14

The Analysis of Rights

Leif Wenar

‘Maine’s reference to Bentham not as discovering or revealing the meaning

of the expression “a right”, but as giving a clear meaning to it is accurate; and
raises a methodological issue of some importance.’ (Hart 1982, 162–3)
‘Faithfulness to the shape of common concepts is itself an act of normative
significance.’ (Raz 1986, 64)

1 Introduction

In some respects investigations into the nature of rights resemble investigations

in the physical sciences. An investigation into the nature of a particular right,
such as the right to remain silent or the right to pass a criminal sentence, will seek
to describe the right in terms of its logical structure and its normative functions.
This is analogous to an investigation of a particular chemical compound, which
will attempt to describe the compound in terms of its physical structure and its
standard causal properties.
Within the philosophy of science it is controversial whether scientists’ gen-
eralizations should be interpreted as causal or nomic.¹ However all sides of this
dispute agree that one scientific theory will be more powerful than another if it
accounts for more phenomena, and if it accounts for the same phenomena using
fewer basic concepts and relations. Comprehensiveness and simplicity are two pri-
mary dimensions along which scientific explanations should be measured. There
is also broad consensus on which scientific theories have more explanatory power,
at least for inter-paradigmatic comparisons. To take an obvious example, there is
near-universal agreement that the explanatory framework of modern chemistry
is superior to the medieval earth-air-fire-water-aether framework that it replaced.
Although the modern periodic table of elements is not as simple as the medieval
diagram of elements, the comprehensiveness of the modern theory makes it more
powerful overall.

¹ Compare Hempel 1965, Salmon 1984, Kitcher 1989.

14-MHKramer-Chap14.indd 251 4/30/2008 5:16:28 PM

 252 Leif Wenar
A theory of the nature of rights will also aim for greater explanatory power,
where two primary measures of explanatory power are again comprehensiveness
and simplicity. All else equal, a theory of rights will be more powerful when it
accounts for more rights, and when it uses fewer basic concepts and relations. As
with scientific theories no one believes that there is an exact schedule for trading
comprehensiveness against simplicity. But there is no reason to think that there
will be more dissensus when it comes to cases than there is in the comparison of
scientific theories.
The ‘phenomena’ that a theory of rights ultimately aims to explain is what
rights there are and what rights there could be. However, what lies within the exten-
sion of that concept is more controversial than what lies within the extension of
the analogous concepts in many physical sciences (eg, ‘what chemical compounds
there are and could be’). This is because what moral rights there are and could be
turns on which moral theory is correct, and what legal rights there are and could
be turns on what is the correct theory of law. Which moral theory is correct, and
which jurisprudential theory is correct, are matters of some dispute.
The theorist of the nature of rights therefore cannot simply set a list of the
phenomena to be explained—the rights that there are and could be—without
making hotly contentious assumptions outside of his domain of inquiry. Indeed
there is doubled trouble here, as it will be controversial not only which moral or
jurisprudential theory is correct, but also which rights are entailed by any such
theory within a given set of circumstances.
In response to these difficulties rights theorists have adopted an indirect
approach to their subject matter. They have tested the explanatory power of their
theories not against what rights there are and could be, but against what rights
people say that there are and could be. Theorists of rights have, in the main, taken
an ordinary understanding of rights to set the phenomena to be explained. For
example, a rights theorist will reject any theory that ascribes rights to tomato
plants, or to ant colonies, because such a theory is incompatible with an ordinary
understanding of rights. The same reason will be given for rejecting any theory
that denies that it is coherent to ascribe rights to women. This indirect approach
is attractive because what informed, thoughtful people believe about rights is
much less contentious than what rights there actually are. Thoughtful people
who are not theorists do have some familiar differences concerning what rights
there are and can be—but moral and legal theorists share all of those differences
and have many more disagreements as well.
The ‘data’ of ordinary understanding are therefore significantly less conten-
tious than the ‘data’ of what rights there really are, and focusing on ordinary
understanding allows debates over the nature of rights to refer to a common set
of facts to be explained. By contrast if some theorist alleged that his theory of
the nature of rights was superior because it fit with his preferred substantive the-
ory of rights, he would immediately be challenged to show that his preferred
substantive theory was correct. A debate on the terrain of ‘conformity with some

14-MHKramer-Chap14.indd 252 4/30/2008 5:16:29 PM

 The Analysis of Rights 253

preferred moral or jurisprudential theory’ would quickly become merely a proxy

for a debate over which moral or jurisprudential theory is best.
For the most part, therefore, theorists have attempted to provide a conceptual
analysis of the concept of a right as this concept is ordinarily understood. In the
language of the Hart quotation that begins this article, theorists have been more
concerned with ‘discovering or revealing the meaning of the expression “a right” ’;
and less concerned with ‘giving’ the concept a meaning. Nevertheless, as we will
see in the final section of this article, rights theorists have also surreptitiously
allowed a desire to bolster substantive theories to pull them toward revisionary
definitions of the concept. Indeed rights theorists have surreptitiously yielded to
this desire in a way that has contributed to making the debate over the nature of
rights permanently intractable up to now.

2 The Will Theory and the Interest Theory

Of the two features of rights that a theory of the nature of rights is meant to
explain—logical structure and normative function—there is much more con-
sensus on the former than the latter. The Hohfeldian framework is by far the
most widely accepted analysis of the logical structure of rights, and it is used by
the majority of contemporary rights theorists. Regarding the functions of rights
however there is a longstanding disagreement. Proponents of the will theory and
the interest theory have struggled for decades if not centuries over which theory
provides the more powerful explanation of what rights do for right-holders.² It is
this debate over the functions of rights that is our main study here.
The question of the function of rights concerns what rights do for those who
hold them. The will theory of rights asserts that the function of all rights is to
give the right-holder choices. According to Hart’s will theory, for instance, the
function of a land owner’s legal right is to give him the legally recognized power
to waive or not to waive the duties that others have not to enter his land. As Hart
describes the core idea of the will theory in the context of legal rights, ‘One who
has a right has a choice respected by the law’ (Hart 1982, 171, 183–5, 188–9).³
The will theorist’s view of the function of rights limits what he recognizes as
a right: where there is no normatively respected choice, there can be no right.
The will theorist’s view also restricts the class of potential right-holders. Only
those beings that have certain capacities—the capacities to exercise choice in
controlling their own actions and the duties of others—are potential will theory
right-holders.

² For some of this history, see Tuck 1997, Brett 1997. See also Simmonds’s reflections on the
history of the jurisprudence of rights (Kramer, Simmonds, and Steiner, 1998, 113–232).
³ Besides Hart, influential advocates of a choice-based approach to rights include Savigny,
Kelsen, Wellman, and Steiner.

