James Gledhill
Procedure in substance and substance in
procedure: reframing the Rawls–Habermas
debate.
Book section
Original citation:
Gledhill, James (2011) Procedure in substance and substance in procedure: reframing the
Rawls–Habermas debate. In: Finlayson, J. G. and Freyenhagen, F., (eds.) Habermas and
Rawls: disputing the political. Routledge studies in contemporary philosophy (23). Routledge,
New York, USA, pp. 181-199.
© 2011 Routledge
This version available at: http://eprints.lse.ac.uk/37091/
Available in LSE Research Online: July 2011
LSE has developed LSE Research Online so that users may access research output of the
School. Copyright © and Moral Rights for the papers on this site are retained by the individual
authors and/or other copyright owners. Users may download and/or print one copy of any
article(s) in LSE Research Online to facilitate their private study or for non-commercial research.
You may not engage in further distribution of the material or use it for any profit-making activities
or any commercial gain. You may freely distribute the URL (http://eprints.lse.ac.uk) of the LSE
Research Online website.
This document is the author’s submitted version of the book section. There may be differences
between this version and the published version. You are advised to consult the publisher’s
version if you wish to cite from it.
Procedure in Substance and Substance in Procedure:
Reframing the Rawls–Habermas Debate
James Gledhill
The suggestion that the Rawls–Habermas debate represents a “failure of two of the greatest contemporary minds
to meet” is not without justification.1 However, both the fact that Rawls and Habermas sometimes seem to be
talking past one another and the relative lack of attention the debate has received should be a source of regret, if
for no other reason than that Habermas offers a prism through which to view Rawls’s work that throws into sharp
relief some of its seldom acknowledged presuppositions. I will claim more than this, however. In forcing
reflection upon these assumptions, Habermas challenges the possibility of the relationship between philosophy
and politics that Rawls assumes.
This notwithstanding, there are significant obstacles to matching up, let alone arbitrating between, the
respective claims of Rawls and Habermas. One point of entry into the issue of procedural versus substantive
approaches on which I focus is the fact that Rawls and Habermas mean different things by the term “justice.”
Rawls is concerned with principles of social justice, that is, principles for ordering the arrangement of basic
social institutions that both assign rights and duties within this institutional framework and regulate the
distribution of the benefits and burdens of social cooperation.2 Habermas, on the other hand, seeks to separate
philosophical problems of equal right, such as equal status before the law, and political questions of distributive
justice. Principles of distributive justice respond to problems in the political domain and require contextual
specification. Principles of equal right, on the other hand, “can be grounded from the standpoint of
universalization, and claim prima facie validity,” even if the application of these principles must be sensitive to
context.3 Habermas’s understanding of justice is consequently more austere than Rawls’s. Justice is “nothing
material, no determinate ‘value,’ but a dimension of validity.”4 More specifically, “justice” is synonymous with
normative validity, the analogue in the normative domain of truth in the theoretical domain.
In examining the way in which Rawls and Habermas both propose a Kantian procedural conception of
practical reason without recourse to Kantian metaphysics I will argue that this shared starting point should not be
allowed to obscure the fundamental divergence encapsulated in their different understandings of justice. Rawls’s
articulation of principles of equal right, or equal liberty, through the Original Position is connected—indeed I
will argue inextricably connected—to a conception of social institutions and of the political relationship of
citizens within those institutions. This should put one in mind of Rousseau and Hegel at least as much as of Kant.
Habermas criticizes this approach in the name of a Kantian principle of universalizability and on the basis of a
conception of philosophy that is synonymous with the defence of the universality of reason. In grounding a
Final draft of article published in James Gordon Finlayson and Fabian Freyenhagen (eds.) Habermas and
Rawls: Disputing the Political, (New York: Routledge, 2011), pp. 181-199.
1
Jonathan Wolff, “In Front of the Curtain,” Times Literary Supplement, 7 March 2008.
John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971; revised edition, 1999), pp. 4/rev. 4.
3
Jürgen Habermas, “Discourse Ethics, Law and Sittlichkeit,” in Autonomy and Solidarity: Interviews with Jürgen
Habermas, ed. Peter Dews (London: Verso, revised and enlarged edition, 1992)., pp. 249-50.
4
Habermas, “Discourse Ethics, Law and Sittlichkeit,” p. 249.
2
1
principle of universalizability in the procedural presuppositions of the social practices of communicative reason,
and in turn distinguishing between the place of procedural rationality in morality and law, Habermas can claim to
show that a Kantian principle of equal right is only contingently connected to particular substantive conceptions
of well-ordered social institutions. If this is right then it challenges the widely held view, evident in the Rawls–
Habermas debate itself, that Rawls’s conception of justice is fundamentally universalistic and Kantian while
Habermas articulates a Hegelian comprehensive doctrine.
Rawls does not engage sufficiently with Habermas’s alternative elaboration of a procedural conception of
reason because he assimilates the question of normative proceduralism to the terms of Anglo-American legal and
democratic theory. On this understanding procedural justice refers to a normatively thin procedural fairness or
neutrality in arbitrating between substantive claims. However, just because Habermas is not a proceduralist in
this sense does not mean that his position is not importantly different from Rawls’s. I begin by arguing for
reframing the debate in terms of substantive versus procedural conceptions of practical reason rather than justice.
Employing this alternative framework, I show how Rawls’s constructivism employs the Kantian procedure of the
Original Position within the Hegelian substance of a shared social world. Habermas’s reconstructivism, on the
other hand, presupposes a social theory according to which Hegelian ethical substance has been fully
“sublimated” into Kantian discursive procedures. Although this analysis is not simply an exercise in
reinterpretation, my conclusion is somewhat ambivalent: Rawls makes a more powerful appeal to ‘our’ moral
sentiments, but Habermas offers a more cogent conception of the place and role of normative philosophical
reflection in complex modern societies. Since I follow Habermas in seeing the issues at stake as fundamentally
philosophical, my conclusion follows my philosophical head rather than my political heart.
