Pietro Maffetone Published by Florida State University Department of Philosophy

Should We Tolerate Benevolent Absolutisms?
Pietro Maffettone
Abstract: In this paper, 1 argue that the real problem with Rawls's view of international
toleration is that, properly understood, il seems not too inclusive, but not inclusive
enough. 1 examine the standing ofwbat Rawls calls "benevolent absolutisms." According
to Rawls, their lack of interna! mechanisms of collective will-formauon means that benevolent ahsolutisms cannot be seen as members in good standing of the Society of Peoples. l claim that if we accept the best reconstruCtion of Rawls's argument for tolerating
decent peoples, then The Law of Peoples does not provide conclusive reasons not to tolerare benevolent absolutisms.
Keywords: Rawls; intemational toleration; benevolent absolutisms
Accord.ing to its critics, John Rawls's account ofintemational toleration1
is excessively inclusive: tolerating decent peoples is tolerating too much.2
In this paper, I go against the thrust of this critique and argue that the real
problem with Rawls's view is that, properly understood, it seems not too
inclusive, but not inclusive enough, More specifically, 1 will take a closer
look at the standing of what Rawls calls "benevolent absolutisms."
1 John
Rawls, The Law of Peoples, Wilh "The Idea of Public Reason Revisited"
(Cambridge, Mass.: Harvard University Press, 1999); hereafter "LP."
2Rawls's critics have claimed: that tbere is no relevant analogy between the tolcration
of reasonable comprehensive doctrines at the domestic leve! and tolerating decent peoples intemationally (Kok-Chor Tan, "lnternational Toleration: Rawlsian vs. Cosmopolitan," Leiden Joumal of lnternational Law 18 (2005): 685· 710; Toleration; Diverstty. and
Global Justice (University Park. Pennsylvania State University Press, 2000); and "Liberal Toleration in Rawls's Law of Peoples," Ethics 108 (1998): 276-95); that the first ínternational original position of LP is artificially rigged in favor of nonliberal societies
(Thomas Pogge, "Do Rawls's Two Theories of Justice Fit Together?" in Rex Martin and
David A. Reidy (eds.), Rawls 's Law of Peoples: A Realistic Utopia? (Oxford: Blackwell
Publishing, 2006), pp. 206-25; "The lncoherence Between Rawls's Theories of Justice,"
Fordham Law Review 72 (2004): 1739-59; and "Rawls on lntemational Justice," Philo­
sophical Quarterly 51(2001):246-53; Charles R. Beitz, "Rawls's Law of Peoples," Eth­
ics 110 (2000): 669-96)); and that the very idea of decent people is internally inconsistent
with Rawls's approach to reasonable pluralisro, as it allows the oppressive use of statc
power (Blain Neufeld, "Civic Respect, Political Liberalism, and Non-Liberal Societies,"
Poiittcs. Philosopny & Economics 4 (2005): 275-99; Simon Caney, "Cosmopolitanism
©Copyright 2016 by Social Theoryand Praaice,Vol. 42, No. 3 (July2016): 525-554
001: I0.5840/se<:theorpracl20!6423 !S
Sbould We Tolerate Benevolent Absolutisms?
527
peoples tbat are able to endorse, for the right.reasons, the eight principies
of LP. Understood in this way, however, the structure of Rawlsian toleration suffers an important shortcoming: it seems unable to specify why ali
three criteria of decency are necessary to endorse the eight principies of
LP. While two of the criteria, respect for human rights and external
peacefulness, are explicitly or implicitly part of the very content of the
eight principies of LP, the third, interna! política! participatíon, is not. In
other words, while it is clear that a people that fails to respect human
rigbts and/or is extemally not peaceful cannot endorse the eight principies of LP, it is less clear why a people that does not allow for internal
political participation would not be able to.
Benevolent absolutisms can be precisely portrayed thus. Tbey are externally peaceful (or, at the very least, we have no principled reason to
believe that they are not) and, by definition, they respect human rigbts. If
this reconstruction of the Rawlsian argument is correct, then LP does not
offer conclusive reasons why we should not tolerate benevolent absolutisms. l consider and reject three main objections to tbis claim. The first
argues that benevolent absclutisms cannot participate in the overlapping
consensus on the principles of LP because these principies are addressed
to a specific type of política! society, namely, a "people." The second
contends that benevolent absolutisrns are not extemally peaceful because
they are not part of democratic peace. The third maintains that benevolent absolutisms cannot respect human rights "as rights."
Finally, note that the argument offered in this paper has broader implications concerning intemational toleration, not simply Rawls's view.
While the paper discusses the position of benevolent absolutisms within
the Rawlsiao framework, it also discusses the general structure of the arguments that are given to support an account of intemational toleration
more broadly. In doing so, it lends support to the view that tolerating a
political society should be based on its ability to follow norrns of intemational conduct that we deem to be acceptable from a liberal perspective.
This does not imply, as we will see below, that such norms wíll tell us
nothing about the interna! organization of a people, but it does suggest a
different way of assessing such interna! política! make-up: in terms of its
ability to sustain congruence with a liberal foreign policy.
t. Toleration in LP: A Brief Overvíew
According to Rawls, representatives of peoples, initially framed as liberal societies, would (behind an appropriately frarned veil of ignorance)
528
Pietro Maffettone
select eight principies to govern their foreigo policies. 4 These principies
largely reflect what we can describe as tbe main nonnative ideals latent
in the public política! culture of intemational society.' They take for
granted the end of colonial rule as a legitimate form of política! relationsbip berween peoples. They ascribe value to collective self-determination
but recognize the limited nature of a state's power over its citizeruy and
the link between the legitimacy of state rule and respect for basic human
rights. They incorporate a basic presumption against outside interference
in the political life of peoples, bar cases of aggressive war and interna!
repression. More controversia!Jy ( at Jeast from the perspective of intemational law), they acknowledge tbe obligations of well-ordered peoples to
aid societies who are not well-ordered because of their unfavorable social, economic, and political circumstances. This is the first part of LP's
ideal theory.
The second part extends this approach to nonliberal societies and
aims to provide an account of intemational toleration. For Rawls, toleration does not simply entail refrainíng from coercíon;6 it means respecting
those who are tolerated, and in the specífic case of LP, respectíng decent
peoples as members in good standing (bona fide) of the Society of Peoples. There are many possíble ways of organízing decent political institutions,' yet ali decent political societies share certain features. Decent
peoples: (a) respect basic human rights; (b) their foreign policy is not aggressive nor expansionist in character; and (c) they consult their members, through appropriate forms of political representation, when irnportant decisions have to be made.8 Tbese three criteria (can them "the
criteria of dccency"), when considered together, mean that the behavior
and features of a decent people meet the standards required to "override
the political reasons we mighr have for imposing sanctions."? Furthermore, according to Rawls, representatives of deceot peoples would
choose to abide by exactJy the same eight principies of international justice endorsed by liberal peoples and to do so for the right reasons, not
simply as a matter of political prudeoce. Decent peoples are thus part of
an international overlapping consensus on the very same principies of
'!bid., p. 3 7.
5Wcnar,
"The UnityofR.awls's Work."
6Rawls, LP, p. 59.
71bid., p. 64.
8Jbid
.. pp. 64-65.
p. 83. Note how this statement is ambiguous in light of this paper's discussion.
To meet the standards aUowing us "to override reasons for imposing sanctions" does not
seern the same thing as to have reasons "to respect" a política) sociery.
9Ibid.,
Should We Tolerate Benevolent Absolutisms?
529
justice that liberal peoples would select as the guide to their own foreign
policy and to structure their mutual undertakings,
2. Overlapping Consensos and the Criteria of Decency: How Are
They Related?
The admittedly brief summary of Rawls's account of intemational toleration presented above emphasizes how both the critería of decency and
allegiance to tbe eígbt principies of LP are relevant in detenníning wbích
nonliberal peoples are to be tolerated. Yet it does not clarify how these
two elements ínteract ( conceptually speaking). How are the criteria of
decency and the allegiance to the eight principies of LP to be related in
the structure of'Rawls's toleration argument?
