Content uploaded by Erik Oddvar Eriksen
All content in this area was uploaded by Erik Oddvar Eriksen on Jul 17, 2014
Content may be subject to copyright.
Democratic or jurist made law?
On the claim to correctness
Erik Oddvar Eriksen
ARENA – Centre for European Studies, University of Oslo.
In democratic societies, legal procedures are to ensure legally correct and rationally
acceptable decisions, i.e. decisions that can be defended both in relation to legal statutes and
in relation to public criticism. But can the legal system via the discretion of the judges itself
really autonomously settle normative questions? On this constitutionalists and proceduralists
disagree. The problem is whether the substantial factors are legitimate, and whether the
judges’ interpretations of the situations are correct. Robert Alexy conceives of the legal
discourse as a special variant of the general discourse but this blurs the distinction between
legislation and application. There is a danger of assimilating law and morality and of
overburdening the legal medium itself. Moral and legal questions point to different audiences,
raise different validity claims and require different procedures for resolving conflicts. The
author favours a variant of constitutional proceduralism hinged on discursive proceduralism
which sets the terms for a fair procedure of reason giving. This standard for correctness is
imperfect but ensures that the substantial, ‘pre-political’ principles entrenched in modern
constitutions as basic rights are subjected to discursive testing in a deliberative process.
The interconnections between law and politics are many and intricate.
For one there is a
mutual relationship as only politics can give the norms that courts act upon. It is the
legislative process that furnishes the legal system with normative inputs. But politicians can
not work unless they observe the legal procedures that judges monitor. No valid law without
politics and no legitimate politics without the law. Law is the lingua franca of democracy and
democracy is the sole remaining legitimation principle in modern societies.
However, also in another sense is there an intricate relationship as majority vote, which is the
operative principle of representative democracy, can not ensure correct decisions, viz., a
rational and just outcome. The majority principle does not ensure political equality. The
phenomenon of permanent minorities is well-known: certain groups are not likely ever to
become a majority or to be a part of majority alliances. Outcomes of majority voting
represent the voice of the winners, not the common will. But can procedures ensure correct
results by themselves?
Discourse theory holds that it is the procedures that warrant the presumption that it is possible
to reach correct decisions: A norm N is correct when it is the result the procedure P (Alexy
1995:110). In a democracy the correctness of decisions depends solely on the procedures
(Habermas 1996b:1495). In pure procedural justice it is fair procedures that ensure the right
result; there is no independent criterion for such. For example, a chance procedure like
gambling is a pure procedural model that ensures just outcomes without any reference to
extra-procedural elements (Rawls 1971:86). But if it is the procedure itself that warrants
correct results, what, then, warrants the procedure? There is a problem with a pure
procedural conception of correctness, and hence with pure procedural conceptions of
democratic legitimacy. Independent standards are required in order to evaluate the process or
the outcome, according to constitutionalists. The latters make use of moral arguments, of
substantive conceptions of what is right or good, in order to solve the problem of rational
adjudication, without this ‘substance’ being neither legitimated nor tested democratically. It
opens for jurist made law. The Supreme Court becomes the final arbiter of constitutional law.
An earlier version of this paper was presented first at the international workshop on Constitutional Rights
through Discourse - on Robert Alexy’s legal theory , Center for the study of Professions, Oslo University
I address the relationship between law and politics when it comes to rational adjudication of
constitutional questions from a discourse theoretical point of view. The question is whether
the substantial factors can be tested democratically so that we can speak of democratic made
law. The specific question is whether the observance of the rules for rational communication
guarantees correctness, and in that case what kind of standard for correctness, pure or
imperfect, is involved. My argument is that the standard is imperfect but is itself an
expression of a normative procedure for justification. This is due to a larger concept of
democracy. It is not merely a voting arrangement constitutionally constrained.
I start out by briefly addressing the schism between constitutionalists and proceduralists in
political theory and the democratic problem involved (2) before a outlining the discourse
model of legal argumentation (3). This is not a pure procedural theory as certain substantial
elements are involved (4). I find that the core morality of discourse theory can be
proceduralised through the rules of argumentation but speech rules does not fully do away
with normative, substantial elements and the indeterminacy problem lingers (5). Here I
address Robert Alexy’s theory of rational adjudication. He sees legal discourse as a special
variant of general practical discourse - die Sonderfallthese – that by itself can ensure
rationality and correctness. But can a judicial discourse intermediate between substance and
procedure in a valid manner (6)? Law and morality are interconnected but is the legal medium
itself capable of handling normative questions adequately; or are additional procedures called
for (7)? Die Sonderfallthese seems to overtax the legal medium (8). Klaus Günther’s proposal
of a distinction between a justification and an application discourse fares better in normative
terms (9), but it cannot ensure correctness (10). The bottom lines is that a core morality is
presupposed in the procedure ensuring a fair process of reason-giving –alluding to a concept
of constitutional proceduralism, which hinges on discursive proceduralism (11). It yields,
however, not more than an imperfect standard of justice (12).
2. Constitution or Procedure?
The so-called constitutionalists build on substantial conceptions of justice and give priority to
fair outcomes over democratic procedures - what are called correctness theories. The
proceduralists put their faith in the procedures, viz. they give priority to the rights that
College April 24-26, 2003, and then at a Cidel seminar at Arena, University of Oslo, March 12. 2004. I am
guarantee political participation and fair processes. Among deliberationists Jürgen Habermas
is supporting such a view and Robert A. Dahl (1989) is a prominent representative of this
view within the ‘voting camp’. The latter holds that majority principle (as well as group
bargaining) is an important part of this because it guarantees equal treatment of all the
members’ interests. Majority rule reflects the principle of equal citizenship and treats
everyone numerically equal. It is, however, insensitive to reasons and argumentation.
Majority rule obeys by numbers not by reasons. Not only can the majority simply be wrong,
the procedure itself can not ensure a rational outcome. As demonstrated in Arrows’
Impossibility Theorem, it is not possible to infer from individual preferences to collective
choices. Voting means the choice between different alternatives is made on the same footing
as the flipping of a coin. However, under certain circumstances, as when all parties are equal,
or when there are only two alternatives and no authoritative “truths”, votes can be used
without raising any problems (cf. Barry 1991). It is when this is not the case, as when the
winner takes it all and the goods are unequally distributed because of it, the relevance of
process-independent standards or “correctness theories” becomes clear. If we want a fair
distribution, we do not decide a case as if it was a lottery. Implicitly this shows that there are
independent measures as to what constitute a correct result. The problem with the majority
procedure is that it can not let any interests or demands be favoured – not even for good
reasons (Estlund 1997:176f). Voting as the primary political action, based on ‘non-
deliberative’ preferences, can never represent real political equality for suppressed or
excluded groups. Constitutionalists oppose the procedural model, which they find wanting as
it can not itself lay down the conditions for a fair procedure:
“The real, deep difficulty the constitutional argument exposes in democracy is that it
is a procedurally incomplete scheme of government. It cannot prescribe the
procedures for testing whether the conditions for the procedures it does prescribe are
met.” (Dworkin 1996:33).
Constitutionalists such as Rawls and Dworkin, give priority to a set of rights, which, by
protecting the individuals’ vital interests, have the task of ensuring fair outcomes. These are
basic rights, which cannot be changed by any occasional political majority, hence the
distinction between principles, referring to reasons for actions and policies relating to
grateful for comments from the participants, in particular to Anders Molander and Agustín J. Menéndez.
collective interests (Dworkin 1977:82ff). This also goes for checks-and-balance mechanisms
such as division of power and judicial review. Such process-independent criteria may
ultimately be said to rest on meta-theoretical, pre-political, moral principles about human
being’s right to life and freedom, that are themselves not subjected to democratic legislation.
They do not emanate from political processes but are prior to them.
Dworkin (1981) arrives
at the fundamental egalitarian principle of treating everyone as equals. Based on the principle
that everybody is entitled to equal concern and respect, the Supreme Court can reason over
the correct application of a norm (Dworkin 1986:397ff).
Rawls presupposes pre-political elements when he gives priority to individual freedom over
democratic self-determination in his interpretation of justice as fairness. With the idea of
equal freedom for all citizens as a point of departure he arrives – on contract theoretical terms
– at a substantial conception of justice. The theory of justice as fairness rests on the thought
experiment of an original position where the actors are placed behind a veil of ignorance.
They have no knowledge of their social situation or personal resources – they have no
knowledge of their future position in society. In this position the actors will be able to agree
on, among other things, the following two lexically ordered principles of justice: they will
only accept solutions which guarantee equal and maximum freedom for all and a distribution
of resources that only favours differences which improve the situation for the least privileged
(Rawls 1971:302). Hence the difference principle: only those economic and social differences
that benefit the least advantageous will be accepted on a free basis.
Here the choice situation
itself is organized in such a way that even self-interested actors will choose morally
acceptable solutions. Morality as such is built into the constraints on the reflection situation.
On whether this is‘natural law’ conceptions see Rawls 1971:505ff, 1995:159, Dworkin 1981, Gosepath 1995
and Forst 1999a:113, 143. Habermas 1998a,b.
