4) in Lorenzo Lotto, ed. Brown, Humfrey tion ‘Giovanni Gerolamo Savoldo tra Foppa, Giorgione
and Lucco, p. 147. e Caravaggio’ (see n.23), Alessandro Nova remarks that,
29 Also a far more complex gure in important according to Marcantonio Michiel, a Venetian noble-
respects. See in this connection the brilliant analysis of man interested in art, Savoldo had four pictures in two
Lotto’s art and its many references to the work of other of the best and most selective Venetian collections (none
artists, and indeed a range of different religious and artis- of the pictures seems to have survived): ‘From Michiel’s
tic sites, in Campbell, The Endless Periphery, esp. chap. 4, description of these Venetian palaces,’ Nova writes, ‘it is
‘Distant Cities: Lorenzo Lotto and Gaudenzio Ferrari’, clear that Savoldo’s paintings were displayed in a sort of
pp. 97–180. Wunderkammer where one could also admire the works of
30 Roberto Longhi, ‘Cose Bresciane del cinquecento’, the best artists of the time (Giorgione, Titian, Cariani,
in Scritti giovanili, 1912–1922, 2 vols, Edizione delle opere Palma, Lotto) mixed with Roman antiquities, precious
complete di Roberto Longhi (Florence, 1961), I, p. 339. crystals, medals and all sorts of natural and unnatu-
31 Alessandro Ballarin, ‘Gerolamo Savoldo: Gli albori ral curiosities. This atmosphere of rarity and exclusive
d’un nuovo umanesimo nella pittura del Savoldo’, in Bal- elitism perhaps relates as well to Savoldo’s slender out-
larin, Gerolamo Savoldo (Milan, 1966), n.p. put, his meticulousness and slow working procedures’
32 Gaetano Panazza, ‘Gian Gerolamo Savoldo: quesiti (p. 433). As Nova also remarks, the multiple Magda-
risolti e problem insoluti’, in Savoldo 1990, p. 37. lenes suggest that Savoldo painted the rst for a patron,
33 Finally, in a later essay, Gilbert notes what he sees and that as the painting became known other collec-
as Savoldo’s ‘lifelong preference for images of single quiet tors requested copies, or rather close variations (p. 433).
 references to pages 160–66 185

