Do I have to be Coherent to be Reasonable?

A Defense of Truth-Abstinence in Rawls’s

Political Philosophy

Abstract

In Political Liberalism, Rawls infamously denies that his political con-

structivism needs to reference the concept of truth, a claim that has been

criticized by Joseph Raz, Joshua Cohen, and David Estlund. In this pa-

per, we argue that these criticisms fail due to the fact that parties to the

overlapping consensus do not have to be coherent in order to be reason-

able. Once it is seen that Rawls’s political constructivism allows such

incoherence to reasonable parties, the inclusion of truth urged by Raz,

Cohen, and Estlund appears to require more of reasonable people than is

necessary for a political consensus.

Keywords: Rawls on Truth, Political Constructivism

Introduction

Central to John Rawls’s Political Liberalism is the method of political con-

structivism. Notoriously, Rawls asserts that this method of justification func-

tions without recourse to the concept of truth: “[Political constructivism] does

not...use (or deny) the concept of truth; nor does it question that concept, nor

could it say that the concept of truth and its idea of the reasonable are the

same. Rather, within itself the political conception does without the concept of

truth.”1 Rawls thus holds that, at the political level, the justification of a con-

ception of justice may proceed without recourse to the concept of truth. This

position has been met with mixed reviews.2 Several commentators think that

Rawls’s position on the exclusion of truth from political discourse is incoherent:

Joseph Raz –

“To recommend [a theory of justice] as a theory of justice for our soci-

eties is to recommend it as a just theory of justice, that is, as a true, or

reasonable, or valid theory of justice.”3

Joshua Cohen –

“The idea of locating a common ground of political reflection and argu-

ment that does without the concept of truth...is hard to grasp. Truth is
1 See Rawls (2005), p. 94.
2 For early criticisms, see Gardiner (1985( and Hampton (1989), especially p. 807. Haber-

mas (1995) is also critical of Rawls’s failing to appeal to the concept of truth.
3 See Raz (1990), p. 15.
Do I have to be Coherent to be Reasonable?

so closely connected with intuitive notions of thinking, asserting, believ-

ing, judging, and reasoning that it is difficult to understand what leaving

it behind amounts to.”4

David Estlund –

“Political liberalism must assert the truth and not merely the reason-

ableness - or acceptability to all reasonable people - of its foundational

principle.”5

The animating concern of all these critiques is that Rawls smuggles in or de-

pends upon the concept of truth all the while claiming to avoid it, and that

this renders Rawls’s account contradictory. In place of Rawls’s avoidance of

talk of truth, these opponents suggest appealing to the concept of truth, yet

nevertheless remaining agnostic when it comes to a theory of truth, whether it

be a correspondence, pragmatist, or deflationary view. The problem with this

recommendation to employ truth yet remain indifferent about how the concept

is analyzed, we will argue, is that it is inconsistent with two key aspects of

Rawls’s project – the notion of reasonableness and the publicity condition:

Concept Indifference – Political constructivism appeals to the truth

of some claims and employs concepts that are conceptually connected to

4 See Cohen (2009), p. 15.
5 See Estlund (1998), p. 253.

2
Do I have to be Coherent to be Reasonable?

truth, but it need not adjudicate between competing understandings of

truth6

Reasonableness – To be reasonable, citizens are only required to be

willing to cooperate under fair terms within a society and tolerate other

viewpoints due to their recognition of the burdens of judgment7

Publicity – All reasonable citizens can access and endorse the political

conception of justice that regulates society as well as the reasons put for-

ward as its justification8

The inconsistency between these three principles results from the possibility of

conceptually incoherent parties to the overlapping consensus – “incoherent” in

that they can have contradictory beliefs concerning their most basic concepts.

The characterization of reasonableness does not require that reasonable citi-

zens’ comprehensive doctrines, or even their most basic beliefs, must satisfy any

coherence conditions. Because Reasonableness allows reasonable people to be

conceptually incoherent, and this is public knowledge via Publicity, it cannot

also be required that all reasonable people satisfy Concept Indifference, or so

we will argue. In Section 2, we present the critiques of Raz, Cohen, and Es-
6 Versions of Concept Indifference are explicitly endorsed by Cohen (2009) and Estlund

(1998).
7 See Rawls (2005), p. 94.
8 See Rawls (2005), p. 66-67.

3
Do I have to be Coherent to be Reasonable?

tlund along with their proposal of Concept Indifference, arguing in Section 3

that all these objections are undermined by our main contention – the possi-

bility of incoherent, but nevertheless reasonable, citizens. Section 4 contains a

fuller response to Rawls’s critics before considering some objections in Section

5. Ultimately, we conclude that Rawls’s political constructivism withstands the

criticisms of Raz, Cohen, Estlund, and others who criticize it on the basis of

some alleged connection between truth and political justification.

1 Rawls’s Critics

There have been several criticisms leveled against Rawls’s exclusion of the con-

cept of truth from political constructivism. The first of these critiques comes

from Joseph Raz, who argues that the acceptability of Rawls’s theory of justice

is inconsistent with remaining agnostic on its truth. In Raz’s view, regarding a

principle of justice as acceptable entails regarding the principle as true:

To recommend [a theory of justice] as a theory of justice for our societies

is to recommend it as a just theory of justice, that is, as a true, or

reasonable, or valid theory of justice. If it is argued that what makes it

the theory of justice for us is that it is built on an overlapping consensus

and therefore secures stability and unity, then consensus-based stability

and unity are the values that a theory of justice, for our society, is assumed

to depend on. Their achievement – that is, the fact that endorsing the

theory leads to their achievement – makes the theory true, sound, valid,

and so forth. This at least is what such a theory is committed to. There

4
Do I have to be Coherent to be Reasonable?

can be no justice without truth.9

According to Raz, if a political conception is acceptable as the focus of an over-

lapping consensus and if this acceptability vindicates its principles, then the

political conception must be “true, sound, valid, and so forth” in virtue of its

ability to serve as the focus of an overlapping consensus. In other words, recom-

mending a theory of justice (according to any given standard) commits one to

asserting that the theory is true. If Raz is correct and if Rawls does, in fact, as-

sert that his theory of justice satisfies the proper normative-political standards

of acceptability, then Rawls also claims–perhaps unwittingly–that his theory of

justice is true, thereby failing to avoid the concept of truth as he had hoped to.

In a similar vein, Joshua Cohen sees a contradiction in being non-committal

with respect to the concept of truth while still employing other concepts closely

related to truth. Many of the activities associated with political deliberation

appear to be conceptually connected with truth – activities such as “thinking,

asserting, believing, judging, and reasoning”10 – and thus it is problematic to

employ these concepts in political deliberation while simultaneously eschewing

all reference to truth. One of these activities that Cohen explores in more detail

is that of believing. It is commonly held that beliefs aim at being true, and that

insofar as a person accepts the falsity of a proposition, they cease to believe it.11
9 See Raz (1990), p. 15. Emphasis in the original.
10 See Cohen, p. 15.
11 See Williams (2002), p. 67. For further work on truth as the aim of belief, see Brandom

(2001), Chan (2013), Railton (1994), Shah (2003), Velleman (2000 and 2005), Wedgwood

5
Do I have to be Coherent to be Reasonable?

If this is correct – and, we must admit, it seems to quite plausible – then by

believing that p, a person undertakes a mental commitment to the truth of p. It

is therefore unclear how parties to the overlapping consensus can believe in the

accepted conception of justice without simultaneously affirming its truth. For

such reasons, Cohen maintains, Rawls’s vision of political deliberation cannot

proceed without some concept of truth.

