COHEN WEBGALLEYS1.DOC 08/04/00 12:10 PM
1589
BEYOND POLITICAL THEOLOGY: COMMENT
ON KALYVAS ON CARL SCHMITT
Jean L. Cohen*
Kalyvas’s paper does us a real service because it draws its
material from Schmitt’s most important piece of work in the Weimar
period: the Verfassungslehre, which, scandalously, remains
untranslated in English. Kalyvas is thus able to set the record
straight by correcting some rather unsophisticated interpretations of
Schmitt’s work. He also aims to demonstrate the relevance of
Schmitt to contemporary discussions of democratic theory and to
theories of democratic constitutionalism. In this, Kalyvas succeeds
admirably. He shows that although Schmitt was not a democrat in
our sense of the word, his thought nevertheless makes important
contributions to democratic theory by revealing some of the major
tensions and paradoxes that modern democracy faces. But Kalyvas
wants to do even more—namely, to distill from Schmitt’s work
elements for the reconstruction of what he calls a substantive model
of radical democracy and of democratic legitimacy. That is where
we begin to part company.
I agree with Kalyvas halfway. I think that democrats must read
Schmitt and the best interpretations of Schmitt, of which Kalyvas’s is
certainly one. It is true that one must confront the issues Schmitt
raises—especially the paradoxes he analyzes in modern democratic
and constitutional theory and practice. However, I do not believe
that one can adequately address, much less resolve, these issues
within Schmitt’s theoretical framework. Indeed, I would argue that
the way in which Schmitt frames the tensions between form and
substance, legitimacy and legality, instituting and instituted, reason
and will, makes them totally antinomic and irresolvable. I assert that
this is not necessary, but due to some of Schmitt’s mistaken
assumptions, which we must abandon and replace in order to
productively address the problems he identified. Tensions will
always remain between the poles just mentioned, but we are not
* Professor of Political Science, Columbia University. Professor Cohen is the author
of CLASS AND CIVIL SOCIETY: THE LIMITS OF MARXIAN CRITICAL THEORY (1982); a
co-author, with Andrew Arato, of CIVIL SOCIETY AND POLITICAL THEORY (1992); and
the author of SEX LAW AND THE CONSTITUTION: DILEMMAS OF REGULATING
INTIMACY (forthcoming 2000).
COHEN WEBGALLEYS1.DOC 08/04/00 12:10 PM
1590 CARDOZO LAW REVIEW [Vol. 21:1589
faced with dichotomous choices in the way Schmitt assumed.
Kalyvas identifies three different levels of democracy in
Schmitt’s analysis—the founding, the normal institutional, and the
extra-institutional1—an important contribution to democratic theory
with which I agree. He is also correct in noting that Schmitt’s
general approach is governed by his analysis of the implications of
the transfer of sovereignty from the king to the people under
conditions of mass democracy. The key issues raised by this shift,
which are certainly still with us today, do indeed concern the relation
between formal and substantive democracy, between inter alia
legitimacy and legality, between instituting/and the instituted,
between rights and democracy, between the constituant and the
constitué, between the constitution as higher law (ratio) and the
constitution as the expression of the will of the people (voluntas),
and between democracy, identity, and representation.
I think the greatest contribution of Kalyvas’s paper is the
interpretation of Schmitt’s concept of sovereignty in terms of the
pouvoir constituant, and the founding of new orders, vis-à-vis the
first moment of democracy.2 The reading of Schmitt’s theory of
popular sovereignty and of the sovereign decision along these
lines—i.e., in relation to the constituting and founding moment of
new powers and new political forms, instead of the usual focus on
the sovereign as the decider of the “exception”— is very
important. Kalyvas is correct in claiming that Schmitt transcended
the traditional theory of the sovereign, construed as an unlimited
exercise of absolute power, as the instance of command, above the
law.3 He does indeed break with this model and with the approach
of Rousseau, the first republican political theorist to use it.