Fascinatingly, the pattern of folds and creases in the 40 See Michael Fried, After Caravaggio (New Haven, CT,
shawls differs sharply from one variant to another. and London, 2016), chap. 2, ‘Toward a Post-Caravaggio
35 Such a claim might better t Veronese, who in Pictorial Poetics’.
1573 defended himself before the Inquisition for includ- 41 Mina Gregori, ‘Caravaggio Today’, in The Age of
ing a depiction of a dog along with various seemingly Caravaggio, exh. cat., New York, Metropolitan Museum
merely secular gures in his magnicent canvas of the of Art; Naples, Museo Nazionale di Capodimonte (New
Last Supper in SS Giovanni e Paolo, or, even more closely, York, 1985), p. 39. In MC I write that the gure in ques-
innumerable artists of later times. But I don’t fault Pardo tion ‘looks back with an expression of dismay, or bitter-
for her proto-modernist leanings – my own private ness, or regret, as if – here one’s powers of description
ahistorical comparison for Savoldo has been Mallarmé. become uncertain – as if he were reluctant to depart? Or
36 See in this connection Keith Christiansen, ‘Gio- as if he regretted his participation, assuming it was such,
vanni Bellini and the Practice of Devotional Painting’, in in the murder? Or, indeed, as if he wished he might have
Giovanni Bellini and the Art of Devotion, ed. Ronda Kasl, interceded to save Matthew? . . . In any case, as an iconic
exh. cat., Indianapolis Museum of Art (Indianapolis, IN, surrogate for the painting’s maker, the eeing bravo’s
2004), pp. 7–58. The other essays in the catalogue (by implication in the murder is undeniable. So what are we
Kasl, Andrea Golden, David A. Miller, and Cinzia Maria to make of his imminent departure from the scene of the
Mancuso and Antonietta Gallone) are also of interest. crime?’ (p. 205).
For general background, see Eugene F. Rice, Jr, Saint 42 And something else. In MC I suggest that Caravag-
Jerome in the Renaissance (Baltimore, MD, 1985). gio’s paintings imply two distinct and notionally succes-
37 Belting’s essay, ‘Poetry and Painting: Saint Jerome sive polar ‘moments’ in their production: ‘The rst is a
in the Wilderness’, appears in Davide Gasparotto, ed., “moment” of extended duration of the painter’s engage-
Giovanni Bellini: Landscapes of Faith in Renaissance Venice, ment in the ongoing, repetitive, partly automatistic act of
exh. cat., Los Angeles, J. Paul Getty Museum, 2017–18 painting; I call that “moment” immersive, imagining the
(Los Angeles, 2017), pp. 1–23. The citations given are painter as so caught up, so immersed, in this phase of
to pp. 25 and 25–6. (The Ufzi picture he mentions is his work on the painting (or rather, simply, in his work
the one briey discussed by me earlier, see n.11, which on the painting) as to be less than fully aware of any
includes the large face-like rock mass directly in front of sharp distinction between the painting and himself. The
the saint.) An essay by Gasparotto, ‘Giovanni Bellini and second is a “moment”, notionally instantaneous, of sepa-
Landscape’, pp. 11–23, is also of interest. See also Belting, rating or indeed recoiling from the painting, of becom-
‘St. Jerome in Venice: Giovanni Bellini and the Dream of ing detached from it, which is to say of no longer being
the Solitary Life’, I Tatti Studies in the Italian Renaissance, immersed in work on it but rather of seeing it, taking it
XVII/1 (Spring 2014), pp. 15–33. in, as if for the rst time; I call that “moment” specular,
38 A claim that goes somewhat against the current. meaning thereby to emphasize the strictly visual or opti-
See Francesco Frangi, ‘Sacre conversazioni e immagini di cal relation between the artist-viewer and the image, or
meditazione, tra Venezia e Brescia’, in T, pp. 118–28, esp. image-artifact, that he has just brought into being. The
pp. 119–20. Interestingly, Frangi emphasizes the origi- contrast between the two “moments”, one might say, is
nality of both the Hampton Court and Turin composi- between the artist’s being “in” the painting (or at least
tions, but stops short of suggesting that they have other “continuous” with it in the ongoing process by which the
than ‘devotional’ ends in view. painted image was laid down on the canvas) and nding
39 On this difcult-to-parse painting, see the cata- himself “outside” the painting, of discovering that he
logue entry by Francesco Frangi in T, pp. 48–50. (The has become not just detached but distanced from it, in a
Milan canvas does not appear in his catalogue raisonné relationship of mutual facing (also mutual freezing) that
of 1992.) It might be noted that the tower at the extreme rst establishes the painted image as an image and with
right also contains a face or mask of sorts, with two it the painting as a picture, as fundamentally addressed
round eyes (windows) and a large open mouth (a squar- to a viewer – in the rst instance, to the artist himself
ish doorway). (or herself, needless to say). In Boy Bitten by a Lizard,
186 references to pages 167–9

