Political Liberalism's Normative Foundations

Political Liberalism’s Normative Foundations

1. Introduction

Many, and perhaps most political liberals resist attempts to

inquire into the normative foundations of political

liberalism.1 John Rawls’s account, for example, famously

1 See in particular John Rawls, “Justice as Fairness: Political Not

Metaphysical”, Philosophy and Public Affairs 14 (1985): 223-251; Political Liberalism,

expanded edition (New York: Columbia University Press, 2005), 89-129.

Henceforth PL. Jonathan Quong, Liberalism Without Perfection (Oxford: Oxford

University Press, 2011), 221-255. Henceforth LWP. Older views include

Bruce Ackerman, Social Justice in the Liberal State (New Haven: Yale University Press,

1980), 1-24; Thomas Nagel, Equality and Partiality (New York: Oxford University

Press, 1991), 154-168. One very important exception might be Charles

Larmore. See “The Moral Basis of Political Liberalism”, The Journal of

Philosophy 96 (1999): 599-625. I think it is possible, however, that

Larmore’s attempt to found political liberalism on the basis of a principle

of ‘respect for persons’ might amount to what Joshua Cohen calls a

‘political conception of truth’, and is therefore compatible with the

spirit of Rawlsian political liberalism. “Truth and Public Reason”,

Philosophy and Public Affairs 37 (2009): 2-42. Stephen Macedo seems to criticise

the ‘reticent liberal’ for being unwilling to engage questions about the

relationship between liberal politics and comprehensive doctrines. Liberal

disavows any commitment to ‘truth’, holding that

‘reasonableness is its standard of correctness, and given its

political aims, it need not go beyond that.’2 Instead of

providing a foundation for the concept of reasonableness,

Rawls holds that ‘it is up to each comprehensive doctrine to

say how its idea of the reasonable connects with its concept

of truth, should it have one.’3 Jonathan Quong names this

justificatory strategy ‘buck-passing.’4

Buck-passing has its virtues. If political liberalism were to

offer a single justification for the concept of

reasonableness, it would risk entangling itself with claims

made by different reasonable comprehensive doctrines, whose

Virtues (Oxford: Clarendon University Press, 1990), 51-64. However, I think

these remarks could be accommodated within a Rawlsian framework, if they

are understood as advocating for arguments from conjecture. See PL, 463f.

Elsewhere, Macedo, claims that political liberalism does not rely on ‘the

justifiability of any particular comprehensive ideal or view of the whole

truth.’ “Liberal Civic Education and Religious Fundamentalism: The Case of

God v. John Rawls?” Ethics 105 (1995): 477.

2 PL, 127.

3 PL, 94.

4 LWP, 230.

endorsement political liberalism requires.5 Doing so would

risk allowing political liberalism to degenerate into a

comprehensive liberalism, where one can endorse liberalism

only by accepting a single, particular comprehensive doctrine

as true. And comprehensive liberalism, so understood, is

incompatible with the liberal principle of legitimacy, which

requires that the constitutional essentials must be acceptable

to all reasonable people, not just those who endorse one

particular comprehensive doctrine over others.6

Buck-passing, I allow, succeeds insofar as it allows political

liberalism to avoid making claims in favour of some reasonable

comprehensive doctrines over others as far as the

specification of the content of reasonableness is concerned. I

argue, however, that the function of reasonableness in political

liberalism draws on claims that do favour some reasonable

comprehensive doctrines and disenfranchise others. In fact, I

will argue that it draws on a claim that some reasonable

comprehensive doctrines necessarily reject.

5 PL, 100n.

6 See PL, 137.

Specifically, I will argue that the function of reasonableness

in political liberalism commits it to a principle that can be

illustrated by the following example. Suppose that Amy

chooses to pursue a comprehensive conception of the good life,

C. According to this principle, it is permissible for Amy to

pursue C just in case she has chosen it freely. The question

of whether C is true or correct has no bearing on the moral

permissibility of her chosen way of life.

Let us call this principle Choice, and formulate it as follows:

Reasonable people are morally permitted to pursue their

conception of the good life if this conception is one

that they have freely chosen.