1. Procedure and substance in justice and in practical reason
Habermas criticizes the adequacy of Rawls’s device of the Original Position as a way of representing a Kantian
moral point of view, arguing that Rawls should keep “the procedural conception of practical reason free of
substantive connotations by developing it in a strictly procedural manner.”5 Habermas’s concern is with the way
Rawls models the fairness of the moral point of view through normatively substantive reasonable constraints on
rational parties’ choice of principles of justice. He offers his discourse theory of morality, or Discourse Ethics, as
an alternative elaboration of a Kantian moral point of view, one that would, however, preclude the derivation of
substantive principles of distributive justice for a well-ordered society. For Habermas, as soon as Rawls seeks to
derive substantive principles he is “speaking as a citizen of the United States” for there are not “any nonprocedural, substantive principles which could be applied to everything at every time.”6 It is for this reason that
Habermas thinks that A Theory of Justice is not, or better is only partially, a philosophical work. The ambition of
Habermas’s alternative approach lies in its elaboration of a communicative conception of reason not restricted to
the domain of the political, a conception moreover concerned with the universality of normative principles.
5
Jürgen Habermas, “Reconciliation through the Public Use of Reason: Remarks on Rawls’s Political Liberalism,” Journal
of Philosophy, XCII, 3 (March 1995): 109–31, p. 116.
6
Jürgen Habermas, “Life-Forms, Morality and the Task of the Philosopher,” in Autonomy and Solidarity: Interviews with
Jürgen Habermas, ed. Peter Dews (London: Verso, revised and enlarged edition, 1992), pp 200–201.
2
After criticizing Rawls’s use of the Original Position from the perspective of his own discourse theory of
morality, Habermas draws upon his political (legal and democratic) theory to criticize Rawls’s understanding of
how the domain of the political is demarcated. He argues that the domain of the political cannot be identified
simply from a moral point of view; in a more orthodox Kantian fashion the “political” should rather be
differentiated in accordance with the criterion of legal regulation. The substantive content of basic liberal rights is
not derived from substantive principles of justice but is inherent in the legal form through which citizens’ public
use of reason is institutionalized. This is the sense in which Habermas’s reconstructivism is more modest than
Rawls’s constructivism, for it “focuses exclusively on the procedural aspects of the public use of reason and
derives the system of rights from the idea of its legal institutionalization.”7 The focus on procedures of rational
opinion- and will-formation leaves substantial questions to be decided by participants in discourse.
The problem with Habermas’s diagnosis of the shortcomings of Rawls’s approach is that Habermas gives
the impression that Rawls is damned if he does and damned if he doesn’t. That is, Rawls will have made a wrong
move both if he makes the full justification of his conception of justice depend upon its capacity to serve as the
focus of an overlapping consensus and if his stance of bare normativity leaves his theory detached from the
institutional realities of complex law-governed contemporary societies.8 The problem is that Habermas’s
preconceived view that Rawls’s theory is an alternative Kantian approach to his own leads him to overlook, or
minimize, fundamental points of difference. Habermas does not subject to scrutiny the way in which the
justification of Rawls’s Kantian procedure of the Original Position depends upon the substantive background of a
shared social world, and how the substantive principles so derived are intended to be realized within the basic
structure that forms that social world. To properly locate the area of disagreement it is necessary to move the
debate to a more abstract level, the point at issue concerning the alignment of principles of justice and social
contexts that is central to Rawls’s constructivism but that Habermas rightly finds problematic. Habermas’s
alternative Kantian reconstructive approach involves both distinguishing between moral theory and legal theory
and within these domains separating a reconstruction of procedural normative principles from a social theoretical
account of their contexts of realization.9
However, if Habermas’s criticisms tend to assimilate Rawls’s approach to his own framework, this is even
truer of Rawls’s reply. When he turns to the topic of procedural versus substantive justice Rawls does not discuss
how and why he characterized the Original Position in A Theory of Justice as a “procedural interpretation of
Kant’s conception of autonomy and the categorical imperative.”10 Neither, more surprisingly, does he take up the
broader and less controversial idea of the Original Position as an instance of pure procedural justice, commenting
that he will leave aside the special case of gambling that he had previously presented as the principal example of
pure procedural justice.11 What Rawls does do is argue forcefully that justice is never simply procedural but
always relies upon substantive principles. However, rather than engaging with Habermas’s view, Rawls adopts a
distinction between the justice of a procedure and the justice of its outcome that he takes from Stuart Hampshire.
7
Habermas, “Reconciliation through the Public Use of Reason,” p. 131.
Jürgen Habermas, Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy, trans. William
Rehg (Cambridge: Polity Press, 1996), pp. 56–66.
9
Habermas, “Reconciliation through the Public Use of Reason,” p. 127.
10
Rawls, A Theory of Justice, p. 256/rev. 226.
11
John Rawls, “Political Liberalism: Reply to Habermas,” Journal of Philosophy, XCII, 3 (March 1995): 132–80, p. 170.
8
3
Against Hampshire, and following Joshua Cohen, Rawls argues that procedural and substantive justice cannot be
separated: the justice of a procedure is partly dependent upon the justice of its outcomes.12
On this basis Rawls offers a two-pronged response to Habermas. First, he argues that Habermas himself
recognizes the dependence of democratic procedures on substantive principles when he argues that the procedure
of the public use of reason can be guaranteed to have reasonable outcomes only insofar as it realizes the
conditions of ideal discourse. Second, to the extent that Habermas’s view is procedural in focusing on legitimacy
rather than justice, it is inadequate. The procedural legitimacy of legislative enactments depends upon the justice
of the constitution. In the long term, legitimacy is self-undermining, for it allows an undetermined range of
injustice that would cause a society to cease to be well–ordered, allowing laws to be passed that would overstep
the bounds of justice that they must respect in order to be legitimate.