the second or specular “moment” is dramatized to the nal, active presence of the painter’s right (brush) hand
extent of largely eclipsing the rst, which can be recov- approximately there.
ered, made intuitable, only by an act of interpretation that 47 This implies a two-part relation between Savoldo
cannot quite point to knockdown evidence in its support and (after a gap of fty-plus years) Caravaggio, though
either inside or outside the painting. By way of staking probably one should introduce at least two more terms
out further reaches of my argument, I want to claim that into the structure: the Bellinesque devotional picture,
precisely such a double or divided relationship between from which Savoldo fundamentally departs; and the
painter and painting – at once immersive and specular, work of Caravaggio’s successors (Ribera, Manfredi, Val-
continuous and discontinuous, prior to the act of viewing entin, Régnier, von Honthorst, Caracciolo and others),
and thematizing that act with unprecedented violence – who establish Caravaggism as a going concern, in effect
lies at the core of much of Caravaggio’s art, including (or guaranteeing its importance for subsequent European
especially) his most radically inventive creations’ (MC, p. painting. (In an obvious sense, the present book, The
40). In other words, in Caravaggio’s art a certain sever- Moment of Caravaggio and After Caravaggio, in that order,
ing and distancing is as it were built into the ‘internal form a trilogy.) One might note, too, a three-part relation
structure of the pictorial act’ (the title of chap. 6 of MC). In precisely with regard to the depiction of hands among
the interest of economy, I have held off introducing these Savoldo, Caravaggio and a major painter not yet men-
considerations in the present book. But they are intrinsic tioned in this book, Giovanni Francesco Barbieri, known
to Caravaggio’s enterprise as I understand it. as Guercino, in whose masterly canvases of 1619–20 the
43 Fried, After Caravaggio, pp. 95–8. conspicuous presence of crossed hands and limbs serves
44 A huge topic. See Michael Fried, Courbet’s Realism at once to solicit the viewer and to keep him or her at
(Chicago, IL, and London, 1990). a certain (near) distance. On this feature of Guercino’s
45 Not surprisingly, then, nothing like the distinction art, see Fried, After Caravaggio, chap. 4, ‘Guercino’s Anni
between immersive and specular ‘moments’, or rather, Mirabiles, 1619–1620’, pp. 135–73.
nothing even remotely like the specular ‘moment’ of dis- As for the three- or four-part relations among the
tancing, freezing and severing as evoked in the passage painters in question, see my discussion of what I call
from MC cited in n.42 is in play in Savoldo’s art. ‘hinge-like structures’ with regard to the relations among
46 In this connection, too, it should be noted that Courbet, Manet and Impressionism, and among Chardin,
although Savoldo seems to have been aware of the right- Greuze and David, in Manet’s Modernism; or, The Face of
angle mirror-reversed self-portrait dispositif, as in the Painting in the 1860s (Chicago, IL, and London, 1996), pp.
Self-portrait as St Jerome, what I have called his entan- 411–13. This is not the place for a fuller discussion of
glement with the painting precluded the later ‘classic’ the recursive nature of such structures, but see the very
rendering of that dispositif that I have associated with interesting remarks by Stephen Mulhall in the last pages
Annibale and Caravaggio (and their successors). Nor, of his essay ‘Deep Relationality and the Hinge-like Struc-
for that matter, in Savoldo’s work do we nd the various ture of History: Michael Fried’s Photographs’, in Michael
emphases on the lower right corner of the canvas that in Fried and Philosophy: Modernism, Intention, and Theatrical-
MC I associate with Caravaggio, a feature of the latter’s ity, ed. Mathew Abbott (New York and Abingdon, Oxon,
paintings that I understand as acknowledging the exter- 2018), pp. 87–103.
AC K N OW L E D G E M E N T S

A num ber of f ri e n d s a n d s c h o la r s have contributed in diverse ways

to this book. My thanks, then, to Jennifer Ashton, Stephen Bann, Stephen
Campbell, Keith Christiansen, Thomas Demand, Francesco Frangi, Anthony
Grafton, the late Ricky Jay, John O’Malley, Walter Benn Michaels, Peter
Parshall, Robert Pippin, Charley Ray, Walter Stephens and Molly Warnock,
for advice and conversation, the latter often in front of fascinating paint-
ings. Special thanks are owed to two Italian friends, Giovanni Careri and
Angela Mengoni, for inviting me to give a series of seminars at the Univer-
sità Iuav di Venezia during the spring of 2018; among other benets, my six-
week stay in that fabulous city enabled me to visit the stunning exhibition
‘Tiziano e la pittura del cinquecento tra Venezia e Brescia’ in the Museo
Santa Giulia in Brescia, which included no fewer than eight important
Savoldos. Happily, Giovanni Careri and Angela Mengoni went through the
exhibition with me, and the exchanges we had in the course of a memorable
day proved extremely fruitful, as I was sure they would.
I am grateful, too, to Michael Leaman of Reaktion Books for welcoming
my project; to Sally Nichols for her invaluable help gathering the illustra-
tions for this volume; and to the nonpareil Gillian Malpass for designing
the book for maximum effectiveness. I might also mention that my involve-
ment with Northern Italian art goes back to my rst semester of graduate
school under a great art historian and teacher of genius, Sydney J. Freedberg.
Finally, two other longtime companions in looking deserve particular men-
tion: my wife, Ruth Leys, and the painter Joseph Marioni, both of whose
insights I have freely turned to my own purposes.
This book is dedicated to Marioni and to another painter of high achieve-
ment, Frank Stella.
P H OTO AC K N OW L E D G E M E N T S

The author and publishers wish to express their thanks to the below sources of illus-
trative material and/or permission to reproduce it. Some locations of artworks are
also given below, in the interest of brevity:

Accademia Carrara, Bergamo/Mario Bonotto – photo Scala, Florence: 62; The

Art Institute of Chicago: 4, 8, 14, 73; Banca Intesa Sanpaolo, La Gallerie d’Italia,
Palazzo Zevallos Stigliano, Naples: 26; Basilica di San Giovanni e Paolo, Venice:
6; Biblioteka Uniwersytecka, Warsaw: 10; The British Museum, London: 56;
Chiesa di San Luigi dei Francesi, Rome: 11, 23, 59; Chiesa di San Nicolò, Treviso/
Mario Bonotto – photo Scala, Florence: 76; Galleria Doria Pamphilj, Rome:
13; Galleria Nazionale dell’Arte Antica, Palazzo Barberini, Rome: 27; Galleria
Sabauda, Turin/photos Sergio Anelli/Electa/Mondadori Portfolio via Getty
Images: 61, 72; Gallerie dell’Accademia, Venice/photo akg-images/Cameraphoto:
82; Gallerie dell’Accademia, Venice/photos Scala, Florence: 44, 74, 81; Gallerie
degli Uffizi, Florence: 83; Gallerie degli Uffizi, Florence/photos Scala, Florence:
79, 80; Gemäldegalerie, Staatliche Museen zu Berlin/photos Scala, Florence/bpk,
Bildagentur für Kunst, Kultur und Geschichte, Berlin: 40, 55; Groeninge-museum,
Bruges: 51; Harvard Art Museums/Fogg Museum, Gift of Edward W. Forbes
in memory of Alice F. Cary, photo © President and Fellows of Harvard College:
15; The J. Paul Getty Museum, Los Angeles: 36, 41; formerly Kaiser-Friedrich-
Museum, Berlin/photo Scala, Florence/bpk, Bildagentur für Kunst, Kultur und
Geschichte, Berlin: 32; Kunsthistorisches Museum, Vienna: 28, 46; The Metro-
politan Museum of Art, New York: 5, 22, 42, 50, 53, 54, 78; Ministero per i beni
e le attività culturali (MiBAC) – Galleria Borghese, Rome: 2, 21, 25, 31, 57, 58;
Ministero per i beni e le attività culturali (MiBAC) – Galleria Sabauda, Turin/
photos Scala, Florence: 63, 85; Musée du Louvre, Paris: 12; Musei Capitolini, Rome/
 photo acknowledgements 189

photos Scala, Florence: 65, 66; Museo d’Arte Sacra San Martino, Alzano Lombardo:
9; Museo e Real Bosco di Capodimonte, Naples: 7; Museum of Fine Arts, Boston/
photo Scala, Florence: 16; The National Galley, London/photos Scala, Florence: 17,
18, 33, 39, 52; National Gallery of Art, Washington, DC: 38, 43, 45, 64, 75; National
Gallery of Ireland, on indefinite loan from the Jesuit Community, Leeson St,
Dublin, who acknowledge the kind generosity of the late Dr Marie Lea-Wilson,
1992/photo © National Gallery of Ireland, Dublin: 24; Pinacoteca di Brera, Milan:
19, 34, 35; Pinacoteca Tosio Martinengo, Brescia: 3, 37, 70; Pinacoteca Tosio
Martinengo, Brescia/photos Scala, Florence: 67, 68, 86; private collection, Milan/
photo Mondadori Portfolio/Electa/Antonio Guerra/Bridgeman Images: 29; private
collections: 30 (Bergamo), 84 (Milan); The Pushkin State Museum of Fine Art,
Moscow/photo Bridgeman Images: 48; Putnam Foundation, Timken Museum of
Art, San Diego, CA: 47, 49; photos © RMN-Grand Palais (musée du Louvre, Paris)/
Franck Raux: 1, 20; Royal Collection Trust/© Her Majesty Queen Elizabeth II 2020:
60, 77; San Giobbe, Venice/Mario Bonotto – photo Scala, Florence: 71; Santa Maria
delle Grazie, Brescia, photo courtesy Diocesi di Brescia/Ufficio per i beni culturali
ecclesiastici: 69.
INDEX

Illustration numbers are indicated in Madonna and Child with SS Catherine

italics of Alexandria and Mary Magdalene
161–2, 82
absorption 36–7, 165–6 St Jerome in the Desert 162–3, 83
Age of Caravaggio, The (exhibition) 12, Belting, Hans 62, 162–3
67 Berenson, Bernard 42, 113–14, 157
Agrippa, Cornelius 151, 152, 153 Bergamo 12, 40, 158
Alzano Lombardo 27, 35 bodily orientation 50
Ancona 40 Bologna 12
angels 35, 97–8, 107, 136 Bosch, Hieronymus 9, 12, 87, 88, 92, 101,
at the tomb of Christ 72 126, 142, 144, 145, 154, 155
of the Annunciation 119, 125–6, 152, Last Judgement 91, 51
153 Temptation of St Anthony 153
Raphael summoning the fish 17–18, Boschetto, Antonio 11
102, 108, 113, 152 Brescia 10, 11, 12, 115, 155, 156
Anguissola, Sofonisba 140 Museo di Santa Giulia 12
anthropomorphic imagery 144 San Barnaba 122
Aretino, Pietro 10 Santa Maria delle Grazie 126
Art Bulletin 91 school of 157
auto-mimesis 64, 83 Val Camonica 154
autonomy 93, 160, 165–6, 167 Bruno, Giordano 151