At least three options are available. First, we can picture the criteria
of decency, by thernselves, as providing sufficient conditions for tolerating a society that meets them (call this the "decency is sufficient" option,
or DIS). Any política! society that meets these criteria should be tolerared, while allegiance to the eight principies of LP would only provide
sorne forro of independent confínnation that a society that meets these
criteria can adopt a foreign policy consisten! with tbe foreign policy of
liberal peoples. Second, we can picture the criteria of decency, again, by
themselves, as providing a set of necessary conditions for toleration. In
this picture, the criteria of decency could be coupled with the endorsement of the eight principies of LP, and then be pictured as jointly suffi­
cíent for tolerating a política! society (call this the "decency is necessary"
option, or DIN). This would entail, for example, that a political society
that meets the criteria of decency and endorses the principies of LP is to
be tolerated, but that a society that <loes not meet those criteria, and yet
still endorses the principies of LP, sbould not be tolerated. Third, we can
imagine the criteria of decency as mínima/ conditions that a society must
satisfy in order to be able to endorse the eight principies of LP and thus
be tolerated (call this the "overlapping consensus" option, or OC). This
would entail that a society that fails to meet tbose criteria cannot be described as being able to endorse the eight principies of LP. Here, note
tbat in OC, the judgment concerning whether we should tolerate a political society ultimately rests on its ability to endorse tbe eight principies of
LP rather tban whether it meets the criteria of decency. The latter, in OC,
only provide evidence of what the possible relationship between a people
that meets those criteria and the eight principies of LP is-not a benchmark for toleration. In other words, in OC, the scope of toleration is coextensive with the scope of those who are able to endorse the eigbt principies ofLP, while the criteria of decency only serve the purpose of'fur-
530
Pietro Maffettone
ther specifying the features of this particular "set" of peoples (more
this below).
011
Tbe Structure oftbe Toleratioo Argument in LP: Three Options
Criteria of
decency
Support for
the eight
principies of LP
Strueture of
the toleration
argument
DIS
Sufficient
conditions
Secondary compatibility check with liberal
foreign policy
Fully based on the
criteria of decency
DIN
Necessary
conditions
Jointly sufficient when
couplcd with the criteria of decency
Partly based on the
criteria of decency
oc
Determine relationship with principies
ofLP
Provides necessary and
sufficient conditions
Based on tbe overlapping consensus
Wbat is tbe best way of understanding tbe structure of tbe Rawlsian
account? DIS strikes me as the least convincing option. 10 To see the criteria of decency as providing, by themselves, sufficient conditions for
tolerating a política) society, independent from allegiance to the eight
principies of LP, would leave unspecified the criteria's origin and justificatory force. The criteria of decency are not the result of the application
of the Rawlsian contractualist methodology to a special class of nonliberal societies. Rawls is clear that in the case of a decent people, "there is
no original positon argument deriving the form of its basic structure.""
Justice as fairness only uses the original position three times, twice for
liberal societies and once for nonliberal but decent ones." It does not atternpt to derive the criteria of decency or any principie of interna! political organization for decent nonliberal societies from the standpoint of an
appropriately frarned nonliberal original position. It is thus unclear wbat
the justificatory force of these criteria would be if we interpreted thern as
sirnply self-standing norrnative criteria for tbe political organization of a
'ºSee also Poner, "Rawls, Reasonableness, and Intemational Toleration."
11 Rawls, LP, p. 70.
12Ibid.
Should We Tolcrate Benevolent Absolutisms?
531
people. From a liberal perspective, the features of a people specified by
the criteria of decency are clearly insufficieut as staodards of political
justice and would not represent an acceptable interpretation of the rights
and prerogatives of liberal citizenship. Tbe criteria of decency, per se,
cannot justify toleration. Tbis is especiaUy the case in the Rawlsian pie.
ture, where toleration is based on mutual respect and entails equal standing in iutemational society, not simply tbe choice of refraíning from tbe
use of coercion.
DIN and OC are probably both more plausible and closer to Rawls's
text. Recall that DIN states that the criteria of decency are necessary
couditions for toleration and jointly sufficient witb allegiance to principies of LP to determine whether a society should, in tbe end, be tolerated
or not, One may initially find DIN unconvincing because it may suggest,
paradoxically, that a liberal society could fail the toleration test. If the
criteria of decency are necessary conditions for toleration, can it be the
case that a liberal (and thus not decent) people fails to meet those criteria? No, because the criteria of decency refer to a subset of tbe rights and
liberties afforded to liberal democratic citizens; if the criteria of decency
represent uecessary conditions, tben, a fortiori, any set of cbaracteristics
containing them (such as liberal democratic rights) can plausibly be said
to meet those conditions too.13 Nonetheless, DIN still suffers from the
same basic problemas DIS. Beyond the mere intuitive fact tbat respectiog human rights, being peaceful, and allowing for política! participation-the críteria of decency-are no doubt "good things" worthy of
sorne form of moral consideration, it simply provides no clear argument
as to why .it is exactly those features that should act as a fonn of threshold for deciding wbicb polities are to be tolerated (and thus deserve equal
standing in intemational society), and which are not.
The choice of the criteria of decency as a tbreshold for toleration suggests a form of ad hoc stipulation that we may use to guide our practica!
behavior in view of the establishment of sorne fonn of modus vivendi,
rather than a clear argument to extend equal respect and equal standing to
tbose who are tolerated. For example, why should we not tolerate a political society that respects human rights and is externally peaceful but does
not allow for interna! political participation, and which nonetheless
pledges allegiance to the same principies tbat liberal peoples give to
themselves?" What would be our intuition in this case? Ifwe sense that
democratic participation is morally relevant (something I will largely as131bid.,
p. 63.
is precisely the case of benevolcnt absolutisms discussed in the following sections of the paper.
14This
532
Pietro Maffettone
sume), then we have to recognize that decent societies are not dernocratic, and that to set the bar at the leve! of a decent consultation hierarchy
does not explain why it is exactly that leve! of departure from democracy
lhat we are prepared to respect, and not something else. Put differently,
just as for DIS, DIN seems unable to explain the justificatory origins of
the criteria of decency, and this impingcs on their ability to provide a
convincing answer to the question of toleration.
Sorne wíll suggest that I am not treading carefully enough here: am 1
implying that political participation should play no role in an account of
international toleration?15 While many believe that democracy is crucial
for tolerating a política! society, and sorne, like Rawls, suggest that a
measure of (nondemocratic) political participation is necessary for tolerating a people, my argument seems to imply that no fonn of political participation is relevan! when it comes to a conception of international toleration. J disagree-this is oot what 1 am arguing. The point l aro making
is that while we have an intuitive understanding of why, hypothetically,
democratic political participation may play a role in deciding which people should be tolerated (and thus have equal standing in intemational society), it is unclear how we can fonn a weaker and yet nonarbitrary
standard of represeatativeness tbat can play the same role. Rawls simply
provides no clear argument to support the idea that a nondemocratic account of political participation can act as a necessary condition for tolerating a política! society. And 1 think that one has good reason to suspect
that such an argument is not forthcoming, especially if we define political participation as weakly as Rawls does for decent peoples, For exarnple, Rawls speaks of decent hierarchical peoples' having a decent consultation hierarchy. While such hierarchy is representative insofar as ali citizens are part of it, both the way in which they are included in the hierarchy and their ability to intluence the political process are very limited. In
a decent consultation hierarchy, persons are seen as members of a group,
and tbey are not entitled to equal representation. In the same way, the
ability to influence tbe political process is limited to tbe idea of consulta­
tion, not the opportunity to have sorne impact on political outcomes. Political participation described in this way is not, per se, very attractive,
morally speaking,
A further possibility is to see the role of political participation as what
we can call a "signaling device" in Rawls's argument for international
toleration. To illustrate, note that my discussion in the previous paragraph assumed that what makes political participation necessary are wbat
we can call its "intrinsic moral properties." But, tbere is more than one
tl¡
wish to thank an anonymous referee for pushing me to consider this objection.
Should We Tolerate Benevolent Absolutisms?