Consider Dworkin: “Justice is a matter of the correct or best theory of moral and political rights, and anyone’s
conception of justice is his theory, imposed by his own personal convictions, of what these rights actually are”
(1986:97). Cf. Michelman: “.…justice ( ..) is unalterably what we may call a “perfectly” process-independent
standard: in judging whether fundamental laws are just (…), no reference can ever be called for to the process of
their legislation.” (1997:147).
Further: “According to laws integrity, propositions of law are true if they figure on or follow from the
principles of justice, fairness, and procedural due process that provide the best constructive interpretation of the
community’s legal practice” (Dworkin 1986:227)
“Because the parties start from an equal division of all social primary goods, those who benefit least have, so to
speak a veto. Thus we arrive at the difference principle.” (Rawls 1999a:131)
Rainer Forst argues that Rawls rather conceives of the private use of public reason in public affairs than of the
public use of reason (Forst 1994: 156). See also Habermas 1998a:51f.
, 1998b:75ff, Benhabib 2002, chapter 5.
The principles of justice is the result of an impartial agreement in an initial situation devised
to ensure that no one’s interest is favoured at the expense of another’s. The concept of justice
as fairness yields an independent or free-standing conception of right and justice. It is
independent of disputed (religious) faiths and beliefs (Rawls 1999b:3). Further, the “political
conception of justice .. specifies certain basic rights, liberties and opportunities, .. assigns a
special priority to these rights, and affirms measures” to make them effective (Rawls
1997a:774). Here public reason is limited to political questions, and more specifically such
that are of a constitutional nature and concern the fairness of the basic structure of society.
Public reason is
1) about the common good as it is embedded in society’s concept of political justice
2) is governed by a reciprocity norm - one appeals to reasons that are convincing
enough to satisfy reasonable people, reasons that are mutually acceptable.
This reason is applicable to questions that can be decided with reference to fairness. John
Rawls understands public reason as an expression of the reason employed by citizens with the
same political rights in democratic states. It characterizes a situation in which equal citizens
in concert exercise political power over one another in the making of statutes and in
amending the constitution.
To liberals such as Rawls and Dworkin it is the discussions and reflections as carried out
among public decision-makers – the judges – that constitute the ideal model of deliberation.
Judges can deliberate against the background of civic virtue and with an insight into what the
reciprocity norm demands. Or in Dworkin’s case lawyers can ‘imitate Hercules’ as the ideal
judge (Dworkin 1986: 263ff). The Supreme Court of a well-ordered society is regarded the
highest body for public reason; it is an exemplar of public reason according to Rawls
(1997b:108). Deliberations in the Supreme Court and the judicial discourse form the basis of
the ‘constitutionalists’ deliberation model. Here the constitution ultimately takes precedence
over the citizens’ self-legislation.
The conflict between proceduralists and constitutionalists
then pertains to the problem of whether there can be a democratic enactment of a constitution.
For criticism along these lines see Alejandro 1996:22, Habermas 1998a:49ff, Forst 1999a,b, 2001. Benhabib
2002, chapter 5. See Rawls 1993: 239 and 1995 for reply.
We may, thus, distinguish between ‘democrats’ who prioritize the political process and
‘consititutionalists’ who claim primacy of law.
Constitutionalists have a theory for interpretation of principles and application of norms
rather than a theory for creating norms and generating principles. In effect, this is an
instrumental justification of democracy: it is necessary in order to realize liberal principles.
“[…] according to contractualist liberalism, political institutions are justified only if they are
effective instruments for enacting laws and politics that promote the justice of society’s basic
structure” (Weithman 1995:316).
Such a justification endangers democracy. A moral
argument can be used to replace democracy – when you know what is RIGHT, you do not
need to consult the people (cf. Maus 1989: 193). Does discourse theory provide a better
3. Infinite regress
One of the problems with such correctness theories is that there is little agreement on the
normative standards that are used in complex and pluralistic societies. How can we know that
e.g., Rawls substantial concept of justice is correct? The original position is after all a mere
thought experiment and as Rawls himself concedes, “Many will prefer another criterion”
(1993:227). If, for example, one should vote over whether a decision corresponds with an
independent standard of justice, some would probably agree, while some will disagree (cf.
Estlund 1997:175). It is an unstable solution. This is not to say that we need to follow
Waldron who does not find more of a common basis for deciding questions pertaining to
justice than he does for questions concerning the good life. “We have also to deal with
justice-pluralism and disagreement about rights.” (Waldron 2001:158) This may be an
empirical fact, but does not exclude the possibility of an inter-subjective basis for justice-
questions. Without siding with moral realism we may maintain that norms that regulate the
realisation of interests can be decided in a rational manner. As consequences for affected
parties can be observed, moral norms can be assessed with regard to impartiality. Such
questions are quite different from deciding questions pertaining to how one should live one’s
Cf. “The best institutional structure is the one best calculated to produce the best answers to the essentially
moral question of what the democratic conditions are, and to secure stable compliance with those conditions”
life, which by their very nature are relative to a culture and a valued way of life. In this
regard one may follow Rawls who prioritizes ‘the right’ over ‘the good’.
According to discourse theory morality is built into the procedures which grant the citizens a
right to participation and guarantee their freedom, but which also force them into a process of
argumentation where they must give as well as respond to reasons. Democratic procedures
themselves substantiate the expectation that decisions are reasonable and fair. The legitimacy
of the laws emerges from the processes and procedures that have created them. Habermas
claims that it is the trust in fair procedures that keep modern societies together politically, not
consensus based on substantive world views, values, or virtues. The universal principles of
justice that Habermas refers to are entrenched in modern constitutions as basic rights but the
only source of legitmation is the autonomous will of the people. Hence the discourse
principle, which claims that only those action norms are valid to which all affected persons
can agree as participants in rational discourses (Habermas 1996a:107, cf.Alexy 1995: 131,
In discourse theory it is the procedures that justify the assumption that it is possible to reach
legitimate outcomes. This does not relieve the citizens of taking a moral stand such as in
Rawls’ original position, but ensures that everyone is treated with equal concern and respect
in order for the better arguments to prevail. It is the process, which, if it is a good one,
guarantees the legitimacy of the laws. On the other hand, it is impossible to argue for
everyone’s right to participate in a debate without presupposing some substantial, normative
and non-procedural argument, of for example, people’s freedom, equality and dignity as
postulated by natural law. Procedural and substantive conceptions of justice interchange in a
justification process in so far as the claims of equal access and participation, inclusiveness
and openness rest on substantive principles of tolerance, personal integrity, guaranteed private
life, etc. In logical terms there is an infinite regress or circular argumentation because rights
that are to ensure the process must be justified procedurally, something which again rests on
substantial lements, which must again be justified etc., etc. (Michelmann 1997:162 f.).
The discourse theory does not totally do away with substantial elements. Procedural
independent standards are needed for securing a fair process. In this sense the discourse
principle is in itself normatively charged, it contains a certain normative content - it
“…explicates the meaning of impartiality in practical judgements” (Habermas 1996a:107). A
minimum of normative content is involved in so far as claims of sound mindedness, of
rationality and an ability to reason form the basis of the concept of a deliberative person,
while everything else is left to the discursive process, according to Rainer Forst (1999a:151).
However, discourse theory may also be seen to build on moral premises – on premises of a
moral person who possesses certain rights and competences. There are inbuilt conceptions of
free and equal citizens that are capable of reasoning about justice and the common good
Such a person participates in a discourse where validity claims are raised; the person takes a
critical stand to his own as well as others’ statements and actions, and substantiates his
standpoints with arguments. He or she has the ability to judge (Alexy 1995:120). The person
is prepared to apply the same norm to him as is applied to others. In other words, a
deliberative person is reflective and responsible prepared for and competent of self-correction
(Günther 1999:87pp). This comes close to Rawls’ concept of reasonableness, i.e., “.. the
willingness to propose fair terms of cooperation and to abide by them provided others do.”..
[…] and “ the willingness to recognise the burdens of judgement and to accept their
consequences for the use of public reason.” (Rawls 1993:54) In addition to the virtue of
reasonableness, citizens have to be equipped with a duty of civility sufficient to be competent
to take a stand on the common good.
Moreover, the outcome of moral discourses is not solely depending on the qualities of the
procedure as the required procedures are considered appropriate only in so far that they lead
to a correct or just outcome. The consensus warranting universalisation principle - the
categorical imperative – is a bridging principle, it “.. ensures that only those norms are
accepted as valid that express a general will” (Habermas 1990:63). Habermas’ idea of a just
outcome that is referring to the likelihood that ‘generalizable interests will be accepted’, is
‘certainly substantive’ (Rawls 1995:173). The concept of generalizable or common interests
is a procedure independent criterion.
Not only in Rawls’s theory, but in discourse theory as well, one has to do with a concept of
correctness that includes substantial elements. It is not procedural all the way down as an
For criticism of this point see Apel 1998. See also chapters in Brunkhorst, Köhler and Lutz-Bachmann (Eds.)
1999, Alexy 1996. Höffe 1996.
independent criterion and a core morality, is presupposed.
Consequently, discourse theory
has not been able to free itself entirely from ‘natural law’. It seems difficult to omit such
elements in establishing a principle of normative justification. In fact, then there is not much
difference between discourse theory and the constitutionalists when it comes to presupposing
an element of substance.