Whereas both Raz and Cohen emphasize the apparent incoherence of engag-

ing in public deliberation while rejecting the concept of truth, David Estlund

takes a different tack, arguing that the inconsistency in Rawls’s thought lies in

the avoidance of truth along with the claim that Rawls’s principles of justice

can create actual moral obligations. On Rawls’s view, reasonableness can play

the role of truth in political deliberation by adjudicating between competing

conceptions of justice. Accordingly, Rawls employs reasonableness as a crite-

rion for any doctrine to be included for consideration. The political conception

that constitutes the focus of an overlapping consensus attains vindication via its

reasonable acceptability, not truth.12 Against this suggestion, Estlund argues

that such acceptance could not ground moral obligations:

Suppose, in order to avoid the truth, we understand political liberalism

not as ordering an account of the true standard but simply as using a

(2002), and Whiting (2010 and 2013).
12 For the sake of brevity and focus, we gloss over tricky ambiguities in Rawls’s justificatory

approach. For alternative ways of interpreting Rawls’s project, see Barry (1995), Dreben

(2003), and Gaus and van Schoelandt (Forthcoming).

6
Do I have to be Coherent to be Reasonable?

standard that is acceptable to all reasonable people (the standard it-

self being acceptability to reasonable people)...The question is whether it

could ground obligation and justify coercion even if the acceptance crite-

rion it uses were not true. Never mind for the moment whether political

liberalism says anything on this question; the answer to the question is

that it could not have those moral consequences irrespective of the truth

on those matters.13

Thus, according to Estlund, if Rawls maintains his ambivalent attitude towards

truth, his theory of justice cannot generate moral obligations for the reasonable

participants in the overlapping consensus.

To drive his point home, Estlund asks us to consider the following thought.

Rawls wants us to believe that what matters in political deliberation is ac-

ceptability to reasonable people. But why should acceptance to such a group

matter? There are plenty of groups we could choose from – acceptability to

all redheads, or to all members of the Branch Davidian cult, for instance – but

what is lacking is a criterion for selecting one of these groups from among the

others. One response, that reasonable people tend to settle on true principles of

justice, is not available to Rawls due to his forbearance of truth. Estlund thinks

that without claiming that reasonable acceptability is the “true” standard of

admissibility in public discourse, Rawls’s “view loses any way to select among

the plurality of insular groups, and it becomes untenable.”14 Without specifying

13 See Estlund (1998), pp. 261-262.
14 Ibid, p. 260.

7
Do I have to be Coherent to be Reasonable?

why acceptability to reasonable persons is a better standard than acceptability

to Branch Davidians, Rawls’s account fails to justify his particular principles of

justice. Insofar as Rawls fails to justify his principles of justice, Rawls also fails

to explain why the overlapping consensus gives rise to moral obligations. An

agreement amongst all Branch Davidians would not give rise to such obligations,

so why think that a consensus of reasonable people would?

All three criticisms of Rawls object to the same, stringent doctrine concern-

ing the role of truth in political justification:

No Concept – Political Constructivism does without appealing to the

concept of truth as well as to any concept that is conceptually linked to

truth

Neither Raz, Cohen, nor Estlund think that Rawls can get by with No Concept.

Raz thinks that the act of recommending a principle of justice is conceptually

linked to truth. Cohen argues that there are several concepts at play in delib-

eration – belief, assertion, and reasoning – that are conceptually connected to

truth. Estlund holds that moral obligations can only be created by a political

foundation that is in fact true. In each case, the conceptual connection between

truth and justification implies that political liberalism, pace Rawls, depends on

the concept of truth.

An attractive alternative to truth-abstinence, suggested by Rawls’s critics,

can be formulated as follows:

8
Do I have to be Coherent to be Reasonable?

Concept Indifference – Political constructivism appeals to the truth

of some claims and employs concepts that are conceptually connected to

truth, but it need not adjudicate between competing understandings of

truth15

The benefit of Concept Indifference is that it can address all of the previous

worries without being too exclusive. Rawls could simply agree with his critics

that truth does have an important role to play in establishing the structure and

role of political discourse yet nevertheless remain noncommital as to which is

the correct theory of truth, whether that be some version of correspondence,

minimalism, pragmatism, etc. This thin concept of truth could then be used to

respond to the critiques of Raz, Cohen, and Estlund, or so the story goes.

2 The Cost of Concept Indifference

Between the criticisms of Raz, Cohen, and Estlund, there seem to be good

reasons to question the eschewal of truth that Rawls advocates. However, to

reject Rawls’s approach before considering the reasons for Rawls’s abstention

from this concept would, of course, be premature. In fact, examining the rea-

sons that Rawls states for avoiding truth reveals an incompatibility between

Concept Indifference, the notion of reasonableness, and Rawls’s publicity condi-

tion. This section begins by examining Rawls’s notion of reasonableness, with

a particular emphasis on why a reasonable person may not be able to endorse

15 Versions of Concept Indifference are explicitly endorsed by Cohen (2009) and Estlund

(1998).

9
Do I have to be Coherent to be Reasonable?

a conception of truth. We then turn to Rawls’s motivation for avoiding the

concept of truth, arguing that involving the concept of truth in the procedure

of political constructivism would undermine key goals of Rawls’s project.

2.1 The Reasonable

Unravelling the meaning of “reasonable” is no simple task.16 The two most

important uses of the “reasonableness” concern (1) reasonable doctrines17 and

(2) reasonable persons. For the topic of political constructivism, it is the sec-

ond of these two notions that plays a dominant role.In Lecture III of Political

Liberalism, Rawls identifies two key features of reasonable persons:

The idea of the reasonable is given in part, again for our purposes, by

the two aspects of persons’ being reasonable: their willingness to propose

and abide by fair terms of social cooperation among equals and their

recognition of and willingness to accept the consequences of the burdens

of judgment.18

In other words, a reasonable person is a conditional, rule-following cooperator,

and tolerates diverse viewpoints in light of the difficulty of coming to conclusions

on matters of faith, morality, and other fundamentals.19 For a reasonable person

then, the most attractive political conception of justice will be one that all could

recognize as an adequate compromise, one that is acceptable ,even if not ideal,

16 For more on the interpretive difficulty surrounding reasonableness, see Wenar (1995).
17 For the three features of a reasonable doctrine, see Rawls (2005), p. 59.
18 Ibid, p. 94
19 Emphasis on the conditional and rule-following aspects of reasonableness can be found in

Rawls (2005), p. 49.

10
Do I have to be Coherent to be Reasonable?

from the perspectives of diverse citizens. Such a conception, Rawls submits, is

one that is built up from values that all endorse, values that are not particular to

any comprehensive doctrine. Therefore, citizens, insofar as they are reasonable,

recognize that a political conception of justice based on commonly held values

and conceptions is one that is worthy of endorsement. Such a conception could

thus operate as the focus of an overlapping consensus among all reasonable

persons.20 Reasonable citizens must therefore satisfy Reasonableness:

Reasonableness – To be reasonable, citizens are only required to be

willing to cooperate under fair terms within a society and to tolerate

other viewpoints due to their recognition of the burdens of judgment

An important feature of the reasonable, or rather, an important omission, is that

reasonable persons need not hold correct beliefs. The conception of a reasonable

person is purely practical, not epistemic or, in a Kantian sense, theoretical.21 So

long as persons meet certain basic requirements that allow us to live peacefully

alongside them–namely, Rawls’s two features of reasonableness–we can come to

political agreement.

20 To simplify the discussion, we focus on the notion of reasonableness as applied to persons

rather than to doctrines. As applied to doctrines, the meaning of reasonableness is obscure.