533
way in which a feature of a society, call such feature F, can be necessary
to the socíety's being tolerated. F can be necessary because tbe absence
of F is, by itse!f, morally intolerable. But it can also be the case that the
absence of F is not what triggers the failure of the necessary condition
directly, or that in any case it is not the absence of F that matters, morally
speaking, but what the absence of F entaíls. For example, call a second
feature of the society in question G, and let the conclusion that a society
should be tolerated be T. It can be the case that:
(1)..,F-.-.G
(2) .., G -. .., T
Here, ex hypothesi, it is the absence of G that matters, moraUy speakíng,
when considering toleration. Yet, note that it is still the case that ..., F implies that we should not tolerate the political society in question:
(3) given (1) and (2), we also have > F
->..,
T
However, our ful! statement of the reasons that lead us to that conclusion
would be very different dependíng on whether F is morally intolerable
per se, or simply acts as a signaling device. In this picture, the role of political participation, even if portrayed as a necessary condition for toleratíon, would be more instrumental than I have so far suggested. For example, it could be argued that it is not polítical participation per se that
matters, morally speaking, for an account of international toleration, Rather, what matters is the type of knowledge that the absence of política!
participation allows us to gain. To illustrate: in LP, the absence of political participation may tell us that the society in question is not wellordered, and, in turn, it could be the fact tbat the society is not wellordered that allows us to conclude that it should not be tolerated. In this
picture, política! participation can still be considered necessary for toleration, logically speaking; however, what is doing the work, morally
speaking, is the idea of well-orderedness of which political participation
is, as Rawls's text tells us, only but one aspect.16
This leads us to consider the following option: should we consider
well-orderedness a necessary condition for toleration? As J have just illustrated, an affirmative answer is compatible with DIN as defined
above. Nonetheless, considering this option allows us to evaluate DIN in
16More
broadly, as one of the anonymous referees has pointed out, I may bave oversimplified the notion of decency, which includes: (a) well-orderedness; but also (b) imposing bona fide moral duties; and (e) the sincere belief on the part of public oñicials tba1
the law is guided by a common good conception of justice. 1 say more about this below.
534
Pietro Maffettone
a different light, namely, through the notion of well-orderedness. My reply here is simply that to consider well-orderedness as necessary for toleration has implausible implications, namely, that we can imagine liberal
political societies tbat are not to be tolerated. In order to see why, let's
start with Rawls's understanding ofwell-orderedness. Rawls tells us that
a society is well-ordered when:
(a) Everyone accepts, and k.nows that everyone else accepts, the same
conception of justice;
(b) The society's basic structure is publicly known, or with good reason
believed, lo satisfy those principies ofjustice;
(e) Citizens have a normally effective sense of justice, which enables
them to understand and apply the publicly recognized principies of
justice and act as those principies require.17
Wíth this definition in mind, imagine a liberal society, call it L, that
has the following features:
(i) L protects the basic liberties of the liberal democratic tradition;
(ii) The conception of justice J that is used to provide a benchmark for
the organization of L's basic structure is rcasonably rejected by at
least sorne of the citizens in L because its justification rests on some
form of eomprehensive understanding of liberalism;
(iii) Frorn (i), (ii), and (a) above, we can conclude that L is a liberal society but not a well-ordered one;
(iv) Oiveo (iii), and taking well-orderedness to be a necessary condition
for toleration, we can conclude that L sbould not be tolerated.
As I said, I believc that the conclusion in (iv) is implausible. One way to
interprct Rawls's view of well-ordercdness is to say that a well-ordered
political society is an ideal: the ideal of political society as a form of justificatory cornmunity where citizens are able to draw on cornmon normative resources in an exercise of mutual reason-giving. The example of the
liberal and yet not well-ordered society simply tells us that the standard
way of describing Rawlsian liberalism as "justificatory liberalism'' is
more complex than we often realize insofar as the "liberal" and the "justificatory" may very well come apart. Indeed, the very definition of a deceot people suggests that mucb: not all well-ordered societies need to be
liberal. Wbat is less intuitive is that not all liberal societies need to be
17John
Rawls, Justice as Faimess: A Restatement (Cambridge, Mass.: Harvard Uni-'
versity Press, 2001 ).
Should We Tolerate Benevolcnt Absolutisms?
535
well-ordered,
Furtherrnore, note that the moral attractiveness of the ideal of a justificatory comrnunity cannot be assessed without sorne knowledge of the
underlying conception of the person and of the content of public reason
in such a community. The moral quality or character of a justificatory
community depends on how it portrays those who are part of its exercise
of mutual argumentation and on the types of reasons that we deem to be
norrnatively acceptable within such a framework.18 In a liberal political
society, individuals are seen as free and equal, and the types of reasons
that they give to one another are reasons that they believe others can reasonably accept as free and equal. This constitutes an important part of
wbat makes tbe liberal model of justificatory community morally attractive, not simply the fact that it is a justificatory community.
FinaJly, note that 1 am not claiming that the idea of well-orderedness
is morally irrelevant. Por example, Rawls tells us that a conception of
justice tbat cannot be used as a public standard of justification in a wellordered society must be considered "seriously defective;"" while otbers
such as Jeremy Waldron-" and Stepben Macedo21 tell us that the idea of
well-orderedness can be morally signifícant in providing mutual assuranee between citizens conceming their shared nonnative commitments to
each other. These are points that I do not wish to deny. However, they
are compatible witb assigning a variety of roles to tbe idea of wellorderedness as an element of the judgment conceming the tolerability of
(and more broadly judgments concerning the moral character of) a political society. For instance, one could see tbe idea of well-orderedness as a
"counting principle"22 or as a "good-making feature"-an element that
we believe allows us to feel more secure wben assessing the moral character of a political society and whether we should tolerate it, but not one
tbat we believe should act as a necessary (or sufficient) condition for
such assessments.
Given the problems highlighted in DIS and DIN, Jet us move on to
OC. Adopting OC, we see allegiance to the eight principies of LP as the
18Here
see Rawls's distinction between the general and particular meaning of wellorderedness (ibid., pp. 8-9).
191bid. Presumably, because such a conception ofjustice cannot be stable for tbe right
reasons given reasonablc pluralism.
20Jeremy Waldron, "What Does a Well-Ordered Society Look Like?" http://www.law
.nyu.edu/sites/default'files/ECM_PRO _ 063313.pdf.
21Stephen Macedo, "Why Public Reason? Citizens' Reasons and the Constitution of
the Public Sphere," http://papers.ssm.com/sol3/papers.cfin?abstract_id= 1664085.
22John Rawls, A Theory o/ Justice, revised ecl (Cambridge, Mass.: Harvard University Press, 1999), p. 364.
536
Pietro Maffettone
detennining feature of Rawls's toleration argument. Yet OC also places
clear emphasis on the criteria of decency, depicting them as minimal
conditions that a society must meet in order to be able to endorse the
eight principies of LP and thus be tolerated. Put differently, OC is able to
combine two importan! elements of a plausible reconstruction of the
Rawlsian view. On the one hand, it does not irnply that the criteria of decency, by themselves, are able to settle the question of internatíoual toleration;" as we have seen, this is not a tenable view. Nonetheless, OC
still grants the criteria of decency an importan! role. It depicts them as
minimal conditions that any fonn of political society must meet for there
to be a plausible claim that it can endorse the eight principies of LP. lo
OC, the criteria of decency provide the widest possible net we can casi to
ensure that the political societies that meet those criteria will also be able
to endorse the eight principies of LP.
The rnost important feature of OC is that the scope of roleration is coextensive with the ability of a people to endorse the eight principies of
LP. Should allegiance to the principies of LP count for so much in view
of Rawls's account of international toleration? lt should. First, imagine
there were further requirernents other than respecting the principies of LP
for the right reasons. Why wouldn't such requirements be part of the
principies themselves in the first place? The principies that regulate a
scheme of social cooperation are used to specify the terms that participants to the scheme can ask one another to abide by. lf there were other
requirements, then those requirements would bave to feature in the principies that regulate the scheme. And if they do not, and participants respect the demands that are contained in the principies, 24 nothing more
can reasonably be asked of them.25 Put differently, what would be the
grounds for not tolerating a society that respects (for the right reasons)
exactly the same conception of international justice that liberal peoples
themselves give to one another? The eight principies of LP are initially
developed precisely to provide a nonnative benchmark for liberal peoples' foreign policies. They are the conception of international justice
that liberals themselves decide to give one another. But then, why should
a nonliberal people abiding by the same rules and principies not be toler23By
this 1 mean that in adopting OC, the criteria of decency do not seem to be pri-
marily concemed with the scope of toleration, but rather with giving conteo! to panicular
set ofpeoples-those who can endorse the eighr principies ofLP.
241ncludiog the fact that their support for the principie be of the right kind, or offered
for the "right reasons." Sorne will add that there are requirements relate-O to public reason.
J consider this point below.
is¡ would like to thank one of the referees for pushing me on this point and for suggesting this line of argument.