4. The imperfection of argumentation procedures
Robert Alexy tries to overcome this problem by showing that the rules of the discourses, as
well as basic human rights do not necessarily depend on discursive justification. They can be
justified empirically with regard to what is ‘the most basic human experience’ as well as with
regard to what purposive rational actors without an interest in rightness, must subscribe to
(Alexy 1996:220). The first argument, that of the nature of human existence, is weak and not
without exception, as consent can be enforced or be the effect of manipulation and
The second argument holds that the sheer cognitive content of the speech rules will do the
work as even a purposive rational person with no interest in correctness, will find it
advantageous to observe the discourse rules. This can also be rebutted, first, because such
rules are binding only for participants in a discourse. This has to do with the so-called
collective action problem; strategists always have the possibility of opting out of
communication when envisaging a better deal.
Secondly, because freedom and equality are
constitutive of the deliberative process, they have priority only in so far as the actors possess
an interest in rightness. An interest in justice is hence to be presupposed. Moreover, only in
so far as deliberative persons are equipped with the capacity to judge, do the rules apply.
The question is nevertheless whether the observance of the rules for rational communication
can guarantee correctness. In other words, even though the argumentation procedure – the
discourse - itself inevitably includes substantial normative elements, can it set the terms for a
fair and rational deliberation process that warrant a right outcome? After all these rules form
It does not amount to pure procedural justice in Rawls’s terminology.
Cp. Wellmer 1986:106ff for the point that there is a decisionistic rest in discourse theory as the commitment
to enter the discussion can not be derived from the principle of non-contradiction. Rationality presuppositions
are different from the presuppositions of morality.
the core element of discourse theory. Regarding norms for how the discussion should be
several rules and prescriptions apply.
Alexy (1989) derives some basic rules from the formal-practical presuppositions of rational
communication. First of all there is a set of process rules, which have to do with securing
participation, equal access, openness, freedom and non-coercion. These rules are based on the
rights ensuring every communicative competent actor real opportunity to participate and that
the members can utter what they want. The rules stem directly from the discourse principle.
But such rules do not ensure a rational discussion, viz., the testing of the standpoints.
Therefore, validation rules that ensure objectivity and coherence in deliberation are required.
These rules pertain to norms such as participants voicing their real opinions without
contradiction, that they consider the arguments of the opponent, and that discussions last long
enough to make sure that the issues are thoroughly discussed, that is, that all stands are made
clear and that all relevant information has been presented (Alexy 1989:188). In addition,
principles for the allocation of the duty to argue are required in order to ensure the fairness of
the deliberative process. Regarding the demand to justify assertions, it is obvious that this
also should involve the opportunity to state the reasons for why one is not willing to provide a
justification, if that is the case. We are here talking about rules for reason giving. Why should
e.g., participants be treated unequally, why is one obliged to produce further arguments only
when met with counter arguments, etc? On the other hand one should stop doubting a specific
point of view once it has been defended adequately. The burden of proof is on the parties who
doubt the arguments of another participant (Alexy 1989: 196f).
These rules operationalize and proceduralize the moral core of the discourse to a large extent
but can not guarantee rational outcomes - one single correct solution. They do not guarantee
agreement because of only partly compliance, unfixed argumentation steps, and because of
historic contingent and changeable normative conceptions (Alexy 1989:206). In other words,
they cannot solve the problem of rational decision-making, because of
1) the burdens of judgement: even reasonable actors may remain at odds with each
other after a rational discussion (Rawls 1993:54ff),
On these points see Alexy 1989:192, 196, and 1981:244ff, Pröpper 1993:138ff, Eriksen and Weigård
2) linguistic uncertainty: grammatical rules underdetermine linguistic meaning
because of multiple contributions of context and usage (Wittgenstein 1978).
However, Alexy maintains that the legal medium itself lays a claim to (moral) correctness in
adjudication. Even though judicial discourses aim at rational outcomes, the question is
whether they accomplish this all by themselves or whether other ‘extra-legal’ procedures are
called upon to redeem the claim to correctness? This pertains to the controversy between
Alexy and Klaus Günther. The former holds on to die Sonderfallthese, the latter to the
distinction between application and justification discourses.
5. Correctness and Indeterminacy
Against legal positivists discourse theorists maintain that law raises a claim to correctness as
it is inescapably linked to the basic stipulations of justice: suum cuique tribuere, to each his
due, equal concern and respect, due consideration of all interests and values, etc. that are
reflective of the impartiality norm of general practical reason, viz, the reason that binds the
free will of autonomous human beings. This moral norm is constitutive of the idea of a justly
organised legal process. Lawyers make up their minds about practical questions, and through
justifying arguments they arrive at presumptively correct answers, i.e., that equal cases must
be treated equally. This is why the kind of reasoning carried out in courts can be referred to as
a practical discourse, and not as logical deduction or strategic interaction. Even in a lawsuit,
where lawyers struggle to obtain the best possible result on behalf of their clients and
themselves, reference is made to objective legal norms and principles when they try to justify
their claims. When appeal is made to impartial judges or to the members of the jury, it must
be made with reference to principles on which rational actors can agree, even if the aim is not
to convince the opponent.
A legal discourse is constituted and regulated by the existing laws. But positive legal norms
are too unclear to give unambiguous, correct answers to normative problems. Rules are
under-specified with regard to action.
They are reflective and subject to interpretation and
contextualisation. This is so because there can be no rule for the correct application of rules –
"Rules are certainly not determinate enough to be predictive; they are not followed in the same way by all
agents, and many often are not followed at all" (Bohman 1991:64).
such a claim opens for an infinite regress. As rules or norms cannot determine their own
application in particular cases there is a cognitive indeterminacy of general practical discourse
But also legal discourses are faced with indeterminacy as positive law as
well has an open texture.
Also the legal language is vague, full of rationality gaps and norm
collisions (Hart 1997:128, Alexy 1998:215).
“Just as norms cannot apply themselves, a legal system as such cannot produce
coherence. To achieve this, persons and procedures are necessary for feeding in new
contents” (Alexy 1999:383)
Lawyers make use of a know-how that is partly determined by the prevailing conceptions of
law and partly by how the law has been practiced earlier (precedent, custom, common values,
etc.). However, legal norms must first and foremost be in accordance with the criteria for
rightness or justice. They must and in fact claim to be correct, even though the criteria are not
given by positive law itself. This is what the discourse theory accomplishes; it specifies the
criteria for correctness (Alexy 1989:11, 18): The rules for legal and practical argumentation
penetrate each other without destroying each other’s respective logics. This is so because
judicial procedures guarantee symmetrical conditions for communication within the legal
community (Habermas 1996a:234). These procedures are not hold to govern practical
argumentation directly, but establish the institutional framework which makes possible a
rational discourse on which norms are appropriate in a given case. How is this possible?
According to discourse theory, both judicial and argumentative procedures aim at rational
outcomes, but none of them can guarantee success because the demanding procedural
presuppositions of such rarely are met. It is difficult to get enough time in legal discourses,
where a decision has to be reached within a given time limit. The problem with
argumentation in a practical discourse is that only the participants can judge if a consensus is
qualified; there is no procedure-independent criterion for the evaluation of a rational
argumentation process. The legal procedure compensates for this weakness, because it
subjects argumentation to temporal, social and substantive constraints. Legal procedures
regulate what topics and questions may be raised, the use of time, the participants, the
distribution of roles etc. The judge as a neutral third party controls that the norms are
followed. These procedures limit the access of premises, they ensure an unambiguous and
binding result, and they connect argumentation to decision-making. Hence, the judicial
procedures compensate for the fallibility of communicative processes and improve their
incomplete or quasi-pure procedural fairness (Alexy 1989:179).
This takes place in two ways:
First, argumentation is disciplined in relation to judicially binding decisions through the
institutionalisation of an expert discourse that can interpret and adapt codified law in a
professional manner, according to internal criteria and specified procedures. Judicial
institutions are designed to systematise and adapt prevailing law to the matter which is to be
regulated. Secondly, correct outcomes can be ensured because the discourse is tied to a legal
public sphere characterised as an open, inclusive and transparent discussion forum. This can
be referred to as the external justification of the principles that are operative in the expert
discourse, as it provides those premises which are not derived from positive law itself (Alexy
In addition to a wide supply of source material – preparatory works, customs,
precedent – moral and political considerations are also taken into consideration, as is
demonstrated in the discussion of so-called hard cases, i.e., fundamental, precedent-forming
cases with a legal political significance.
These are extra-legal factors that are used to decide
which norms should have priority, on the basis of substantial conceptions of justice and
freedom (Habermas 1995:55, Dworkin 1977:81 ff.).
The judges cannot avoid evaluating the validity of approved norms because only a uniform
and consistent legal system can ensure a rational decision. The conclusion must be inferred
from rational normative premises. This is a prerequisite for solving norm-collisions and
practical-ethical dilemmas in the court-room. Legal norms must be interpreted and
operationalised, and even a valid legal norm, for example a constitutional norm, must be
given a legal interpretation in relation to validity criteria before its correctness and relevance
can be established (Alexy 1989:23, cf. Dworkin 1986, Hart 1961).