Rawls is certainly right that Habermas’s procedural idea of the public use of reason is not purely formal
and that it depends upon substantive content. Indeed, it relies upon the value of autonomy, or freedom as equal
intersubjective self-legislation, implicit in Habermas’s conception of communicative reason. The crucial question
is, though, whether the substantive content of normative procedures requires regulation by substantive principles
or whether such substantive principles are implicit within the procedures of moral discourse and democratic
procedures of the public use of reason. The approaches of Rawls and Habermas are both substantive in the sense
that they reject majoritarian views that seek to rely upon a normatively thin conception of democratic process that
is impartial between, and does nothing to shape, the pursuit of substantive ends. But while for Rawls it is a
corollary of this rejection of majoritarianism that one must endorse liberal constitutionalism, Habermas’s Kantian
Republicanism rejects this dichotomy.
That Rawls’s discussion of the contrast between procedural versus substantive justice misses the mark is
evident in Habermas’s comment that Rawls’s observations “do not capture the sense in which I use the
expression ‘procedure’ and ‘procedural rationality’ when I assert that a practice of argumentation instituted in a
certain way tends to support the assumption that its results are rationally acceptable.”13 To understand
Habermas’s conception of procedural reason, and Rawls’s, it is necessary to move from a contrast between
procedural versus substantive justice to one between strictly procedural and substantively informed philosophical
conceptions of practical reason. One will only be in a position to assess Rawls’s and Habermas’s contributions to
legal and democratic theory if one recognizes that the work of both begins from an understanding of
proceduralism as this is understood in moral theory. In this sense, Habermas’s more strictly Kantian procedural
approach may be contrasted with Rawls’s more Hegelian approach in which procedural Moralität is realized
within the Hegelian Sittlichkeit of the basic institutions that form our social world.
12
See Joshua Cohen, “Pluralism and Proceduralism,” Chicago-Kent Law Review, LXIX (1994): 589–618.
Jürgen Habermas, “‘Reasonable’ versus ‘True’, or the Morality of Worldviews,” in his The Inclusion of the Other: Studies
in Political Theory, eds. Ciaran Cronin and Pablo De Greiff, trans. Ciaran Cronin (Cambridge: Polity Press), p. 278n18.
Rawls’s idea of pure procedural justice has been used as a standard against which to assess Habermas’s discourse theories of
morality and law: see Christina Lafont, “Procedural Justice: Implications of the Rawls– Habermas Debate for Discourse
Ethics,” Philosophy and Social Criticism, XXIX (2003): 163–81; and Michel Rosenfeld, “Can Rights, Democracy, and
Justice Be Reconciled through Discourse Theory? Reflections on Habermas’s Proceduralist Paradigm of Law,” in Michel
Rosenfeld and Andrew Arato, eds., Habermas on Law and Democracy: Critical Exchanges (Berkeley: University of
California Press, 1998). However, neither the place of pure procedural justice in Rawls’s approach nor the basis of
Habermas’s normative proceduralism in a conception of procedural rationality have been examined sufficiently.
13
4
2. Constructivism, reconstructivism and the problem of self-grounding
In making the case for reframing the Rawls–Habermas debate I can only offer a brief overview of the wider
framework that informs my analysis, one that will be indirectly justified to the extent that it sheds light on the
procedure–substance distinction. I begin from the assumption that Rawls and Habermas work within and take
forward a modern rational agency tradition comprising Rousseau, Kant, Fichte, Hegel and Marx, a tradition
concerned with freedom as autonomy and the deliberative capacity of practical reason through which freedom
can be realized.14 Moreover, both Rawls and Habermas are concerned with the problem of self-grounding that
Hegel bequeaths to this tradition. This is the problem of applying the value of autonomy that principles of justice
seek to realize to the philosophical justification of those principles, and requires that these principles be shown to
be implicit within existing social practices. As Charles Larmore recognizes, Rawls and Habermas share the aim
of developing a self-standing or autonomous conception of the guiding principles of modern democracy.15
Larmore emphasizes the negative sense in which Rawls differentiates a political conception of justice from
comprehensive doctrines in describing the former as “freestanding.” By contrast, I seek to follow through the
implications of Rawls’s more positive way of putting the point when he himself refers to a political conception as
self-standing.16 A political conception of justice is grounded in the values of the domain of the political. Rather
than seeking to rely upon a foundational ‘neutral’ principle of equal concern and respect, Rawls’s political
constructivism presupposes these values for practical political purposes and “aims to generate the justification of
justice, not to impose it from the start.”17
Two levels of analysis need to be distinguished.18 First there are normative grounds, the procedural
principles of equal right, which are constructed or reconstructed. On this first level both Rawls and Habermas are
advocates of a procedural ethics in the sense that both are concerned with a modern morality of intersubjective
principles of the right rather than conformity with an external substantive good.19 Second, though, there is the
question of the ground of these grounds. Both Rawls and Habermas argue that the full justification of such
procedural grounds rests on their being shown to be implicit but not fully realized within existing social
practices, an account of such social practices involving a social ontology and an account of moral motivation. On
this second level, Rawls’s substantive normative conception of social institutions and of the motivation of
reasonable citizens contrasts with Habermas’s formalist conception of a constitution-making practice in which
citizens agree to regulate their lives through positive law. The understandings of the self-grounding relationship
between normative grounds and the institutional and motivational ground of these grounds are correspondingly
different. The most significant difference for present purposes is that while Rawls’s constructivism seeks to
14
Robert B. Pippin, Hegel’s Practical Philosophy: Rational Agency as Ethical Life (Cambridge: Cambridge University
Press, 2008), p. 21.
15
Charles Larmore, The Autonomy of Morality (Cambridge: Cambridge University Press, 2008), p. 153.
16
John Rawls, “The Idea of Public Reason Revisited,” reprinted in his The Law of Peoples with, “The Idea of Public Reason
Revisited” (Cambridge, MA: Harvard University Press, 1999), pp. 175, 179.
17
Catherine Audard, John Rawls (Stocksfield: Acumen, 2007), p. 7.
18
These two levels correspond to Charles Taylor’s distinction between advocacy and ontological questions in his “CrossPurposes: The Liberal-Communitarian Debate,” in Nancy Rosenblum, ed., Liberalism and the Moral Life (Cambridge, MA:
Harvard University Press, 1989).