Wenar (1995) argues that ts original definition is far too permissive to achieve what Rawls

wants. When drastically modified by Rawls (1997) in his later work, it becomes too stringent,

since a reasonable doctrine must countenance “the essentials of a constitutional democratic

polity” . This second characterization is as problematic as the first, since it threatens to

render Rawls’s project trivial (see our Section 3).

21 See Rawls (2005), p. 93.

11
Do I have to be Coherent to be Reasonable?

The critical importance of this characterization of reasonableness becomes

evident upon considering concrete cases. Consider, for example, that Reason-

ableness does not exclude a citizen who holds Political Noncognitivism from the

overlapping consensus:

Political Noncognitivism – Political assertions are neither true nor

false.

What is the motivation for such a view? Propositions pertaining to political

life are often normative in nature, discussing norms of justice along with politi-

cal legitimacy and obligation. Philosophers who endorse non-cognitivism about

normative domains generally would thus be drawn to such a view, and they

would be no less capable of satisfying Reasonableness for this reason. Certainly,

persons endorsing such a doctrine could be reasonable: they could be conditional

cooperators who recognize the burdens of judgment. Hence, if Rawls required

that all the comprehensive doctrines included in the overlapping consensus re-

gard political propositions as true, then a reasonable view held by reasonable

persons, including many philosophers, would be excluded.

Political Noncognitivism is one doctrine that reasonable people can hold

concerning the truth of the principles that are selected within the overlapping

consensus. By itself, Political Noncognitivism does not lead to incoherence,

unlike our next case:

Pragmatist Truth – It is true that p if and only if p would be believed

12
Do I have to be Coherent to be Reasonable?

at the ideal limit of inquiry22

Let’s suppose that the pragmatist holds that Pragmatist Truth, is a correct

description of natural language uses of ’true.’ As critics of pragmatism point

out, many pragmatists also believe that there will be truths that will not be

believed at the limit of inquiry because they remain undecidable, propositions,

for instance, about events in the distant past.23 Suppose this criticism is cor-

rect. Then, several pragmatists, C.S. Peirce included, hold inconsistent beliefs

and are committed to a conceptual incoherence. It is not possible both that

Pragmatist Truth is correct a priori and that any true propositions will remain

undecidable at the limit of inquiry. Nevertheless, these conflicting commitments

would not disqualify Peirce from political reasonableness, for he would still be a

tolerant, rule-following cooperator with operative practical reasoning faculties

and would still endorse certain widespread conceptions and values. Indeed, it

would seem preposterous to exclude a cooperative, law-abiding philosopher from

22 See Peirce (1902), p. 565. Peirce gives this now famous characterization saying, “Truth is

that concordance of an abstract statement with the ideal limit towards which endless investi-

gations would tend to bring scientific belief.”

23 Those who have leveled this objection against the pragmatist conception of truth include

Plantinga (1982) and Wright (1992 and 2001). Plantinga originally forwarded the conditional

fallacy against Putnam’s pragmatist account of truth, an argument that Wright adapted in

(1992) and then directed towards Peircean pragmatism (2001). Peirce (1878) uses the phrase

“buried secrets” in anticipating precisely this objection: “But I may be asked what I have to

say to all the minute facts of history, forgotten never to be recovered, to the lost books of the

ancients, to the buried secrets. Do these things not really exist because they are hopelessly

beyond the reach of our knowledge?” (p. 207).

13
Do I have to be Coherent to be Reasonable?

the political consensus simply because they had a conceptual incoherence in his

or her philosophical theory. The surprising upshot is that reasonable people, in

Rawls’s terminology, can be conceptually incoherent.

2.2 Why Avoid Truth? The Publicity Condition

Perhaps this should not be surprising. After all, throughout Political Liberalism,

Rawls describes the central problem that he intends to address as concerning the

possibility of attaining a wide consensus in a diverse population: “How is it pos-

sible that deeply opposed though reasonable comprehensive doctrines may live

together and all affirm the political conception of a constitutional regime?”24

Presumably, this wide consensus aims to include the wide swath of citizens

whose worldviews are not fully or systematically worked out.

The cases of Political Noncognitivism and Pragmatist Truth, combined with

the stated aim of Political Liberalism, together clarify why Rawls sought to

avoid appealing to the concept of truth. If citizens endorse diverse comprehen-

sive doctrines, then insofar as the task of a political conception of justice is to

reconcile these disparate doctrines, to gain unanimous assent, and to function as

the focus of an overlapping consensus – to this extent, a political conception of

justice must aim to be neutral with respect to divisive philosophical or religious

commitments. One such commitment is the nature of truth and its relation to

normative propositions, including political propositions. Since, by hypothesis,

24 See Rawls (2005), p. xviii. See also pp. 24, 90, 97, and 100-101.

14
Do I have to be Coherent to be Reasonable?

all citizens are committed to achieving a mutually acceptable conception of jus-

tice, using the reasonable as a standard by which to test conceptions of justice

is less divisive than using truth as a standard. In other words, various compre-

hensive doctrines may not countenance the concept of truth, and avoiding the

concept of truth in political constructivism allows those who hold such doctrines

to endorse the focus of the overlapping consensus without contradicting their

own particular comprehensive doctrines. As Rawls writes:

One thought is that the idea of the reasonable makes an overlapping

consensus of reasonable doctrines possible in ways the concept of truth

may not.25

And again:

Political constructivism doesn’t use this idea of truth, adding that to

assert or deny a doctrine of this kind goes beyond the bounds of a political

conception of Justice framed so far as possible to be acceptable to all

26
reasonable comprehensive doctrines.

Rawls’s strategy of truth-avoidance comes in the form of “political construc-

tivism,” a procedure wherein Rawls leverages certain conceptions of the person

and of society that he takes to be implicit in the culture of any liberal democratic

society. Due to their latent presence in the public culture, such conceptions are–

at least implicitly–endorsed by all citizens. In this way, any principles of justice

that emerge from these commonly held values could also be endorsed by all
25 Ibid, p. 94.
26 Ibid, p. 114.

15
Do I have to be Coherent to be Reasonable?

citizens in the liberal democratic society.

By employing his constructivist procedure, Rawls is able to satisfy a desider-

atum that he calls the “publicity condition.”27 A society can satisfy the publicity

condition on three distinct levels of increasing demandingness:

1) Citizens know and accept a single conception of justice. In addition,

they accurately and justifiably believe, as a part of common public knowl-

edge, that society’s institutions satisfy the demands of this conception of

justice.

2) Citizens affirm the same empirical, social facts that are relevant to

political justice.

3) The full justification (i.e. the argument in support of) the political

conception of justice is publicly known or publicly available.28

To appreciate the importance of this condition, recall the aim of political liberal-

ism and the function of political constructivism.29 As mentioned above, Political

Liberalism seeks to offer an account of how “a plurality of reasonable doctrines,

both religious and nonreligious, liberal and nonliberal, may endorse” a single

political conception of justice. To this end, a constructivist procedure draws

solely from society’s stock of shared values and conceptions, namely those that
27 Ibid, p. 66.
28 Ibid, pp. 66-67
29 Rawls cites several reasons for the importance and desirability of satisfying the publicity

condition. We focus on the reason that lies closest to the core of his project. For a deeper

discussion of the importance of publicity, see Kogelmann (forthcoming).