In democracies legal procedures are meant to guarantee decisions are legally correct and
rationally acceptable, i.e. decisions that can be defended both in relation to legal statutes and
I return to the problem of who are the participants in the legal public sphere.
Cp Habermas 1996a:232. Although the law has rules that regulate collisions between norms, for example lex
superior, lex posterieour, and lex specialis, these are too unspecified to solve disputes in a consequent and
in relation to public criticism. But can the legal system through the discretion of the judges
itself really autonomously settle normative or moral questions which by their very nature are
political? Moral and legal questions point to different audiences, raise different validity
claims and require different procedures for the resolving of conflicts. The problem is whether
the substantial factors are legitimate, and whether the judges’ interpretations of the situations
6. Law and Morality
Law and morality, which both concern practical questions and claim to regulate interaction in
the interest of all parties involved – the resolving interpersonal conflicts, refer to different
contexts of cooperation and have different validity bases. Whereas the law applies to a
concrete community of people which can be subjected to the same duties, viz., a political
community, morality refers to humanity as such. Law is also different from morality in that it
only regulates external behaviour. The law says nothing about the citizens’ motives for
abiding by the laws; it only tells us what actions are illegal and indictable. Morality
presupposes freedom to make one’s own choices. Moral duties can not be enforced. A final
distinction between legal and moral questions is that the law is also a means to realise
collective goals, while moral norms are ends in themselves.
Legal norms apply to
territorially demarcated communities and regulate behaviour that has consequences for
different sets of interests, values and concerns. Hence, they are too concrete merely to be
justified on a moral basis.
With regard to the claim to correctness there is a clear difference between moral and legal
norms, as the latter claim validity for the members of a particular legal society, while moral
norms claim are universalistic as they claim to be valid for everyone – on the basis of free and
But if legal norms are binding only for a community of citizens how
can they be trans-culturally valid, absolute, deontological principles? This determination of
legal norms as relative to their spatio-temporal embeddedness does not sit very well with the
See Kant 1785, Habermas 1996a, and consult Wellmer 1986:114ff.
“The moral universe, which is unlimited in social space and historical time, includes all natural persons and
their complex of life histories; morality itself extends to the protection of the integrity of fully individuated
persons (Einzelner). By contrast, the legal community, which is always localized in time and space, protects the
integrity of its members precisely in so far as that they acquire the artificial status as rights bearers”. (Habermas
concept of legal norms as deontological – as absolutely binding (Alexy 1998b:228). Alexy
proposes to see principles as non-conclusive optimization requirements.
Principles such as
freedom, equality, rule of law, democracy are not merely deontological norms, functioning
like trumps in collective decision making processes; rather they are to be seen as norms to be
weighted and balanced in the adjudication of particular conflicts, they are to be optimized to
the greatest extent possible. Principles only take the character of a trump “.. when some
competing principle has a greater weight in the case to be decided“ (ibid p. 58, cf p. 102). In
the wording of Alexy: “A principle is trumphed when some competing principle has greater
weight in the case to be decide. … Principles represent reasons which can be displaced by
other reasons” (Alexy 2002:58, 57). To be able to reach an optimal decision principles must
be weighted and balanced. They are not anti-utilitarian trumps. Principles neither protect the
separation of power unconditionally – the red-tape argument – nor human rights – the fire
wall argument. I return to this. For the moment the point is that law cannot itself establish a
sufficient account of the normative premises at work in legal argumentation.
On the one hand legal argumentation is faced with the requirement of making right decisions,
on the other hand due to the contested and open texture of positive law – a law is never
absolutely lucid - correctness can not be provided by legal standards alone. “Therefore only a
recourse to standards other than legal standards is available, such as general reflection on
utility, traditional and common ideas of what is good and evil, as well as principles of justice”
(Alexy 1998:216). Legal norms certainly are premised not only on moral reasons, but on
utilitarian considerations, prudent reasons, collective values, etc., as well. But how is it
possible to ensure correctness in adjudication? Alexy claims that this requirement can be
fulfilled by conceiving of legal argumentation as a special case of a general practical
”Under the conditions of a non-positivist concept of law nothing is left to be
connected because substantial correctness, and therefore morality, are already part of
law.“ (Alexy 1998:219)
“… characterised by the fact that they can be satisfied to varying degrees, and that the appropriate degree of
satisfaction depends not only on what is factually possible, but also on what is legally possible.” (Alexy
The special case thesis is objected to by legal positivists, legal realists (such as Niklas
Luhmann) as well as discourse theorists such as Jürgen Habermas and Klaus Günther: It is
not for the legal discourse to justify normative claims. Günther has introduced the distinction
between a justification and an application discourse maintaining that it is in the particular
ruling, in the individual case, that the weight of rules, norms and values can be determined.
The norms that have passed the test of justification have to be implemented correctly.
7. Justification and application
Practical rationality is about normative questions or reasons for actions and refers to what is
obligatory, prohibited or permitted. Such decisions are so to say a compartment of moral
obligations and conscience. The relation between autonomous morality, that refers to
irreplaceable human beings or humanity as such, and positive law, which regulates actions
that have diverse consequences, is a complementary one; moral norms must be implemented
and the law must be justified. Moral norms can only be realised if they are formulated in legal
categories which allow sanctions, and the law can only obtain legitimacy to the extent that
there is equality before the law. It is the legal medium that transforms such obligations and
concerns of conscience that are legally relevant valid into practical results. It entails
institutional mechanisms for connecting validity and facticity such as secondary rules for
While justification discourses require that all interests are considered and judged impartially,
application discourses require a procedure where all relevant features of the situation are
given equal treatment (Günther 1989:155). Application discourses are due to the limitedness
of our actual knowledge and of the interests and values of the members. In a legal debate
where concrete matters are to be decided, the question of the universal validity of the norms is
put into parenthesis. This validity is simply presupposed, and one proceeds to ask which
norms are relevant and should apply to a particular case.
“Everybody who uses a valid moral norm as a reason for a justification of singular normative propositions has
to observe at least two principles: (1) A justification of a singular normative proposition requires a complete
description of the concrete case with respect to all those moral reasons which are relevant. (2) A justification of
a singular normative proposition requires a coherent interpretation of those morally valid reasons which are
directly or indirectly relevant to the concrete case.” (Günther 1993b:151)
It is the consequences which are at issue in application discourses. The point of interest is
what the effect will be if justified moral norms are implemented. In an application situation,
we are faced with a choice between justified norms, which must be decided in relation to a
number of factors: empirical elements, the type and quality of the information available,
actual power relations, the important values at stake as well as the balancing of non-
generalisable interests. Legal norms have a much more complex validity basis than have
moral norms. The law realises political values and ethical and moral norms by positivising
them. Through this procedure, they become sanctionable and collectively binding. Morality,
on its part, tests and justifies positive legal norms. Whether legal decisions are correct,
ultimately depends on whether the decision process has made possible impartial judgment,
which requires that certain higher-ranking conditions for argumentation have been met.
While discourses of justification pertain to justifying general legal norms in abstracto and in
light of the consequences its observance may have for affected parties, discourses of
application have an hermeneutic structure and relate to the justification of concrete
judgements given already justified norms.
These two discourses are dealing with two kinds of questions, but the question of what is
correct in a situation and which moral norm is correct can not easily be distinguished.
practice they are intertwined, and to give up the claim of moral justification with regard to
situational features is to renounce rational decision-making.
Hermeneutics takes everything
into consideration but can not tell what is valid. This strategy, according to Alexy, means
giving up the claim that the application of norms should be right. “It is empty, because it does
not say which aspects are to be considered in what way” (Alexy 1998b:231). Hence, it
cannot solve the problem of rational adjudication. Günther can not uphold the claim to
correctness. This can only the accomplished by seeing that every application necessarily
involves a justificatory discourse, complying with the rules of rational argumentation.
“From the fact that those two questions have to be distinguished it does not follow that there exist two
essential different kinds of discourse“ (Alexy 1993:167).
“The fundamental moral demand of equal treatment runs empty because in the sparsely furnished normative
universe of this construction there is nothing that could grant equal treatment” (Alexy 1993:163).
This is possible as “the law is not only open to moral criticism from the outside. The critical dimension is
replaced into the law itself” (1999:382). “General practical arguments have to float through all institutions if
the roots of these institutions in practical reason shall not be cut off. “(1999:384)
8. Overtaxing the legal medium
For Alexy (1999: 375ff) a practical discourse does not merely comprise moral questions of
justice and universalisation (as in Habermas’ and Günther’s set up), but rather is a discourse
encompassing the whole spectre of normative, non institutional reasons, that have to be
employed to settle what ought to be done, to reach a correct decision in case of interpersonal
conflicts. A legal discourse differs from a practical discourse in that it deals with what is
correct within the legal framework and within a time frame and not with what is ultimately
But the judge can not omit making use of political or moral arguments in order to
justify the balance he strikes when it comes to the basic principles and their sub-principles.