14-MHKramer-Chap14.indd 253 4/30/2008 5:16:29 PM

 254 Leif Wenar
The interest theory, by contrast, maintains that the function of all rights is
to further their holders’ interests. The most prominent interest theory analysis
is Raz’s: ‘ “X has a right” if X can have rights, and, other things being equal, an
aspect of X’s well-being (his interest) is a sufficient reason for holding some other
person(s) to be under a duty.’ (Raz 1986, 166) Here rights do not give choices;
rather rights are claims on the actions of others that are justified by the interests
of the right-holder.
The advocates of the each theory are deeply entrenched in their positions. At
times the debate appears to be one—like the debate over Newcomb’s problem—
where each side can scarcely imagine that the other side has a reasonable view.⁴
Thus Hart in laying out the will theory claims that ‘It is hard to think of rights
except as capable of exercise’ (Hart 1982, 184) while Williams in opposition insists
that ‘No one ever has a right to do something; he only has a right that someone
else shall do (or refrain from doing) something. In other words, every right in the
strict sense relates to the conduct of another.’ (Williams 1956, 1145)
These positions are entrenched despite the widely acknowledged fact that each
theory is too narrow when judged against an ordinary understanding of rights.
Each theory accounts for an insufficiently comprehensive range of rights, leaving
large areas of commonly accepted rights unexplained. The ways in which each
theory is too narrow are by now well understood. Indeed the debate between
the two theories has generated a standard account of the shortcomings of each
approach. Here I will just summarize this standard account of these shortcom-
ings so as to set up a discussion of the strategies that the two groups of theorists
have taken in response to the problem of narrowness.⁵
The narrowness of the will theory is apparent, first, in the types of actions that
the theory can recognize as rights violations. Many important legal rights do cor-
respond to Hart’s legally-respected choices. But many do not. For example, you
have no legal discretion to alter your entitlement against being enslaved, or your
entitlement against being tortured to death. The will theory therefore does not
recognize that you have a legal right against being enslaved, or against being tor-
tured to death. Yet most would regard these unwaivable claims as rights, indeed as
among the more important rights that individuals have.⁶ Indeed the will theory
does not recognize that the criminal law confers any rights on citizens, since the
power to enforce the law rests not with citizens but with state officials (Kramer,
Simmonds, and Steiner 1998, 230, Wellman 1985, 85). Yet most citizens would

⁴ ‘I have put this problem to a large number of people . . . To almost everyone it is perfectly clear
and obvious what should be done. The difficulty is that these people seem to divide almost evenly
on the problem, with large numbers thinking that the opposing half is just being silly.’ (Nozick
1969, 117)
⁵ This summary draws from Wenar 2005.
⁶ Despite not being able to recognize a person’s claim against being tortured as a right, the will
theory does recognize many negligible claims (such as your waivable claim not to be patted on the
head) as rights (MacCormick 1977, 197).

14-MHKramer-Chap14.indd 254 4/30/2008 5:16:29 PM

 The Analysis of Rights 255

find it surprising to hear that the criminal law did not ascribe to them a right
against being murdered or raped.
The limitations of the will theory are also evident in its inability to account for
the rights of certain kinds of right-holders: for example, for the rights of incom-
petent (eg, comatose) adults, and of children (MacCormick 1982, 154–66). The
will theory can acknowledge rights only in those beings competent to exercise
choices—which incompetent adults and children are not. Incompetent adults
and children therefore are not possibly right-holders on this view. This result
diverges significantly from an ordinary understanding of rights. Few thought-
ful laymen would insist that it is a conceptual impossibility, for example, for the
comatose to have rights against bodily mutilation.
Since the interest theory turns on interests instead of choices, it can recog-
nize unwaivable rights against enslavement and torture. The interest theory can
also accept children and incompetent adults as right-holders, since children and
incompetent adults have interests that rights can protect.
Yet the interest theory is also inadequate to an ordinary understanding of
rights. Staying with Raz’s version of the interest theory, there are many rights
for which the interests of the putative right-holder are not sufficient to hold other
person(s) to be under a duty. For example, Raz himself allows that the interest of
a journalist in protecting his sources is not itself sufficient reason to hold others
to be under the corresponding duty (Raz 1986, 179, 247–8). It is rather the inter-
ests of the general public in an active and independent media that grounds the
journalist’s right to protect his sources. Yet as Kamm observes, ‘If the satisfaction
of the interests of others is the reason why the journalist gets a right to have his
interest protected, his interest is not sufficient to give rise to the duty of non-
interference with his speech’ (Kamm 2002, 485). Nor does this difficulty only
affect the rights of office-holders like journalists, as Raz admits that weighty rights
such as the rights of free expression and freedom of contract are not justified solely
by the interests of the individual citizens who hold them (Raz 1996a, 30–43, 131).

3 Three Strategies for Fitting Theory and Data

Both the will and the interest theories are, in their standard forms, too narrow.
Each theory accounts for some but not all of the rights that any ordinary under-
standing of rights will accept.
Will and interest theorists have wrestled with this mismatch between the
scope of their theories and the range of the phenomena for a long time. They
have adopted three strategies in response to the problem of narrowness. The first
strategy is to claim that their theories are only meant to describe a more limited
range of the phenomena than was originally supposed. The second strategy is
to attempt to expand the scope of their theories in order to explain more of the
phenomena. The third strategy is to replace the set of phenomena to be explained

14-MHKramer-Chap14.indd 255 4/30/2008 5:16:29 PM

 256 Leif Wenar
with a different set. The three strategies could be summarized as narrowing the
data, broadening the theory, and replacing the data-set.
We next examine some examples of each of the three types of strategies taken
by the will and the interest theories in response to the problem of narrowness.
While the three strategies are perfectly reasonable responses to the problem, the
results of pursuing each of the strategies have been unsatisfactory. The reason for
this repeated failure, as it turns out, is not contingent. As will be shown in the dis-
cussion afterwards, the two theories share a premise which prevents them from
yielding an adequate understanding of rights, regardless of how the theories are
recast.

3.1 Narrowing the Range of Phenomena to be Explained

The first response of rights theorists to the problem of narrowness has been to
cut the domain of rights to be explained to fit the ambit of the explanatory the-
ory. Thus Hart, when faced with counterexamples to the will theory involving
constitutional immunities, confesses that his theory is satisfactory ‘only at the
level of the lawyer concerned with the working of the “ordinary” law’, and is not
adequate to handle individual rights at the level of constitutional law. Still less,
Hart says, is his theory equipped to explain rights as they are understood by indi-
vidualistic critics of the law and by social theorists (Hart 1982, 185–6, 192–3).
Hart attempts to make a virtue of this limitation by asserting that his will
theory is only intended to explain rights within the ‘ordinary’ law. Yet even were
we to grant Hart that his theory accounts for this limited domain, this first strat-
egy makes a major concession. As with scientific theories, a more comprehen-
sive account of the subject matter is always preferable to a less comprehensive
account. This is particularly clear with rights. A theory that is adequate only to
rights within one part of the law will at best satisfy certain specialists, and will
not provide an analysis that is useful for understanding rights as a central concept
in morality, in politics, and in the law viewed more broadly.
Interest theorists face the analogous difficulty that people’s rights frequently
outrun their interests. To take one type of example, because of ignorance or care-
lessness people often enter into promissory relations that vest in them rights to
receive goods or services that they have no interest in receiving. Imagine, for
instance, a budding auto enthusiast who finds in a newspaper what he thinks is
a fine deal on a second-hand engine, and pays to have this engine delivered to his
house the following week. As it turns out, this enthusiast has erred in buying this
engine: as he does not yet realize, it does not fit his (or indeed any extant) car. The
enthusiast has no interest whatsoever in having this useless and hard-to-dispose
engine winched onto his driveway. As he will discover when he returns home
from work and tries to install the engine, the delivery has made him significantly
worse off. Yet as the lorry rumbles toward his house with the bulky engine in
the back, there is no doubt that the lorry drivers are fulfilling the enthusiast’s

14-MHKramer-Chap14.indd 256 4/30/2008 5:16:29 PM

 The Analysis of Rights 257

claim-right. The example shows that a person’s well-being does not go up merely
in virtue of promises to him being kept. A promisee can be in every way better
off if the deal he has foolishly entered into goes unconsummated, and so his right
remains unfulfilled.
In response to counterexamples where rights outrun interests, several interest
theorists have framed their theories around weak generalizations, which only
attempt to explain some but not all rights. MacCormick, for example, phrases
his central thesis in terms of ‘normal circumstances’: ‘To ascribe to all members
of a class C a right to treatment T is to presuppose that T is, in all normal cir-
cumstances, a good for every member of C.’ (MacCormick 1982, 160, emphasis
added) Similarly, Kramer presents his theory in terms of what is ‘generally benefi-
cial for any typical human being or collectivity or non-human creature’ (Kramer
and Steiner 2007, 290, emphasis added). The qualifications ‘normal circum-
stances’, ‘generally’, and ‘typical’ limit the domain of rights that these theories
will attempt to explain to a domain with certain rights (specifically, the coun-
terexamples) removed.⁷
Weak generalizations are unsatisfying because of their lack of comprehensive-
ness. A linguist studying English will not rest content with the rule that in all
normal circumstances, ‘ i’ comes before ‘e’. Nor will a toxicologist be satisfied with
the thesis that mushrooms are generally harmless when eaten. It is unlikely that
the best theory of rights takes the form: ‘All rights have feature F (except for
those that lack feature F).’⁸ A weak generalization can be better than another
generalization that is weaker still, or better than no generalization at all. Yet a the-
ory based on a weak generalization will always be discarded once a theory with
greater explanatory scope is found.