Ballarin, Alessandro 159 Campagnola, Giulio, Astrologer, The 101,

Baroque 32 147, 153, 56
Bayer, Andrea 12, 67–71, 96, 107 Campbell, Stephen 126
Bellini, Giovanni 6, 78, 144, 157, 161–3 Endless Periphery, The: Towards a
 index 191

Geopolitics of Art in Lorenzo Lotto’s Catholic Church 37, 40, 116, 142, 153,
Italy 115–16 158, 171
Caravaggio 48–52, 164–9 Mass 151, 154
and Lombard tradition 12 Cavell, Stanley 36–7, 38
as predecessor of Savoldo 8, 11–12, 48, Cecco del Caravaggio 142
141–2, 157 Chicago Art Institute 22, 38
decapitation in art of 166 Christiansen, Keith 11–12, 34, 67
depiction of hands 167 Cima da Conegliano 62, 156
handling of drapery 94, 145–7 Clark, Stuart, Thinking with Demons
Longhi on 96 153–4
Savoldo’s affinity with 32, 34–5, Corinth, Lovis 140
164–9 Correggio 6, 12, 26
self-portrayal 48–9, 52, 59, 140 Couliano, Ioan P. 151
Bacchino Malato 49, 21 Courbet, Gustave 51, 52, 140, 167
Bacchus 50 Cranach, Lucas, the Elder 12, 87, 91, 142
Boy Bitten by a Lizard 44, 49, 139, 166, Temptation of St Anthony 91, 50
167, 18 Cremona 12
Calling of St Matthew 108–9, 136, 59
David with the Head of Goliath 49, 166, decapitation 26, 52, 166, 167
25 Degas, Edgar 56
Death of the Virgin 36, 166, 12 demons, demonic imagery 9, 92, 116,
Judith and Holofernes 51–2, 27 151–6, 158, 159, 166, 171
Lute Player 67 Derrida, Jacques 94
Martyrdom of St Matthew 32–5, 108, Descartes, René 39
166, 11, 23 ‘devotional’ paintings 10, 160, 161–7
Martyrdom of St Ursula 34, 49, 51–2, distancing 8, 32, 57, 125, 127–8, 163,
26 164, 167, 169
Musicians, The 49, 50, 22, Dominican order 22, 154, 155
Penitent Magdalene 36, 37, 166, 13 ‘dragon’ imagery, and St Margaret 152
St Jerome Writing 166 Dürer, Albrecht 12, 69, 87, 142, 144
Supper at Emmaus 113 faces appearing in work 143–4
Taking of Christ 49, 51, 24 Self-portrait at Age Twenty-two 142
see also Fried, Michael, After Six Pillows 142–4, 78
Caravaggio; Moment of Caravaggio,
The Ebert-Schifferer, Sybille 11
Cardano, Girolamo 151 Elsner, Jaś 16–17
Carracci, Annibale, Self-portrait with empathic projection 37, 38, 57, 84, 95,
Other Figures 46, 59, 139, 19 134–5, 150, 165–6
Carracci family 12 entanglement 128, 167–9
192 index

Eucharist 115–16, 125, 151, 152, 154, and other art historians 11
171 on Penitent St Jerome 63
Eucharistic naturalism 115, 119 on Portrait of a Young Man (St John the
Evangelist) 59–62
facing into the picture 51, 165 on Savoldo and Lotto 158
Ficino, Marsilio 151–2, 153 on Savoldo’s early paintings 78
De via coelitus comparanda 151 on Savoldo’s individual figures 14–16,
Flemish painters 12, 13, 62 132
Florence 10, 16, 156 on Savoldo’s self-portraits 8, 35, 47,
Foppa, Vincenzo 12, 157 52, 54–5, 137–40
Francesco II Sforza, Duke of Milan 10, on Savoldo’s signatures 159
96, 149, 156 on Savoldo’s training 156
Frangi, Francesco 11, 12, 139 on smallness of Savoldo’s oeuvre 157
Frankfurt 11 on Tobias and the Angel 104, 107
Freedberg, Sydney J. 36, 41, 42, 64–6, 85, Giorgione 6, 12, 16, 47–8, 78, 104, 107,
116, 155 148, 157
Painting in Italy, 1500–1600 13–17, Giotto 144
22–6, 29 Giovanni Gerolamo Savoldo tra Foppa,
Fried, Michael Giorgione e Caravaggio (exhibition) 11,
After Caravaggio 164–5 15
Moment of Caravaggio, The 8, 9, 36, 44, Golden Legend 84, 85, 98
48, 49, 59, 109, 136, 139, 164–5, Gregori, Mina 11, 27, 29–31, 34, 166
166, 170 Grimani, Cardinal Domenico 92