We should note that Choice is not something that Rawls

explicitly endorsed, nor would he have likely accepted it as a

requirement of political liberalism. Although he does

acknowledge that the conception of the person in political

liberalism entails that citizens make about their conception

of the good are to be regarded as ‘self-authenticating’, it

seems that they should only be regarded as such ‘from a

political point of view’.7 If, however, political liberalism

is committed to Choice, it follows that this qualification is

unnecessary. Political liberalism holds that citizens – or at

least, reasonable citizens – are self-authenticating sources

of value, period.

Note also that Choice is incompatible with many, and perhaps

most familiar comprehensive doctrines, generally regarded by

political liberals as reasonable.8 A Catholic person who lives

out a Catholic way of live will hold that it is morally

permissible (and indeed, obligated) to live this way, because

the claims that Catholicism makes about God and the world are

true. Likewise, consider a secular comprehensive doctrine

such as utilitarianism. Utilitarians must necessarily hold

that it is morally permissible to live according to the7 PL, 32f.

8 On the reasonableness of Catholicism, at least in its post-Vatican II

form, see PL, 438, 477n; Leslie Griffin, "Good Catholics Should be Rawlsian

Liberals", Southern California Interdisciplinary Law Journal 5 (1997): 297-371..

Regarding the reasonableness of utilitarianism, see PL, 170. Quong,

however, appears more sceptical of the reasonableness of utilitarianism.

See LWP, 185f.

principle of utility just because utilitarianism is in some

way true or correct as a moral doctrine. In both cases, the

fact that a person chooses to be a Catholic or a utilitarian

is neither here nor there; their lives are conducted in a

morally permissible way insofar as they conform to a

comprehensive doctrine that is true.

One further clarification concerning the meaning of Choice is

appropriate. Rawls and all reasonable comprehensive doctrines

would agree that Amy, as a reasonable person, is free to

pursue C in the sense that her pursuit of C ought to be safe

from the interference of coercive state power. Catholics and

utilitarians may ascribe value to some notion of ‘freedom of

conscience’ in this respect. But this does not, by itself,

imply anything about the more complete moral status of Amy’s

pursuit of C – that is, whether it is actually morally

permissible for her to do so. My point, however, is that the

way in which reasonableness functions as a justification for

the former implies the latter. That is, the justificatory

foundations of political liberalism entail a commitment to

Choice, even if political liberalism makes every effort to

avoid such a claim.

In the next section, I briefly outline the content and

function of reasonableness in political liberalism.

Importantly, the function of reasonableness centres around its

identification of a selection of people, namely, embodied

reasonable people, who form the constituency of political

liberalism. The third section considers Jonathan Quong’s

“spare wheel” objection to this claim. The fourth section

shows how the fact that real reasonable people are the

constituency of political liberalism entails that it is

committed to Choice. In the fifth section, I briefly explore

what significance Choice might have for the viability of

political liberalism.

2. Reasonableness: content and function

Rawls outlines the concept of reasonableness in terms of its

content. This concept consists of two features:

Persons are reasonable in one basic aspect when, among

equals say, they are ready to propose principles and

standards as fair terms of cooperation and to abide by

them willingly, given the assurance that others will

likewise do so.9

The second basic aspect . . . is the willingness to

recognise the burdens of judgment and to accept their

consequences for the use of public reason in directing

the legitimate exercise of political power in a

constitutional regime.10

Some have argued that familiar comprehensive doctrines will be

unable to accept the concept of reasonableness, so understood.

Samuel Scheffler, for instance, casts doubt over the capacity

of utilitarians to abide by fair terms of cooperation, since,

as the early Rawls acknowledged, utilitarianism rejects the

idea of reciprocity.11 Leif Wenar, argues that most major

9 PL, 49.

10 PL, 54.

11 Samuel Scheffler, “The Appeal of Political Liberalism”, Ethics 105 (1994):

10; John Rawls, A Theory of Justice (Harvard: Belknap University Press, 1971),

religious traditions reject the burdens of judgment, since a

religion ‘characteristically presents itself as universally

accessible to clear minds and open hearts.’12 In other words,

the fact of reasonable pluralism is incompatible with a view

that reason renders a decisive verdict in favour of one’s own

doctrine.