“It follows, then, that the claim to legal correctness necessarily attached to the decision
includes a claim to moral correctness.” (Alexy 2002a:77)
Alexy (2002b:370-71) analysis is based on the reconstruction of a four-stage procedural
model of legal argumentation, which combines law and rational practical arguing in such a
way that rational decision-making is possible. His point of departure is the rules of rational
practical discourse, which are easy to justify but comes at a price – they are “compatible with
very different outcomes”. Hence practical discourse needs to be complemented by law, first
by the legislative procedure in which decisions are made but which “…is not capable of
establishing in advance just one solution”, then by a legal discourse which has to “…respect
statue, precedent, and legal doctrine … but … uncertainty of outcome is not totally
eliminated”. There is thus need for a fourth procedure – the court process – in which one and
only one decision is made and which can claim to be rational because of the process.
The problem that emerges here pertains to the limits of the law and that of the judges’
competence. How can we know that the judges are right when they do not merely apply
politically laid down laws and reasons but make use of their own discretionary know-how and
value-base? In the theory of Alexy there are the dangers of assimilating law and morality and
of overburdening the legal medium itself. These dangers are interlinked:
“It is not concerned with what is absolutely correct but what is correct within the framework and on the basis
of a validly prevailing legal order” . . it “… is bound to the statutes and to precedents and has to observe the
system of law elaborated by legal dogmatics.” (Alexy 1999:375)
1) The first problem is simply this: How can the imported substance, the norms and values
that are needed for reaching a single correct decision, be tested? How can we know they are
right? Only norms that have been accepted by the parties in a free and rational debate can in
principle be considered legitimate. As well as it is only in a discourse among affected that we
can know what justice really consist in, we can not now whether the adoption of rights
clauses in particular situations are right unless the ones concerned have been heard. This is
underpinned by the fact that the right is permeated by the good, as also Alexy himself points
to. One’s conception of justice is affected and can be changed according to altered self-
understandings and revised self-descriptions. Even human rights require democratic
legitimation and public deliberation to be correctly implemented. They are unfulfilled until
they have been codified and interpreted (and subsequently transformed to basic rights). This
means there are no fixed moral precepts, principles or concepts of justice that judges in a non-
controversial way can appeal to in order to adjudicate in case of conflicts.
According to legal positivists
as well as legal realists
this can not be settled by the law:
The medium of law, based on the binary coding legal/non-legal can not justify normative
questions themselves. What is correct/right have to be settled outside of the legal media. The
law itself simply can neither deny nor confirm the validity of a moral argument. This directs
us to the overtaxing problem.
2) The latter pertains to the question of the addressees. There is a claim to correctness as the
judges both out of conceptual and moral reasons have to contend that their decisions are right.
But can the lawyers really establish the premises for redeeming this claim all by themselves?
Are they so to say the right addressees? Alexy distinguishes between institutional and non-
institutional circles of addressees. The former are the ones affected by respective legal acts,
the latter contains relevant participants who can use all arguments allowed in a legal
discourse – in the legal public sphere, as mentioned. The latter includes everyone who takes
the point of view of a participant in the respective legal system. It is an inclusive public forum
Cf, the following statements from one of the founding fathers of legal positivism: “Legal decision consists of a
given case’s subsumption under valid law; like any other logical conclusion it is independent from will. There is
no freedom on the resolution whether the consequence should take place or not; it is produced-as it were-by
itself, by intrinsic necessity” (Paul Laband 1911:178, cited from La Torre 2002:379)
”Das, was richtig is, wird nun durch eigenen Programme fixiert” … and “ Die frage nach der Richtighkeit der
Programmatik als solcher hat keinen erkennenbaren Sinn- es sei denn, … , im context einer Rejektion des
Rechtscodes.” (Luhmann 1995:192)
which can bring forward the interests and viewpoints of legal consociates that are relevant for
passing a correct judgement. The premises that can not be derived from the actual law are to
be justified by the external ‘jury’.
But who are the participants in such a process and what
arguments are permitted? It follows from the constraints of a legal discourse that only judicial
arguments are allowed and that it is not the participants but the judge who has the last word.
Admittedly the equality criterion can be approximated by making more open and inclusive
forms of participation possible, but the legal argumentation is from its inception constrained –
as an impartial judge is present and the participants are legal consociates:
“No matter how many schemes we conceive in order to increase the number of
interests and arguments to which judges are exposed, the fact that it is they and not the
parties who have the last word precludes any direct reference to participation as a
source of legitimacy” (Menéndez 2001:240)
In a practical discourse there is no neutral judge, no limited number of participants or
constraints on time, the themes for debate, etc. In a justification discourse there can only be
participants. Only the participants themselves can pass judgment over their equal interest and
their common good. There can be no procedure independent criteria of correctness. From a
democratic point of view these two questions, that is what is legal and what is normative or
politically valid, branches out in two kinds of procedures – the legal one where the addresses
are confined to the circle of legal consociates, and the political one referring to circle of
citizens, which are not merely bound by the legal medium but also authorized as le pouvoir
constituant to constitute power and give the law new normative content: Common action
norms can only be legitimately tested in the wider public sphere where competent citizens
and all affected parties are present.
“Once the judge is allowed to move in the unrestrained space of reasons that such a
“general practical discourse” offers, a “red line” that marks the division of powers
between courts and legislation becomes blurred. In view of the application of a
particular statute, the legal discourse of the judge should be confined to the set of
“It is the task of the external justification to justify those premises which cannot be derived from positive law”
(Alexy 1989: 228, cf. Hart 1961: 89).
“.. one cannot deny that it is the court which is to decide and argue in the last instance” (Alexy 1999:377).
reasons legislators either in fact put forward or at least could have mobilized for the
parliamentary justification of the norm” (Habermas: 1999:447)
9. Deliberative democracy and legal uncertainty
Concerning the question whether legal argumentation can satisfy the conditions for legitimate
law-making, as the special case thesis holds, or whether other procedures are called upon as
Günther and Habermas suggest, from a democratic point of view the latter solution fares
better. Even though Alexy may be right in maintaining that application discourses can not
solve the problem of rational adjudication of interpersonal conflicts, the procedures for
tackling normative content, substance, are better specified in the formers suggestion.
“Procedural rights guarantee each legal person the claim of a fair procedure that in
turn guarantees not certainty of outcome but a discursive clarification of the pertinent
facts and legal questions.” (Habermas 1996a:220)
The general problem with the special case thesis in this perspective is that judges should
apply norms, not make them. Conceiving of the legal discourse as a special variant of the
practical discourse blurs the distinction between legislation and application because it allows
the judge to make use of normative reasons in general, not only the ones given by the
legislators. Now Alexy may defend his thesis by pointing to the fact that in modern societies
heavily strained by complexity the political system become overburdened and does not
produce the required set of norms. When the legislator does not fulfill its function the courts
have to intervene and upgrade the legal system so that it becomes possible to handle the
complexities facing it. It leaves the generation of norms to be handled autonomously by the
discretion of the judges - hence the prevalence of delegation and framework legislation. The
politicians are not doing their job in furnishing the legal system with the required normative
premises and leaves to the discretion of the judges to find the ‘correct’ normative basis for
Legal discourses refer “from the outset to democratically enacted law … ” (Habermas 1996a:234). On the
problems of principles as trumps see Alexy 2002b: 102, a 77ff, Günther 1993a: 214ff, 219,227
As far as this is the actual situation of the legal system in a modern welfare state, the
‘Sonderfalltese’ is plausible in descriptive terms. But it can not be sustained normatively. It
does not sit with the basic principle of democratic legitimacy that the people should be the
final arbitrator of constitutional law (subjected only to a limited set of constraints). Alexy’s
solution gives too much leeway to the discretion of the lawyers as they become authorized to
choose the decisive reasons themselves. But how can the problems of legal uncertainty be
avoided when judges have to make use of extra-legal premises in order to decide in conflicts
over the law? How can affected parties know that the decision is correct and how can they
check that the premises of a ruling are the right ones?
These queries makes us aware that another concept of democracy is required than the one
based on majority vote. As mentioned the majoritarian model of democratic politics is
inadequate. The voting procedure cannot ensure correctness because of its un-attentiveness to
reasons. Rather we should conceive of it as deliberative democracy where citizens are
involved in inclusive legislative processes and where inputs as well as outcomes are subjected
to public critical debate. The essence of the deliberative conception of democracy is that
citizens have a right to justification of those action norms that affect them. It is based on ‘the
rule of reasons’ and on the moral principle of reciprocal and general justification (Forst
2001). It defines popular rule as legitimate only if stringent prerequisites are satisfied such as
equal access, autonomy, full information, openness. These prerequisites may be hard to fulfil
in any real existing form of democracy, and should be conceived more of as regulative ideas.
Deliberative democracy is, nevertheless, consistent with the republican idea of popular
constitutionalism conceiving of the people and not the judge as the interpreter of last resort.
In this perspective the whole complex of demanding communicative arrangements in the
political sphere of action comes into consideration. In this sphere proposals are subjected to
public debate prior to law enactment. Prior to adjudication there are comprehensive processes
of norm-testing deliberation. In the general public sphere in civil society - in media and
newspapers - in networks, social movements, popular assemblies etc., problems are
discovered, thematized and dramatized, and social concerns are verbalised and claims are
justified. Here a moral discussion over what norms should prevail and how different and
similar cases be treated, take place. The general public sphere, the generic principles of which
are participation, inclusion, equality, freedom, open agenda, is the locus focus of popular
sovereignty and practical reason. Here people can address the political questions of the day,
can assemble and try to influence the political system and in rare moments – in constitutional
moments - also change its basis rules (cp. Ackerman 1998).