16
Do I have to be Coherent to be Reasonable?

are implicit in the public political culture. All three levels of publicity concern

the understanding and endorsement of the political conception, its realization,

and the reasons that underlie and justify it. Without satisfying all three levels,

some citizens in such a society cannot fully, cognizantly endorse the governing

political conception of justice. Therefore, satisfying the publicity condition is

necessary for fully realizing the aim of Political Liberalism:

Publicity – All reasonable citizens can access and endorse the political

conception of justice that regulates society as well as the reasons put

forward as its justification

The method of political constructivism is devised as a means of making the

political conception understandable and justifiable to the citizenry as a whole

– i.e. as a means of satisfying the publicity condition. It is this desideratum

that drives Rawls to avoid truth, and to instead employ the public conception

of reasonableness as the standard by which to judge a political conception of

justice. The concept of truth, being denied or doubted by many reasonable

citizens, appears to be inconsistent with the ideal of publicity, and therefore

with the aim of Political Liberalism.

2.3 Joint Inconsistency

We have struck upon an inconsistency between Concept Indifference, Reason-

ableness, and Publicity. If citizens must endorse the methodology of political

constructivism and need only be social cooperators who recognize the burdens

of judgment, then people who do not agree with the using truth in political

17
Do I have to be Coherent to be Reasonable?

constructivism are possible. How should Rawls proceed?

Rejecting Concept Indifference and continuing to eschew the concept of

truth, leaving Rawls’s project open to the criticisms of Raz, Estlund, and Cohen.

Leaving behind No Concept means either adjusting the conception of ‘reason-

able doctrine’ to be more stringent, thereby requiring a less broad consensus, or

giving up on the publicity condition and accepting that some reasonable per-

sons will be unable to endorse the political conception of justice. In other words,

Rawls must either defend his method of truth-avoidance or he must sacrifice the

core aim of Political Liberalism: providing a justification, endorsed by all rea-

sonable doctrines in a pluralistic democratic society, for a political conception

of justice.

In the remainder of this paper, we will defend the retention of No Concept as

more desirable than either strengthening the criteria of reasonableness or weak-

ening the publicity condition. We will argue that, given the aims of a political

conception of justice, going beyond the reasonable to include truth as a concept

in political constructivism would deface Rawls’s project. In addition, the diffi-

culties of avoiding truth are largely dissolved once we note the radical, though

intentional, incompleteness of the political conception. The key contention on

which we base our defense of Rawls and of No Concept is that reasonable people

can be incoherent.

18
Do I have to be Coherent to be Reasonable?

3 Reasonable People can be Incoherent

Does Rawls drop No Concept, or does he remain steadfast in excluding truth

from his constructivist argument for the political conception of justice? Let’s

evaluate the relative costs of these two options in light of our previous discussion.

3.1 Dropping No Concept

We have identified two ways to pay the price of incorporating truth into the

justification of a political conception of justice. The first is by tightening the

conception of reasonableness so that only persons or doctrines that affirm some

concept of truth receive full justification from their own point of view. The

second is by dropping or weakening the publicity condition, so that not every

reasonable citizen can view the governing political conception of justice as ac-

ceptable and fully justified.

First, consider Rawls’s conception of reasonableness. The appropriateness

of this construal of reasonableness is apparent when one recalls that Rawls’s

project is to achieve a political conception of justice and a justification thereof

that a deeply diverse citizenry can endorse. The conception of reasonableness

that Rawls posits is a plausible answer to the question: what are the most basic

requirements that citizens must exemplify in order for them to agree on and

abide by a single conception of justice? As the stringency of reasonableness

increases, the diversity of those who must endorse the governing conception of

justice decreases. If we consider reasonable only those who hold appropriately

19
Do I have to be Coherent to be Reasonable?

coherent or true beliefs with respect to normative political statements, then

the reasonable constituency to which our justification appeals is smaller, less

diverse, and less realistic.30

Recall our two concrete cases: Political Noncognitivist and Pragmatist Truth.

By hypothesis, both are reasonable in Rawls’s sense, yet (we have assumed) both

are mistaken or even incoherent when they endorse a conception of justice. The

cost of satisfying Raz, Cohen, and Estlund by tightening the conception of rea-

sonableness is the exclusion of such persons from the justificatory constituency.

Given the aspirations of Political Liberalism, this cost may be prohibitive. Be-

fore making this conclusion, however, let us examine Rawls’s other options.

The second way in which Rawls could purchase the concept of truth is by

weakening the publicity condition. However, this approach also has a high price.

As we have seen, the three levels of publicity all describe ways in which citizens,

from their own standpoints, can understand and endorse the institutions and

political conception of justice that prevail in their society. Recall that with-

out satisfying each of the three levels of publicity, there will be some subset

of reasonable persons who cannot endorse the prevailing political conception of

justice. Either they cannot accurately affirm that their society and its institu-

tions satisfies a conception of justice they endorse (first level), they do not agree
30 Salvaging Rawls’s position by such means would therefore exacerbate the problem of

stability. To appreciate the severity of this problem, see Huemer (1996) and Klosko (1997).

20
Do I have to be Coherent to be Reasonable?

with the empirical facts or methods of inquiry that support the justification of

the prevailing conception of justice (second level), or they cannot know or do

not have access to the argument used to justify the prevailing conception of jus-

tice (third level). In all cases, a set of reasonable persons is unable to endorse

the political conception of justice and Rawls fails to achieve the stated aim of

Political Liberalism.31

Estlund realizes that the reasonableness of the noncognitivist presents a

challenge to Concept Indifference. He responds by arguing that parties to the

overlapping consensus need merely accept a minimal “truth schema” — p is true

if and only if p — and that such a schema is sufficiently minimal to function

as a public conception of truth. Estlund, however, claims that this political

conception of truth serves an important justificatory purpose even if noncog-

nitivists must reject it on account of their general rejection of the concept of

truth.32 Estlund’s view, however, does not take into account the importance

of transparency in political constructivism. Publicity requires that the noncog-

nitivist acknowldege and endorse the justification of the principles of justice.

31 Cohen (2009) attempts to do away with No Concept without limiting participation in the

overlapping consensus, thus trying to avoid strengthening the conditions for reasonableness.

On Cohen’s view though, a political conception of truth must hold that believing, asserting,

and judging are all believing, asserting, and judging to be true (pp. 26-27), something that we

have seen the Political Noncognitivist explicitly denies concerning the principles of justice. For

this reason, we disagree with Cohen that a political conception of truth can avoid deviating

from the Rawlsian project.

32 See Estlund (1998), p. 270.

21
Do I have to be Coherent to be Reasonable?

Yet, on Estlund’s account, this entails acknowledging and endorsing the con-

cept of truth, since truth plays an important role in justifying the principles of

justice. The noncognitivist cannot accept this metholodogy. Despite Estlund’s

efforts at selecting the most minimal form of a theory of truth, a conflict still

arises between Concept Indifference, Reasonableness, and Publicity, effectively

excluding the noncognitivist from the political consensus.33

3.2 Dropping Truth

The cost of satisfying Raz, Cohen, and Estlund and of avoiding incoherence thus

appears to be quite high. In placating his critics, Rawls would undermine the

very aim of his project. What is the cost of continuing to eschew truth? Raz,

Cohen, and Estlund’s basic strategy is to find a contradiction in the approach

taken in Political Liberalism by claiming that Rawls smuggles in or depends

upon the concept of truth all the while claiming to avoid it. Rather than ex-

amining each position separately, as we will in the next section, we consider a

general formulation of the objection. Recall that we understand incoherence as

a contradiction in the beliefs of a particular individual; one is in a state of con-

ceptual incoherence when one’s beliefs are ultimately contradictory. The basic

strategy of Raz, Cohen, and Estlund is to show that political liberalism cannot
33 Raz (1990) entertains the possibility of the Political Noncognitivism, but fails to note

how the noncognitivist provides an argument for No Concept. This is because Raz’s main

target is a purported inconsistency in Rawls, not reasons for endorsing No Concept or Concept

Indifference (p. 15).