19
Charles Taylor, “The Motivation behind a Procedural Ethics,” in Ronald Beiner and William James Booth, eds., Kant and
Political Philosophy: The Contemporary Legacy (New Haven, CT: Yale University Press, 1996).
5
contribute to the public political culture in a manner that brings about its substantive conception of a well-ordered
society,20 Habermas’s reconstructivism is limited to uncovering an existing reason always already at work in
discursive practices.
Habermas sees a procedural understanding of legitimacy as originating with Rousseau and as being worked
out more fully as a conception of practical reason by Kant. But Habermas criticizes Rousseau for mixing up this
procedural principle of legitimacy with substantive proposals for institutionalizing just rule.21 The significance of
a fully developed discursive conception of procedural legitimacy is that it is a reflexive level of justification
whose validity is independent of specific substantive institutional forms. The fairness of procedural legitimacy
depends not upon conformity to substantive values, but on the fact that procedures of discourse may be turned
back upon themselves to question their background substantive presuppositions. Habermas’s approach therefore
differs from Rawls’s in developing further the implications of a procedural conception of autonomy for the
relationship between philosophy and politics. It is not, as Habermas implies, that Rawls allows substantive
elements to intrude into a procedural approach in a way that could be eliminated without revising the basic
structure of the Rawlsian architectonic.22 This analysis both brings Rawls’s methodological presuppositions
closer to Habermas’s and distances Rawls’s conception of Political Liberalism from Habermas’s Kantian
Republicanism. What is at issue in the Rawls–Habermas debate is not so much a Kantian “familial dispute” as a
family resemblance between two different branches for taking forward the rational agency tradition.
3. Rawls’s constructivism: Procedure in substance
I will analyze the relationship between procedure and substance in Rawls’s constructivism by first identifying the
role that pure procedural justice plays within the Rawlsian architectonic and then highlighting the dependence of
pure procedural justice on a substantive background context. In approaching Rawls’s theory of justice we may
ask of it the same four questions that Rawls asks of Rousseau’s theory of justice:
1. What does the conception say are the reasonable or true principles of political right and justice;
and how is the correctness of these principles established?
2. What workable and practicable political and social institutions most effectively realize these
principles?
3. In what ways do people learn principles of right and acquire the motivation to act from them so
as to preserve stability over time?
4.How might a society realizing the principles of right and justice come about; and how has it in
some actual cases, if any such exist, come about?23
20
John Rawls, Lectures on the History of Political Philosophy, ed. Samuel Freeman (Cambridge, MA: The Belknap Press of
Harvard University Press, 2007), pp. 1–7.
21
Jürgen Habermas, Communication and the Evolution of Society, trans. Thomas McCarthy (Cambridge: Polity Press,
1979), pp. 185–86.
22
Habermas comes closer to accurately characterizing Rawls’s approach when, following Rainer Forst, he recognizes that
with the idea of an overlapping consensus “Rawls salvages a valuable insight of Hegel’s critique of Kant.” Habermas,
“‘Reasonable’ versus ‘True,’” p. 100.
23
Lectures on the History of Political Philosophy, p. 237.
6
These are questions that Rawls argues must be considered by any political conception of right and justice, and his
own conception of Justice as Fairness is no exception. The first three questions correspond to the three parts of A
Theory of Justice—Theory, Institutions and Ends—with the third question, concerning the problem of stability,
containing two sub-parts. Rawls’s perception of his failure to adequately address the second part of the third
question prompts him to re-present Justice as Fairness as a “political not metaphysical” conception while the idea
of an overlapping consensus responds both to this problem and the fourth question of the possible realization of
Justice as Fairness.
In considering the place of pure procedural justice within this framework it is important to recognize that
pure procedural justice is distinct from what Rawls calls formal justice.24 In A Theory of Justice Rawls defines
formal justice as the “impartial and consistent administration of laws and institutions, whatever their substantive
principles.”25 He postpones discussion of whether substantive and formal justice always go together until after he
has considered the most reasonable substantive principles of justice. Given that Rawls turns next to presenting his
two principles of justice before then considering the topic of procedural justice, different forms of procedural
justice may be interpreted as establishing a variety of relationships between formal and substantive principles of
justice, the adequacy of formal justice depending upon the maintenance of a fair substantive background context.
Rawls distinguishes three main forms of procedural justice (Table 1).26
Form of
procedural
justice
Perfect
Procedureindependent
criterion?
Yes
Procedure guarantees
correct outcomes?
Example
Yes
Fair division of a cake
Imperfect
Yes
No
Criminal trial
Pure
No
Yes
Gambling
Table 1. Procedural justice27
In the case of perfect procedural justice there is an independent criterion of justice separate from and prior to the
procedure. For example, assuming that a fair division of a cake is an equal division, a procedure in which the
person who divides the cake is the last to take a slice is guaranteed to lead to the independent standard being met.
This example depends upon the existence of a social convention about what it means to treat people fairly; in this
case that when something is to be shared, and there are no reasons to believe anyone has a stronger claim than
anyone else, a fair division is an equal division. Second, in the case of imperfect procedural justice there is again
an independent criterion of justice, but this time there is no feasible procedure guaranteed to lead to the correct
answer being identified. A criminal trial is an example, designed to declare a defendant guilty only if he has
24
On the novelty of Rawls’s idea of pure procedural justice, see William Nelson, “The Very Idea of Pure Procedural
Justice,” Ethics, LXXX (1980): 502–11.
25
Rawls, A Theory of Justice, p. 58/rev. 51.
26
Rawls, A Theory of Justice, pp. 83–90/rev. 73–78.
27
See Robert Dahl, Democracy and its Critics (New Haven: Yale University Press, 1989), p. 165 and Bernhard Peters, ‘On
Reconstructive Legal and Political Theory’ in Habermas, Modernity and Law, ed. Mathieu Deflem (London: Sage, 1996), p.
116.