Sbould \Ve Tolerate Benevolent Absolutisms?
537
ated? Wouldn't it be inconsistent for liberals to argue that peoples who
are prepared to respect, for the right reasons, exactly the same norrns
they think should organize intemational society are not to be tolerated
because they are not intemally liberal=-somethiug that is not required by
Rawls's liberal conception of international justice? Furthermore, liberal
peoples cannot claim decent ones are (intemationally) unreasonable.26 To
be reasonable means to be prepared to offer, and abide by, fair terms of
cooperation, and to recognize the equal standing of tbose witb whom we
cooperate.27 According to OC, decent peoples are prepared to honor a
liberal conception of intemational justice, and by honoríng such a conception tbey explicitly accept that other well-ordered peoples are free,
equal, and independent.28
3. Three Objections to OC
There are at least three objections to OC as 1 bave outlined it above.29
First, why have criteria of decency at all if the scope of toleration is ultimately decided in terms of a society's ability to endorse tbe eight principies of LP? Are criteria of decency redundan! after ali? Even accepting
26See
Porter, "Rawls, Reasonableness, and International Toleration."
Boencher, "What is Reasonableness?" Phiíosophy & Social Criticism 30
(2004): 597-621.
28The domestic account of reasonableness also features the readiness to accept the
burdens of judgment. To wbat extent this requirement applies to collective agents is unclear, but if it does, we have no basis to claim that deccnt pcoples would not accept iL
29Some will suggest a fourth objection, namely, that OC cannot be convincingly supported by Rawls's text, For example, in one of tbe most important passages in LP, wben
Rawls outlines the basic reasoning tbat grounds bis view of toleration, be writes that
"provided a nonliberal society's basic institutions meet certain specified conditions of
political rigbt and justice ami lead its people to honor a reasonable and just law for tbe
Society of Peoples, a liberal people is to tolerate and accept tbat society" (pp. 59-60; my
emphasis). The latter quote is consistent with DIN, and seems to suggest precisely what
OC rejects, namely, tbat what Rawls here calls "cenain specified conditions of rigbt and
justice" and include wbat 1 have called the criteria of decency, do play sorne independent
role in the argument for ioleration. My reply is that even accepting tbat the passage is
closer to DIN, 1 believe that the text can be pul in a more positive light as a wbole, if interpreted along the lines r have suggested. Finally, note that Rawls's first version of LP,
his 1993 Oxford Amnesty Lecture (see his Collected Papers, ed. Samuel Freeman (Cambridge, Mass.: Harvard University Press, 1999), pp. 529-64), features tbe same passage
but with an importan! grammatical change tbat aliers the structure of the argument. There
Rawls writes that "a liberal society must. respect other societies organized by cornprehensive doctrines, provided their política! and social instinnions meet certain conditions that
lead the society to adhere to a reasonable law ofpeoples" (p. 530; my emphasis). Replacing "and" with "that" cbanges the nature of the claim even in tbe specific passage under
consideration and makes tbe latter closer to OC.
27James
538
Pietro Maffettone
the primary role of the eight principies of LP, the principies themselves
cannot give us a concrete picture of a kind of people that would be a ble
to endorse them. This seems particularly important for Rawls's account
of international toleration, because it is precisely by specifying the conteo! and meaning of the criteria of decency that we are able to give shape
to the possibility that a nonliberal people that endorses the eight principies of LP can in fact exist. Neither the fact that such peoples exist, nor
what they could look like, can simply be inferred from the eight principies themselves. The criteria of decency, then, give shape to the scope of
intemational toleration in a way that the principies of LP do not. They
provide content to our picture of a decent society, and thus allow us to
confirm in reflective equilibrium that our argument for international toleration is sound. By providing an understanding of how citizens of a decent people are treated by its govemment and institutions, the criteria can
show us if the general principies that guide our account of international
toleration are compatible witb our considered convictions about minimal
human welfare and dignity. The criteria of decency, then, do not provide
reason enough to extend equal standing to decent peoples; rather, they
provide reason enough to allow that an argument that does is not necessarily to be rejected. Furthermore, note how this picture is also consisten!
with my comments conceming the role of well-orderedness in our assessments of a political society. Jf, as l have suggested above, we consider tbe idea of well-orderedness as a counting principie or a good-making
feature, we can use the fact that decent peoples are well-ordered (including sorne of the other features that Rawls ascribes to decent peoples) as
ways of confirming our judgment, in reflective equilibriurn, tbat we
should tolerare a decent society that respects the eight principies of LP
for the right reasons.
Second, it can be claimed that accepting OC confuses the scope of
toleration with the scope of public justíñcarion." In other words, OC assumes that the scope of toleration is co-extensive with the scope of public justification insofar as the object of toleration, decent peoples, are also depicted as members in good standing of international society and as
participants in the justificatory process of the principies of international
justice. This is a víew that Rawls himself rejects at the domestic leve!.
For example, while unreasonable citizens of a liberal democratic polity
are not part of the constituency of public justifícation,31 they are nonethe30See
Jonathan Quong, Liberahsm Without Perfection (Oxford: Oxford University
Press, 2011 ), pp. 297, 298 n. 26.
31Though sorne ihink they should be: Erin Kelly and Lionel McPherson, "Oo Tolerating the Unreasonable," Journal ()f Poliucal Philosophy 9 (2001): 38-55. 1 will remain
Sbould \Ve Tolerate Benevolent Absolutisms?
539
less part of the constituency of toleration insofar as they are afforded, bar
exceptional circumstances, the same rights and prerogatives of reasonable citizens, l think this is an important point, but ultimately 1 believe it
signals sorne fonn of asymmetry between the domestic and international
conceptions of toleration rather than problems with OC. Simply put, at
the international leve) it seems clear that Rawls believes tbat the constituency of toleration and the constituency of public justification are coextensive.
This can be defended both exegetically aod conceptually. The fonner
defense is twofold: Rawls is clear that decent peoples are equal members
ofthe Society of Peoples; and that they too participate in the justificatory
process for the eight principies of LP through a second run of the international version of the original position.32 These two features seem to imply that the eight principies of LP do not simply "apply" to decent peoples, but that the latter are also part of the international constituency of
public justification.33 However, the exegetical justification alone is insufficient. The reader is entitled to ask: What justifies this asymmetry beyond mere textual support? Why should there be a difference between
the way a liberal theory deals with "unreasonableness" domestically and
"decency" intemationally? The explanation is to be found in the way in
which liberal peoples themselves articulate their mutual undertakings in
the first intemational original position of LP. As 1 have already noted,
liberal peoples endorse the eight principies of LP as their liberal conception of intemational justice. Yet liberal peoples also realize that decent
peoples can endorse these very same liberal principies of intemational
justice too. Liberal peoples are thus confronted with agents tbat, while
not internally liberal, are still prepared to abide by liberal international
principies. But, according to liberal political morality itself, liberals are
required to provide a justification to nonliberal agents who are willing to
comply (for the rightreasons) with a liberal conception of justice." Thus,
the asymmetry (i.e., the fact that decent, not fully reasonable, societies
are part of the constituency of justification intemationally while unreasonable citizens are not part of such constituency domestically) is based,
conceptually, on the fact that the eight principies of LP are permissive
enough to allow even intemally nonliberal societies to endorse liberal inagnostic on this point. Wbat 1 want to claim is that, even accepting Quong's argument,
there seem to be good reasons to think that LP deals with this problem differently.
32Rawls, LP, pp. 68-70.
33See
also Porter, "Rawls, Reasonableness, and lntemational Toleration." Avila, "Defending a law of Peoples," however, offers an argument that could support a different
conclusion on this point.
34Here 1 follow Poner, "Rawls, Reasonableness, and Intemational Toleration."
540
Pietro Maffettone
temational principies for the right reasons."
The third objection to OC relates to the explanation I have provided
as to why decent peoples should be par! of the constituency of justification at the intemational level. Why are liberal peoples prepared to agree
to the eigbt principies of LP instead of a more demanding set of principies that requires ali peoples to be intemally liberal? One may legitimately suspect that the only reason for doing so is to artificially alter the conclusions of the first (liberal) intemational original position witb tbe aim
of accommodating decent peoples.~ This would be problematic, because
the princíples of LP would not constitute a liberal conception of international justice, but a conception of inremational jusrice already geared to
reflect a concem for intemational toleration. If this were to be the case,
allegiance to the principies of LP would presuppose an account of toleration and would thus not be able to provide one (as OC suggests).