3.2 Expanding the Scope of the Theory

The second strategy of rights theorists for overcoming the problem of narrowness
has been to attempt to modify their theories so as to capture more rights. Both
will and interest theorists have taken this path.
For example, Steiner has found new rights within the scope of the will theory
in response to the charge that the will theory fails to recognize the rights of the
‘powerless’ (Kramer, Simmonds, and Steiner 1998, 258–62). Steiner’s will theory
cannot recognize the rights of children, comatose adults, and animals, because
these beings lack the power to waive or enforce the duties of others. Nor can the
will theory recognize the rights of criminal defendants, since such defendants do
not have the power to waive others’ duties against assaulting them, killing them,

⁷ See also Raz 1986, 173–6.

⁸ It is of course possible to try to recapture universal application by specifying some prop-
erty specially rigged for the purpose, for example ‘All rights share the property of belonging to a
set whose members generally have feature F.’ But the weak spot in the generalization will always
remain.

14-MHKramer-Chap14.indd 257 4/30/2008 5:16:29 PM

 258 Leif Wenar
and so on. Yet Steiner argues that his will theory can acknowledge rights that
protect these unempowered beings—so long as those rights are seen to reside in
beings besides those who have traditionally been taken to be the right-holder.
On Steiner’s interpretation of the will theory the right-holder in these cases
will not be the citizens, children, or animals who are protected by certain duties.
Rather, the right-holders will be the ‘power-possessors’ who have the authority to
waive or enforce those duties. So, for example, Steiner says that the right that a
citizen not be assaulted is held not by that citizen, but by the magistrate who can
decide whether to charge an assailant of that citizen. The right that the citizen not
be assaulted is the magistrate’s right. Similarly, the right that a particular child
not be abused is vested not in that child, but in the judge who will decide whether
to punish a person convicted of child abuse. It is the judge, not the child, who has
the right that the child not be abused.
In response to the unnatural feeling of this location of rights in the case of chil-
dren, the comatose, and animals, Steiner writes:
What scintilla of a practical or analytical difference can it make if we construe the rights
correlative to those protection duties as one held by those power-possessors rather than one
held by unempowerable creatures? As far as I can see, none. And if those power-possessors
are indeed the holders of those rights, then, as we saw in the previous discussion of crim-
inal law, the rights they hold are none other than will theory rights. (Kramer, Simmonds,
and Steiner 1998, 261)
The difficulty that Steiner faces here is that he is both appealing to an ordinary
understanding of rights to support his interpretation of the will theory, and also
claiming that it makes no difference that this interpretation continues to conflict
with such an ordinary understanding. Steiner’s will theory acknowledges a greater
number of rights than do other versions of the will theory, since on his interpret-
ation it can be said that there are rights that protect citizens, children, and so on.
This fits Steiner’s will theory more closely to an ordinary understanding of rights.
Yet Steiner’s revised theory locates these extra rights in what are, on any common
appreciation of rights, entirely the wrong places. Within Steiner’s theory citizens
still have no rights against criminal assault, children have no rights against abuse,
and so on. So here there are more rights, which is welcome from an ordinary per-
spective; yet these rights are said to be vested in the wrong individuals, which is
not. Steiner can legitimately commend his interpretation of the will theory for
the ways in which it better captures an ordinary understanding of rights, but he
can hardly then maintain that the ways in which his interpretation still conflicts
with an ordinary understanding make no difference.
Kramer similarly considers expanding his interest theory to accommodate
rights that seem beyond its reach.⁹ Kramer suggests that within a ‘capacious’

⁹ I hesitate to address Kramer’s interest theory, as Kramer has not yet had the chance to set out
his evolving theory fully. This is evident in Kramer’s last published writing on his interest theory
(Kramer and Steiner 2007). For example, Kramer makes philosophically significant modifications

14-MHKramer-Chap14.indd 258 4/30/2008 5:16:29 PM

 The Analysis of Rights 259

version of his interest theory (which he discusses but does not endorse) the powers
and privileges belonging to various offices are properly classified as rights because
those powers and privileges usually promote certain interests of the office-holders.
This suggestion does not have much immediate appeal. For example, consider the
fact that a judge’s power to sentence criminals is properly classified as a right. It
would sound odd to say that this fact is explained by the fact that the possession
of this power is generally beneficial for the judge. Similarly with the fact that a
policeman’s liberty to detain a suspect is properly classified as a right. One might
not think that this fact is explained by the fact that the possession of such a lib-
erty is generally beneficial for the policeman.¹⁰
Yet that is what the ‘expansive’ version of Kramer’s interest theory holds.¹¹
Within this expansive theory, if a norm (here a role) bestows a normative ability
(a Hohfeldian power or privilege) on a person, then the fact that that normative
ability is properly classified as a right is explained by the fact that their having

to his theory without having space to explain fully why he has done so or what the further implica-
tions might be, such as when he declares that a large class of interests (which he labels ‘vicarious’)
are irrelevant to what rights there are (Kramer and Steiner 2007, 302–4). More broadly, Kramer
has yet to set out a usable method for applying his theory’s distinctive test for locating claim-right
holders. Kramer’s test says that X holds a claim-right if X’s detriment is sufficient to establish a
breach of a duty, yet Kramer has not yet said how one can tell whether X’s detriment is in fact suf-
ficient to establish a breach of a duty (without begging the question by surreptitiously relying on
one’s beliefs about whether X holds a claim-right). A full treatment of Kramer’s interest theory
must wait until Kramer has had the opportunity to present a complete statement of his revised the-
ory, by explaining what he believes are the necessary and sufficient conditions for the ascription of
a right, and by offering some systematic account of how one can tell whether these conditions have
been met.
¹⁰ Unlike Raz’s interest theory, on the expansive version of Kramer’s interest theory the interests
of a right-holder are not necessarily what justifies the establishment of a right: they are not neces-
sarily what justifies the imposition of duties or the creation of norms or roles. This can be seen from
Kramer’s test for right-holding (on the expansive interpretation of his view): ‘If a norm or decision
bestows a Hohfeldian entitlement on Q, and if the possession of that entitlement would usually
be beneficial for someone in Q’s situation, then Q is a right-holder under the norm or decision.’
(Kramer and Steiner 2007, 290) Here the interest (what ‘would usually be beneficial for someone in
Q’s situation’) does not necessarily have any justificatory relation whatsoever to the norm. Rather
the interest (partly) explains the fact that Q has a right by usually being present when the norm that
bestows the relevant Hohfeldian entitlement is present: that is, by satisfying the second conjunct in
the antecedent when the first conjunct is also satisfied. This is the sense in which, for example, the
fact that a policeman’s liberty to detain a suspect is properly classified as a right is explained by the
fact that the possession of such a liberty is generally beneficial for the policeman.
¹¹ Kramer sets out the ‘expansive’ version of his interest theory at some length, and says it
may be worth developing further, but in the end he does not accept it (Kramer and Steiner 2007,
290–5). Th is expansive version of the interest theory would need further explication in any case.
For example, in setting out this theory Kramer appeals to a distinction between ‘intrinsic’ and
‘extrinsic’ effects of legal norms without defining these terms or saying how one might distinguish
one from the other (293). Kramer also appeals to the purpose of legal norms to make sense of these
cases (293), where a few pages earlier he had said that purposes had no ‘determinative bearing’ and
were ‘quite immaterial’ in his theory (289, 290). I discuss the expansive version of Kramer’s inter-
est theory here because it is a serious attempt to broaden an interest theory so as to encompass the
rights of offices and positions. Without some modification like this one, there seems little hope that
Kramer’s interest theory will be able to capture these rights.