The strategy of buck-passing enables political liberalism to

provide these objections with a response, of sorts. If

comprehensive doctrines are capable of accepting the two

concepts of reasonableness, so much the better for them. But

if they are not, then they are unreasonable doctrines. The

fact that this would render many familiar comprehensive

doctrines unreasonable may strike us as highly counter-

intuitive13, but it does not, by itself, render political

liberalism incoherent, in the sense that it ensures that

political liberalism remains acceptable to those (select few)14.

12 Leif Wenar, “Political Liberalism: An Internal Critique”, Ethics 106

(1995): 44.

13 See for example David Enoch, “Against Public Reason”, in Oxford Studies in

Political Philosophy, ed. David Sobel, Peter Vallentyne, Stephen Wall (Oxford:

Oxford University Press, 2013), 122.

whom it regards as reasonable. Buck-passing saves political

liberalism from taking a stance on its own foundations, albeit

at the cost of an ever-shrinking constituency of

justification.

So much for the content of reasonableness. But what is its

function in political liberalism? First and most obviously,

it provides a standard of acceptability that forms part of the

liberal principle of legitimacy. Rawls formulates this as

follows:

Our exercise of political power is fully proper only when

it is exercised in accordance with a constitution the

essentials of which all citizens as free and equal may

reasonably be expected to endorse in the light of

principles and ideals acceptable to their common human

reason.14

The concept of reasonableness thus sets a standard of

acceptability for principle selection and institutional

design, where the constitutional essentials are concerned.14 PL, 137.

But it has functions that go beyond this. The concept of

reasonableness also sets a standard of acceptability for the

selection of the theory itself. That is, the procedure or

‘framework’, in accordance with which political liberalism

selects principles must also be acceptable to reasonable

people.15 This seems to be a requirement of Rawls’s publicity

condition, according to which it is not enough for a political

conception of justice to be acceptable to reasonable people;

rather, its justification, too, must always be ‘publicly

known, or at least . . . publicly available’.16 Hence, the

liberal principle of legitimacy itself must be acceptable to

reasonable people. Reasonable people must be able to accept

the claim that political legitimacy requires that the

constitutional essentials must be acceptable to other

reasonable people.

Now this may seem like a requirement that stretches the

concept of reasonableness beyond its content, as specified by

the two features that Rawls identifies. But this does not

15 See Thomas Besch, “Political Liberalism and Public Justification: The

Deep View”, 2. Available at https://whu-cn.academia.edu/ThomasBesch

16 PL, 67.

seem to be the case. As Quong points out, the idea of

fairness implies reciprocity, meaning that the first feature

tells us that reasonable people must be committed to a

justification of the constitutional essentials that is

acceptable to all other reasonable people.17 The second

feature, meanwhile, informs us that reasonable people accept

the fact of reasonable disagreement, and so limits the kind of

values and principles that can be acceptable to all other

reasonable people.

However, the function of reasonableness as a standard of

acceptability does not end here. Rawls’s characterisation of

the point of view from which justification ought to take place

is not. In response to communitarian criticism over the

conception of the person presupposed by the idea of the

original position in A Theory of Justice, he distinguishes three

different points of view:

that of the parties in the original position, that of

citizens in a well-ordered society, and finally that of

ourselves - of you and me who are elaborating justice as17 LWP, 38f.

fairness and examining it as a political conception of

justice.18

Rawls is careful to stress that the first two points of view

are ‘merely the artificial creatures inhabiting our device of

representation.’ It is the third point of view – that of you

and me as real, existing, embodied people as we are – that

political liberalism is addressed to. Now, if we read this in

such a way as to be consistent with the restricted scope of

justification found in the liberal principle of legitimacy, we

should take it to mean ‘you and me as reasonable people.’19

This tells us two important things. First, the constituency

of political liberalism consists of real, existing, embodied

reasonable people (as opposed to, say hypothetical or

idealised versions of reasonable people) are the constituency

of political liberalism. This does not mean, of course, that

all embodied people are reasonable, but rather, that the kind

of reasonable person to whom Rawls thinks political liberalism

18 PL, 28.

19 On this point, see Thomas Besch, “Political Liberalism, The International

Conception, and the Problem of Public Dogma”, Philosophy and Public Issues 2

(2012): 161.

needs to justify itself is an embodied reasonable person.