The reply here is that the content of the principies of LP and the structure of the account of toleration l am defending are, strictly speaking, orthogonal to one another.37 The content of the principies only affects the
scope of international toleration. Put differently, imagine that Rawls is
wrong about what representatives of liberal peoples would agree to in an
appropriately framed intemational original position. Instead of the eight
principies of LP, representatives of liberal peoples would choose a much
more demanding set of fully liberal principies. lf tbis were tbe case,
clearly we would not be able to tolerate decent peoples. However, the
correctness of OC is unaffected by this conclusion. OC only specifies tbe
reason for not tolerating these societies, narnely, that they would not be
able to support the allegedly correct specification of a liberal foreign policy for the right reasons,
Nonetheless, I still want to say something about the content of the
principies. The reason for doing so is tbat in the next part of the argument, I will challenge the Rawlsian account of the scope of toleration for
not being broad enough. Claiming tbat it should be even narrower than
Rawls suggests is something that would affect the conclusions reached at
35Note
that this is an unusual problem given the Rawlsian concern with stability in
liberal democratic socíeties. In the domestic case, Rawls investigares the possibility of
overlapping consensus given the fact of reasonable pluralism, Instead, in LP, we tend to
approach things frorn the opposite direction, so to speak, and reflect on our altitudes toward nonliberal societies given their allegiance for the right reasons to a liberal conception of justice.
.
l6pogge, "Do Rawls's Two Theories of Justice Fit Together?"; "The locoherence Between Rawls's Theories of Justice": and "Rawls on lnternatíonal Justice": Beitz,
"Rawls's Law of Peoples."
37[ want to thank an anonymous referee for pushing me to address tbe third objection
to OC more thoroughly.
Should \Ve Tolerate Benevolent Absolutisms?
541
the end of tbe paper. The standard reason used to defend the content of
the principies of LP without claiming tbat they are already implicitly
aimed at tolerating nonliberal societies is to recall tbat the rights protected by a liberal constitutional regime are already guaranteed within tbe
different domestic jurisdictions of liberal peoples.38 In the first original
position in LP, representatives of peoples know that they are representatives of liberal peoples. Thus they know tbat the main aspects of a liberal
democratic regime are alreadypart of the domestic basic structures of the
peoples they represent. What would be the gain in duplicating tbese features at the international level? I think that this standard reply is, as it
stands, incomplete. To see why, note that its soundness depends on
whether we take representatives of liberal societies to be aware of the
fact that there is goiog to be a second run of the intemational posítioo
featuring nonliberal societies. If representatives of liberal peoples know
this, then it is unclear why they would want to take for granted that liberal rights are guaranteed within ali the relevant domestic jurisdictions. So,
should the representatives of liberal peoples know about the second run
of the intemational original position featuring nonliberal peoples? While
I cannot provide a full argument here, my intuition is that they should. If
representatives ofliberal peoples know that they represent a specific type
of society, then it seems reasonable to believe that they will also know
that there are other types of societies. Otherwise, specifying that they
represent liberal peoples makes little sense.
There is, however, a second type of argument for establishing the
plausibility ofweakening the content of the principies. The difference
with the first argument is that tbe latter, unlike the former, is unaffected
by the knowledge of the existence of decent peoples. Liberal peoples realize that reference to the basic internal features of tbeir political institutions (beyond the requirernent that they respect human rights) would essentially give otber liberal peoples greater say over tbeir domestic affairs-it would turn their constitutional politics into a matter of international concern, and thus curtail one of their defining interests as liberal
peoples, namely, política! self-determination.'? Consider the implications
for liberal peoples' self-determination ofthe inclusion in the principies of
LP of the basic liberties covered by the first principie of justice as faírness, If we included these liberties in LP, the way in which they are implemented, the interpretation of their concrete meaning, and how they are
38See
also Samuel Freeman, Justice and the Social Contraer Essays on Rawlsian Po­
litical Philosophy (Oxford: Oxford University Press, 2007), p. 276; and Rawls (New
York: Routledge, 2007).
39Rawls, LP, p. 29.
542
Pietro Maffettone
entrenched in the domestic constitution of a liberal people would ali be
transformed into matters of international concern and partiaHy removed
from the democratic process of each liberal people. To illustrate, in A
Theory of Justice. Rawls is clear that the first principie of justice can only provide broad guidance on how the basic liberties to which it refers
sbould be adjusted to cohere into a single scheme. In tum, such adjustment sbould track the specific conditions of a given society and cannot
be sirnply specified ex ante at the intemational level." For example, one
can conjecture that different basic liberties may have comparatively different values according to prevailing cultural, social, economic, and political conditíons in different societies. In the same way, when Rawls discusses the curtailment of basic liberties, he states that it may happen only
for the sake of the basic liberties themselves.41 Yet, decisions concerning
the curtailment of basic liberties require sorne sensitivíty to local circumstances, It would not be irnplausible to think that a liberal people that
faces recurren! terrorism could decide to put in place more restrictive
controls on political speeches supporting the terrorists, and that sucb controls sbould (in tbis specific case) ultimately track the collective deliberations of tbe particular public in question, not those of the members of the
Society of Peoples. In short, a version of LP requiring intemally liberal
política! institutions would largely duplicate what representatives of liberal peoples take for granted (that they ali represen! liberal peoples) and
would also negatively affect one of the defining interests of those they
represen! (liberal peoples' interest in self-determination).
4. A Final Comment Concerning OC
1 want to conclude tbe discussion of OC by illustrating a clear advantage
that OC has when compared to DIN once we recall sornething that is
crucial about any account of tolcration.42 As David Heyd famously put
it, toleration is an elusive virtue. More than that, following Bemard
Williams, we can say that toleration has an almost paradoxical character,
since it effectively asks us to give normative authority to principies and
ideas (I use the terminology quite loosely here) that we deem to be
morally wrong or defective. Of course, we can have prudential or
"°Rawls, A Theory of Justice, pp. 52 ff.; Thomas Pogge, John Rawls: His Lije and
Theoryof Justice (Oxford: Oxford Univcrsity Press, 2007), pp. 85 ff.
41Rawls,
A Theory of Justice.p. 54.
42Catriona
McKinnon, Toleration:A Critica/ Introduction (Abingdon: Routledge,
2006); David Heyd, Tolera/ion: An Elusive Virtue (Princeton: Princeton University Press,
1998).
Should We Tolerate Benevolent Absolutisms'?
543
skeptical reasons to tolerate an agent or doctrine.43 However, if we understand toleration as compatible with mutual respect and equal standing,
as Rawls does, we clearly have to confront the tension: how can we possibly respect and bestow equal standing to an object that we deem, ex
hypothesi, to be seriously defective? I think that once we accept this tension, it becomes clearer why, among other reasons, OC is an attractive
alternative to the standard interpretations of the Rawlsian conception of
intemational toleration. OC effectively provides a specific strategy to
defuse the tension 1 have just highlighted. Substantively, it does so by
allowing Iiberals to be consistently faithful to their liberal commitments.
Conceptually, it does so by providing two distinct standpoints from
which to assess nonliberal political societies, a domestic one and an
intemational one.
Let me start with what l have called the substantive part of the strategy. Liberal peoples are, by definition, intemally liberal. Furthermore,
when it comes to their foreign policies, the principies of justice they subscribe to are, as Rawls tells us, the principies that representatives of liberal peoples wou1d endorse as a reasonable interpretation of what a liberal foreign policy requires. However, liberal peoples realize that, for reasons that are connected to their desire to protect their (liberal) selfdetermination, there is a "gap" between their interna) views about how a
política) society should be structured and tbe principies of justice for international relations they endorse. In tum, this gap allows them to conjecture that there could be nonliberal political societies that wbile not internally liberal could still subscribe to a fully liberal conception of ínternational justice. In this picture, our assessment of nonliberal and yet decent well-ordered political societies is not what guides our views about
toleration. lnstead, we see decent political communities as equal members in good standing of intemational society sirnply because they follow
exactly tbe same type of political principies we believe are correct from a
liberal standpoint. This <loes not entail, of course.• that principies of political justice for a liberal law of peoples should necessarily be silent on the
intemal political make-up of those who participate in the overlapping
consensus. And fuere is certainly room for a process of reflective cquilibrium in which the principies, and what endorsing them for the right reasons allows a people to do and to be internally, adjust to one another in
light of what we find acceptable ali things considered. However, it is irnportant to stress that from the perspective of OC, liberal and decent peoples are equals before the law of peoples.