What is more, politicians increasingly find the situation of jurist made law in the modern
welfare state problematic and have initiated a plethora of quasi-legal and quasi-political
bodies out side of the political system, in order to settle the problem of legitimate normative
inputs to the political and the legal system. These bodies are so to say involved in the
generation of the norms. On the input side of the political system we find different kinds of
public hearings, popular meetings, citizens juries, experiments with deliberative reflection
groups, ethical committees, consensus-conferences, etc. which all are oriented towards the
enlightening and the resolving of controversial value- and norm-questions. On the output side
of the political system as well we find a plethora of bodies where citizens are involved in the
concretization and operationalisation of the political laws and regulations.
The new participatory forms of governance of a deliberative kind is important because here
legitimacy is not merely a matter of the citizens’ preferences as they are expressed by the
voting slip, but what reasons the participants have for agreeing or disagreeing with a
standpoint. Some of these deliberative sites may be seen as forms of delegation of the norm-
production of the society as they involve the citizens in reflective moral deliberation. Their
democratic value pertains to the proviso that only through a fair process of argumentation
between affected can one tell whether a norm is unbiased or not. A legitimate decision is not
an expression of preconceived ideas of what is just or the common good but the outcome of
everybody’s deliberation in a free, open and rational process. Consequently, there is, in
principle, an alternative to the legal public sphere when it comes to the settling of normative
10. Proceduralism and judgement capacity
But also on a deeper level there is a problem with Alexy’s architectonic. This has to do with
the aforementioned notion of principles as optimization requirements and the implied
weighing criterion, which rule out the possibility of being able to criticise the principles
Ranging from Conflict councils and law mediators, family courts to user-participation in a broad range of
public service institutions, cp. Eriksen 2001.
themselves in light of higher ranking norms and validity claims. Principles conceived of as
optimizing requirements are subject to weighting and balancing and hence are framed on
merely axiological values and weak evaluations reflective of wants and preferences and not
moral norms in the Kantian sense.
What happens then when rules, which are definitive
norms related to permissions/prohibitions collide, and when it is necessary to recur to higher
ordered principled debates on what norms should have priority? The problem is how to solve
collisions of norms rationally when the procedures for discursive meta-legal assessment are
While Alexy, on the one hand, subscribes to the discourse principle he holds, on the other
hand, certain moral norms as valid prior to a discourse. The procedural concept of correctness
say that a norm N is correct when it is the result of the procedure P. However P is not
reserved for discursive justification: “P kann durch ein Individuum durchgefürt werden,..”
(Alexy 1995: 96, 104, 110). Norms can so to say be justified in a monological way. As we
have seen Alexy 1996 justifies basic human rights in a way that does not depend on
discursive justification. I this regard Alexy sides with the constitutionalists. His concept of
procedure is not connected to the criterion of a rational consensus as is the justification
program of Habermas (1983), but to the carrying out of the discourse procedure. He
disconnects consensus and correctness. The norm N is not correct because of a rational
acceptance in actual or ideal discourses but because of the way the procedure is carried out
enables judgement capacity, viz., the way the members manage to distinguish good from bad
reasons. “Was ein guter Grund ist, kan sich erst im Prozess der diskursiven Überprüfung
zeigen.” (Alexy 1995:121) The achievement of the procedure depends on the judgement
capacity of the members. It is the purpose of the procedure to develop such. The procedure is
to enhance the capacity to judge by the particular discourse members, to distinguish between
good and bad reasons with regard to what should be done. But, then, what is the criterion of
correct decisions? There is no basic procedure for testing the fallible outcomes of actual
discourses. In other words, there is no recurs to a more basic justification procedure. What is
at stake with this conceptual strategy is the space for discretion left over to the judges, and
hence the expansion of “constitutional legal reasoning” to the detriment of the political
process itself and of the citizens self-government.
Alexy so to say presupposes “ a teleological reinterpretation of principles” and he reduces “the justification
problem to justifying decisions of preference” (Günther 1993:227).
The substance involved, the restrictions within legal discourses needs to be checked, and this
check cannot take place at the same level – it can only be tested by a higher order,
justificatory discourse (Pavlakos 1998: 153).
Because the rules of the discourse are basic to
justification, their correctness must be justified on a deeper level (Gril 1998: 158, 162). Thus
there is a problem of relativism in Alexy’s theory. He has not provided an ‘objective’
correctness theory. His concept of correctness is relative: It depends on the (arbitrary)
judgement capacity of the judges. Alexy is a procduralist who is not a ‘democrat’ in the
present use of the term. The substantial rest in Alexy’s program is larger than in Habermas’s,
in which the rational consensus establishes a criterion for correctness, i.e., when affected
parties comply with identical reasons under free and equal conditions. Here constitutional
proceduralism hinges on discursive proceduralism. The task is therefore to establish a
procedural standard for legitimacy where the democratic deliberation process itself makes the
result right. But does the discourse theory really accomplish this?
11. Epistemic or moral justification?
According to Habermas any substantial standpoint raises claims of justification and makes
demands on knowledge which can only be met argumentatively. On the one hand, only
substantial standpoints can justify outcomes. We do not know whether a case has been
correctly dealt with until we know the parties’ arguments – their views, reasons, valuations,
etc. On the other hand, no grounds are beyond justification in a political context, something
that is due to the fact that there is no reality-check of normative statements. Norms are only
‘valid’ to the extent that the actors agree on them in a rational discourse.
All contents, no matter how fundamental the action norm involved may be, must be made
to depend on real discourses (or advocacy discourses conducted as substitutes for them)
Discourse theory brings together substance and procedure when it is maintained that in a free
argumentation process the validity of the reasons can be tested, viz. they can be assessed in
“If one allows a picture of the discourse, where these is no room for any substantive norms at its starting
point, and where consequently all discourse-rules are simply speech rules (….,) then one should renounce the
idea of a substantive claim for rightness; in this case the rightness-claim would remain a sheer formal concept,
which would be nevertheless too weak to justify the whole enterprise of practical discourse” (Pavlakos 1999:9)
On this point see also Habermas 1999:302ff.
relation to autonomy and universality. It is therefore the process that justifies the outcomes
(Habermas 1996b:1508). Neither the formal qualities of the procedure nor substantial grounds
justify the outcomes. Rather it depends on whether the process has included objections and
counter-arguments in such a way that outcomes can stand up to public criticism. “ …
(L)legitimacy through procedure does not result from the structure of procedure itself, which
guarantees the right of participation, but rather from the quality of discursive processes which
they make possible” (Maus 1996:880-81). It is a qualitative good process of argumentation
that is able to test the justifiableness and reasonability of claims and interests involved that
bear the burden of legitimation in discourse theory. Thus it is neither the participation of all,
nor expressions of will and the respect for preferences that give the democratic process its
legitimating power, but rather the deliberative process. This power lies in the access to a
process where standpoints can be tested in such a way that one can expect reasonable and
publicly acceptable, i.e., rational results.
The problem with this solution is that the standard for evaluating the quality of the outcomes
is given independently of an actually performed deliberation process. The standard is
constituted by an ideal procedure, which specifies the contra-factual conditions for a public
discourse where all limitations on time and resources have been suspended, and where the
only authority is that of the better argument. Real communication processes can only
approximate such ideals. To Habermas the rational consensus is the standard by which the
correct outcome can be defined. It is the criterion of legitimacy. By observing the ideal
conditions for conversation one should arrive at the correct decision – one that everyone can
approve of. The problem is that this does not explain why the reasons for an actual decision
are good reasons. How can public deliberation be both moral and epistemic, i.e., how can
features of the process justify the outcome at the same time as it has good effects? Under non-
ideal conditions the problem with justifying the epistemic value of deliberation arises. This is
due to the fact that actual deliberations will not generally meet ideal requirements: they will
be marked by, for example, ignorance, asymmetric information, power and strategic action.
Estlund (1993) and Gaus (1997) therefore ask the question of whether the reasons that can be
stated publicly also are good (convincing or correct) reasons.
Ideal proceduralists attempt to avoid these problems of justification. Their claim is that the
ideal deliberative procedure is constitutive for correctness as long as certain conditions are
met. But if correctness is seen as what the actors will support under ideal conditions, it will be
difficult to prove the epistemic qualities, i.e., that actual deliberation leads to better and fairer
decisions (Bohman and Rehg 1997a:xix). This is only possible when one can appeal to a
process-independent standard of truth or correctness. This is what the constitutionalists put
forth when they operate with a substantial measure of justice as it is formulated, for example,
in the theory of justice as fairness as we have seen.