Second, real reasonable people provide the ultimate standard

of justification in political liberalism, meaning that

political liberalism in its entirety must be acceptable to

this constituency. Not only questions regarding institutional

design or principles of justice, but every aspect of the

theory, including its framework, must meet this standard of

justification. As Rawls puts it:

Here the test is that of reflective equilibrium: how well

the view as a whole articulates our more firm considered

convictions of political justice, at all levels of

generality, after due examination, once all adjustments

and revisions that seem compelling have been made.20

That this limits the extent to which Rawls’s political

liberalism can be read as exclusively concerned with ideal theory

should be clear.21 To be sure, Rawls focuses considerable

20 PL, 28.

21 For a fuller explanation of the way in which Rawls’s political liberalism

transcends the limitations of ideal theory, see Besch, “Political

Liberalism and Public Justification: The Deep View”.

attention on the kinds of principles that might be acceptable

to idealised citizens in a well-ordered society, where they

are assured of full compliance.22 But if acceptability by

embodied reasonable people is truly the ultimate standard of

justification in political liberalism, the choices that these

idealised citizens might make in these idealised circumstances

carry normative authority only if real reasonable people agree

that they would make similar choices in similar circumstances.

This, then, is the real function of reasonableness. It

identifies a particular selection of real existing people, to

whom political liberalism accords nothing less than a genuine

‘veto’ power. Each aspect of political liberalism –

institutional design, principles, and justificatory framework

– can be justified only through their acceptance of it. This

means that political liberals can only rely on buck-passing

where the function of reasonableness is concerned if each

aspect of its function continues to be acceptable to all

reasonable people.

22 PL, 35-40.

In section 4, I will argue that this is not the case. The

fact that reasonableness has this extraordinary function in

political liberalism means that political liberalism is

committed to Choice – a claim which many, and perhaps most

reasonable people reject. First, however, it is important for

us to address Quong’s “spare wheel” objection, since it so

directly challenges what I have claimed thus far.

3. The spare wheel objection

According to Quong, the appeal to real reasonable people is

redundant in political liberalism, since it does not perform

any normative work:

It is the two aspects of the reasonable as defined by

Rawls that ensures a distinctively liberal outcome; the

assent of real citizens does not appear to be essential

to the project at all. The idea of justification aimed at

real citizens is thus nothing more than a spare wheel,

and so we can call this the spare-wheel objection to the

external conception.23

23 LWP, 146.

For Quong, it seems, it is enough that political liberalism

can meet the standard of justification that Rawls identifies

as the second standpoint, namely, that of idealised citizens

of a well-ordered society. Since we modified the third

standpoint to refer only to real reasonable people, it seems

fitting that we should do the same with this second

standpoint, so that it refers only to idealised reasonable

citizens of a well-ordered society.

Let us now consider how political liberalism is changed if

Quong’s standard of justification is adopted by political

liberalism. We could distinguish between two standards as

follows:

S1: Justification requires acceptability to all real

reasonable people.

S2: Justification requires acceptability to all reasonable

idealised citizens of a well-ordered society.

S1 represents the standard to which, I am arguing, Rawls’s

political liberalism is committed. S2, on the other hand,

represents the standard that Quong believes is sufficient.

Admittedly, S2 has an important role for Rawls, insofar as he

is at least partly interested in what principles would be

agreed to where the parties can be assured of full

compliance.24 But if S1 is the ultimate standard of

justification in political liberalism, any role that S2 might

have results from the fact that it itself meets the standard

of S1.

Now the point of Quong’s objection, I think, is that there is

no significant difference between S1 and S2. If we formulate

S2 in such a way that requires acceptability to idealised

reasonable citizens, then whatever will meet the standard of S2

will also meet S1, and vice versa. Both standards set a

standard of reasonable acceptability, according to which X

meets the standard just in case X is at least consistent with

the two features of the reasonable person. The design of the

constitutional essentials, the principles of justice, and the24 See PL, 35-40.

liberal principle of legitimacy itself seem entirely

unaffected by the selection of a standard of justification.