Galileo Galilei 39 hands 157, 163

gallery pictures 53, 165, 166–7, 169 grasping/gripping 27, 33, 39, 40, 48,
Gaston de Foix, duc de Nemours 47, 57, 58, 64, 66, 94–5, 97, 110, 115,
139 134, 150, 167
Gentileschi, Artemisia 140 left and right 150, 170
Germany 142 Hegel, G.W.F. 174n10
Gilbert, Creighton Herri met de Bles 144
admiration for Savoldo 159 hinge-like structures 186
on Adoration of the Shepherds (Turin)
116 imaginative identification 9, 37, 38, 57,
on Death of St Peter Martyr 31, 32 71, 77
on donor figures 112 immersion vs specularity 185–6n18,
on Giorgione 157 186n45
on hands in Savoldo’s works 35, 56–7, Impressionism 56
61–2, 67, 86–7, 121 influence 152, 156, 183n20
and Longhi 13, 56–7 internal mimesis 83, 86
 index 193

Jesi 40 hands in work of 42–3, 113–14

Jesuits 142 and paintings of St Jerome 62
Jews 178n15 portraits 38, 84, 158
‘Jheronimus Bosch e Venezia’ (exhibition) Savoldo’s closeness to 12
91–2 Crucifixion 158
Juan Diego, vision of 182n19 A Dominican Friar as St Peter Martyr
22, 40–43, 15
Koerner, Joseph, Moment of Self- Holy Family with St Catherine of
portraiture in German Renaissance Art, Alexandria 113–14, 62
The 143–4 Madonna and Child with St Peter Martyr
22, 7
Laocoön and His Sons (sculpture) 25 St Jerome (two versions) 163
Leonardo da Vinci 6, 12, 16, 82, 96, 107, Virgin and Child with SS Jerome and
148, 157 Nicholas of Tolentino 41–2, 16
Last Supper 114 Luther, Martin 115–16
Lombardy
Lombardy/Venice divide 13 magic, magical imagery 151–4
tradition 12–13, 16, 22, 157 in Adorations 113, 116–19, 126, 152
Longhi, Roberto and Mass 151, 154
on Gilbert 13, 56–7 miracle-working 126
and Lombard connection 13, 32–3 and mirrors 150
on Lotto 158 in Portrait of a Woman as St Margaret
on Savoldo’s treatment of hands 8, 35, 121, 152–3
52, 55–7, 58, 141, 157 quasi-magical operation 18, 104, 115,
on St Matthew and the Angel 96 116
on Savoldo 159 and science 142
on Savoldo as predecessor of in Tobias and the Angel 104, 106–7, 113,
Caravaggio 109, 145–7, 157, 169 152
on Savoldo’s use of drapery 85, 94, see also demons, demonic imagery
145–7 magnetic gesture 18, 104, 108–9, 113,
‘Cose Bresciane del Cinquecento’ 52 116–17, 136, 152
‘Quesiti caravaggeschi, II: I precedenti’ Magus 152, 163
12 Manfredi, Bartolomeo 51, 142, 165, 167
Loreto 40 Mantegna, Andrea 6, 144
Lotto, Lorenzo Mantua 12
altarpieces 157–8 martyrdom 28–9, 31–2, 34, 35, 39, 42–3,
contrasted with Savoldo 10–11, 15, 85, 109, 137
84 Michelangelo Buonarroti 6, 10, 25, 109
faces and masks in work of 144 Milan 10, 12, 34, 95, 157
frescoes 158 Zecca 107, 156, 159
194 index