22
Do I have to be Coherent to be Reasonable?

be coherently held and should therefore be rejected or modified.

The defender of Rawls might think that, in order to avoid this difficulty, the

burden is to show that Rawls’s political conception is a fully coherent system.

However, such an approach misses the point. Given the aims of Political Liber-

alism, conceiving of its doctrines as a systematic whole is a mistake. Instead,

political liberalism and its argumentative strategy, political constructivism, aim

to provide a radically incomplete doctrine, one that is so minimal and so restric-

tive in its assumptions and in its scope that a deeply diverse society can come

to endorse it. Its incompleteness allows the political conception to function as

a “module”34 that can find a basis of justification among various, incompati-

ble comprehensive doctrines. As such, there exists a vast array of philosophi-

cal questions on which political liberalism need not–indeed, should not–take a

stand. One of these is the question of truth and its connection to interpersonal,

political justification.35

Contrast the sparsity of political liberalism with the nature of comprehen-

sive doctrines, which include “conceptions of what is of value in human life,

and ideals of personal character. . . ideals of friendship and of familial and as-

sociational relationships,” as well as metaphysical and religious beliefs.36 It

34 See Rawls (2005), p. 12
35 For a concise summary and criticism of Rawls’s method of constructivism, see Klosko

(1997). For an account of the evolution of Rawlsian constructivism over Rawls’s life, see

O’Neill (2002).
36 See Rawls (2005), p. 175 and p. 374.

23
Do I have to be Coherent to be Reasonable?

is a comprehensive doctrine, not a political conception, that must clarify the

conceptual connection between truth and political justification. Therefore, it is

only a comprehensive doctrine that can be conceptually incoherent with respect

to the relationship between truth and political justification. Political liberalism

is simply too incomplete to be incoherent on this subject. It does not propose or

deny any theory of truth or any theory of the justificatory importance of truth,

choosing instead to keep its hands out of such matters and to leave them within

the realm of the non-political.

Raz, Cohen, and Estlund therefore take aim at the wrong target. In claiming

that political liberalism is incoherent in its eschewal of truth, they miss the fact

that truth may enter into a full justification of a political conception. Rawls’s

contention is simply that it must do so at the level of the comprehensive doc-

trine, not at the level of political constructivism. The criticism should therefore

target doctrines like Political Noncognitivist or Pragmatist Truth, which don’t

simply eschew truth, but positively deny its relevance or hold contradictory

views about its nature. If rejecting truth renders political justification incoher-

ent, then it is the comprehensive doctrine that refuses to accept the concept of

truth that is accountable, not the political conception that takes no stance on

the issue. In principle, Rawls, qua philosopher, could even agree with the criti-

cisms of Raz, Estlund, and Cohen – but qua political philosopher, searching out

a public basis of political consensus, Rawls might well view such criticisms as

irrelevant. The task at hand is to establish a conception of justice that will allow

24
Do I have to be Coherent to be Reasonable?

deeply diverse citizens to cooperate on fair terms, not to discover the panacea

of final truth. This is part of what Rawls means when he says that political

constructivism is an exercise of practical reason, not of theoretical reason.37

In sum, the critiques of Raz, Cohen, and Estlund apply at the level of the

comprehensive doctrine, and at this level, their correctness or incorrectness is

simply not germane to Rawls’s project. Therefore, the defender of Rawls can

affirm that the cost of maintaining No Concept is low. Some comprehensive

doctrines may espouse conceptually coherent and factually accurate beliefs, they

may supplement the bare-bones justification offered by Rawls with a theory of

normative truth.38 But participants in the overlapping consensus can also hold

positions that are incorrect, even on pain of conceptual incoherence, without

ceasing to be reasonable. The aim of Political Liberalism is to reconcile these

disparate doctrines, to find a conception that a diverse citizenry “religious and

nonreligious, liberal and nonliberal” – coherent and incoherent – “may endorse

for the right reasons.”39

4 A Final Response to the Truthers

With this understanding in place, what should we make of Raz, Cohen, and

Estlund’s particular criticisms? We have already shown that the charge of

incoherence can be assuaged by the simple fact that parties to the overlapping
37 Ibid, p. 93.
38 Ibid, pp. 144-145.
39 Ibid, p. xxxix.

25
Do I have to be Coherent to be Reasonable?

consensus can be conceptually incoherent. To drive the point home, we examine

each critic’s view in greater detail.

4.1 Raz

Recall that Raz asserts that recommending a principle of justice entails re-

garding that principle as true. Thus, by recommending the outcomes of the

overlapping consensus, Rawls involves himself in a commitment to the truth of

the principles selected by this procedure. One way of construing Raz’s thesis is

that the norm governing the practice of normative assertion is that of knowl-

edge: when one asserts p, one is also be affirming that one knows p, and of

course, knowledge entails truth.40

A defender of Rawls might wish to challenge Raz’s thesis by denying that

knowledge is the norm of assertion. Given the task of a political conception of

justice, however, the debate is not over what the norm of assertion actually is,

but what notion of assertion all parties in an overlapping consensus could agree

on. The aim of a certain kind of discourse should have a bearing on the norms

that govern it.41 In political discourse, especially of the justificatory type, the
40 For endorsements of the knowledge norm of assertion, see Adler (2002), DeRose (1991,

1996, and 2002), Hawthorne (2003), Stanley (2005), Unger (1975), and Williamson (1996 and

2000). That knowledge entails truth has long been the standard view, see Bonjour (2002), p.

32, Tienson (1974), p. 289, and Williamson (2000), p. 42. For an important rejoinder, see

Hazlett (2010).
41 Again, we see that fundamental to Rawls’s approach is presupposing a shared political

task. Arising from the constraints imposed by this task is a standard of public reason, which

26
Do I have to be Coherent to be Reasonable?

goal is not to describe a mind-independent reality.42 Instead, the task of po-

litical discourse – for us “here and now” – is related to solving the problem of

political liberalism, viz., “to work out a political conception of political justice

for a constitutional democratic regime that a plurality of reasonable doctrines,

both religious and nonreligious, liberal and nonliberal, may endorse for the right

43
reasons.”

If political discourse has this practical task, then the norms that govern po-

litical discourse must not generate unnecessary faction or controversy. Different

reasonable doctrines have radically different positions regarding the status and

relevance of truth. Thus, a more sensible norm would be one that all parties

can endorse as in conformity with the values and concerns that motivate po-

litical discourse. Reasonableness, understanding a “reasonable assertion” to be

one which all involved parties can accept insofar as they are reasonable, is (by

definition) the norm that fulfills this requirement. Because the reasonableness

norm is premised on appealing to the reasoning faculties of all involved parties,

it fosters consensus rather than discord, thereby fulfilling the task of political

discourse. Therefore, even if we accept Raz’s claims about normative-political

truth, it does not follow that political constructivism should endorse his view of

the norm of assertion. Political constructivism provides an argument that may

allows for interpersonal justification in a society of citizens that share a set of normative

conceptions related to this task. This is another implication of Rawls’s invocation of “practical

reason” as a political idea when developing his constructivist method.

42 See Rawls (2005), pp. 91-93.
43 Ibid, p. xxxix.

27
Do I have to be Coherent to be Reasonable?

be supplemented in coherent or incoherent ways by diverse comprehensive doc-

trines. Whether it is true or false, coherent or incoherent, will depend on how

a comprehensive doctrine decides to fill out its incomplete doctrines. Again,

political liberalism is not meant to be taken alone. It is, instead, a “moduleâĂę

that in different ways fits into and can be supported by various reasonable com-

prehensive doctrines that endure in the society regulated by it.”44

4.2 Cohen

The case of the Political Noncognitivist should not mislead us into thinking

that Rawls is endorsing a noncognitivist understanding of political assertions.