Their content under a political liberalism where S1 is the

standard would seemingly be identical to their content under a

political liberalism where S2 is the standard. For this

reason, it seems, Quong regards the appeal to real reasonable

people as an unnecessary ‘spare wheel’.25

There is, however, an important difference between S1 and S2.

If S1 is truly the ultimate standard of justification, then any

role that S2 might have is contingent on its own capacity to

meet the standard of S1. If S2 was the ultimate standard of

justification in political liberalism, then S2 would be exempt

from the standard of acceptability to real reasonable people

(S1). If real reasonable people were dissatisfied with the

idea that the standard of acceptability should be identified

with idealised reasonable citizens of a well-ordered society,

then this would be of no consequence as far as the

justification of S2 is concerned.

25 LWP, 146.

True, Quong might argue that real reasonable people cannot

remain reasonable without accepting S2, and that the

distinction between it and S1 has no practical significance.

Let us assume, perhaps optimistically, that S2 itself is

capable of meeting the standard of S1. Even if this is so (and

I’m not so sure that it is), the fact that S1 is the standard

of justification in political liberalism means that real

reasonable people possess a kind of authoritative status that

they would not otherwise have.

I think Thomas Besch’s recent work on the idea of discursive

respect can help us see how this is the case. A theory

accords people discursive respect when it attempts to relate

to others on the basis of reasons that must be, in some

relevant sense, acceptable to them.26 Political liberalism is

clearly one such theory. Importantly, however, the value of

discursive respect – or rather, the extent to which discursive

respect really conveys respect – can vary for individuals to

whom it is accorded. In particular, it can be affected by two

26 Thomas Besch, “On Discursive Respect”, Social Theory and Practice 40 (2014):

207.

main factors: purchase, and depth.27 The greater degree of

purchase and depth, the greater the value of discursive

respect, other things being equal.28

The idea of purchase refers to the particular notion of

acceptability at hand. An actualist conception, for example,

requires acceptability to real people as they are, while a

counterfactualised conception requires acceptability to some

idealised version of people, such as a more rational or

coherent version of themselves.29 Needless to say, an

actualist conception generates a greater degree of purchase

and, other things being equal, a form of discursive respect

that is more valuable for those to whom it is addressed, than

a counterfactualised conception.

27 Besch calibrates discursive respect in terms of three factors: depth,

purchase, and scope. A theory has a broad scope of discursive respect if

its constituency includes a wide constituency of people, as opposed to a

very select group. “On Discursive Respect”, 214-216. I do not discuss

scope here, since a wider scope might increase the overall value of

discursive respect, but does not necessarily increase its value for those

who are already included within the scope.

28 Besch, “On Discursive Respect”, 224.

29 Besch, “On Discursive Respect”, 216-220.

The idea of depth, meanwhile, refers to the material that must

be acceptable to the constituency. A deep form of discursive

respect will hold that not only the reasons for a particular

action must be acceptable, but also the principles governing

these actions. Discursive respect is deeper still if it holds

that all standards of practical justification must be

acceptable, while it becomes less deep if it only holds that

reasons for action must be acceptable.30

If S1 is the standard of justification for political

liberalism, then real reasonable people are the beneficiaries

of a deep, purchase-rich, form of discursive respect. It is

rich in purchase because it requires acceptability by real

reasonable people, rather than some idealised version of

themselves. It is deep, because this standard of

justification is the ultimate standard of justification. Both

the content of political liberalism and its framework need to

meet it.

30 Besch, “On Discursive Respect”, 213-214.

What about S2? It could be interpreted as suggesting that the

kind of acceptability required by political liberalism is only

a counterfactualising kind, addressing idealised reasonable

citizens of a well-ordered society, as opposed to real

reasonable people. However, if we assume that S2 can meet the

requirements of S1, then there is still a sense in which Quong

might claim that political liberalism offers an actualised

conception of acceptability. Real reasonable people can

accept S2, and hence, the entirety of a political liberalism

built on S2 as an ultimate standard of justification. In this

case, however, depth is compromised. Real reasonable people

do not possess full authority over the standards of

justification. Their acceptance or rejection of S2 has no

bearing on its becoming the ultimate standard of justification

in political liberalism. The fact that they might all happen

to accept it any way is neither here nor there. As a result

of this, the depth of discursive respect is compromised by

making S2 the standard of justification.