mimesis 84, 87, 106, 136, 165 Pesaro 10

auto-mimesis 64, 83, 107, 145 San Domenico 64
internal 83, 86 Peter Martyr, St 22
mimetic dynamic 165–6 Peterzano, Simone 109
mimetism 38 Pico della Mirandola, Gianfrancesco 151
mirror-reflection 44–50, 56, 58, 59, 64, pictorial anagrams 57
109, 127, 135, 139, 148–50, 160, 167, 170 pictorial density 165
right-angle 44–6, 49, 50, 58–9 Piero di Cosimo 144
mirror-reversal 38, 40, 44–6, 48, 71, 139, Pietro d’Argenta 10
150, 166 Pino, Paolo 85, 159, 160, 167
models 30, 31, 33–4, 84, 149 Dialogo di pittura 10, 11, 47–8, 148–51,
Montagna, Bartolomeo 62 156
Moretto da Brescia (Alessandro Pittori della realtà in Lombardia, I
Buonvicino) 12, 115, 157, 159 (exhibition) 12
Nativity with the Shepherds, St Jerome Pordenone 25–6
and a Hieronymite Donor 127, 70 Protestantism 142

nature, imitating itself 148–9, 150 Raphael 6

nearness 137, 141 Fire in the Borgo 91
New York, Metropolitan Museum of School of Athens, The 139
Art 12 Realist paintings 167
Northern art 87–92, 142, 145, 158, 159 reciprocity 174n10
Reformation 39
Painters of Reality: The Legacy of Leonardo Régnier, Nicolas 51, 142, 165, 167
and Caravaggio in Lombardy (exhibi- relational field 82
tion) 12, 96 religious crisis 9, 115, 142
palm branch of martyrdom 2, 31–5, 39, Rembrandt van Rijn 140
42, 85, 137 Ribera, Jusepe de 142
Palma Vecchio 25, 112 right-angle dispositif 139
Martyrdom of St Peter of Verona 27–9, Romanino (Girolamo Romani) 62, 115
31, 35, 9 Rome 16, 25
Panazza, Gaetano 107, 159 Palazzo Barberini 100
paragone 48 S. Luigi dei Francesi 32
Pardo, Mary 77, 103, 148–9, 159–60 Roustang, François, Influence 183n20
‘Subject of Savoldo’s Magdalene, The’
72–6 sacred naturalism 115
Parma 10, 12, 156 Savoldo, Giovanni Gerolamo
Patinir, Joachim 12, 87, 142 affinity with Caravaggio 164–9
Pensaben, Marco 137 altarpieces 164
 index 195

art historians and 11–17 Adoration of the Shepherds (Venice) 10,

autonomy 160 128–9, 71
drapery 71–2, 85–6, 93–101, 107, 109, Adoration of the Shepherds
118, 135–6, 142, 145–7, 150, 155–6, (Washington) 118–19, 125, 152–3,
170 163, 64
exhibitions 11–12 Crucifixion 11, 13
grotesque images (faces and masks) Death of St Peter Martyr 11, 14, 22–43,
9, 85–6, 89–91, 95–6, 99–102, 107, 48–9, 57, 58, 64, 71, 85, 98, 127, 137,
125, 129, 135–6, 142–8, 144–8, 139, 141, 145, 150, 154, 158, 159,
150–56, 159, 160, 169, 170 163, 167, 4, 8, 14, 73
hands, depiction of 8–9, 35–40, 42, Drawing of a Man’s Head 30, 10
52–77, 94–5, 98, 110–16, 118, 121, Elijah Fed by Ravens 13, 14, 78–85, 87,
126, 127–8, 134, 135, 137, 141, 150, 91, 95–6, 112, 132–4, 145, 154–5,
157, 163, 167 170, 171, 43, 45
‘independent’ paintings 10, 11, 161, Evangelist 53–4, 58, 134, 29
164–5 Head of a Man, The 139, 77
interest in Northern art 87–92, 126, Lamentation over the Dead Christ 59–62,
142, 145 86–7, 32, 46
life and work 9–11 Madonna and Child Enthroned among
Lombard roots 13, 33 SS Nicholas, Dominic, Thomas, Jerome
magical effects 106–7, 113, 115, and Liberale (Treviso altarpiece) 87,
116–17, 119, 121 137–9, 167, 76
self-portraits 8, 40, 63, 137–42, 157 Madonna and Christ Child and Two
signature 52, 137, 159 Angels in Glory with SS Peter,
singularity 6, 128, 158–9 Dominic, Paul and Jerome 64–7, 34,
smallness of oeuvre 157 35
and work of Caravaggio 8 Magdalenes (as group) 8, 9, 33, 72–7,
Savoldo, Giovanni Gerolamo, wo r k s 100–103, 137, 141, 145, 147, 150,
Adoration of the Christ Child with 153, 156, 158, 159–60, 39–41
SS Jerome and Francis 111–16, 118, Penitent St Jerome 62–4, 83, 93–4, 96,
163, 61, 72 138, 141, 145, 155, 158, 163, 17, 33,
Adoration of the Shepherds (Brescia) 9, 52
122–9, 132, 145, 152–3, 156, 169, Portrait of a Gentleman as St George
171, 67, 68 71–2, 135, 38, 75
Adoration of the Shepherds (Terlizzi) 10, Portrait of a Man in Armour 9, 35, 46–7,
128–9 49, 71–2, 134, 139, 148, 150, 158,
Adoration of the Shepherds (Turin) 10, 166, 170, 1, 20
107, 116–18, 122, 152–3, 158, 63, Portrait of a Woman as St Margaret
85 119–21, 137, 152–3, 65, 66
196 index