7
committed the offence with which he is charged but unable to guarantee this will occur. Third, in pure procedural
justice, while there is no independent criterion for the right result, there is a fair procedure that translates its
fairness to the outcomes provided it has been properly followed. Rawls gives the example of gambling, or, more
specifically, the distribution of cash that is the outcome of a series of fair bets. It is significant here both that this
is a series of fair bets in which the fairness of the procedure is defined by the background circumstances and that
it is a series of fair bets. The procedure has to actually be carried out, it not being enough to just say that a
particular outcome could have been produced by a series of fair bets.
Justice as Fairness involves the idea of pure procedural justice at the highest level, exhibited in the
reasonable constraints on the choice of principles imposed by the veil of ignorance in the Original Position. It is a
corollary of the rational autonomy of the deliberating parties, the fact that they recognize no standpoint external
to their own point of view that would constrain them by prior and independent principles of justice.28 Second,
pure procedural justice plays the role, in connection with Rawls’s first principle of justice, the liberty principle,
of securing the fair value of the political liberties. The idea is to “incorporate into the basic structure of society an
effective political procedure which mirrors in that structure the fair representation of persons achieved by the
original position.”29 It is this, together with the second principle of justice, including the difference principle,
which explains why the basic liberties are not merely formal. We therefore turn, third, to the role of pure
procedural justice with respect to distributive shares. This depends upon the fair equality of opportunity secured
through the establishment and maintenance of a just basic structure that maintains background justice: “Only
against the background of a just basic structure, including a just political constitution and a just arrangement of
economic and social institutions, can one say that the requisite just procedure exists.”30 The principles of Justice
as Fairness are most effectively realized by the institutions of a property-owning democracy.31 Finally, Rawls’s
commitment to pure procedural justice has consequences for how we understand the idea of justification through
reflective equilibrium. The argument from the Original Position does not have independent justificatory force
that can be turned against persons’ considered judgments.32 Rather, as Rawls puts it, “reflective equilibrium
works through the original position” and its conception of pure procedural justice.33
Rawls sees Kant as giving a deeper foundation to Rousseau’s conception of autonomy and he argues that
generality and universality should not be seen as central to Kant’s ethics.34 In calling the Original Position a
procedural interpretation of Kant’s conception of autonomy and the categorical imperative, Rawls means that it is
set within a reasonable empirical framework of social institutions rather than within the framework of
28
Rawls, A Theory of Justice, pp. 120, 136 /rev. 104, 118.
John Rawls, Political Liberalism (New York: Columbia University Press, paperback edition 1996), p. 330.
30
Rawls, A Theory of Justice, p. 87/rev. 76.
31
John Rawls, Justice as Fairness: A Restatement, ed. Erin Kelly (Cambridge, MA: Harvard University Press, 2001), pp
135–80.
32
David Lyons recognizes that the most faithful interpretation of Rawls is to see the “contract argument” as a branch of the
“coherence argument,” but rejects this for compromising the justificatory force of the Original Position. “Nature and
Soundness of the Contract and Coherence Arguments,” in Norman Daniels, ed., Reading Rawls (Stanford: Stanford
University Press, 1989).
33
Quoted in Samuel Freeman, “Original Position,” The Stanford Encyclopedia of Philosophy (Spring 2009 Edition), ed.
Edward N. Zalta. http://plato.stanford.edu/archives/spr2009/entries/original-position/.
34
Rawls, A Theory of Justice, p. 251/rev. 221.
29
8
transcendental idealism.35 It is not that ‘Kantian’ principles of justice are independently derived and then applied
to the basic structure. Rather, a social theory of the basic structure drives the construction of the Original Position
choice procedure. “Given the unique features and role of [the basic] structure, the idea of an agreement [in social
contract views] must be appropriately transformed if the intent of the Kantian form of the contract doctrine is to
be realized.”36 Equality will then be “supported by the general facts of nature and not merely by a procedural rule
without substantive force.”37
In assigning priority to the social and the role of the basic structure, Rawls reverses the direction of Kant’s
ethics. While Kant moves from the application of the categorical imperative to personal maxims to the creation
of a systematic order, Rawls interprets the categorical imperative as a procedure of unanimous collective
agreement for regulating the basic structure of society.38 Behind the veil of ignorance the parties are constrained
only by Kantian “formal constraints of the concept of right,” constraints that are justified by the role of principles
in “adjusting the claims that persons make on their institutions and one another.”39 Indeed, the Original Position,
with the specific facts it veils and the general social facts it admits, already contains in nuce a substantive
conception of democracy as “a political relationship between citizens within the basic structure of society,”
which Rawls proceeds to unfold.40 According to Rawls, “[a] purely procedural theory that contained no structural
principles for a just social order would be of no use in our world, where the political goal is to eliminate injustice
and to guide change toward a fair basic structure.”41
I have shown how the substantive constraints on the choice of principles in the Original Position lead to the
choice of fair principles of justice and, in turn, how when these fair principles provide the substantive
institutional constraints on the operation of the market that maintain background justice, the distributive
outcomes that result will themselves be fair. But what is the ultimate justification for the substantive reasonable
constraints in the Original Position? Rawls describes Justice as Fairness as a theory of the moral sentiments,
setting out the principles that are regulative of, in the sense of being implicit within, our sense of justice.42 In
endorsing a conception of justice in reflective equilibrium, we “understand how the moral sentiments can be
regulative in our life and have the role attributed to them by the formal conditions on moral principles” in the
Original Position.43 In Political Liberalism Rawls seeks to show how a shared political sense of justice is
compatible with a plurality of reasonable comprehensive doctrines. This will be shown if a political conception
of justice represented by the Original Position can serve as the focus of an overlapping consensus, a procedure in
which citizens achieve wide and general reflective equilibrium. In reflecting on why Rawls feels he can draw
upon a shared fund of democratic ideas, a promising place to look is at the way that, in Hegelian fashion, Rawls
35
Rawls, A Theory of Justice, p. not present/rev. 226; Rawls, Political Liberalism, p. 285.
Rawls, Political Liberalism, p. 288.
37
Rawls, A Theory of Justice, p. 510/rev. 446.
38
John Rawls, “Kantian Constructivism in Moral Theory,” reprinted in Samuel Freeman, ed., Collected Papers (Cambridge,
MA: Harvard University Press, 1999), p. 339.