This is important to how Rawls would respond to Cohen’s criticism. Cohen

argues that by using concepts that are conceptually connected to truth, Rawls

commits his political constructivism to something beyond No Concept. Again,

as Cohen puts it, “Truth is so closely connected with intuitive notions of think-

ing, asserting, believing, judging, and reasoning that it is difficult to know what

leaving it behind amounts to.”45 As in the Rawlsian response to Raz sketched

above, the issue is not that Cohen is wrong about the nature of belief or of

assertion. Rather, it is that many reasonable people disagree, and therefore an

understanding of belief or assertion that involves truth instead of mere reason-

ableness is unduly exclusive. One class of such people is noncognitivists, who

hold that normative statements are not truth-apt. If a noncognitivist were per-

suaded by Rawls’s constructivist argument to endorse the political conception

44 Ibid, p. 145.
45 See Cohen (2009), p. 15.

28
Do I have to be Coherent to be Reasonable?

of justice and to abide by its demands, then whether they think the principles

of justice are “true” is beside the point. This is why Rawls aims to construct

the conception out of materials that are drawn from a public culture, rather

than from some particular view of what is morally worthy or true.

Instead of forwarding reasons for his principles of justice on the basis of their

truth, Rawls proposes a procedure of construction by which each citizen can see

the principles as issuing from their own practical reason and normative concep-

tions. Doing so does not require that Rawls commit himself to any theory of

the truth-aptness of normative claims. In his critique, Cohen exhibits a serious

confusion on this point by arguing that Rawls is committed to a cognitivist view

of normative political statements. First Cohen argues that:

The claims made by a political conception... must be truth-apt ... They

must be, if there is to be a common ground of argument under conditions

of doctrinal disagreement. To deny the truth-aptness of the claims made

on the terrain of public reason would offend against the essential idea of

public reason. That is because the very propositions advanced in public

political argument, even if not taken as or presented in that context

as true, might be judged to be true by the religious or moral doctrine

affirmed by a citizen.46

Cohen is correct to say that Rawls is not free to deny the truth-aptness of nor-

mative claims. Doing so would alienate moral or religious doctrines that judge
46 Ibid, p. 18.

29
Do I have to be Coherent to be Reasonable?

such claims to be true or false. But from this, Cohen goes on to infer that

“Rawls’s proposal is to endorse a cognitivist understanding of political concep-

tions of justice and political argument on which notions of judgment, reasoning,

and argument are fully in play, while denying the availability of the concept

of truth within such conceptions.”47 This inference, however, is mistaken. It

would be just as illicit for Rawls to endorse and argue from a cognitivist view

as a noncognitivist view. In either case, there are reasonable comprehensive

doctrines, affirmed by reasonable citizens, that reject the metaethical view in

question. Accordingly, the proper path for political constructivism is to avoid

taking a stand on this metaethical question.48 Doing so would undermine the

project of Political Liberalism and thwart the task of political constructivism by

rendering the procedure unpersuasive, unacceptable, or even incomprehensible,

to a large group of reasonable citizens. In fact, Rawls explicitly rejects the need

to analyze or critique the theories of truth espoused by the comprehensive doc-

trines that constitute the overlapping consensus or to endorse some one theory

of truth:

Political constructivism does not criticize, then, religious, philosophical,

or metaphysical accounts of the truth of moral judgments and of their

validity. Reasonableness is its standard of correctness, and given its

47 Ibid, p. 19.
48 Illustrating this idea is the point of introducing Political Noncognitivist and Pragmatist

Truth. As Rawls (1999) says, “It is important to notice here that no assumptions have

been made about a theory of truth. A constructivist view does not require an idealist or a

verificationist, as opposed to a realist, account of truth.” (p. 351).

30
Do I have to be Coherent to be Reasonable?

political aims, it need not go beyond that.49

4.3 Estlund

Finally, the constructivist argument that we sketched above also brings into

focus the Rawlsian response to Estlund’s critique. Estlund points out that

acceptance by a particular group is insufficient for a doctrine to gain norma-

tive import. Estlund points out that Rawls’s standard of permissibility–viz.

acceptability by reasonable persons–must be reflexive. That is, reasonable per-

sons must all accept that acceptability to reasonable persons is necessary, and

perhaps sufficient, to allow a certain doctrine or consideration into political dis-

course. Yet, this standard cannot be sufficient to establish itself as the correct

standard of acceptability. The fact that reasonable persons all hold it to be

correct is irrelevant. After all, there are many such “insular” groups, like the

Branch Davidians, who might hold that self-acceptability is an acceptable stan-

dard. In order to justify privileging one standard over others, one must hold

that it is true, not merely reasonable.

In response to Estlund, this paper has suggested that we do have good rea-

son to favor reasonable-acceptability over Branch-Davidian-acceptability, but

this reason is not its truth. Estlund believes that truth is the only autho-

rizing feature, i.e. the only feature that could allow us to pick out one insular

group from among the plurality and to assert its special justificatory status. Yet
49 See Rawls (2005)p. 127, emphasis added.

31
Do I have to be Coherent to be Reasonable?

Rawls goes to great lengths to demonstrate that this is not the case. In Lecture

III Rawls discusses at length how political liberalism upholds a conception of

objectivity that yields a standard of correctness,reasonableness, that can be un-

derstood apart from any comprehensive notion of truth.50 Political liberalism

accomplishes this task by employing practical, rather than theoretical, reason,

and representing its principles as issuing from a procedure that models this type

of reason.51 While practical reason does not utilize the idea of “an independent

order of moral values,”52 it does assume a shared aim or goal, taken as presup-

positions of political deliberation itself: “As reasonable and rational we must,

as it were, suitably construct the principles of right and justice that specify the

conception of the objects we are to produce and in this way guide our public

reason by practical reason.”53 The public political culture provides the objects

we aim to produce, which Rawls summarizes as “conceptions of society and per-

son, and the public role of principles of justice.”54 We want a fair, cooperative

society, populated by free and equal citizens, which is governed by conception

of justice that all citizens, as such, can endorse.

Given this presupposed set of shared normative conceptions, the reasonable

50 Ibid, pp. 91-93.
51 Ibid, p. 117. “PC. . . holds that there are different conceptions of objectivity appropriate

for theoretical and practical reason.”

52 Ibid, p. 91.
53 Ibid, p. 117.
54 Ibid, p. 110. Rawls (2005) also says “Without the ideas of society and person, conceptions

of the right and the good have no place. They are as basic as the ideas of judgment and

inference, and the principles of practical reason” (p. 110).

32
Do I have to be Coherent to be Reasonable?

acceptability standard emerges as uniquely suited to fulfill the role of political

justification. If the political conception of justice aims to make mutually ac-

ceptable the institutions and claims to which each of us is subject, as citizens,

then Rawls need not consider acceptability to groups that lack the essential fea-

tures of the citizen, namely reasonableness and rationality. Rawls seeks only to

establish that given certain shared aims and conceptions, reasonableness is the

most suitable standard of acceptability. The structure of the argument begins

with an ideal (the well-ordered society), and works backwards to show how such

an ideal could be possible. The conditions of its possibility, as Rawls identifies

them, include shared conceptions of society and person. Given these shared

conceptions, practical reason can provide an objective standard of correctness

which all persons, insofar as they endorse the shared conceptions of society and

person (which includes the moral power of reasonableness), can use to mod-

erate political debate and determine answers to essential political questions.