The sovereign as the pouvoir constituant—as the locus of the
genesis of power, new constitutions, and new political forms—is
quite distinct from the sovereign as the one who exercises power
with no limits. Moreover, the theory that the popular sovereign is
the sole legitimate source of the constitution and of the shape given
to political life does seem to be the heart of any theory of democratic
legitimacy.
However, I will argue that Schmitt does not really transcend the
more basic assumptions undergirding the traditional and the
Rousseauian concept of sovereignty, and that many of the dilemmas
he constructs which plague his own theory are traceable back to
these assumptions. There are deep-seated theoretical reasons why
1 Andreas Kalyvas, Carl Schmitt and the Three Moments of Democracy, 21 CARDOZO
L. REV. 1525 (2000).
2 Id. at 1532.
3 Id. at 1533.
COHEN WEBGALLEYS1.DOC 08/04/00 12:10 PM
2000] BEYOND POLITICAL THEOLOGY 1591
Schmitt’s approach leads to antidemocratic, illiberal outcomes,
despite the existence of the rudiments of a theory of democratic
legitimacy in his work, and despite his embrace of constitutionalism.
Unless we challenge his core assumptions—which continue to
inform many contemporary theorists of “radical democracy”— we
will be doomed to follow where we do not want to go.
In particular, three key assumptions have to be challenged:
(l) The first assumption is Schmitt’s purely substantive
definition of democracy as equality among a restricted circle of
equals involving a series of identifications (between governed and
governors, sovereign and subject, ruler and ruled, etc.). This equality
is understood in substantive terms as homogeneity or sameness
along some key dimension. What the definition omits from the
concept is the discursive, deliberative, and procedural dimensions.4
While a purely procedural concept of democracy involves selfmisunderstanding,
insofar as there are always, as Kalyvas stresses,
substantive value commitments undergirding any institutionalization
of democracy and every procedure, a purely substantive
understanding of democracy is even more flawed and dangerous.
One can always point to the substantive values undergirding
procedures to show that procedures are not neutral; but this does not
mean that such procedures cannot be impartial or that they are
extrinsic to democracy. Only if democracy and democratic
legitimacy are construed without the moment of reasoned
deliberation would procedures for ensuring the fairness of the
process of deliberation be irrelevant to democracy. But one cannot,
in my view, define deliberation and procedure or, for that matter,
institutionalized public spheres (civil and political), out of the
concept of democracy, however central the ideas of identity and
equality are. The choice is not substance or process, identity or
talking, democracy or liberalism. We can never, in a democratic
polity, get out of the spiral of procedure/substance/procedure.
Rather, one must see that the heart of democracy is the possibility of
shifting to the level of deliberation about substantive values, to
procedures, and back again, within procedural forms. Two
“procedures”—or rather, processes—matter most: one involving the
circulation of power (elections and voting), the other involving the
circulation of opinion (uncensored civil publics and receptivity to
their influence on the part of decisional political instances).
Representative government in liberal democracies requires
autonomous elected representatives free to deliberate about policy
4 See CARL SCHMITT, THE CRISIS OF PARLIAMENTARY DEMOCRACY (Ellen Kennedy
trans., 1985) (attributing deliberation and proceduralism to liberalism, and restricting the
concept of democracy to a series of identities or identifications).
COHEN WEBGALLEYS1.DOC 08/04/00 12:10 PM
1592 CARDOZO LAW REVIEW [Vol. 21:1589
and autonomous (uncensored) public discussion able to generate
public opinions on key issues and able to influence representatives.
Institutionalized public spaces which allow deliberation to occur
among a plurality of views, perspectives, and opinions, which permit
reflection on substance and process, and which promote statutes
ensuring responsiveness and receptivity to public opinion by
decision-makers, are also central to democracy.
(2) The second, related problem with Schmitt’s theoretical
framework is that he operates with the basic elements of subject
philosophy—the most fateful dimension of which is the analysis of
democracy, legitimacy, and sovereignty in terms of the concept of
“the will” and the metaphor of the body politic. Hannah Arendt
criticized him for this long ago.