39
Rawls, A Theory of Justice, p. 131/rev. 113.
40
Rawls, Political Liberalism, p. 217.
41
Rawls, Political Liberalism, p. 285.
42
Rawls, A Theory of Justice, p. 51/rev. 44.
43
Rawls, A Theory of Justice, p. 478/ rev. 418.
36
9
sees the background institutions of the basic structure as forming our social world.44 Rawls takes all forms of
social cooperation to exist within “some more or less clearly specified situation embedded within the background
institutions of the basic structure” and the goal of the Original Position is to “extend the idea of agreement to this
background framework itself.”45 The conception of justice so derived is directed towards citizens who “desire for
its own sake a social world in which they, as free and equal, can cooperate with others on terms all can accept.”46
Paul Ricoeur correctly identifies the hermeneutical circularity of Rawls’s argument. Rawls first sets out his
two principles of justice and then argues that they would be chosen in the Original Position, the whole argument
being a progressive systematization or rationalization of our preunderstandings about justice in the search for
reflective equilibrium. In posing the question of whether a purely procedural theory of justice is possible, Ricoeur
concludes that Rawls’s “procedural conception of justice at best provides a rationalization of a sense of justice
that is always presupposed.”47 The procedural deontological perspective of morality cannot ground itself and
remains grounded in the substantive perspective of ethics. The question is though whether, pace Ricoeur, Rawls
would disagree with this contention. Rawls acknowledges that Justice as Fairness is substantive in the sense that
it “springs from and belongs to the tradition of liberal thought and the larger community of political culture of
democratic societies,” and that it therefore fails to be formal and universal in Habermas’s sense.48 I outline
Habermas’s alternative conception of procedural rationality in the next section and show how in contrast to
Rawls Habermas takes deontological procedures to have severed their link to determinate substantive ethical
contexts.
4. Habermas’s reconstructivism: Substance in procedure
A normative conception of procedural rationality is central to Habermas’s understanding of the requirements of
post-metaphysical thinking in modernity. According to Habermas, the possibility of thinking philosophically of
society in holistic terms was undermined by the procedural rationality that has developed in the empirical
methods of the natural sciences since the seventeenth century and the formalism in moral and legal theory and
the institutions of the constitutional state that has developed since the eighteenth century. Both science and
morality come to rely solely on the rationality of their procedures: “Rationality [Rationalität] is reduced to
something formal insofar as the rationality [Vernünftigkeit] of content evaporates into the validity of results.”49
Habermas’s discourse theory of morality is a formalist ethics insofar as it designates a procedure for
judging morally relevant conflicts from a moral point of view, but this does not mean that it is formal in the sense
44
See Rawls, Political Liberalism, pp. 11, 23, 41, 43, 77, 229, 262–71. Michael Hardimon employs the term social world as
equivalent to Hegel’s “ethical world” and recognizes that as such it is “very closely related to Rawls’s idea of the ‘basic
structure.’” See his Hegel’s Social Philosophy: The Project of Reconciliation (Cambridge: Cambridge University Press,
1994), p. 16.
45
Rawls, Political Liberalism, p. 23.
46
Rawls, Political Liberalism, p. 50.
47
Paul Ricoeur, “Is a Purely Procedural Theory of Justice Possible?” in his The Just, trans. David Pellamer (Chicago, IL:
University of Chicago Press, 2000), p. 50.
48
Rawls, Political Liberalism, p. 432. Nor is it plausible to see this is a revision of the view of A Theory of Justice. See Sibyl
A. Schwarzenbach, “Rawls, Hegel, and Communitarianism,” Political Theory, XIX (1991): 539–71.
49
Jürgen Habermas, “Themes in Postmetaphysical Thinking,” in his Postmetaphysical Thinking, trans. William Mark
Hohengarten (Cambridge: Polity Press, 1992), p. 35.
10
of scientific instrumental rationality. Rather, it “deprives practical reason of all specific normative contents and
sublimates it in the form of a procedure for justifying possible normative contents.”50 Kant’s categorical
imperative is employed not as a test of maxims for action but as a principle of justification that can be given a
discursive procedural form. Like Rawls, Habermas argues that participants in a moral practice have lost their
metaphysical guarantees and must derive normative principles by drawing on a common practice they already
share. However, Habermas goes further than Rawls, arguing that in post-traditional societies not only can one not
expect a substantive consensus on the good but that these shared features “shrink to the fund of formal features of
the performatively shared situation of deliberation.”51 Joint practical deliberation that justifies moral norms
convincing to all participants because of their impartiality becomes a post-traditional equivalent to the traditional,
substantive grounding of a normative consensus. “The missing transcendent good can be replaced in an
immanent fashion only by appeal to the intrinsic constitution of the practice of deliberation.”52 For Habermas, the
priority of the right over the good means the uncoupling of the horizontal perspective of the moral regulation of
interpersonal relationships from the vertical perspective of ethical reflection on individual and collective lifeprojects. The “remnant of the good at the core of the right” ensures an internal relation between justice and
solidarity, the two being two sides of the same coin.53
When it comes to law, Habermas develops his position out of a critique of the social contract tradition and
bourgeois formal law, in Max Weber’s sense. Habermas criticizes Weber for assuming that all forms of
procedural rationality are analogous to the formalism of instrumental rationality and ignoring the possibility of
ethical formalism. According to Habermas:
Weber does not sufficiently distinguish between structural and substantive— or formal and
material aspects [of the social contract] . . . He falsely identifies the procedural properties of a posttraditional level of justification with substantive values. Therefore, he does not see that the model
of the social contract (in a way similar to the categorical imperative) can be understood as
proposing a procedure whose rationality is supposed to guarantee the correctness of whatever
decisions come about in a procedural manner.54
In the balance sheet of modernity, morality has lost its ability to justify substantive ethical positions, but it has
gained an integral place within formal rules of social organization. There is a procedural morality ingrained in
law that “has rid itself of all specific normative contents and has been sublimated into a procedure for the
justification of possible normative contents.”55
Habermas’s procedural conception of democracy is not merely formal, but it does assume that, under
conditions of post-metaphysical thinking, we cannot expect a consensus on substantive values that reflect a
shared substantive conception of equal citizenship, only a commitment amongst citizens to legitimately regulate
50
Jürgen Habermas, “Towards a Communication-Concept of Rational Collective Will-Formation. A Thought-Experiment,”
Ratio Juris, II (1989): 114–54, p. 149.