Let me explain this further by way of the following example.

Suppose a father takes his daughter to an ice cream stand.

The father knows that his daughter wants chocolate-flavoured

ice cream, and his daughter will also choose chocolate ice

cream if given the opportunity to do so. It seems, then, that

whatever decision-procedure is followed, the outcome will be

the same: his daughter gets chocolate ice-cream. But this

does not necessarily make the selection of a decision-

procedure redundant. Three options seem available:

i) The father picks an ice cream flavour for his daughter.

ii) The father lets his daughter choose a flavour for

herself.

iii) The father asks his daughter whether she would like to

choose the flavour herself, or whether she would prefer

him to choose one for her.

In this case, ii) displays a greater degree of depth of

discursive respect than i), while iii) is deeper than ii). In

iii), other things being equal, the daughter is maximally

enfranchised by the decision-procedure. She has veto-power,

not only over the choice of ice-cream, but over the choice of

decision-procedure for the choice of ice-cream. Now consider

the case of real reasonable people in political liberalism.

If S2 is the standard of justification, their status is

somewhat analogous to ii) in the example above: they have the

ability to select principles of justice, but their influence

over the selection of the procedure for principle-selection is

limited at best, given that they have no authority to accept

or reject S2. If S2 is capable of meeting the requirements of

S1, then they still ultimately get what they want, no less than

the daughter gets the ice cream she wants. But they receive a

substantially greater degree of enfranchisement if the

standard of justification is S1, which is analogous to iii).

Hence, greater discursive respect is shown to real reasonable

people if S1 is adopted as the ultimate standard of

justification for political liberalism, rather than S2.

It might still be an open question whether the most compelling

version of political liberalism adopts S1 over S2. The

advantage of the former is that it offers a greater degree of

discursive respect to real reasonable people. The idea that

this is an overwhelmingly important value for political

liberalism might be challenged. What I hope to have

demonstrated, however, is that the appeal to real reasonable

people is not an unnecessary spare wheel, as Quong suggests.

And if Rawls’s political liberalism holds that real reasonable

people are the constituency of justification for political

liberalism as a whole, then it must hold that S1 is the

ultimate standard of justification in political liberalism.31

4. S1 and Choice

We have seen that political liberalism, at least in its

Rawlsian form, is committed to taking S1 as the ultimate

standard of justification. I now intend to show that this

entails that it is committed to Choice. Recall that Choice holds

that a reasonable person is morally permitted to pursue her

chosen conception of the good life, as long as she freely

chose it. Recall also that most political liberals would31 Other prominent political liberals seem to share this same commitment.

Larmore’s attempt to found political liberalism on a moral principle of

‘respect for persons’ seems similar to the idea that real people ought to

be the recipients of a valuable form of discursive respect. See Larmore,

“The Moral Basis of Political Liberalism”. See also Macedo, Liberal Virtues,

39-77; Thomas Nagel, “Moral Conflict and Political Legitimacy”, Philosophy

and Public Affairs 16 (1987): 221.

likely resist the claim that political liberalism is committed

to Choice, even those who, like Rawls, present political

liberalism in such a way that it is committed to S1. As we

have seen, Rawls thinks that reasonable citizens are ‘self-

authenticating’, but only ‘from a political point of view’.

Macedo similarly holds that ‘the goodness of good reasons, for

a public moral theory, becomes entirely a function of their

capacity to gain widespread agreement among reasonable

people.’32 Both are slightly ambiguous in their language here,

but I think it likely that both would reject the idea that

political liberalism holds that good reasons are those that

reasonable people would accept, when it comes to questions

about comprehensive doctrines.

Why, then, should we think that S1 commits political liberalism

to Choice? The answer lies in the fact that S1 is the ultimate

standard of justification in political liberalism. In order

to maximise depth, and accord real reasonable people maximal

discursive respect, it provides a standard of justification

for all aspects of the theory. There can be no deeper

principle that governs S1; if there were, then it would no32 Macedo, Liberal Virtues, 46f.

longer be the case that the entirety of political liberalism

needs to be justifiable to real reasonable people. It would

also no longer be the case that political liberalism accords

reasonable people as high a level of discursive respect, since

its depth would now be substantially limited.