14-MHKramer-Chap14.indd 259 4/30/2008 5:16:29 PM

 260 Leif Wenar
this ability will normally make them better off. Within this theory the fact that a
judge’s power to sentence criminals is properly classified as a right is explained by
the fact that having this power allows a judge ‘to carry out his judicial responsi-
bilities with smoothness and efficiency’. Within this theory the fact that a police
patrolman’s liberty to detain a suspect is properly classified as a right is explained
by the fact that his having this liberty enables the patrolman to ‘fulfill his duty to
detain [the suspect], without exposing himself to penalties for so doing’ (Kramer
and Steiner 2007, 290, 291). Judges and policemen are made better off, the
theory holds, by having the normative abilities to carry out their responsibilities.
So these normative abilities further the interests of these role-bearers, and are
therefore rights.
Such appeals to smooth and punishment-free discharge of responsibilities can-
not help an interest theory to explain why the powers and privileges of offices are
rights. For this line of reasoning simultaneously relies upon and misunderstands
the norms that are roles. It is not as though there is a coherent role described as
‘judge who has the responsibility to sentence but no power to sentence’—and then
we make a separate determination that the life of someone filling that role would
go better if they gained the power to sentence. The original description makes no
sense: a role that assigns a responsibility to ␸ but with no normative ability to ␸
is not a role that fits into any recognizable human practice.¹² Similarly with the
patrolman. There is no intelligible job that is ‘policeman who has the duty not to
detain suspects whenever he has the duty to detain suspects’. Such a job could be
imagined only at the edges of a fantasy, if there; speculation about the interests
of such a job-holder is moot. In reality, offices such as ‘judge’ and ‘policeman’
always pair the responsibilities of office with the normative abilities appropriate
for carrying out these responsibilities. ‘Rights of office’ are not optional add-ons
that help a person do a job; rather, rights of office are an integral part of every job’s
description.¹³

¹² Kramer might rather claim that a judge’s responsibility is not to sentence but rather to see
that sentences are passed. Yet imagining someone who has a responsibility to see that sentences are
passed, but who has no power of his own to sentence, would not fit Kramer’s words: such a person
would not have ‘judicial responsibilities’ in this respect. Such a person would rather be the holder of
some sort of administrative office. The powers of that administrative office (to see that sentences are
passed) would presumably include powers to appoint, or perhaps simply to discipline, the judges
who have the power pass sentences. Whether it would be in the interest of someone holding such
an administrative office for his office to be redefined so that he himself gains the power to sentence
depends entirely on how we imagine his office to be currently defined (eg, how many judges he is
responsible for overseeing, whether he is expected to have detailed understanding of sentencing
procedures, what oversight he himself faces for discharging his responsibilities, etc). Within any
well-designed system of roles there will be no presumption that overseeing officers have any such
interest in their roles being redefined. (This reasoning applies also to the cases of the traffic warden
and the army captain (Kramer and Steiner 2007, 290).)
¹³ It might be noted here that this discussion is not concerned with whether it is in any individ-
ual’s interests to hold a particular office in the first place. It may be beneficial for an individual to
occupy some office, or it may be entirely a burden. It may or may not be in Jane’s (or anyone’s) inter-
est to be a judge, but in either case Jane will have the rights of a judge if she is a judge. It might also

14-MHKramer-Chap14.indd 260 4/30/2008 5:16:29 PM

 The Analysis of Rights 261

The expansive version of Kramer’s interest theory cannot account for the fact
that role-bearers’ powers and liberties to do what they have duties to do are rights.
Nor can it account for the fact that their discretionary liberties are rights. For
example, a parent has the liberty either to punish or not to punish her child.
Kramer attempts to explain why this first liberty (the liberty to punish) is a right
as follows. Imagine a world in which a parent has no liberty to punish her child
(ie, she has a duty not to punish), but in which she retains the liberty not to pun-
ish her child. In such a world, Kramer says, the parent would be better off if she
gained the liberty to punish her child because then she would no longer be liable
to penalties whenever she did punish (Kramer and Steiner 2007, 291–2).¹⁴
In Kramer’s imagined world parents are prohibited from punishing their
children. In this world parents also have no duty to punish their children. Why in
this world would a parent be better off if she gained a liberty to punish her child?
She will be better off gaining a liberty to punish her child only if she has some
reason to punish her child. Yet there is no reason that Kramer can depend upon
here.
Kramer appears to suggest that in his imagined world a parent would have a
role-based reason to punish her child, and so would be better off with the liberty
to punish. He says that with this liberty she would ‘not have to worry about being
penalized for taking steps which she reasonably deems necessary for the effective
performance of her role as a parent’ (Kramer and Steiner 2007, 292). Yet within
this imagined world parents are prohibited from punishing their children, so
their role is quite different than in our world. In this imagined world, the role of
parent could at best be described as ‘raising one’s children well, so far as one can
do so without punishing them’. Punishing one’s child could not be a step reason-
ably deemed necessary for the effective performance of that role. In the imagined
world, discipline is no part of a parent’s job description. So Kramer has no role-
based reason available to explain an interest in gaining the liberty to punish.
Kramer might instead venture that a parent in his imagined world would have
some non-role-based reason to discipline her child.¹⁵ Yet this depends entirely
on how we imagine this world to be. For example, in this imagined world where
parents have no duty to discipline their children and indeed are prohibited from
doing so, the responsibility for disciplining children might well be (indeed likely

be mentioned that this discussion remains neutral concerning what justifies the creation of offices
that are defined by specific duties and rights.
¹⁴ Kramer notes that his reasoning is the same across two variants of this example. The dis-
cussion here concerns the variant in which the parent has a Hohfeldian privilege not to punish
the child, and Kramer considers why the parent’s privilege to punish is a right. The analysis of the
other variant, where the parent has a duty to punish the child, is captured by the police patrolman
example above.
¹⁵ As mentioned in n 9 above, Kramer does not allow appeals to ‘vicarious’ interests within his
theory (ie, interests that ‘reside wholly in the furtherance of somebody else’s interests’ (Kramer and
Steiner 2007, 303)). So whatever reason to discipline Kramer might posit here, it cannot be the rea-
son that a parent has to discipline her child for the child’s own good.

14-MHKramer-Chap14.indd 261 4/30/2008 5:16:29 PM

 262 Leif Wenar
would be) assigned to someone else. Were we to posit that parents have an interest in
punishing their children even when someone else is effectively doing so, we would
seem to be making parents out to be simply cruel. Or again: parents in Kramer’s
imagined world might be just as glad, all things considered, to be legally prohib-
ited from punishing their children. They might think that gaining the liberty to
punish would result in their having endless headaches (familiar from our world)
that they would just as soon avoid.¹⁶
An interest-theoretical approach such as this one cannot be the path toward
understanding the rights of offices and positions. Rights of office cannot be
explained by the interests of the individuals who occupy the office as currently
defined. An interest theory needs to be expanded so as to capture these rights, but
the interest framework does not provide the resources for the theory to do so.