Second, let's consider what 1 have called the conceptual part of the
43$ee
McKinnon, Toleration.
544
Pietro Maffettone
strategy. Sorne wi 11 object tbat tbe last sentence in tbe previous paragraph
suggests thar decent peoples are fully just and thus jusi equal to liberal
peoples. This is oot tbe case. From tbe perspective of a liberal political
cooceptioo of social and distributive justice, they clearly are defective.
Indeed, this is precisely what creates the tension that l mentioned in the
beginning of this section. lnstead, we should think of decent nonliberal
peoples as "objects" that take on different shapes depending on the
standpoint we use to look al tbem: from tbe standpoint of a liberal conception of domestic social justice, they are but a pale shadow of what we
think is just, and yet looked at from the standpoint of a liberal foreign
policy, they endorse the same liberal principies we (liberals) think are
just and reasonable. In turn, this idea invites the following question: Why
should the fact that decent peoples comply with a liberal law of peoples
be sufficient and oot be overriddeo by the fact that decent peoples are internally morally defective? Tbe answer is that it will depend on what we
take to be the balance between tbe values tbat are in play. And such values
can be liberal ali the way down: tbey tell us something about the type of
"price" we are prepared to pay to protect the political self-determínation
of liberal democratic citizens. Sorne may believe that such a price, given
how weak tbe eight principies of LP are, is set too high by Rawls. However, simply accepting tbat tbere is at least a price worth paying creates
the conceptual space for considering whether or not nonliberal peoples
can in fact adopta liberal foreign policy, and sheds light, if this happens,
oo why we should tolerare them.
S. Tolerating Benevolent Absolutisms?
1 have argued that the best option for conceptualizing the strucmre of roleration in LP is to see it as co-extensíve witb an overlapping consensus
on the eight principies of LP. In tbis picture, the criteria of decency are
conceptualized as minimal conditions tbat a society must meet if it is to
·be able to endorse the eigbt principies of LP. In the following sections, 1
argue that accepting the account of toleration 1 have sketched weakens
the case for not tolerating benevolent absolutisms. In short, this is because in adopting OC, we cannot fully explain why (ali) the criteria of
decency are required for supporting the eigbt principies of LP. More specifically, recall thar we have identified three criteria of decency: (a) respect for human rights; (b) a nonaggressive foreign policy; and (e) a system of interna) political participation. While the first two criteria of decency are explicitly or implicitly required by the eight principies of LP,
Should We Tolerate Bencvolent Absolutlsms?
545
the third one, interna] political participation, is not. The eight principies
of LP44 require tbat well-ordered peoples respect human rigbts (principie
6); wbile principies 4 and S articúlate a plausible definition of what a
nonaggressive foreign policy is. If we see the criteria of decency as requirements for a people's allegiance to the eight principies of LP, tben a
people's respect for human rigbts and its having a nonaggressive foreign
policy are features that are either explicitly or implicitly required by the
very content of the eight principies of LP; but the fact that a decent people must provide for internal political participation is not.
Yet, if we drop the requirement of political participation from the criteria of decency, the scope of intemational toleration changes. In LP,
Rawls states that benevolent absolutisms cannot be members in good
standing of the Society of Peoples because, while they respect the human
rights of their citizens, .the latter "are denied a meaningful role in making
political decisions."? But, given the overall strategy on toleration offered
by OC above, and unless the eight principies of LP themselves explicitly
or implicitly require it, it is unclear wby the fact that benevolcnt absolutisms do not pennit meaningful political participation would imply that
thcy are not to be tolerated. Why should the lack of interna! política! participatioa detenniae the chances of benevolent absolutisms being able to
endorse the cight principies of LP? Three arguments can be made against
the assertion that a proper understanding of the structure of the toleration
argument in LP atlows for the inclusion of benevolent absolutisms. One
could argue that benevolent absolutisms cannot endorse the eight principies of LP, which are addressed to peoples, because they are nota "people." One conld argue that benevolent absolutisms cannot endorse the
eight principies of LP because they are not extemally peaceful, while LP
requires ali peoples to be nonaggressive in their foreiga policy. Finally,
one could argue that benevolent absolutisms cannot respect the principies
of LP because, given their interna! política! make-up, they cannot secure
human rights as rights, In the following three sections ofthe paper I explore and refute ali three argumentative strategies.
6. A Law Far "Peoples"?
In LP, Rawls uses tbe term "peoples" only for (decent or liberal) wellordered political societies, Furtbennore, the eight principies of LP are
explicitly addressed to peoples, not outlaw states, burdened societies, or
44Forthe
wording ofthe eight principies, see Rawls, LP, p. 37.
<S]bid., pp. 4, 63.
Pietro Maffettone
546
benevolent absolutisms. A possible objection to the inclusion of benevolent absolutisms in tbe scope of intemational toleration, therefore, is that
they cannot endorse the eight principies of LP because the principies are
addressed to a different type of agent, namely, "peoples." This argument
is at best inconclusive. To call one política! society a "people" and another a "benevolent absolutism" will not by itself determine the scope of
toleration. If we attach a label to a type of agent, and if attaching the labe! itself implies not tolerating tbe agent in question, then we are bound
to provide a nondescriptive argument as to why the label has been allocated in this way, and such an argument has to provide independent normative grounds for not tolerating the agent in question. What is tbe reason for attaching the label "peoples" to a political society? Initially,
Rawls adopts the terminology to distinguish peoples from states. The latter are rational actors, while the fonner are also reasonable and thus have
a moral nature." Yet, to have a moral nature (in this specific context)
simply suggests the ability to restraio ooe's pursuit of rational ínterests
througb a reasonable law of peoples. Given tbat we are trying to establish
whether benevolent absolutisms can in fact follow the principies of LP, it
would be circular to exclude them because we assume that they cannot.
Alternatively, Philip Pettit has suggested that the term "peoples" as
used by Rawls can only be employed to describe a political unit if there
is a specific type of representative relationship between the rulers and the
ruled.47 That may be tbe case, but note that, as Pettit recognizes, Rawls
(mostly) develops this idea when descríbing liberal societies, not all
well-ordered societies generally.48 Therefore, accepting this line of argumeot, we may have to question why Rawls extends toleratioo to decent
peoples. Furtherrnore, even if we accept Pettit's argument about what a
people "is" and allow the term to apply to both liberal and decent peoples, the ascription of the tenn does not provide an argument for deciding
the scope of international toleration. As we have just seen, ·defining what
a people is, even in tenns of a morally praisewortby characteristic such
as interna! political participation, cannot per se explain why we should
not tolera te those who are not "peoples" or why a conception of international justice that will implicitly decide the scope of international tolera461bid.,
p. 29.
47Philip
Pettit, "Rawls's Peoples," in Martin and Reidy (eds.), Rawls's law o/ Peo·
ples, pp. 38-56; and "Legitimare lntemational lnstitutions: A Neo-Republican Perspective," in Samantha Besson and Jobn Tasioulas (eds.), The Philosophy o/ International
Law (Oxford: Oxford University Press, 201 O), pp. 139-62, at p. 154.
'8Pettit's text ("Legitimate International Institutions," p. 154 nn. 24-25) references p.
24 and p. 38 in LP: in both instances Rawls is outlíning the relationship between a people
and its govcmmeni in a liberal democratic rcgirne.
Should We Tolerate Benevolent Absolutisms?
547
tion should be addressed to "peoples" as defined by Pettit.
A further argurnentative strategy would be to ground the distinction
between well-ordered political societies ("peoples") and non-wellordered ones on the norrnative ideal of genuine social cooperatioa. Rawls
distinguishes between genuine systems of social cooperation organized
by a liberal ( or common good) conception of justice featuring democratic
procedures (ora decent consultation hierarchy) and social systems based
on comrnand by force (a putative description of benevolent absolutisms)." Unfortunately, this line of argument does not even get off the
ground, since Rawls is clear that it is the violation of basic human rights
that qualífies a form of political rule as a systern of cornrnand by force,
not intemal political participation/" However, benevolent absolutisms do
respect the human rights of their members (but see below for discussion
of this point).