We have seen how the role of ideal theory in political

liberalism, as specified by S2, depends on its ability to be

acceptable to real reasonable people, and thus meet S1. The

same goes for the idea, so central to political liberalism,

that the domain of the political is unique and requires a

special standard of justification. Although Rawls and Macedo

sometimes seem to imply that S1 is a standard of justification

for the domain of the political only, this is a view that is

inconsistent with S1’s status as the ultimate standard of

justification in political liberalism. What they seem to be

doing, in effect, is invoking a principle that is lexically

prior to S1, which we might call P:

P: The exercise of political power requires a special

justificatory standard that applies exclusively to this

domain.

It is only by invoking P, or something like it, as a lexically

prior claim to S1 that Rawls and Macedo can claim that S1

applies exclusively to politics. But if S1 really is the

ultimate standard of justification in political liberalism,

operating at a maximally deep level, then it is the fact that

P meets the requirements of S1 that generates special

justificatory standards for politics. S1 is not itself

constrained by P; rather, P follows from S1.

Now the only way in which political liberalism could avoid

committing itself to Choice would be to constrain S1 with P. If

P does not provide such a constraint, S1 simply states that

acceptability to real reasonable people is what justifies,

period. It does not limit its applicability to politics. It

therefore includes the decisions that real reasonable people

make about which comprehensive doctrine they wish to pursue.

One might claim that the content of reasonableness itself

implies P. Fairness, according to the first feature of

reasonableness, is identified as necessary for ‘terms of

cooperation’, rather than as a broader, comprehensive value.33

Similarly, the second feature, concerning the burdens of

judgment, requires that reasonable people ‘accept their

consequences for the use of pubic reason in directing the

legitimate exercise of political power in a constitutional

regime.’34 It seems, then, that the concept of reasonableness

itself has some idea of politics as constituting a unique

domain, requiring a special standard of justification.35 This

does not, however, enable political liberalism to avoid

committing itself to Choice. It would simply mean that P, and

whatever else is entailed by the two features of

reasonableness, carries normative weight in political

liberalism only because it meets the justificatory standard of

S1.

33 PL, 49.

34 PL, 54.

35 I owe this point to an anonymous reviewer.

Another, perhaps more vexing problem for my claim that

political liberalism is committed to Choice concerns

paradoxical circumstances where Choice is rejected by certain

real reasonable people. Indeed, as I have earlier noted,

Choice is incompatible with many familiar comprehensive

doctrines that political liberals commonly take to be

reasonable. I do not deny that this poses a considerable

difficulty for political liberalism – in fact, this is the

very point of my argument here – but it arises because

political liberalism is committed to taking S1 as the ultimate

standard of justification. In fact, it is only because S1 has

this status in political liberalism that it is so concerned

with what real reasonable people might think in the first

place. Hence, I do not think that political liberalism can

avoid committing itself to Choice by pointing out that Choice is

rejected by many real reasonable people.

5. Two options

I have argued that political liberalism is committed to Choice,

in virtue of its commitment to taking S1 as its ultimate

standard of justification. Since many, and perhaps even most

real reasonable people reject Choice, this puts political

liberalism in a problematic position. Buck-passing, as we

have seen, presumes that the claim for which justification is

sought must be acceptable to reasonable people, and that a

person’s inability to provide herself with a justification for

this claim simply reveals her own unreasonableness. The fact

that reasonable people can coherently reject Choice therefore

shows that buck-passing cannot work in this particular case.

I will finish here by considering two ways in which political

liberalism might deal with this.

First, it might explore the possibility of abandoning S1, or at

least, not taking it to be the ultimate standard of

justification. The motivation for doing so might come from

the objection we considered at the end of the previous

section, which might be rephrased in the form of a reduction

ad absurdum. If political liberalism is takes S1 as its

ultimate standard of justification, then it arrives at an

absurdity. On the one hand, it says that any claim that

political liberalism makes is justified if and only if it is

acceptable to real reasonable people. On the other hand, it

is committed to a particular claim (Choice) that many real

reasonable people reject. The upshot of this, it might be

suggested, is that political liberalism cannot really take S1

as the ultimate standard of justification.