3.3 Redefining the Phenomena to be Explained

The third strategy of will and interest theorists has been to assert that their the-
ories are intended to account only for ‘rights’ in some technical sense of that term
instead of in an ordinary sense. This third strategy resembles the first strategy in
altering the domain over which the theory is meant to apply. Yet this strategy,
unlike the first, applauds the fit between the theory and some artificially con-
structed concept that is given the name ‘rights’. Interest theorists who take this
path stipulate that their theories are not meant to account for rights as commonly
understood, but only for ‘rights’ defined as Hohfeldian claim-rights. Will the-
orists who take this tack say that their theory is only intended to explain ‘rights’
defined as Hohfeldian claims accompanied by Hohfeldian powers of waiver or
enforcement.
Thus when these theorists present their ‘theories of rights,’ the term ‘rights’
is intended to refer to a technical explanandum (such as ‘claims’ or ‘claims-with-
powers’). ‘Rights’ no longer refers to the ordinary explanandum, which is the
full catalogue of rights as commonly understood. As we will see, these theorists
do not in the end repudiate ‘fit with an ordinary understanding’ as a criterion of
success. Yet at least initially they take the phenomena to be explained as ‘rights’ in
some specially defined sense.

¹⁶ This example here resembles the case of John (Kramer and Steiner 2007, 292–3). Kramer
addresses this case by invoking his weak generalization about what is generally in people’s interests,
and then asserting that it will be ‘extremely rare’ for people to be better off for facing a legal prohib-
ition that disinclines them from doing something risky or difficult that they would otherwise do.
But as the parent example shows, Kramer must face this kind of question all the time. Moreover,
reflection on paternalistic legislation gives further reason to doubt Kramer’s assertion. Paternalistic
legislation is just an attempt to make people better off by instituting a legal prohibition that disin-
clines them from doing something risky or difficult that they would otherwise do. For Kramer to
establish that it is extremely rare for a legal prohibition to further the interests of those restricted by
it, he would need to show that it is extremely rare for paternalistic legislation to achieve its aims.

14-MHKramer-Chap14.indd 262 4/30/2008 5:16:29 PM

 The Analysis of Rights 263

This substitution of the reference of ‘rights’ means that theorists using this
third strategy speak a different dialect than that used by ordinary speakers such
as judges, lawyers and laymen. Thus when Wellman is confronted with the result
that his will theory cannot recognize the rights of infants, he says:
Surely it is confusing for me to insist that infants could not have legal rights as I conceive
of a right but to admit that infants can and do have rights as judges and lawyers con-
ceive of rights. Still, this confusion can be minimized, although probably not completely
avoided, by distinguishing carefully between two spheres of discourse, the language of
the law and the language of the philosophy of law. (Wellman 1995, 135)
Wellman here distinguishes the ordinary language of the law from the lan-
guage in which his technical explanandum occurs: the ‘language of the philoso-
phy of law’. Kramer makes the same kind of distinction to justify a move to his
own preferred technical explanandum. Kramer wishes to work within a sphere of
discourse in which the referent of ‘rights’ is claim-rights. Like Wellman, Kramer
faces the difficulty that ordinary discourse does not line up with his technical
definition. In Kramer’s case, one conflict is that claim-rights concern only the
actions of others, while ordinary usage accepts many rights that give the right-
holder themselves rights to act (eg, speak, worship, promise). Faced with obvi-
ous cases where ordinary usage acknowledges rights to act, Kramer responds by
disparaging ordinary language: ‘Our ordinary ways of speaking about rights as
entitlements to do various things are loose’ (Kramer, Simmonds, and Steiner
1998, 13–14).
In Wellman’s ‘language of the philosophy of law’, the term ‘rights’ refers to
the technical concept ‘claims-with-powers’. In Kramer’s ‘strict sense’, the term
‘rights’ refers to the technical concept ‘claims’. The question for the theorists of
rights who deploy a technical concept is whether they can offer a rationale for
moving the analysis away from ordinary language and toward their favoured
technical explanandum in particular. One obviously inadequate justification
would be to point out that redefining the explanandum makes the phenomena to
be explained fit better with their favoured explanans. A will theorist, for instance,
should not simply say that he prefers his technical definition of the term ‘rights’
because the will theory is so good at explaining ‘rights’ so defined. Yet putting
such special pleading to the side, how else could a move to some particular tech-
nical explanandum be motivated?
Rights theorists who work within a technical discourse characteristically allege
that this move is necessary because the ordinary discourse of rights is hopelessly
vague, or because ordinary language speakers are prone to fall into contradictions
when discussing rights (eg, Kramer and Steiner 2007, 295). As Kramer puts it
in the quote above, ordinary ways of speaking about rights are ‘loose’. No the-
ory, these theorists say, can hope to explain a set of assertions if the assertions
have extremely indeterminate or contradictory content. So, these theorists allege,
they must aim their theories at an artificially-defined domain of ‘rights’ instead.

14-MHKramer-Chap14.indd 263 4/30/2008 5:16:29 PM

 264 Leif Wenar
Interest theorists are particularly likely to cite Hohfeld in motivating their move
to a technical discourse of rights, as Hohfeld was the originator of the thesis that
in the ‘strictest sense’, all rights are claims (Hohfeld 1919, 36).¹⁷ Will theorists
take a differently-defined technical concept to be the object of their explanatory
theory, which is more congenial to their thesis that rights endow their bearers
with discretion over the duties of others.
Yet how then to decide which subset of the Hohfeldian incidents is the one that
theories of rights should take as their focus? In fact, and surprisingly, the clearest
and most repeated justification offered by the technical rights theorists for using
their own favoured technical definition of ‘rights’ is that their favoured definition
fits better with ordinary language (eg, Kramer and Steiner 2007, 296–7). Having
spurned ordinary understanding in order to motivate the move to a technical
concept, these theorists then emphasize the overlap between their favoured tech-
nical definition and an ordinary understanding in order to validate that defin-
ition. Indeed the debate between the technical will and interest theories has not
infrequently turned into a debate over whether the area of overlap with common
usage is greater using one technical definition of ‘rights’ rather than the other.
This appeal to ordinary language leaves the technical rights theorists in a pre-
carious position. On the one hand, they attempt to cast enough aspersions on
ordinary ‘rights-talk’ that there will appear to be no option but to shift from an
ordinary to a technical explanandum. On the other hand each camp of tech-
nical theorists appeals to the fit between their favoured technical explanandum
and ordinary understanding as the reason to judge their favoured explanandum
superior. Interest theorists argue that the technical characterization of rights as
claims is ‘more acceptable to ordinary understanding’, while will theorists argue
that this honour is more fittingly given to their technical characterization of
rights as claims-with-powers (Kramer, Simmonds, and Steiner 1998, 74).
The title character in Oliver Sacks’s The Man Who Mistook His Wife for a Hat
suffered from a certain kind of aphasia which led him to attempt to embrace his
wife with his right arm at the same time as he attempted to push her away with his

¹⁷ Hohfeld’s curious, unargued stipulation that ‘in the strictest sense’ all rights are claims dis-
ables anyone who accepts it from giving a straightforward analysis of many commonly-asserted
rights. For example, any analysis of legal rights should be able to explain a judge’s legal right to sen-
tence a convicted criminal. Some who adhere to Hohfeld’s stipulation set aside the obvious analysis
that this right consists in the judge’s Hohfeldian power to impose duties on the convict, and say
rather that the right consists in the claims protecting the judge from the interference of others when
he exercises such a power (see for example Raz, who separates the power to promise from the ‘right’
against interference with one’s promising (Raz 1986, 173–6)). However, construing such rights
as claims against interference is strained. One can see this by imagining situations in which inter-
ference is literally impossible (eg, where judges pass sentences from impregnable strongholds,
communicate telepathically, etc). Here a claim-right against interference makes no sense, but
judges would still have a right to sentence. Similarly, anyone who accepts Hohfeld’s stipulation
will have difficulties explaining many (power- and privilege-) rights that religious believers have
for centuries attributed to God (God has a right to make promises, to command his creations, to
punish sinners, etc). These rights cannot possibly be construed as God’s rights against interference,
for such interference is literally unimaginable.