Fiaally, one may wish to ground the distinction between peoples and
other typeS of political society on the difference between social cooperation and "merely coordinated social action. "51 Social cooperation is not
an "activity coordinated by orders issued by an absolute central authority"; rather, social cooperation, for Rawls, is "guided by publicly recognized rules and procedures which those cooperating accept as appropriate."52 This is a more promising line of inquiry, as the language employed may suggest that social cooperation requires sorne form of consultation procedure if it is to avoid being described as merely coordinated
social action. However, once again, this strategy would need to be supplemented by an argument that explains why the scope of international
toleration should be determined by the distinction between social cooperation and rnerely coordinated social action. That a society displays a
morally relevant feature that another lacks does not, per se, tell us much
about toleration. A further argument is needed, one that explains the relationship between social cooperation and toleration. One possibility is that
describing a group as a form of social cooperation makes it appropriate
to consider the leadership of that group as legitimare, and thus in a posi49
Reidy, "Cosmopolitanism: Liberal and Otherwise"; Reidy, "Rawls on lntemational
Justice."
50Rawls,
U', p. 68.
Justice as Faimess: A Restatement, p. 6.
52Ibid.
The idea of social cooperation also includes a reference to fair tenns of cooperation that each person can reasonably accept anda reference to each person's rarional
good. 1 do not discuss these features, because 1 see them as less salient in the present context, The first may be effeetively violated by decent peoples given that decent citízens are
not seen as reasonable within the common good conception of justice. The second is a
weak requirement plausibly met by bcnevolent absolutisms,
51Rawls,
548
Pietro l\·taffettone
tion to impose genuine duties on its members." However, even if successful, this argument would ground a duty not to preven! citizens a "le.gitimate" política! society to follow their prima facie moral obligations to
comply with the duties imposed by the leadership of their country, not
one to respect them as equal members in good standing of international
society. Prima facie, the duty not to prevent compliance is compatible
with endorsing noncoercive measures (for example, public criticism and
the use of positive incentives) meant to influence the interna! political
lífe of decent societies (both of which are incompatible, according to
Rawls, with toleration as understood in LP). This suggests that the legitimacy of the leadership is not sufficient to justify what is required by the
conception of toleration in LP. Conversely, even a society whose leadership does not impose genuine duties on its members may rule according
to principies and ideas that are morally acceptable ali things considered,
The leadership of such a society may not have a right to rule, and citizens
may not havc content-independent reasons to obey the law. However .•
conceptually, the possibility is still open that the content of the laws is
such that citizens bave reasons to comply with them despite the fact that
these are not strictly speaking legitimare." This implies that to sce the
ability to impose germine duties on citizens as a necessary condition for
toleration is too strict a requirement for toleration.
1 would like to thank an anonymous referee for pushing me to consider this option.
For example, David Reidy has recently argued tbat one reason to tolerare decent peoples
is tbat their interna! political structure is "a system of cooperation between persons governed by public right" capable of imposing genuine duties. See Reidy, "Cosmopolitanism: Liberal and Otherwise," p. 181. 1 should note that 1 cannot possibly do justice to the
complexity of Reidy's argument given the space available. The referee also suggests that
the idea of social cooperation may be instrumental to claiming that a group can be considercd a collective ageot. 1 am not convinced by Ibis claim. Most of the literature on collective agents (both intentional and action-based accounts) does not mention the idea of
representation as a necessary (or sufficient) condition. For a recent example, sce Margaret
Moore, A Political theory of Territory (Oxford: Oxford University Press), pp. 46 ff. Representativcness may play a larger role when it comes 10 collective responsibility rather
than collective agency.
54This option is related to the idea, discussed above, of a non-well-ordered and yet
liberal política! society. For example, we can imagine a dccply content-liberal political
society lacking appropriate fonns of democratíc participation and where the use of political power is justified by nonpublic (and yet fully content-liberal] reasons, Note that tbe
possibility of such a view is not completely remete. For example, if we shifted, for the
sake of argumenr, to a consent-based account of legitimacy, we would find rcal-lifc supporters for it: indeed, this seems to be the auitude of sorne philosophical anarchists to·
ward contemporary liberal democracies: sce A. Jobn Simmons, Justificationand Legiti­
macy: Essays on Rights and Obligations (Cambridge: Cambridge University Press,
2000).
53
Should \\le Tolerate Benevolent Absolutisms?
549
7. Benevolent Absolutisms and Human Rights as Ríghts"
Rawls provides a relatively short list of (basic) human rights;56 however,
respecting them is clearly a necessary condition to be a mernber in good
standing of the Society of Peoples, no matter which account of toleration
among the ones 1 have discussed one subscribes to. In LP, human rights
set a precise limit to the scope of interna! political autonomy of a people
aud explicitly feature in the eight principies of justice that both decent
and liberal peoples endorse. Can benevolent absolutisms respect human
rights? At first sight, the question invites a relatively simple answer: they
can because this is how Rawls defines them. Unfortunately, this answer
is not very interesting for two reasons. First, it salves the conceptual
question by definitional fiat. Second, and more subtly, because there is a
conceptual gap between the two following ideas: (a) a political society
!hat honors the substance of basic human rights; (b) a political society
that secures human rights as rights.57 The gap can be illustrated as follows: while in both cases no form of outside intervenrion seems to be
justified (at least if one follows Rawls and believes that human rights
have this function), only the second type of society provides reliab/e as­
surance lhat it can effectively protect human rights and endorse the principies of LP for the right reasons. The conceptual gap can be illustrated
by the following argument:
(i) benevolent absolutisms honor the substance of human rights;
(ii) benevolent absolutisrns do not secure human rights as rights;
(iii) from (ii): benevolent absolutisms cannot reliably protect hwnan
ríghts;
(ív) from (ii) and (iii): benevolent absolutisms cannot fully comply with
the eight principies of LP;
(v) from (ii), (iii), and (iv): benevolent absolutisms cannot be considered as members in good standing ofthe Society of Peoples;
(vi) from (i) and (v): benevolent absolutisms are sovereign over their
people and territory but should not be tolerated.
This characterization of benevolent absolutisms would allow us to retain
551
would like to thank an anonymous reviewer for pressing me on this point. Tbe
substance of my analysis is greatly indebted to Walter Riker's excellent paper, "Human
R.ights Without Political Participation," Human Rights Review 15 (2014): 369·90.
56Tbough he probably is implicitly committed to a somewhat "Ionger" Jist. See Reidy,
"Human Rights and Liberal Toleration."
57Sec Henry Sbue, Basic Rights: Subsistence, Afjluence, and US. Foreign Poiicy,
2nd ed, (Princeton: Princeton University Press, 1996).
Pietro Maffettone
550
much of the substance of Rawls's definition without having to accept
that they can become part of an overlapping consensus on the principies
of LP. Why should we believe that benevolent absolutisms cannot respect human rights as rights'! Because, following Henry Shue, we can
argue that the very idea of an absolute fonn of political rule means that
those who enjoy the substance of basic rights do so only at the discretion
of those who rule. However, we normally see rights as entitlements that
cannot be changed at the discretion of those who are in power. Shue explicitly considers tbe compatibility between the idea of a benevolent dietatorship and respect for basic rights "as rights." He concludes that while
a dictator may actually allow his subjects to enjoy the substance of basic
rights, "[ w ]hat is missing that keeps people under the ideally enlightened
despot we are imagining from enjoying rights, and that would be supplied by rights to participate in the control of the substances of other
rights, is social institutions for demanding the fulfilment of the correlative duties, especially the duty to provide one vital form of protection:
protection from the government itself."58
1 have two replies. First, as Shue bimself recognizes, his theoretical
case for saying that a benevolent despot cannot allow his subjects to enjoy certain rights as rights is not, in his own view, final insofar as "it is
not utterly inconceivable that sorne set of institutions without a basic
right to participation might turn out to protect people's other rights as
well as, or better than, the most effective participatory institutions do.
However unlikely this is in practice, it is not evident how to rule out the
theoretical possibility. "59 For example, again following Shue, there is no
conceptual impossibility in imagining a benevolent dictator establishing
sorne forrn ofrule oflaw that is independent from the dictator's will.60 In
other words, at least conceptually, we can conceive of a rights-respectingbenevolent-absolutism.