51
Jürgen Habermas, “A Genealogical Analysis of the Cognitive Content of Morality,” in his The Inclusion of the Other:
Studies in Political Theory, eds. Ciaran Cronin and Pablo De Greiff, trans. Ciaran Cronin (Cambridge: Polity Press), p. 41.
52
Habermas, “Genealogical Analysis,” p. 41.
53
Habermas, “Genealogical Analysis,” pp. 28–29.
54
Jürgen Habermas, “Law and Morality,” The Tanner Lectures on Human Values, trans. Kenneth Baynes (1986), p. 228.
http://www.tannerlectures.utah.edu/lectures/documents/habermas88.pdf.
55
Habermas, “Law and Morality,” p. 247.
11
their lives by positive law. This constitution- making practice presupposes law as the medium of binding
regulations and the discourse principle as a way of understanding reasonable deliberation and decision making
and it is not therefore free of all normative content.56 Rawls takes this to imply that the practice includes
substantive elements, pointing to Habermas’s acknowledgment that the procedural legal paradigm retains a
dogmatic core in the idea of autonomy.57 But Habermas argues that the idea of citizens giving themselves laws,
which is explicated in constitutional principles and a system of rights, “already contains as a doctrinal core the
(Rousseauian-Kantian) idea of the self-legislation of voluntarily associated citizens who are both free and equal.”
This can be “fully developed in the course of constitution-making processes that are not based on the previous
choice of substantive values, but rather on democratic procedures.”58 This conception of popular sovereignty as
procedure contrasts with Rawls’s substantive conception of democracy, well described by Samuel Freeman:
In the end, Rawls seems to commit himself to the view that the ‘people’ is an ideal implicit in
democratic political culture: that of free and equal persons united together as one legal body, the
body politic, which exercises constituent power to make the higher law in such a way that it
expresses the political values of public reason, thereby enabling them to realize the (moral) powers
that make them free and equal democratic citizens. This conception of the person and the people
seems to be the basis for the substantive conception of democracy that Rawls sees as implicit . . . in
the public political culture of which [the] constitution is an integral part.59
I have outlined the basis of Habermas’s procedural conception of reason, but there remains Rawls’s
criticism that no institutional procedure without substantive guidelines for admissible reasons can “cancel the
maxim ‘garbage in, garbage out.’”60 This raises two issues to do with first the outcomes of discursive procedures
and second whether there are substantive standards beyond discourse.
Regarding the outcomes of democratic procedures, Rawls argues that to ensure just outcomes majority rule
must be constrained by antecedently defined principles of justice. For Habermas, by contrast, constraints on
outcomes are constitutive of procedures of democratic discourse. Majoritydecisions are “only caesura in a
process of argumentation that has been (temporarily) interrupted under the pressure to decide.”61 As provisional
resolutions of ongoing discourses, majority decisions are both shaped by these discourses and their legitimacy
depends upon them leaving open the possibility of their being questioned in further discourses. Rather than
seeing the intrinsic fairness of democratic procedures as being restricted to the formal fairness of giving each
participant an equal chance to present their arguments and cast their vote, which would invite the need for
external substantive principles to regulate outcomes, Habermas argues that procedures of democratic discourse
derive their meaning from the idea of the continual search for better substantive reasons. Adapting Rawls’s
56
Habermas, Between Facts and Norms, pp. 82–131.
Rawls, Political Liberalism, p. 426.
58
Jürgen Habermas, “Reply to Symposium Participants, Benjamin N. Cardozo School of Law,” in Michel Rosenfeld and
Andrew Arato, eds., Habermas on Law and Democracy: Critical Exchanges (Berkeley: University of California Press,
1998), p. 406.
59
Samuel Freeman, “Political Liberalism and the Possibility of a Just Democratic Constitution,” in his Justice and the
Social Contract: Essays on Rawlsian Political Philosophy (Oxford: Oxford University Press, 2007), p. 210.
60
Rawls, Political Liberalism, p. 431.
61
Habermas, “Reply to Symposium Participants,” p. 415.
57
12
terminology, Habermas says majority rule owes its legitimacy to an “‘imperfect’ but ‘pure’ procedural
rationality.”62
We can see the basis of this claim by looking second at Rawls’s argument for the need for substantive
guidelines for governing admissible reasons in discourse. This recalls Rawls’s claim that the procedure of the
public use of reason can be guaranteed to have reasonable outcomes only insofar as it realizes the conditions of
ideal discourse. But Habermas rejects Bernhard Peters’s similar claim that it is substantive reasons on the basis of
which outcomes are judged as correct, reasons independent of any procedure and that mean that procedures are
of only instrumental importance. This would make ideal discourse a form of perfect procedural legitimacy, since
under perfect conditions it would isolate the correct answer. However, while Habermas agrees that procedures
cannot guarantee right answers, he argues that there are no criteria for right answers given prior to argumentation.
It may be that it is substantive reasons that convince us that an outcome is right but “the soundness of these
reasons that can be demonstrated only in real processes of argumentation.”63 The substantive content of a moral
principle resides in the anticipation that in an ideal procedure it could rebut every possible objection.