In short, because Schmitt remained caught in the discursive
field of subject philosophy, he did not cut off the head of the king by
transferring sovereignty, understood as the pouvoir constituant, to
the people. Instead, with his discourse of the sovereign people as a
collective subject with a single will, he did indeed fall prey to what
Lefort (and Schmitt himself) called the totalitarian logic of unity.
His work, and his understanding of the political itself, is a Herculean
effort to shape the fragmented masses, brought into the polity via
democratization, into a unified, homogeneous, unitary, self-identical,
and identified collective subject—the “friend” component of his
political theory. Sovereignty works through installing clear
boundaries between “us” and “them,” and representation involves,
for Schmitt, a logic of incarnation. The whole conception is a form
of political theology whereby the boundaries, unity, and identity of
“the people” are ensured by the sovereign and the discourse of
sovereignty.
However, once the popular sovereign is construed as a
collective subject with a collective will, it becomes impossible, as
Schmitt himself pointed out, to avoid or resolve the ancient dialectic
in the theory of the will of the people that plagues democratic
theory—namely, that the minority might better express the true will
of the collective sovereign than the majority, because the people can
be deceived. That is why Schmitt says that everything depends on
how the will of the people is formed. Accordingly, the only practical
questions are who shapes the will of the people, who has the control
over the means with which the will of the people is constructed, and
who can claim to embody or to be identical with it? This flawed
theory of representation as a form of incarnation leads Schmitt to
construe the will of the people as that which is embodied by the
successful political actor (the sovereign representative) who manages
to get the people to “identify” (their will and their very identity as a
COHEN WEBGALLEYS1.DOC 08/04/00 12:10 PM
2000] BEYOND POLITICAL THEOLOGY 1593
people) with him. Thus, neither the form of democratic decisionmaking
nor the presence of deliberative civil publics matters once
past the age (nineteenth century) of parliamentarism. The only
thing that matters is the ability of a political leader or instance to get
the people to identify their will with his, be it via plebiscites,
referenda, or other forms of acclaim. Representation, in Schmitt’s
hands, thus becomes collapsed to the logic of identification and of
the production of identity: it becomes absolute representation,
whereby the gap between the civil and the public opinion, on one
side, and political representatives, on the other, is abolished. In his
model it is again possible to claim, “I am the pouvoir constituant, I
embody the will of the people.” But this, in my view, spells the end
of democracy, not its modern form.
That is also why, in Schmitt’s approach, dictatorship is not
antithetical to democracy and why every political power, whatever
its form, can hope to achieve this identification and claim to be
“democratic.” We will remain caught in this trap unless democratic
legitimacy is tied to communication rather than will—to discourse,
deliberation, publics, plurality, heterogeneity, and their influence in
addition to numbers and vote-counting (fifty-one percent); unless
procedure and institutions are brought back into the concept of
democracy. The subject of the sovereignty has to be
disincorporated. The metaphor of the people as a subject with one
will that is represented, one identity that is embodied in a sovereign,
is both theological and dangerous. The figure of the enemy is
intrinsic and central to Schmitt’s understanding of politics, given the
centrality of will and the focus on identity and homogeneity. As
Arendt pointed out, only from the standpoint of having to fight a
common enemy, understood existentially, can the people appear as a
unity. Political theology, as Lefort argues, locates knowledge, right,
and power in the same place: in the “head” of the body politic,
(re-)incarnated in the sovereign. This metaphor of the body politic,
of the people as one, of the sovereign representative, identical to the
people’s will, is deeply antidemocratic, in my view.
(3)The third misleading assumption that we must question is the
notion that representation and democracy, or representation and
popular sovereignty, are antithetical—another core Rousseauian
idea which Schmitt never abandoned. Schmitt construes
representation as an exclusively aristocratic principle in political life,
which entails a substitution of the active will of a political elite for
the will of the people in decision-making and law-making. And
indeed, will cannot be represented, since it is the very moment of
decision. Either our will or someone else’s will decide.