14-MHKramer-Chap14.indd 264 4/30/2008 5:16:29 PM

 The Analysis of Rights 265

left (Sacks 1985, 8–22). The technical rights theorists have something of the same
disposition toward the ordinary understanding of rights. These will and interest
theorists are aware that their theories of the functions of rights cannot account
for a large number of rights that are commonly accepted. They therefore insist
that ordinary rights-talk is vague and inconsistent in order to attempt to shift
attention toward their favoured technical recharacterizations of the term ‘rights’.
Yet when attention is so shifted, they once again attempt to embrace an ordinary
understanding of rights as the guarantor of their theory’s superiority. This pos-
ition holds the ordinary understanding of rights to be so irremediably vague and
inconsistent as to be useless as the object of theoretical explanation, but not so
corrupt as to be useless in mediating between artificially constructed definitions
of ‘rights’.
This is an awkward posture to maintain, and in fact there is no need to assume
it. The technical theorists’ stated rationale for moving to a technical explanan-
dum was to overcome the vague and contradictory nature of ordinary assertions
about rights. However, vagueness and contradictoriness in ordinary language
can be no reason whatsoever to switch to an artificial concept of rights.
First, technical rights theorists have not in fact established that ordinary dis-
course is frequently vague and contradictory, instead of being a discourse that
systematically assigns different meanings to the same word in different contexts.
Such systematic variation in meaning is familiar in common speech. For example
there is nothing vague or contradictory in the statement that ‘In a free market one
is free to lend money interest-free.’ ‘Free’ in this statement takes three different
but determinate meanings, the meaning of each occurrence being determined by
the context. Similarly in ordinary discourse one often hears the word ‘right’ used
to refer variously to privilege-rights, claim-rights, power-rights, and immunity-
rights, with the intended referent made clear by the context. Interpretations of
ordinary speech that find vagueness or inconsistency instead of precise and deter-
minate variation in usage are often just poor interpretations of ordinary speech.
Second, any vagueness and inconsistency within ordinary discourse about
rights, insofar as it exists, does not justify a radical shift to a technical explanan-
dum. For there is a straightforward solution to any problems with vagueness and
contradiction, which is for rights theorists to use the Hohfeldian framework to
discuss rights (as indeed most already do).
Using the Hohfeldian framework of privileges, claims, powers, and immun-
ities gives maximum specificity to statements about rights, while simultaneously
insuring against contradictions. So long as rights theorists use the Hohfeldian
language correctly, they cannot commit errors of vagueness or inconsistency.
Therefore once theorists are using this analytical framework, there is no fur-
ther need for them artificially to designate some subset of the Hohfeldian inci-
dents as the referent of the term ‘rights’. Once theorists have agreed to use the
Hohfeldian terminology, there would need an extra argument to motivate a
redefinition of ‘rights’ either as ‘claims’, or as ‘claims-with-powers’, or indeed as

14-MHKramer-Chap14.indd 265 4/30/2008 5:16:30 PM

 266 Leif Wenar
anything else. Any attempt by technical rights theorists to draw on the author-
ity of (what these theorists allege is vague and contradictory) ordinary language
at this point cannot provide that required extra argument, and can only cause
confusion. Rights theorists utilizing the Hohfeldian framework make precise
and consistent statements about rights as they are commonly understood, so
there is no further rationale for these theorists to replace the object of analysis
with any technical concept.
It might nevertheless be thought that there exists an extra argument for will
and interest theorists to move to a technical explanandum, beyond the limp
argument just discussed of avoiding vagueness and contradiction in ordinary
language. It might be thought that a move to a technical explanandum could be
justified on the ground that a theory directed toward it would be not only more
comprehensive and simpler but also more fruitful. As Carnap says in the context
of scientific explananda, ‘a scientific concept is the more fruitful the more it can
be brought into connection with other concepts on the basis of observed facts; in
other words, the more it can be used for the formulation of laws’ (Carnap 1950, 6).
A scientific concept is more fruitful, in other words, the better it fits into a larger
system of explanatory generalizations. Perhaps a technical concept of rights could
also be more fruitful in this way.
It is fairly common for scientists to move from an ordinary to a technical
explanandum on grounds of fruitfulness. For example consider the concept of
‘fruit’ itself. Within a botanist’s conceptual scheme, and in contrast to ordinary
usage, a tomato is a ‘fruit’ but a stalk of rhubarb is not. This is because botanists have
substituted a technical concept of ‘fruit’ for the ordinary one; to a botanist ‘fruit’
means a ‘seed-filled ripened ovary of a flowering plant’. The botanist will prefer
to work with his technical concept rather than the ordinary concept, because
the technical concept fits better into the larger botanical theory of the life-cycle
of plants. The botanical concept of ‘fruit’ is, given these general theories, more
fruitful.
Will and interest theorists might analogously argue that their move from an
ordinary to a technical explanandum is justified by the fruitfulness of their favoured
concept within more general theories of morality or the law. Thus Wellman in
motivating the move to his favoured technical explanandum suggests that ‘a more
restricted application of the language of rights may be theoretically required in
order to provide a clearer and more revealing map of the law’ (Wellman 1995,
136). While he does not elaborate upon this idea, what Wellman appears to
mean is that his favoured technical characterization of rights fits better within his
preferred jurisprudential theory—that is, it fits better with his view of what the
nature of the law is.
However unlike in the scientific case, such appeals to fruitfulness must be
illegitimate and for reasons we have already seen. As noted above, there is no
agreement on which substantive theory of morality or the law is correct. Will
and interest theorists cannot say that their favoured concepts are more fruitful

14-MHKramer-Chap14.indd 266 4/30/2008 5:16:30 PM

 The Analysis of Rights 267

given the correct theory of morality or of the law, because unlike the theory of the
life-cycle of plants these theories are not ‘given’. Indeed, which substantive the-
ory of morality or law is correct is a matter of fundamental dispute. An interest
theorist who is told that a will-based definition of ‘rights’ fits more fruitfully into
some controversial moral or jurisprudential theory will not believe that he has
been given a reason to accept that definition (and vice versa). Of course it is vital
that philosophers continue their debates within normative theory about which
substantive theory of morality and of the law is the correct theory. However,
within a conceptual analysis any appeal to a contentious substantive theory will
merely infect that analysis with the contentiousness of the substantive theory
appealed to.
All of the efforts of technical will and interest theorists to move away from ‘fit
with ordinary understanding’ as the criterion for a successful analysis of the con-
cept of a right are, finally, unsuccessful. The allegedly vague and contradictory
nature of ordinary usage cannot justify such a move, nor can an appeal to fruit-
fulness. The only usable criterion for success in analysis remains that the ana-
lysis tracks the ways that informed, thoughtful speakers of the language use the
concept—which is in fact the criterion that even technical rights theorists use,
despite themselves, most frequently.