Therefore, acceptability to reasonable persons is a special standard because it

best achieves the aims of the political project: realizing freedom, equality, and

fairness through consensus and stability. In this respect, as we have seen, it

outperforms truth. Thus, it is the political task that we are engaged in and the

moral bases that such a task implicates that favor the criteria of reasonableness

over that of Branch-Davidianism.55

55 Note the Kantian flavor of this argument: We reason backwards from the task at hand to

the requirements of its possibility. This is yet another way to understand Rawls’s invocation

of practical reason in the constructivist argument.

33
Do I have to be Coherent to be Reasonable?

Estlund could protest that such a justification fails to extend to large swaths

of society.56 Reasonableness, he may admit, is an attractive standard of accept-

ability to those who desire a society like the one Rawls envisages–a society

governed by a shared conception of justice that secures basic rights, liberties,

and means–but what about those who do not have an interest in such things?

What about the Branch Davidian that believes the end is near and that estab-

lishing peace on earth is therefore not worth the trouble? More pointedly, what

about the Neo-Nazi who rejects the political values of freedom, equality, and

fairness?

Rawls’s curt response is that such persons are not reasonable.57 At first

glance, this response fails miserably. Of course Rawls can characterize reason-

ableness such that Branch Davidians and Neo-Nazis do not qualify. If Rawls is

free to characterize persons however he wants, then any utopian dream can be

shown to attain unanimous consent and stability. The question that Estlund

has posed is how to address the unreasonable, not how to show that a specially

defined group can endorse a specially tailored doctrine of justice.

To view Rawls’s response as trivial, however, is to misunderstand the nature

and goals of his project. Indeed, if Rawls were to place no constraints on his

characterization of citizens or doctrines, then the task he aims to accomplish

would be trivial. But Rawls does place constraints on this characterization.

56 See Huemer (1996).
57 See Rawls (2005), p. 138.

34
Do I have to be Coherent to be Reasonable?

Citizens are characterized by a deep but reasonable diversity, where “reasonable”

is fixed as a minimal, albeit loose, concept. The goal of Rawls’s project is to

show the possibility of a well-ordered society under such conditions.58 Given the

kind of diversity Rawls has in mind, especially with respect to diverse religions,

this is not a trivial task. If Rawls can show how the reasonable Muslim and the

reasonable secular atheist can come to mutually acceptable terms, he will have

accomplished a great deal. Of course, as with all possibility proofs, Rawls’s

starts with certain assumptions.59 Progress is made when we can weaken these

assumptions, but that does not detract from the achievement made by the

original proof.60 Therefore, Rawls’s project may succeed even if it does not

attempt to convince the unreasonable–e.g. the Branch Davidian or the Neo-

Nazi–on terms that the unreasonable would find persuasive. Indeed, it would

succeed even if it were merely a stepping stone towards proving the possibility

of fair cooperation among an ever-deeper diversity of citizens.

58 Under some readings of Rawls’s project, the notion of well-orderedness is untenable. It

may be replaced with a society characterized by public legitimacy. See Gaus and van Schoe-

landt (Forthcoming).
59 See Rawls (2005), p. 104 – “Not everything, then, is constructed; we must have some

material, as it were, from which to begin.”

60 Rawls (2005) himself describes his project in the language of possibility proofs, stating that

justice as fairness can be thought of as “the defense of the possibility of a just constitutional

democratic regime” (p. 101).

35
Do I have to be Coherent to be Reasonable?

5 Two Objections

An important concern might be that what we have described within political

constructivism is simply a functionalist concept of truth. On this view, part

of what constitutes the truth role is acceptable belief and assertion after de-

liberation. Thus, regardless of how various comprehensive doctrines actually

characterize the relationship between reasonable acceptability and truth, they

should be viewed as two names for the same concept since they play the same

functional role.61 In accordance with what we have said so far, a functionalist

view of truth is just one more position that argues that political constructivism

has made claims or has employed principles that are conceptually connected to

truth, this time identifying reasonable acceptability with truth. Our response

is the same as before. Even though truth functionalism may be a correct view

of the concept of truth, political constructivism need not take a position on

whether it is correct. There are those within the overlapping consensus that

will disagree with truth functionalism, and since our task is specifically polit-

ical, it would be ineffective for creating political principles that are approved

by all reasonable peoples to exclude proponents of a particular comprehensive

doctrine based on its incorrect views on truth. Thus, even if truth functional-

ism is true, it should not prevent the Rawlsian constructivist from endorsing No

Concept within the overlapping consensus.

61 This is one possible interpretation of Joseph Raz’s (1990) position, that to deem a principle

of justice acceptable is simply to affirm it “as a true, or reasonable, or valid theory of justice”

(p. 15).

36
Do I have to be Coherent to be Reasonable?

Another important worry concerns a distinct interpretation of the challenges

of Raz, Cohen, and Estlund. Suppose that Raz, Cohen, and Estlund are correct

in asserting that, taken alone, the argument that Rawls provides for a concep-

tion of justice is incoherent in virtue of its eschewal of truth. Now, conceptual

incoherence is a very strange state to require of reasonable people. The over-

lapping consensus should, of course, expand to both coherent and incoherent

comprehensive doctrines, so political constructivism cannot itself be incoherent.

In other words, if political constructivism is incoherent, then a society in which

its constructed political conception of justice prevails will fail to satisfy the pub-

licity condition. There will be individuals and doctrines that cannot endorse

the public justification in virtue of its incoherence.

Do Raz, Estlund, and Cohen’s criticisms have any more bite once it is admit-

ted that political constructivism cannot be incoherent? No. The reason why is

that, as noted above, political constructivism is a radically incomplete doctrine.

It is, as Rawls put it, “political, not metaphysical,” and it is therefore up to each

particular comprehensive doctrine to flesh out the foundations of the political

justification.62 They can do so in a way that is coherent or incoherent, while

still remaining reasonable. We have already seen that political constructivism is

not detailed enough to rule out a plurality of viewpoints ranging from different

analyses of truth to different understandings of the grounds of moral obliga-

62 See Rawls (1999), p. 388.

37
Do I have to be Coherent to be Reasonable?

tion and norms of assertion. Given that these diverse viewpoints will flesh out

the connection between the prevailing political conception and truth in diverse

ways, the claim that political constructivism is incoherent cannot hold. Some

doctrines may complete the justification in incoherent ways. Others may sup-

plement it with coherent theories of truth that satisfy Raz, Cohen, and Estlund.

Critically, however, we want individuals and doctrines of both types to be party

to the overlapping consensus insofar as they are willing to cooperate according

to fair terms.

In fact, coherence (of a certain type) characterizes the overarching goal of

Political Liberalism, which, recall, seeks to show how a unifying political con-

ception can be compatible with a deeply diverse, though reasonable, citizenry.

In his “Reply to Habermas” Rawls details his account of justification and the

crucial role it plays in fostering consensus and stability (“for the right reasons”).

Justification, on Rawls’s account, occurs on three levels. First, pro tanto justi-

fication occurs when a conception of justice is derived from political values and

offers reasonable answers to all pressing political problems. Second, full justifi-

cation, which occurs on the individual level, is realized when citizens are able to

support or reconcile the political conception with their private comprehensive

doctrines. Finally, public justification occurs when citizens take account of one

another’s endorsement of the political conception and ascribe normative weight

to the fact that the political conception functions as the focus of an overlapping

consensus. The additional normative weight engenders increased stability and

38
Do I have to be Coherent to be Reasonable?

legitimacy, thus helping to fulfill the aspirations of Rawls political project. This

account of justification makes clear that coherence is not only a possibility for

political liberalism, due to its incomplete nature, but that that the coherence of

the political conception with a wide array of diverse comprehensive doctrines is

actually a critical goal, or perhaps an inviolable constraint, of Rawls’s project.