In order to defend the epistemic qualities of deliberation, process-independent standards are
needed. An epistemic justification of outcomes will in that case become independent of ideal
deliberative conditions, but dependent on what the deliberation leads to with regard to
rational decisions – independently defined. We are therefore faced with the following
paradox: if deliberative democracy defends its claims on moral qualities via an ideal process,
it cannot legitimate its claims on epistemic value. On the other hand, if deliberative
democracy claims to have epistemic qualities, it can only be defended by standards that not
only are process-independent, but also independent of actual deliberation (cf. Bohman
Gutmann and Thompson (1996:27) argue that deliberative justification is neither entirely
procedural nor substantial (constitutional) - none of them have priority, but both are
necessary. They argue that substantial principles should be included but these are to be seen
as moral and political provisional (Gutmann og Thompson 2002). Then little is won. David
Estlund also argues for a mixed principle that does not make the legitimacy of the result
dependent on its correctness, but on the epistemic value of the process that created it:
“Democratic legitimacy requires that the procedure is procedurally fair and can be held, in
terms acceptable to all reasonable citizens, to be epistemically the best among those that
are better than random.” (Estlund 1997:174)
This is a weak principle. The only requirement is that the citizens agree that there are more or
less favourable ways of evaluating arguments. Consequently a “thin substance” is involved
(Estlund 1993). The normative problem with this solution is that it lacks a theory of criteria
specification and of conflict resolution based on a theory of public justification (Bohman
1998:407). The analytical problem is that this solution does not pick up on the moral
authority of the process, i.e., that the process itself has the ability to bring forth results and
which by itself induces respect. It is only the moral qualities of the procedure that can explain
that political decisions have a binding power also on those who disagree (cf. Tyler 1990). It
is hard to see how discourse theorists can maintain that it is the epistemic value of
deliberation that makes results correct, when the procedure itself is constituted by moral (non-
deliberative) norms. But what kind of standard are we talking about?
12. Discursive proceduralism
The tension between substance and procedure that occupies such a prominent place in present
day political theory reflects the very constitutional structure of modern democracies in which
the citizens have the opportunity to give themselves the laws that they have to obey, but
where the participation is regulated by non-disposable procedures and inalienable rights.
Fundamental norms like equality, autonomy, human dignity, legal certainty, democracy,
which make up the basic procedures of the modern constitutional state, are higher-ranking
principles, belonging to the deontological realm. The procedures themselves are normatively
charged and are subjected to judicial review of one sort or the other, monitored by judicial
bodies. The tension then is involved in the medium of law itself as it claims to hand down
correct solutions but can not accomplish this without importing substantial reasons of what is
right or wrong, good or poor. This tension repeats itself in the discourse theory as only the
procedure can tell which argument is a good one, but the speech rules constituting the process
reflect substantial elements. So are the Constitutionalists right after all?
The standard for democratic justification that is at work here is not an objective one. It
involves actors, their arguments and fallible opinions and it refers to the idea of a justly
organised process. Hence a concept of justice is presupposed. This standard is an intrinsic
critical standard for assessing existing imperfect democratic procedures: it can be used in
assessing every actual institutionalisation of political deliberation and decision-making.
“What it can argue for are improved forms of justification if there are grounds to
assume that good reasons have been neglected, but there is no independent way to
“find” an objective truth beyond reciprocal and general argumentation” (Forst
Regarding this it becomes clear that the alleged problem of infinite regress is only a problem
for foundationalists – it may be seen as a “… product of an old providential model of
authority that leads us to look for authorization prior to action rather than the other way
round.” (Honig 2001:796). In discourse theory, which builds on a pragmatic conception of
validity there is a departure from induction and deduction as rightness warranting principles.
Argumentation replaces propositional inferences as the central area of logic in practical
contexts. The theory is based on the connection between a speech act and its pragmatic
presuppositions. There are basic rules of argumentation that cannot be opposed, because they
are presupposed as valid already when the discussion starts. The constitutive rules of the
game can be criticized, ridiculed, changed, modified but can not be disposed of if the game is
to be played at all. It is this idea of procedure that makes criticism and correction of
established processes possible. Thus an independent standard – but not a pure procedural
justice standard - is given for the evaluation of actual procedures and political decisions. This
standard is not objective in a simple scientific sense as it includes normative elements (such
as a notion of justice and democracy) and cannot yield correct results as the ideal conditions
of discourse can only be approximated. It is only a question of better answers compared to
those that can be reached by means of democratic processes, in the sense that an
argumentative process will bring about more justified standpoints.
The standard is thus imperfect, but is itself is an expression of a normative procedure for
justification, which is entrenched in the legal and moral nexus of modern constitutional
democracies. Discourse theory reads the constitutional rights as procedural arrangements for
substantiating the presumption of popular made law. In this specific meaning procedures
overrule correctness theories, but the procedure itself is constituted by substantial claims to
freedom and equality. The latters cannot be explained procedurally, because the very
procedural claim of equal right to participation itself presupposes the basic substantial claim
of everybody’s equality and the right to equal freedom. Such basic moral or human rights
have priority over other claims - this is why they are constitutionalised in the first place - and
they also must have priority in order for democracy to be acceptable for all (Gosepath
2001:389). Consequently, in discourse theory there is a substantial rest that leaves room for
judicial review and for judges as guardians of the constitution in ensuring correct results.
Generally, discourse theory holds that correctness is more a question of what can be justified
in a process, than what is right or fair according to some external (theoretical) standard. It is
the citizens’, not the judges or the political philosophers’ moral judgement that is decisive.
But the process must be secured by substantial elements, because it is required that the
procedure is fair. In contrast to contractarian constitutionalism the task of philosophy is,
according to discourse theorists, merely to clarify the moral point of view and the criteria for
democratic legitimacy through an analysis of the procedural requirements for a rational
debate. It does not include deducing certain norms or criteria for distribution of resources.
Neutrality in this context has to do with ensuring mutual respect and equal conditions for
communication, so that everyone can express their opinion and that arguments can be tested.
The debate on substance and procedure, is not merely a theoretically one because it has to do
with whether there can be a democratic enactment of a constitution. What can people
sovereignty mean under the rule of law in modern society? In short, should the legislators or
the judges have the upper hand? Should the Supreme Court’s or the politicians be the final
arbiter of constitutional law? This article has established that proceduralists cannot do without
substance in conceiving of popular rule. For the people to be free to make their own
constitution some moral properties are presupposed. Hence, the rights-based perspective of
the constititutionalists have a case against proceduralists. But what is meant by procedure? If
it comes down to solely majority vote, a procedural approach cannot be sustained. If, on the
contrary, it designates the procedures for ensuring a fair deliberative process – a fair
procedure of reason giving – the case is a bit different. Deliberationists like the discourse
theorists plead for a larger concept of democracy than the one depicting it as merely a voting
arrangement and hold that substantive and procedural conceptions of justice interchange in
In this regard one may conceive of politics as a special variant of practical reason. It is
specialised on collective goal realisation revolving on common good considerations
constrained by the principle of justice. Moral and legal principles are seen merely as the
constraints that ensure equal access, autonomy and equality. The constitutive power - le
pouvoir constituant - always lies with the people; all law and public authority stems from the
citizens. Moreover, it is the deliberative process itself that bear the burden of legitimation in
so far it manages to define and mobilize support for collective goals and secure a fair
treatment of all actors’ interests and needs. In this model, judging the legitimacy of the
arguments is in the last instance left to the participants involved in the discussion. A
legitimate decision is not an expression of preconceived ideas of what is just or the common
good but the outcome of everybody’s deliberation in a free, open and rational process. Such a
procedure is constituted by substantial moral norms that can not themselves be
proceduralised. What then from a democratic point of view seems to be the contribution of
discourse theory is the way it does not fix substantial conceptions of what should be done in
final categories, left over to the politicians, judges and bureaucrats to implement, but
constantly subject them to argumentative testing and higher-ordered procedural constraints
based on idealized deliberative preconditions.
Ackerman, B. (1998): We the People: Transformations, Cambridge, Mass.: Harvard University Press.
Alejandro, R. 1996: "What is Political about Rawls’s Political Liberalism?” In The Journal of Politics
Alexy, R. 1981: “Die Idee einer prozeduralen Theorie der juristischen Argumentation.” In A.
Aarnio et al. (Eds.): Methodologie und Erkenntnistheorie der juristischen Argumentation. Berlin:
Duncker & Humbolt.
Alexy, R. 1978/1989: A Theory of Legal Argumentation. The Theory of Rational Discourse as Theory
of Legal Justification. Oxford: Clarendon Press.
Alexy, R. 1993: "Justification and Application of Norms". In Ratio Juris Vol. 6 No.2 July, pp.157-70
Alexy. R. 1995: Recht, Vernunft, Diskurs. Frankfurt: Suhrkamp.
Alexy, R. 1996: “Discourse Theory and Human Rights”. In Ratio Juris. Vol. 9. No.3: 209-35.
Alexy, R. 1998: “Law and Correctness.” In Current Legal Problems 51: 205-21
Alexy, R. 1998b:”Jürgen Habermas' Theory of Legal Discourse,” in Michel Rosenfeld and Andrew
Arato (eds.), Habermas on Law and Democracy: Critical Exchanges, Berkeley, University of
California Press, pp. 226-31.
Alexy, R. 1999: “The Special Case Thesis”. In Ratio Juris. Vol.12 No.4. (374-84)
Alexy, R. 2002a: The Argument from Injustice. A Reply to Legal Positivism. Oxford:
Alexy, R 2002b: A Theory of Constitutional Rights. Oxford: Oxford University Press
Apel, K-0. 1998: Auseinandersetzungen in Erprobung des transzendental-pragmatischen Ansatzes.