4 The Standoff

Neither the will nor the interest theory provides a comprehensive enough
account of an ordinary understanding of rights. Neither theory succeeds, there-
fore, in achieving a fundamental goal of a theory of rights. Will theorists and
interest theorists have explored three strategies of response to this problem in
some detail. Yet as we have seen none of these lines of response—restricting the
domain of application, attempting to expand the scope of the theory, or resort-
ing to a technical discourse—has proved adequate. In this situation neither side
of the debate is able to prevail conclusively in the main arena, nor can either
side garner more resources or shift the field of play to one more favourable to
its view. The result, as Wayne Sumner has said, is a kind of standoff (Sumner
1987, 51). Th is is a battle in which, despite the deployment of great ingenuity
on each side for many years, there seems no chance that either side can emerge
victorious.
Lacking the means to prove that their preferred theory is superior, will and
interest theorists have resorted to turning up the volume in pointing out how
the rival theory conflicts with ordinary understanding. Steiner is relatively civil
in casting aspersions on the interest theory, pointing only to the ‘grave implaus-
ibility’ of its implications and how it ‘places considerable strain on our ordin-
ary understanding of rights’ (Kramer, Simmonds, and Steiner 1998, 285, 287).
MacCormick’s frustration with the standoff between his own interest theory and

14-MHKramer-Chap14.indd 267 4/30/2008 5:16:30 PM

 268 Leif Wenar
the will theory leads him to more irritated pronouncements, wondering whether
we must accept a theory that ‘does such violence to common understanding’:
We are entitled to ask somebody who stipulates that there shall be held to be ‘rights’ only
where there are choices, whether that stipulation does not go wholly against common
understanding, and whether there is any profit derived from it. (MacCormick 1977, 197)
Kramer’s exasperation in being unable to dispose decisively of the rival theory
results in an all-out high-decibel assault. ‘One can scarcely help being puzzled’,
he says, about the will theory’s ‘arresting’ claims and ‘bizarre stipulations’ that are
‘needlessly odd’ and ‘flout too many entrenched linguistic intuitions to be very
powerful’. The will theory is guilty of ‘gratuitous contraventions of ordinary pat-
terns of usage’, and ‘yields some results that tend to strike the ordinary observer
as ridiculous’. Finally, he reviews the will theory in language usually reserved by
critics for the year’s worst film:
Many people would shrink from a theory which defines ‘right’ in a way that commits the
proponents of the theory to the view that children and mentally infirm people have no
rights at all. Even when stripped of its ghastliness by being carefully explained, such a
view tends to sound outlandish when stated.¹⁸
When a long-running debate reaches this level of acrimony without coming
any closer to producing a conclusive result, we may conclude that the debate is no
longer progressing.

5 The Shared Restrictive Premise

In one way, the debate over the functions of rights is presently in a bad state.
Theorists of the two leading views have been contesting for so long, and have
become so familiar with the limited resources available on each side, that even
the most stalwart defenders seem resigned to battle for the minor honor of hold-
ing the less starkly counter-intuitive theory.¹⁹ In such a debate, as Schopenhauer
said about diplomatic squabbles, each side complains about the other, and both
sides are correct. Even more disheartening is when outsiders to this debate import
either the will or the interest theory as a premise from which to derive further
conclusions about rights in their normative theorizing. One cannot blame these
outsiders for reaching for a leading theory in an area outside of their specialism.

¹⁸ Kramer, Simmonds, and Steiner 1998, 72, 69, 73, 75, 69. In another essay Kramer describes
the conclusions of the will theory as ‘bizarre’, ‘preposterous’, and ‘jarringly and gratuitously at odds
with ordinary patterns of discourse’ (Kramer 2001, 71).
¹⁹ Thus Steiner: ‘Theories of rights don’t come cheap. Buying either of them [the will theory
or the interest theory] involves paying some price in the currency of counter-intuitiveness. Nor,
I should add, has this centuries-long debate about the nature of rights ever revealed any distinct
third theory that even approaches their levels of generality, let alone promises to undercut their
prices.’ (Kramer, Simmonds, and Steiner 1998, 298)

14-MHKramer-Chap14.indd 268 4/30/2008 5:16:30 PM

 The Analysis of Rights 269

Yet these theorists do so without realizing the genetic weaknesses that their
normative arguments thereby inherit.
There is, however, cause for optimism, as a broader view of the deadlock
between the will and the interest theories shows where progress must come. The
structure of the debate that we have seen is as follows. Each of the two theories
explains some but not all of the relevant phenomena—in this case, an ordinary
understanding of rights. Each of the theories has attempted predictable responses
to its own narrowness. None of these responses has been adequate, even after
many variations have been advanced. It seems quite likely that the correct diag-
nosis of this situation is that each theory captures part of the truth about the
nature of rights, but that each also has within it some unremovable premise that
prevents it from capturing the whole truth.
What could this premise be? We get an initial indication of the location of the
premise by recalling two of the formal desiderata that these theories are attempting
to fulfil: comprehensiveness and simplicity. Both the will and the interest theor-
ies, we have found, are insufficiently comprehensive. The natural suspicion must
be, therefore, that they are excessively simple. There must be some oversimplified
view of rights entrenched within these theories that prevents them from fram-
ing a thesis that would account for all of the phenomena to be explained. To put
this the other way around, there must be some complexity in the nature of rights
that these theories cannot acknowledge while they remain will or interest theories.
If there were some way to rework these theories to capture this complexity, it
seems that will or interest theorists would have found it by now.
Where more specifically is the restrictive premise within these theories? In
this debate there are two theories, each professing that rights have a single func-
tion. Each of these theories appears to capture part—but only part—of the truth
about what rights there are. The erroneous shared assumption must be that rights
have a single function. The correct assumption therefore must be that rights have
more functions than one. The difficulty faced by both will and interest theorists
throughout their long debate is that they have each been advancing a monis-
tic theory to account for pluralistic phenomena. This explains why the debate
between them has been unresolvable. Each side can claim a certain domain as its
own, and cast counter-examples at the other side. But neither side can give up its
focus on just one function of rights without giving up the basic character of its
theory. Thus the theories are stuck in the stalemate.
This situation has precursors in the history of physical theory. The pre-Socratics
put forward contending monistic theories of the physical world. The debate
between Thales’ thesis ‘all is water’ and Anaximenes’ thesis ‘all is air’ resembles
the modern debate between the will thesis ‘all rights give choices’ and the interest
thesis ‘all rights further well-being’. Progress in scientific theory came only with
the abandonment of the shared monistic premise. What post-Socratic scientific
theories gave up in simplicity, they more than made up for in comprehensiveness.
Progress in rights theory can be expected to come along the same path.

14-MHKramer-Chap14.indd 269 4/30/2008 5:16:30 PM

 270 Leif Wenar
The truth about the functions of rights is that there must be more functions
than one. And as the scientific example shows, the complete set of functions
may not be merely a concatenation of the monistic functions. The Pre-Socratic
monistic theories eventually gave way to the five-element Aristotelian framework
(earth, air, water, fire, aether), which was itself then replaced by the modern 113-
elemented table of chemical elements. The Aristotelian framework is quite a bit
less simple than its monistic precursors, just as the periodic table is considerably
less simple than it. Yet in each case the later theory yields more powerful explana-
tions. We are willing, it seems, to sacrifice a good deal of simplicity in order to
find a theory that captures all of the phenomena. If we look for a general answer
to how many basic theoretical posits will render a theory insufficiently simple, the
answer seems to be ‘one more than in the simplest theory that explains all of the
data’. In physical theory five posits were better than one, and as it turns out, 113
posits are better than five.