For Rousseau this meant that democracy, also on the second
COHEN WEBGALLEYS1.DOC 08/04/00 12:10 PM
1594 CARDOZO LAW REVIEW [Vol. 21:1589
level of ordinary legislation, must be direct democracy; thus,
representative democracy is an oxymoron. But as Kalyvas points
out, Schmitt knew (as did Rousseau) that pure absolute direct
democracy is impossible.5 He also knew that representation is
unavoidable in a modern polity. Thus, the deep message is that
democracy, in the sense of self-government by the popular sovereign
who literally decides and legislates, is impossible.
Instead, following Schmitt, we must enter into the dynamics of
democratic identification with those who really decide, while
providing for a figure (a president perhaps) to re-present the unity of
the whole. We should remember that the representation that is
unavoidable for Schmitt is darstellung—the symbolic and
performative representation of unity—not vertretung, or being
someone’s delegate. Schmitt argues that there once was a form of
government that appeared to reconcile liberalism, representation,
and democracy—nineteenth-century parliamentarism—but this has
entered into crisis and become an anachronism, in part due to
democratization and the demise of virtual representation. Once
again, we are led to plebiscitary democratic forms as the only ones
possible today. The plebiscite generates identification with some
project articulated by the instance representing unity and
sovereignty.
Here, too, Schmitt’s theoretical assumptions have to be
challenged. If one does not recognize the democratic and
egalitarian in addition to the aristocratic dimensions of
representation—if one does not see that there is more than one
conceivable form of representative government to be had, and that
parliamentarism of the nineteenth century was only one of them—
then there is once again no answer to Schmitt’s link between
democracy and dictatorship, or to his reduction of democratic
assemblies to the passive role of acclaim.
It is this set of assumptions that leads one to frame the tensions
always present between liberalism and democracy, legitimacy and
legality, substance and process, form and content, and reason and
will, as a set of irresolvable antinomies among which we are forced
to choose. I submit that these choices need not be ours.
I close this comment with several questions to Andreas Kalyvas:
(l) Does Kalyvas accept the claim that the sovereign founding
moment is a decision, an expression of the will of the people
emanating from nothingness and created ex nihilo? To say that a
radical founding, creating something truly new, is not determined by
a prior legal normativity is one thing; to say it occurs ex nihilo is
5 See Kalyvas, supra note 1, at 1554.
COHEN WEBGALLEYS1.DOC 08/04/00 12:10 PM
2000] BEYOND POLITICAL THEOLOGY 1595
quite another. Is there no prior tradition, social bond, experience—
no element of legal continuity and identity—that is drawn upon in
acts of founding? Are there not traces of the past and some
continuities in every new constitution? Is there no way to solve
paradoxes of new beginnings other than through an existential
decision or with the mere exercise of power? Shouldn’t founders be
constrained to act “as if” there is continuity, as Arato has argued?
(2) There is another problem of infinite regress that Rousseau
at least addressed regarding foundings: that of the necessity for a
people able to create a democratic political form of life prior to
having experienced one—the effect being taken for the cause.
Rousseau solved the problem in the classic way, by invoking the
foreigner, the legislator. Schmitt, of course, had constitutional
conventions in mind in place of the figure of the legislator. But on
his radical democratic substantive conception of foundings the
question remains: Who decides who is a member of the people that
legislates for itself and what the friend component of identity
consists of? How can “we” democratically decide who “we are” and
who belongs to the circle of equals—can this be an existential
decision ex nihilo?
(3) Finally, what does Kalyvas mean by substantive radical
democracy? How can the identity of the constituent subject be the
normative standard by which constitutions and institutional
arrangements should be judged? Without institutions, specified
procedures and structures, without publics—some institutionalized
and some not—without constituted bodies and procedures even
for the pouvoir constituant—as Arendt well knew—the reference
to the actual manifestation of the will of the popular constituent
subject and the direct participation of the people in the process of
founding a constitution will only be “democratic” in the
plebiscitary sense. This, to a republican, liberal, democratic
constitutionalist like myself, will not do.