The immediate consequence of this move would be to diminish

somewhat the value of discursive respect for the constituency

of political liberalism, on account of a reduction of its

depth. Certain aspects of political liberalism, such as P,

which stipulates the importance of politics as a domain

requiring a special standard of justification, would no longer

be subject to the acceptance of real reasonable people. Yet

strangely, I would conjecture that many reasonable people

would respond more warmly to a political liberalism that was

not committed to Choice. Of course, this goes without saying

for those reasonable people who reject Choice.

This might seem like a relatively simple adjustment to the

order of justification in political liberalism. Rawls’s claim

that political liberalism as a whole needs to be justifiable to

real reasonable people should be modified slightly, so that

this standard of justification applies exclusively to the

domain of politics. But the sheer fact that this adjustment

might be necessary has some potentially disturbing

implications. It seems to suggest that many people, including

many reasonable people, do not regard discursive respect as an

overriding good, or even one that might rank alongside of

their comprehensive commitments. Perhaps many of these people

might think that Raz was right after all: they think they are

most properly treated with respect, not when others relate to

them on the basis of reasons that they can accept, but rather,

on the basis of reasons that reflect the truth.36 At the very

least, they seek a form of social cooperation that does not

deny that the authority of moral judgments can be determined

on acceptability-independent grounds. For a theory like

political liberalism that seems to put a premium on discursive

respect, this would be a strange concession, though perhaps

not fatal.

At the same time, it is worth remembering that there are some

reasonable people who may find Choice unobjectionable, and even36 Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press, 1986), 157.

accept it. Take individual subjectivists for example.

Although Choice does not strictly imply that individual

subjectivism is true, since it speaks only of moral

permissions, and not about moral truth or falsity, it does

seem plausible to expect that an individual subjectivist will

be quite comfortable, other things being equal, with a

political theory committed to Choice. The individual

subjectivist may well prefer this as an alternative to

abandoning S1 as the ultimate standard of justification, and

diminishing the value of discursive respect.

We are thus led towards a second option for political

liberalism: retain S1 as the ultimate standard of

justification, acknowledge a commitment to Choice, and deal

with the consequences. As we have just noted, some, such as

the individual subjectivist, will likely be untroubled by

this. But for the many reasonable citizens who reject Choice,

it seems to place them in a difficult position. It may be an

exaggeration to say that they are now forced to choose between

political liberalism and their comprehensive doctrine. But

the extent to which they would then identify with the project

of political liberalism might well now be called into

question. Their status as reasonable seems to compel them to

acknowledge the legitimacy of political institutions that are

designed in accordance with a proper ordering of political

values, and as such, it is unlikely that they would seek to

overthrow them. But it seems just as unlikely that they will

continue to view themselves as citizens in the fullest sense.

The notion of a resident alien, who, as Andrew March puts it,

‘does not seek to share political sovereignty with his

compatriots, does not identify with the political system, and

resents or is indifferent to the contribution he makes to

society’s welfare and security’, may be a more appropriate

description of their status.37 Political liberals should find

the prospect of a society full of reasonable people who see

themselves as resident aliens extremely unattractive, to say

the least. It is hard to imagine a scenario further from the

spirit of public justification.

37 Andrew March, “Liberal Citizenship and the Search for an Overlapping

Consensus: The Case of Muslim Minorities”, Philosophy and Public Affairs 34

(2006): 383.

Ultimately, I think political liberals will be inclined

towards the first option over the second, given that the

latter pays an extraordinary price for retaining S1 as the

ultimate standard of justification. But I think the fact that

this is the case should prompt further reflection about how

political liberalism can best meet its deepest normative

commitments. The motivation behind public justification, it

seems is to treat reasonable people with respect. But perhaps

what reasonable people want most from politics is to be

treated not just in ways that they can accept – important

though this may be – but in ways that can allow for the

intelligibility of robust moral commitments that are justified

on acceptability-independent grounds.