(Savoldo, wo r k s continued) separation of onlookers 126–7, 166–7,

Portrait of a Young Man with Flute 67, 169
157, 158, 37 severed presence 52, 165, 166
Portrait of a Young Man as St George shawls
120 Elijah’s 82, 83
Portrait of a Young Man (St John the Magdalene’s 8, 9, 76–7, 100–101, 137,
Baptist?) 157 145, 147, 150, 160
Portrait of a Young Man (St John the Shearman, John 112
Evangelist) 59–62, 134, 31 style, stylistic considerations 8, 109, 112
Prophet (or Apostle) 52–3, 57, 134, art historians and 13, 16–17, 27, 31,
137–8, 150, 158, 28 34–5, 52, 104, 141
Rest on the Flight to Egypt 163–4, 84 classical 14, 16
St Matthew and the Angel 8, 33, 95–100, surcharged colour 14, 155
104, 107, 136, 145, 153, 156, 157, surface, ‘this’ side of and beyond 58, 135,
158, 159, 163, 42, 53, 54 141
SS Paul and Anthony 13, 14, 78–85, Switzerland 142
87, 95, 112, 132–4, 154–5, 171, 44,
74, 81 Terlizzi, Santa Maria la Nova 10, 128–9
Self-portrait in the Costume of St Jerome Tintoretto 6, 157
54–5, 58–9, 64, 134, 139, 150, 30 Titian 6, 12, 14, 40, 78, 157, 158
Shepherd with a Flute 67–71, 94, 145–7, Martyrdom of St Peter 25–6, 29, 31,
158, 36 38
Temptation of St Anthony (Moscow) 9, Martyrdom of St Peter, copy by Carlo
12–13, 87–91, 142, 145, 154, 155, Loth 22, 6
171, 48 Martyrdom of St Peter, engraving by
Temptation of St Anthony (San Diego) Martino Rota 22, 5
9, 12–13, 87–91, 95, 142, 145, 154, Noli me tangere 76–7
155, 171, 47, 49 Pesaro Madonna 112
Tobias and the Angel 6, 17–18, 34, Titian e la pittura del cinquecento tra
104–7, 109, 126, 135–6, 152, 156–7, Venezia e Brescia (exhibition) 12
158, 163, 2, 57, 58 Tobit, Book of 104
Transfiguration 147, 79, 80 transactional field 141
Virgin Adoring the Child with Two Trent, Council of 39, 142
Donors 109–16, 118, 122, 152–3, Treviso 10, 40, 87
163, 60 Trithemius of Würzburg 151
Schongauer, Martin 12, 87, 91, 142
Second World War 61 unknown artist, Adoration of the Child
self-portraits, self-portrayal 8, 40, 44–52, 126–7, 69
63, 137–42, 148, 157, 166
 index 197

Valentin de Boulogne 51, 142, 165, 167 San Giobbe 10, 128
van der Goes, Hugo 87 SS Giovanni e Paolo 22, 63
Van Gogh, Vincent 140 Venturi, Adolfo and Lionello 11, 157
Vandenbroek, Paul 154 Verona 10, 22
Vasari, Giorgio 95, 107, 156, 159, 160 Veronese, Paolo 6, 157
Lives of the Painters 10, 77 Vivarini, Alvise 156
Veca, Alberto 54
Venice 6, 9, 10, 13, 14, 22, 39, 40, 155, Walker, D. P. 151
156, 161 Washington, DC, National Gallery of
Fondamente Nuove 63 Art 170
Palazzo Ducale 91 witches 119, 126, 152, 153, 154
Riva degli Schiavoni 163–4
San Barnaba, Bargnani chapel 10, 128 Zenale, Bernardo 144