Second, it is by no means self-evident that the correct way of picturing the relationship between rights to political participation and other
basic rights is the one suggested by Shue. Shue claims that rights to political participation are instrumentally necessary to secure other basic
rights, One way to describe the instrumental relationship between political participatioo and human rights is to see political participation as a
58Jbid.,
59Jbid.,
p. 76.
p. 87.
"°For example, Reidy suggests that "the bencvoleni absolutism migh; fultill the old
natural law conception of rightful rule as a public dictate of reason aimed by she who has
authority over the community at the common good of the community." See David A.
Reidy, "On the Human R.ight lo Democracy: Searching for Sense Without Stilts," Journal
o/Social Philosophy43(2012):177-203, p. 192.
Should We Tolerate Benevolent Absolutisms?
551
"lever?" that citizens can use to effectively constraint political leaders
and make sure that they respect basic human rights. Yet, as Walter Riker
argues, this argurnent is implausible for the simple reason that it is very
unlikely that política! leaders who are willing to violate basic human
rights will nonetheless respect rights to political participation. Furthermore, while we can argue that política! participation protects basic human rights, we still need to explain wbat exactly will protect political
participation in the first place wben citizens are facing political leaders
who are willing to disregard their basic rights. 62 According to Riker, it is
the shared nonnative cornmitments ofthose who participate in social cooperation in the basic welfare of their fellow citizens, rather than political
participation rights per se, that actually constitute the core defense such
individuals have against the violations of their basic human rights.63
However, if this is the case, there is no clear reason to believe that such
shared commitments to the welfare of citizens cannot be found witbin a
benevolent absolutism. 64 If this picture is correct, then the lever argument
is at best inconclusive.
8. Benevolent Absolutisms and Oemocratic Peaee
Can it be the case that benevolent absolutisms are unable to be consistently (extemally) peaceful? Rawls does not sayas much, but bis description of well-ordered peoples as "satisfied peoples" and his sui-generis interpretation of democratic peace might signa! this conclusion. Rawls's
analysis starts from liberal peoples. Liberal peoples are different from
states. While states are simply guided by what they perceive as their rational interests, liberal peoples have a moral nature and they set limits to
what their interests can legitimately be.65 Given their reasonable ínterests, liberal peoples do not want more power, territory, wealth or glory.
According to Leif Wenar and Branko Milanovic, liberal peoples • peacefulness in LP is based on their commercial character, their indifference to
perpetua! economic growth, and their tolerant attitude toward pluralism
of religious beliefs.66 More controversially, Rawls also extends this anal61Hcre
6l(bid.
I follow Riker, "Human Rights Without PoliticaJ Participation," p. 381.
.
pp. 383 and 388.
to Riker, we can picture Rawls as endorsing what he calls the "Rights
View." According to such a view, basic human rights actas preconditions to political participatíon. lbid., p. 3 75.
65Rawls, LP, p. 29.
66Leif Wcnar and Branko Milanovic, "Are Liberal Peoples Peaceful?" Journal of Po­
litical Philosophy 17 (2009): 462-86, p. 469.
6l(bid.,
64 According
552
Píerro Maffettone
ysis to decent peoples and claims that even decent peoples are "satisfied
peoples." As Wenar and Milanovic point out, the extension of democratic peace to nondemocratic but decent peoples is problematic." The empírica! support for tbe idea of democratic peace cannot readily be deployed to tell us tbat even nondemocratic but decent peoples can behave
in tbe same way.68 However, even if we were to accept Rawls's extension of the democratic peace hypothesis, the very features of wellordered peoples that Rawls gives as the basis for their being "satisfied
peoples" (that is, their indifference to economic growth, their tolerance,
and their commercial character) can be plausibly extended to benevolent
absolutisms. There seems to be no principled reason to state tbat benevoJent absolutisms cannot sbare these tbree features. Rawls's definition only tells us that they respect human rights but do not allow political participation-it tells us nothing about their altitudes toward growth, commerce, and religious diversity.
Even more importantly, wbatever the merits and effective boundaries
of Rawls's interpretation of democratic peace, it can only establish that
certain peoples (liberal and decenr, for example) are more likely to be
peaceful toward each other; it cannot establish a negative relationship between non-well-ordered institutions (the kind that would belong to benevolent absolutisms) and aggressive behavior in foreign policy. This is
the result of the very structure of the claims made by the democratic
peace thesis. The democratic peace thesis makes two types of claims.69
The first is descriptive and highlights a pattem in foreign policy behavior. lt holds that armed conflict between two democratic countries is less
likely to occur than armed contlict between a democratic country and a
nondemocratic country or between two nondemocratic countries. The
second type of claim is causal and pertains to the explanation of the first,
descriptive, claim. It tries to ascertain which features (or set of features)
of democratic countries and institutions can provide an explanation for
their more peaceful behavior toward each other."
For our purposes here, it is important to note two things in connection to the dernocratic peace thesis. First, no matter which causal account. one adopts, it will only be able to explain the reasons for tbe externa! peacefulness of democratic countries toward each other-not
67But
see Walter Riker, "The Democratic Peace is Not Democratic.: On Behalf or
Rawls' Decent Socieries," Political Studies 51 (2009): 617-38, for an argument in favour
of the exrension.
68Michael W. Doyle, "Three Pillars of the Liberal Peace," American Polittcal Science
Review 99 (2005): 463-66.
69Here 1 follow Riker, "The Dcmocratic Peace is Noi Democratic," p. 622.
70Sec ibid. for a good discussion of the different causal accounts.
Should We Tolerare Benevolent Absolutisms?
553
why nondemocratic societies cannot be consistently peaceful." Second,
and relatedly, the descriptive thesis of democratic peace can only tell us
that democracies are statistically more peaceful toward each other
compared to other forms of regimes, but it cannot tell us whether a given nondemocratic regime type is externally peaceful or not. In orher
words, the democratic peace thesis can only establish (at best) sufficient
conditions for reliable peace between certain types of collectives; it
cannot, logically, establish necessary conditions as well. In tum, if
democratic peace theory cannot establish necessary conditions for externa! peacefulness, then even assuming that benevolent absolutisms
are not within the remit of democratic peace (something that we can legitimately doubt given Rawls's understanding of the bases of democratic peace ), it will be silent on whether benevolent absolutisms are externally peaceful or not. W e thus have no reason to state that they would
not be able to endorse principies 4 and 5 oí LP. It could be argued tbat
while democratic peace theory cannot establish necessary conditions
for externa! peacefulness, it provides support for the conjecture that
sorne regimes are "much more likely" to be externally peaceful than
other types of regimes. That may very weU be tbe case, but recall that
we are looking for reasons to conclude that benevolent absolutisms
cannot be members in good standing of the Society of Peoples. In other
words, we are looking for reasons to exclude an entire class of societies
from the scope of toleration, The reasons we pul forward need to be
categorical rather than probabilistic in character. Simply arguing that a
type of regime (decent or dernocratic) is more reliable in developing a
peaceful attitude toward other weU-ordered peoples is not reason
enough to exclude an entire class of nondecent or nondemocratic societies a priori.
9. Conclusion
LP's critics have been concemed with the excessively permissive nature
when it comes to toleration. In this paper, 1 have shown that if we adopt
the best reconstruction of the Rawlsian account of international toleration, LP does not provide conclusive reasons not to tolerate benevolent
absolutisrns. This may suggest that benevolent absolutisms should be
seen as members in good standing of intemational society; yet, however
convincing my arguments have been in this paper, their implications are
'1Formally expressed: lf A - Bit does not follow that - A - - B. Where A stands
for "being a democratic country" and B stands for "being statistically more pcaceful toward other democratic countries." The latter would only follow if A ... B.
554
Pietro l\1affettone
not this far-reaching. What that they do imply, however, is that unless we
can find a better argument to explain the ímportance of interna! political
participation for allegiance to the eight principles of LP, we should be
prepared to revise the scope of Rawlsian intemational toleration. Unfortunately, neither Rawls nor bis more sympathetic readers have laid out an
explicit versión of such an argument. 72
School of Govemment and lnternational Affairs,
Durham University
pietro.maffettone@durham.ac.uk
121
would like to thank two anooymous referces for their detailed aod constructive
reviews, and lan Cárter, Mat Coakley, Peter Jones, Laura Valentini, Leif Wenar, and
John Willia.ms for written comments at various stages of the writing process. Last but not
least, 1 would like to thank the editors ofthe journal for their support,