We may note finally how this analysis of Habermas’s conception of procedural rationality challenges
Rawls’s characterization of it as a Hegelian comprehensive doctrine. Rawls suggests that the relationship
between procedure and substance in Habermas’s approach is such that once the procedural presuppositions of
discourse have been set out “then all the alleged substantial elements of those religious and metaphysical
doctrines and the traditions of communities have been absorbed (or sublimated) into the form and structure of
these presuppositions.”64 But Habermas’s understanding of the requirements of post-metaphysical thinking
contradicts Rawls’s characterization. Habermas employs the term post-metaphysical to refer not only in a
methodological sense to a concern with procedural rationality but also to a substantial “agnostic” position that
sharply distinguishes between belief and knowledge without denying the possible cognitive content of religious
traditions. Habermas distinguishes “rationalist approaches that (in the Hegelian tradition) subsume [aufheben] the
substance of faith into the philosophical concept, from dialogical approaches that . . . adopt a critical attitude
towards religious traditions while at the same time being open to learning from them.”65
5. Conclusion
In entitling his article on Rawls “Reconciliation through the Public Use of Reason,” Habermas appeals to the idea
of universality in Kant’s idea of the public use of reason. But Rawls refers not to reconciliation by the public use
of reason but to “reconciliation by public reason.”66 For Rawls, as for Rousseau, public reason is the reason of a
democratic people and its subject is the public good, namely what a political conception of justice requires of
62
Habermas, “Reply to Symposium Participants,” p. 397.
Habermas, “Reply to Symposium Participants,” p. 409.
64
Rawls, Political Liberalism, pp. 431–32.
65
Jürgen Habermas, “The Boundary between Faith and Knowledge: On the Reception and Contemporary Importance of
Kant’s Philosophy of Religion,” in his Between Naturalism and Religion: Philosophical Essays, trans. Ciaran Cronin
(Cambridge: Polity Press, 2008), p. 245.
66
Rawls, Political Liberalism, pp. lx, 157; Rawls, “The Idea of Public Reason Revisited,” p. 177; italics added.
63
13
society’s basic structure of institutions and what ends they should serve.67 One of the limits to reconciliation by
public reason is that “comprehensive doctrines are, politically speaking, unreconcilable.” Economic and social
conflicts can, however, be reconciled when we accept reasonable principles of justice and know our social
institutions conform to them.68 In reconciliation by public reason we become reconciled to our social world and
the existence of reasonable pluralism. The political relation of citizens in a constitutional democratic regime can
still be one of civic friendship.69 Habermas criticizes Rousseau for seeking a consensus of hearts not of
arguments, and the same criticism can be applied to Rawls.70 Rawls’s conception of public reason is a natural
development of the “purity of heart” towards which we may aspire when we reinterpret the point of view of
eternity as a form of thought and feeling that persons can adopt within their social world.71 Habermas’s discourse
theory of law, by contrast, seeks an understanding of the common good without a Rousseauian expectation of
citizen virtue: “practical reason withdraws from the hearts and heads of collective or individual actors into the
procedures and forms of communication of political opinion- and will-formation.”72
In making a final assessment of Rawls’s constructivist “procedure within substance” and Habermas’s
reconstructive “substance in procedure” we can compare these approaches in terms of their contrasting
economies of philosophical ambition. Both Rawls and Habermas claim that their approach is more modest than
the other’s, Rawls on the basis that it is political rather than comprehensive and Habermas on the basis that it is
strictly procedural rather than substantive. Habermas says he has “never had any ambition of sketching out a
normative political theory,” of designing “the basic norms of a ‘well-ordered’ society on the drafting table.”73 But
in thinking of their respective approaches as systems of thought in which what philosophy supplies seeks to
respond to what the social world demands, in each case one should keep in mind the ambitions that motivate this
modesty. As a result of this philosophical modesty, each claims that their theory encompasses that of the other.
Rawls argues that Habermas’s account of democratic legitimacy would be acceptable within public reason,74
while Habermas seeks a synthesis of Rawls’s normative approach to social institutions and the insights of the
sociologist and systems theorist Niklas Luhmann.75 In comparing Luhmann’s approach to that of Ronald
Dworkin, Habermas asks: “Who can embrace whom—the sociologist the legal philosopher, or vice versa?”76 We
can ask an analogous question with respect to Habermas and Rawls.
I have argued that Rawls presupposes a substantive shared social world and I follow Habermas in seeing
the facts of growing social complexity as rendering this assumption problematic and undermining the normative
conception of social institutions it assumes. It is for this reason that Habermas states that a “sceptical evaluation
of current world conditions” is what distinguishes his approach from “purely normative conceptions such as John
67
Rawls, Political Liberalism, p. 213.
Rawls, Political Liberalism, pp. lx, 612.
69
Rawls, “The Idea of Public Reason Revisited,” p. 137.
70
Jürgen Habermas, The Structural Transformation of the Public Sphere, trans. Thomas Burger and Frederick Lawrence
(Cambridge: Polity Press, 1989), p. 98.
71
Rawls, A Theory of Justice, p. 587/rev. 514.
72
Habermas, “Reply to Symposium Participants,” p. 385.
73
Jürgen Habermas, The Past as Future, trans. and ed. Max Pensky (Cambridge: Polity Press, 1994), p. 101.
74
Rawls, “The Idea of Public Reason Revisited, p.142.
75
Habermas, Between Facts and Norms, pp. 56–66.
76
Jürgen Habermas, “Ronald Dworkin—A Maverick among Legal Scholars,” in his Europe: The Faltering Project, trans.
Ciaran Cronin (Cambridge: Polity Press, 2009), p. 39.
68
14
Rawls’s theory of justice, admirable as it is in itself.”77 But we should not lose sight of the greater expectations of
moral and political agents that contemporary social practices make, and that philosophical reflection upon such
practices can presuppose, that is the flip side of this sceptical evaluation of the possible role of philosophy in
relation to politics. If Habermas’s procedural reconstructivism cannot sustain a realistic utopia of a well-ordered
society, it can remind us of the utopian dimension always already at work in our communicative social practices.
In this changed guise the interdependence of philosophy and democracy continues to provide the basis for
addressing basic normative questions regarding the well-ordered society.78
77
Jürgen Habermas, “A Conversation about Questions of Political Theory,” in his A Berlin Republic: Writings on Germany,
trans. Steven Rendall (Cambridge: Polity Press, 1997), p. 132.
78
Jürgen Habermas, “The Relationship between Theory and Practice Revisited,” in Truth and Justification, trans. Barbara
Fultner (Cambridge: Polity Press, 2003)
15