6 The Relation between Analytical and

Substantive Theories of Rights
As we have seen both the will and the interest theories are based on a monistic
premise, and each fails because it is so based. So far as an ordinary understanding
of rights is concerned, any adequate analysis of the functions of rights must be
pluralistic. All rights perform at least one function, but there is no single function
that all rights perform.
The idea of a pluralistic analysis of ordinary rights-talk is not a difficult one.
Indeed for a concept such as the concept of rights, which has been deployed in so
many different contexts through a long history, it might be thought that a plural-
istic analysis would be the assumption by default. There is no reason, after all, that
we should think that the term ‘rights’ is in this respect different from other major
normative terms that have a variety of senses, such as ‘freedom’ or ‘justice’.²⁰ The
struggle to claim each of these concepts for one ideology or another has been a
feature of political debate throughout the modern era, and indeed even in earl-
ier times. These struggles have left us with concepts stretched over a complex
of overlapping senses, instead of with concepts that mind the strictures of some
one-factored definition. Why, then, have rights theorists repeatedly presented
monistic analyses?
The answer, it seems, lies in monistic theorists’ desire to advance some contro-
versial moral or jurisprudential theory of rights. Many theorists who put forward
a theory of the nature of rights have done so not as an independent exercise in
conceptual analysis, but as a prelude to introducing a substantive theory of what

²⁰ For ‘freedom’ see Wenar (2007).

14-MHKramer-Chap14.indd 270 4/30/2008 5:16:30 PM

 The Analysis of Rights 271

rights there really are. So, for example, Steiner presents the will theory of rights as
a preparation for his left-libertarian political theory, and Raz advances the inter-
est theory as a step in the argument for his perfectionist account of social justice.
The strategy here is to use a monistic analysis of an ordinary understanding of
rights to relieve some of the justificatory burden from the substantive theory that
will follow. On this strategy, if Steiner is accused of putting forward a substantive
account of rights within which animals have no rights, he can reply that on the
best analysis of an ordinary understanding of rights it is impossible for animals
to have rights. If Raz is confronted by the objection that his political theory quite
controversially rests the justification of rights on the interests (instead of, say, on
the intrinsic dignity) of the right-holder, he can reply that on the best analysis
of an ordinary understanding of rights the function of rights is to further the
interest of the right-holder.
Monistic theories have continued to attract theorists, then, because such the-
ories are useful for supporting one or another controversial substantive theory of
rights. Within moral and political theory, the will theory has been used to sup-
port Kantian normative theories (which emphasize autonomy), while the interest
theory has been used to support welfarist normative theories (which emphasize
individual well-being). ‘Fit with the theorist’s preferred substantive theory’ has
in this way been a suppressed desideratum in presenting theories to account for
an ordinary understanding of the nature of rights. As Raz himself puts it, ‘Moral
and political philosophy has for long embraced the literary device (not always
clearly recognized as such) of presenting substantive arguments in the guise of
conceptual explorations’ (Raz 1986, 16).
This strategy is, of course, illicit. The fact that a monistic theory can be used to
bolster a controversial moral or jurisprudential theory is no reason to accept such
theories as a superior account of an ordinary understanding of rights.
Indeed the susceptibility of rights theorists to the invisible gravitational pull of
their substantive theories has contributed the continuing deadlock in the debate
over the functions of rights. The pull of such substantive theories has dragged
these theorists toward defending one of the two monistic theories as an account
of an ordinary understanding of rights. Having been pulled into these positions,
the debate over the function of rights has then become a proxy debate in the bat-
tle between the substantive (Kantian and welfarist) theories. Such a proxy debate
has made no more progress than has the debate between the two substantive
theories themselves.
As a subject of scholarly inquiry, an analysis of an ordinary understanding of
rights has its own integrity. This integrity requires that the analysis be conducted
independently of the pull of controversial substantive theories. If such an analysis
is so conducted then, and only then, it can become useful as part of an inquiry
into which substantive theory of rights is best.
For an unbiased analysis of an ordinary understanding of rights will be
useful for weighing the justificatory burden that any substantive theory of rights

14-MHKramer-Chap14.indd 271 4/30/2008 5:16:30 PM

 272 Leif Wenar
must bear. Suppose we had such an unbiased analysis on hand. Then when a
substantive theorist claimed that the set of rights that there really are or could be
differs from the set of rights that is commonly acknowledged, we could ask him
to demonstrate that his substantive theory is so compelling that the common
understanding of rights must be adjusted where there are conflicts. We might
ask Steiner, for example, to argue on the strength of his Kantian political the-
ory that the common idea that animals have rights must be abandoned. Or a
Millian might be asked to show that Mill’s reforming definition of ‘rights’ should
be accepted on the strength of Mill’s utilitarian theory, despite its incomprehen-
sion of many rights assertions that would ordinarily be regarded as innocuous.²¹
Once we have achieved an analysis of the common concept of ‘a right’, we will
be able to judge how much of our ordinary understanding of rights we are being
asked to modify by theorists who advance some substantive account of rights or
other. We will, that is, be able better to assess the proposals of those theorists who
wish (as Hart said of Bentham) to give the term ‘rights’ a new meaning. We will
be honouring Hart’s method for the conceptual analysis of rights:
Hart’s method implies, first, that conceptual analysis is a mode of inquiry that is dis-
tinct from and logically prior to substantive theory; and, second, that conceptual analysis
aims at recovering some, perhaps idealized, common understandings, in the sense that
it articulates but can never transcend the understanding already implicit in ordinary use
and reflection. (Stavropoulos 2001, 71)
A reliable assessment of the ways that people do think about rights is the only
common starting point for arguments about how people ought to think about
rights. I have argued that a pluralistic analysis of an ordinary understanding of
rights will be superior to any monistic account. If that is correct, then any sub-
stantive theory of rights will need such a pluralistic analysis as the background
against which to present its own arguments for revisions in usage.
It is not enough, of course, simply to say that a pluralistic analysis must be
the correct one. It is unreasonable to expect a single-function theorist of either
variety to give up his theory until a plural-function theory of rights is available.
There must be, that is, some place to jump. In other work I have set out an ana-
lysis of an ordinary understanding of rights, in which I argue that rights have sev-
eral specific functions (Wenar, 2005). That analysis is I believe a first step towards

²¹ In Utilitarianism Mill presented a famous reforming definition of ‘rights’ as that which one
has ‘a valid claim on society to protect [one] in the possession of ’ (Mill 2002, 54). This definition
of ‘rights’ fits very well within Mill’s normative theory: it is ‘worth it’ in utilitarian terms to protect
possession of certain things even at the cost of imposing social sanctions. However when we retain
a grip on an ordinary understanding of rights we notice that Mill’s reforming definition rejects
many seemingly innocent rights as incoherent. There would ordinarily seem nothing amiss for
example in attributing rights to people in society-less state of nature, or even in saying that every
individual has the right to be free from society’s protection. Yet neither of these ascriptions of rights
could make sense within Mill’s definition. To establish his definition of rights as the correct one, a
Millian would need to show why the attractions of his normative theory are great enough to lead us
to give up our ordinary understanding of rights in cases such as these.

14-MHKramer-Chap14.indd 272 4/30/2008 5:16:31 PM

 The Analysis of Rights 273

a theory that is adequate to an ordinary understanding of rights. It might be

(though I do not believe it so) that this particular pluralistic analysis is mistaken,
and that rights have multiple functions different than those I have suggested.
However this may be, pluralistic analyses of an ordinary understanding of rights
should become the main topic for investigations into the nature of rights. The
history and structure of the debate between the two monistic theories of the
functions of rights show that these theories will always lack adequate explanatory
power. Only a pluralistic theory can provide a sufficiently comprehensive analysis
of rights.

14-MHKramer-Chap14.indd 273 4/30/2008 5:16:31 PM

14-MHKramer-Chap14.indd 274 4/30/2008 5:16:31 PM