Barry, B. 1991: Democracy and Power. Oxford: Oxford University Press.
Benhabib, S. 2002: The Claims of Culture. Equality and Diversity in a Global Era. Princeton:
Princeton University Press.
Bohman, J. 1991:New Philosophy of Social Science: Problems of Interderminacy. Cambridge, Mass.:
The MIT Press
Bohman, J. 1998: ”Survey Article: The Coming of Age of Deliberative Democracy”. The Journal of
Political Philosophy, vol. 6, no. 4.pp.400-425
Bohman, J. and W. Rehg (eds.) 1997: Deliberative Democracy. Cambridge, Mass.: The MIT Press
Brunkhorst, H., W.R. Köhler and M. Lutz-Bachmann 1999 (eds.): Recht auf Menschrechte. Frankfurt:
Dahl, R.A. 1989: Democracy and its Critics. New Haven: Yale University Press.
Dworkin, R. 1977: Taking Rights Seriously. London: Duckworth.
Dworkin, R. 1981: ”What is Equality? Part 2: Equality of Resources”, In Philosophy and Public
Dworkin, R. 1986: Law’s Empire. Cambridge, Mass.: Harvard University Press.
Dworkin, R. 1996: Freedom’s Law. Cambridge: Harvard University Press
Eriksen, Erik O. 2001: Demokratiets sorte hull. Oslo: Abstrakt Forlag.
Eriksen, E. O. and J. Weigård 2003. Understanding Habermas. Communicative Action and
Deliberative Democracy. London: Continuum.
Estlund, D. 1993: "Who's Afraid of Deliberative Democracy? On the Strategic/Deliberative
Dichotomy in Recent Constitutional Jurisprudence". In Texas Law Review
Estlund, D. 1997: "Beyond Fairness and Deliberation: The Epistemic Dimension of Democratic
Authority.” In J. Bohman og W. Rehg (red.): Deliberative Democracy. Cambridge, Mass.:
The MIT Press.
Forst, R. 1994: Kontexte der Gerechtigkeit. Frankfurt:Suhrkamp
Forst, R. 1999a: “Die Rechtfertigung der Gerechtigkeit. Rawls’ Politischer Liberalismus und
Habermas’ Diskursteorie in der Diskussion”. In H. Brunkhorst and P. Niesen (red.) Das
Recht der Republik. Frankfurt: Suhrkamp
Forst, R. 1999b: “The Basic Right to Justification: Toward a Constructivist Conception of Human
Rights”, In Constellations 6(1):35-60.
Forst, R. 2001: ”The Rule of Reasons. Three Models of Deliberative Democracy”, In Ratio Juris. Vol.
14. no. 4. pp. 345-78.
Gaus 1997: ”Reason, Justification and Deliberation: The Epistemic Dimension of Democratic
Authority.” I J. Bohman . og W. Rehg (red.): Deliberative Democracy. Cambridge,
Mass.: The MIT Press.
Gosepath, S. 1995: "The Place of Equality in Habermas' and Dworkin's Theories of Justice" In European
Journal of Philosophy. vol. 3. No. 1, pp.21-35.
Gosepath, S. 2001: “Democracy out of Reason. Comment on Rainer Forst’s “The rule of Reasons””. In
Ratio Juris, vol. 14.no 4. pp. 379-89.
Gril, P. 1998: Die Möglichkeit praktischer Erkenntnis aus der Sicht der Diskursteorie. Berlin: Duncker and
Günther, K. 1989: "A Normative Conception of Coherence for a Discursive Theory of
Legal Justification". In Ratio Juris 2:155%166.
Günther, K. 1993a: The Sense of Appropriateness: Application Discourses in Morality and Law.
Albany: State University of New York Press.
Günther, K. 1993b: ‘Criticial Remarks on Robert Alexy’s “Special-Case Thesis”’ Ratio Juris. Vol. 6.
No. 2 July, 143-56)
Günther, K. 1999: “Welchen Personenbegriff braucht die Diskurstheorie des Rechts? ” in H.
Brunkhorst and P. Niesen (eds.) Das Recht der Republik. Frankfurt: Suhrkamp.
Gutmann A. and D. Thompson 1996: Democracy and Disagreement. The Belknap Press of
Harvard University Press, Cambridge, Mass. 1996.
Gutman, A. 1999: “Deliberative Democracy and Majority Rule”. I H.H. Koh and R.C. Slye
(red.): Deliberative Democracy and Human Rights. New Haven: Yale University Press.
Habermas, J. 1983/1990: “Dicsourse Ethics: Notes on a Program of Philosophical Justification”, in
Moral Consciousness and Communicative Action. Cambridge, Mass.: The MIT Press.
Habermas, J. 1996a: Between Facts and Norms. Cambridge, Mass.: The MIT Press.
Habermas, J 1996b: “Reply to Symposium Participants, Benjamin N. Cardozo School of Law”, In
Cardozo Law Review 17:1477-1558.
Habermas, J. 1998a: “Reconciliation through the Public Use of Reason.” In The Inclusion of the
Other: Studies in Political Theory. Cambridge, Mass.: The MIT Press.
Habermas, J. 1998b: “”Reasonable” versus “True”, or the Morality of Worldviews". In The Inclusion
of the Other: Studies in Political Theory. Cambridge, Mass.: The MIT Press.
Habermas, J. 1998c: “On the Internal Relation between the Rule of Law and Democracy.” In J.
Habermas: The Inclusion of the Other: Studies in Political Theory. Cambridge, Mass.: MIT
Habermas, J. 1999: "A short Reply“. In Ratio Juris 12. pp.447-8
Hart, H.L.A 1961/1997: The Concept of Law. Oxford: Clarendon Press
Honig, B. 2001: “Dead Rights, Live Futures. A reply to Habermas’s “Constitutional Democracy””. In
Political Theory, vol. 29 No.6 pp792-805
Höffe, O. 1996: Vernunft und Recht. Bausteine zu einem interkulturellen Rechtsdiskurs.
Kant, I. 1785/1970: "Grunnlegging til moralens metafysikk." In I. Kant: Morallov og frihet.
Oslo: Gyldendal Norsk Forlag.
Laband, P. 1911: Das Staatsrecht des Deutschen Reiches, vol. 2. Tübingen: Mohr
La Torre, M. 2002: "Theories of Legal Argumentation and Concepts of Law. An Approximation”. In
Ratio Juris Vol. 15 No. 4, pp.377-402
Luhmann, N. 1995: Das Recht der Gesellschaft. Frankfurt: Suhrkamp.
Maus, I. 1989: “Die Trennung von Recht und Moral als Begrenzung des Rechts.” In Rechtstheorie 20,
Maus, I. 1996: “Liberties and Popular Sovereignty: On Jürgen Habermas’s Reconstruction of the
System of Rights”, Cardozo Law Review 17:825-882.
Menéndez, A.J. 2001: Justifying Taxes. Dordrecht: Kluwer Academic Press
Michelman, F.I. 1997: „How can the People ever make the Laws? A Critique of Deliberative
Democracy.“ I J. Bohman og W. Rehg (red.): Deliberative Democracy. Cambridge, Mass.: The MIT
Pavlakos, G. 1998: “The Special Case Thesis. An Assessment of R. Alexy’s Discursive Theory of
Law”. In Ratio Juris. Vol. 11.no. pp.126-54
Pavlakos, G. 1999: “Persons and Norms: On the normative groundwork of Discourse-Ethics.” In
Archiv für Rechts- und Sozialphilosophie, pp..-22
Pröpper, I.M.A.M. 1993: “Argumentation and Power in Evaluation-Research and in its Utilization in
the Policy Making Process.” In R. von Schomberg (Ed.): Science, Politics and Morality.
Scientific Uncertainty and Decision Making. Dordrecht: Kluwer Academic Publishers.
Rawls, J. 1971: A Theory of Justice. Oxford: Oxford University Press.
Rawls, J. 1993: Political Liberalism. New York: Colombia University Press.
Rawls, J. 1995: ”Reply to Habermas”, In The Journal of Philosophy 92(3):132-180.
Rawls, J. 1997a: “The Idea of Public Reason Revisited”. In The University of Chicago Law Review.
Vol. 64, no 3. pp. 765-807
Rawls, J. 1997b: “The Idea of Public Reason”. In J. Bohman and W. Rehg (eds.): Deliberative
Democracy. Cambridge, Mass.: The MIT Press.
Rawls, J. 1999a: A Theory of Justice. Revised Edition. Cambridge, Mass.: Harvard University Press
Rawls, J. 1999b: The Law of Peoples. Cambridge: Harvard University Press
Tyler, T. 1990a: Why People Obey the Law. New Haven: Yale University Press.
Wellmer, A. 1986: Ethik und Dialog. Frankfurt: Suhrkamp
Waldron, J. 2001: Law and Disagreement. Oxford: Oxford University Press
Weithman, P. J. 1995: “Contractualist Liberalism and Deliberative Democracy.” In
Philosophy & Public Affairs 24:314-43.
Wittgenstein, L. 1978: Philosophical Investigations. Oxford: Basil Blackwell.