Conclusion

Raz, Cohen, and Estlund may be right that assertion, belief, and moral grounds

are conceptually connected to truth, but we have seen that this does not under-

mine Rawls’s avoidance of truth in his political constructivism. On the level of

a political conception, the question of truth and its connection to justification

is best left untreated. On the level of comprehensive doctrines, we must ac-

knowledge that reasonable people can be incoherent, and because of this, they

may fail to grasp various conceptual connections to truth or may endorse in-

correct theories of such conceptual connections. This failure, however, does not

prevent them from participating in the overlapping consensus, and thus Rawls’s

constructivism may stand as an inclusive method for political deliberation.

39
Do I have to be Coherent to be Reasonable?

References

[1] Adler, Jonathan. 2002. Belief’s Own Ethics. Cambridge, MA: MIT Press.

[2] Barry, Brian. 1995 “John Rawls and the Search for Stability.” Ethics 105:

pp. 874-915.

[3] Bonjour, Laurence. 2002. Epistemology: Classical Problems and Contem-

porary Responses. Lanham, MD: Rowman and Littlefield.

[4] Brandom, Robert. 2001. “Modality, Normativity, and Intentionality.” Phi-

losophy and Phenomenological Research 63, no. 3: pp. 611-623.

[5] Chan, Timothy (ed.). 2013. The Aim of Belief. Oxford University Press.

[6] Cohen, Joshua. 2009. “Truth and Public Reason.” Philosophy and Public

Affairs 37, no. 1: pp. 2-42.

[7] DeRose, Keith. 1991. “Epistemic Possibilities,” Philosophical Review 100:

pp. 581âĂŞ605.

[8] DeRose, Keith. 1996. “Knowledge, Assertion, and Lotteries,” Australasian

Journal of Philosophy 74: pp. 568âĂŞ580.

[9] DeRose, Keith. 2002. “Assertion, Knowledge, and Context.” Philosophical

Review 111, no. 2: pp. 167-203.

[10] Dreben, Burton. “On Rawls on Political Liberalism.” In The Cambridge

Companion to Rawls. Edited by Samuel Freeman. Cambridge University

Press.

40
Do I have to be Coherent to be Reasonable?

[11] Estlund, David. 1998. “The Insularity of the Reasonable: Why Political

Liberalism Must Admit the Truth.” Ethics 108, no. 2: pp. 252-275.

[12] Gardiner, Barry S.. 1985. “Rawls on Truth and Toleration.” The Philosoph-

ical Quarterly 38, no. 150: pp. 103-11.

[13] Gaus, Gerald and Chad van Schoelandt. Forthcoming. “Consensus on

What? Convergence for What? Four Models of Political Liberalism.”

Ethics.

[14] Gaus, Gerald and John Thrasher. “Rational Choice in the Original Position:

The (Many) Models of Rawls and Harsanyi.” In The Original Position.

Edited by Timothy Hinton. Cambridge University Press.

[15] Habermas, Jurgen. 1995. “Reconciliation Through the Public Use of Rea-

son,” Journal of Philosophy 92: pp. 109–31.

[16] Hawthorne, John. 2003. Knowledge and Lotteries. Oxford University Press.

[17] Hazlett, Allan. 2010. “The Myth of Factive Verbs.” Philosophy and Phe-

nomenological Research 80: pp. 497-522.

[18] Hampton, Jean. 1989. “Should Political Philosophy Be Done without Meta-

physics?” Ethics 99, no. 4: pp. 791-814.

[19] Huemer, Michael. 1996. “Rawls’s Problem of Stability.” Social Theory and

Practice 22, no. 3: pp. 375-395.

[20] Klosko, George. 1997. “Political Constructivism in Rawls’s Political Liber-

alism.” American Political Science Review 91, no. 3: pp. 635-646.

41
Do I have to be Coherent to be Reasonable?

[21] Kogelmann, Brian. Forthcoming. “Justice, Diversity, and the Well-Ordered

Society.” Philosophical Quarterly.

McKinnon, Rachel. 2015 The Norms of Assertion: Truth Lies and Warrant.

Palgrave Macmillan.

[22] O’Neill, Onora. 2002. “Constructivism in Rawls and Kant.” In S. Freeman

(Ed.), The Cambridge Companion to Rawls. Cambridge: Cambridge Uni-

versity Press, pp. 347-367.

[23] Peirce, C.S. 1878. “How to Make Our Ideas Clear.” In The Nature of Truth:

Classic and Contemporary Persectives. Ed. Michael Lynch. MIT Press,

2001: pp. 193-210

[24] Peirce, C.S. 1902. “Truth and Falsity and Error.” The Collected Paper of

Charles Sanders Peirce, vol. 5, ed. Charles Hartshorne and Paul Weiss.

Harvard University Press, 1965: pp. 565-573.

[25] Plantinga, Alvin. 1982. “How to be an Anti-Realist.” Proceedings and Ad-

dresses of the American Philosophical Association 56: pp. 47-70.

[26] Railton, Peter.1994. “Truth, Reason, and the Regulation of Belief.” Philo-

sophical Issues 5: pp. 71-93.

[27] Rawls, John and Freeman, Samuel. 1999. John Rawls: Collected Papers.

Harvard University Press.

[28] Rawls, John. 1997. “The Idea of Public Reason Revisited.” The University

of Chicago Law Review 64, no. 3: 765-807.

42
Do I have to be Coherent to be Reasonable?

[29] Rawls, John. 2005. Political Liberalism. Columbia University Press.

[30] Raz, Joseph. 1990. “Facing Diversity: The Case of Epistemic Abstinence.”

Philosophy and Public Affairs 19, no.1: pp. 3-46.

[31] Shah, Nishi. 2003. “How Truth Governs Belief,” Philosophical Review 112:

pp. 447-482.

[32] Stanley, Jason. 2005. Knowledge and Practical Interests. Oxford University

Press.

[33] Tienson, John. 1974. “On Analysing Knowledge.” Philosophical Studies 25:

pp. 289-293.

[34] Unger, Peter. 1975. Ignorance: A Case for Skepticism.

[35] Velleman, David. 2000. “On the Aim of Belief.” In The Possibility of Prac-

tical Reason. Edited by David Velleman. Oxford University Press.

[36] Velleman, David. 2005. “Doxastic Deliberation,” Philosophical Review 114:

pp. 497-534.

[37] Wedgwood, Ralph. 2002. “The Aim of Belief.” Philosophical Perspectives

16: pp. 267-297.

[38] Whiting, Daniel. 2010. “Should I believe the Truth?” Dialectica 64, no. 2:

pp. 213-224.

[39] Whiting, Daniel. 2013. “Nothing but the Truth: On the Norms and Aims of

Belief.” In The Aim of Belief. Edited by Timothy Chan. Oxford University

Press.

43
Do I have to be Coherent to be Reasonable?

[40] Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy.

Princeton University Press.

[41] Williamson, Timothy. 1996. “Knowing and Asserting,” Philosophical Re-

view 105: pp. 489–523.

[42] Williamson, Timothy. 2002. Knowledge and its Limits. Oxford University

Press.

[43] Wenar, Leif. 1995. “Political Liberalism: An Internal Critique.” Ethics 106,

no. 1: pp. 32-62.

[44] Wright, Crispin. 2001. “Minimalism, Deflationism, Pragmatism, Plural-

ism.” In The Nature of Truth. Edited by Michael Lynch. MIT Press.

[45] Wright, Cripsin. 1992. Truth and Objectivity. Harvard University Press.

44