of 26
8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
1/26
http://ptx.sagepub.com
Political Theory
DOI: 10.1177/0090591704268372
2005; 33; 218Political TheoryJohn S. Dryzek
Deliberative Democracy in Divided Societies: Alternatives to Agonism and Analgesia
http://ptx.sagepub.com/cgi/content/abstract/33/2/218The online version of this article can be found at:
Published by:
http://www.sagepublications.com
can be found at:Political TheoryAdditional services and information for
http://ptx.sagepub.com/cgi/alertsEmail Alerts:
http://ptx.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/cgi/alertshttp://ptx.sagepub.com/cgi/alertshttp://ptx.sagepub.com/subscriptionshttp://ptx.sagepub.com/subscriptionshttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsPermissions.navhttp://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsReprints.navhttp://ptx.sagepub.com/subscriptionshttp://ptx.sagepub.com/cgi/alerts8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
2/26
10.1177/0090591704268372ARTICLEPOLITICALTHEORY/April2005Dryzek/DELIBERATIVEDEMOCRACY
DELIBERATIVE DEMOCRACY
IN DIVIDED SOCIETIES
Alternatives to Agonism and Analgesia
JOHN S. DRYZEK
Australian National University
For contemporary democratic theorists, democracy is largely a matter of deliberation. But the
recent rise of deliberative democracy (in practice as well as theory) coincided with ever moreprominent identity politics, sometimes in murderous form in deeply divided societies. This essay
considershowdeliberative democracycan process the toughest issuesconcerning mutually con-
tradictory assertions of identity. After considering the alternative answers provided by agonists
andconsociationaldemocrats,the author makes thecase fora power-sharing state withattenu-
atedsovereigntyand a more engageddeliberativepolitics in a public sphere thatis semidetached
from the state and situated transnationally.
Keywords: deliberative democracy; consociational democracy; agonism; identity politics;
ethnic conflict
I. DEMOCRACY CONFRONTS IDENTITY CONFLICTS
Democracyis todaya near-universal validating principle for political sys-
tems.And according to contemporary democratic theorists, at least since the
early 1990s, democracy is largely, though not exclusively, a matter of delib-
eration. Democratic practice too has witnessed a range of deliberative
innovations.
The same decade that saw the rise and rise of deliberative democracyalso
sawidentitypolitics prominent, sometimes in murderous form.Identitypoli-
218
AUTHORSNOTE: Previousversions of this article were presented to the Conference on Delib-
erative Democracyand Sensitive Issues at the University Of Amsterdam,March25-26,2003,the
Social and Political Theory Program at Australian National University, and the Department of
Political Science at the University of North Carolina. For comments I thank Tjitske Akkerman,
John Forester, Bora Kanra, Ilan Kapoor, StephenLeonard, Christian List, Gerry Mackie, Claus
Offe, Benjamin Reilly, Mark Warren, and Iris Young.POLITICAL THEORY, Vol. 33 No. 2, April 2005 218-242
DOI: 10.1177/0090591704268372
2005 Sage Publications
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
3/26
tics, including the murderous variety, are hardly new. But as the cold war
world order fell apart, the political gap was often filled by assertions and
denials of identity. Religious fundamentalisms showed renewed vigor, in
opposition toeach other aswell as what BenjaminBarbercallsMcWorld.1
I consider how deliberative democracy can process what are arguably the
toughest kinds of political issues, the mutually contradictory assertions of
identity that define a divided society. The assertions in question might
involve nationalism (Republicans and Unionists in Northern Ireland; any
number of separatist movements), combinations of religious and ethnic con-
flicts (Palestinians versus Israelis), and religious versus secular forces
(Islamic fundamentalism against Western liberalism on the global stage;
Islamistsversus secularists in Turkeyand Algeria; Christianfundamentalists
versus liberalism in theUnitedStates). Thebasic problem in allthesecases isthat one identity can only be validated or, worse, constituted by suppression
of another. Radical Islamists cannot live in or with a McWorld. A state that
was no longer a Jewish state forged in struggle would be anathema to many
Israelis. Christian fundamentalists regard the political presence of gays and
lesbians not justas an irritantbut asa standingaffront towho theyare. A mul-
tinational society is not just a policy opposed by militant Serb nationalists, it
is a perceived attack on their core political being.
Deliberationacross divided identities is hard. On a widelyshared account,
deliberation is what Bessette calls themild voiceof reason2exactlywhat
is lacking in tough identity issues, at best an aspiration for how opponents
might one day learn to interact once their real differences are dissolved.
Deliberative democrats influenced by Rawls might follow him in excluding
the background culture from the purview of public reason. But, asBenhabib points out, issues generated by the background culture and its
comprehensive doctrines can be especially pressing.3 Gutmann and
Thompson believe that deliberation can be extended to deep moral disagree-
ments, but the precondition is commitment on all sides to reciprocity, the
capacity to seek fair terms of cooperation for its own sake, such that argu-
ments are made in terms the other side(s) can accept.4 Again, mutual accep-
tance of reasonableness is exactly what is lacking in divided societies.
Gutmann and Thompson require adoption by all sides of a particular
moral psychologyopenness to persuasion by critical argumentthat is in
fact not widely held, and explicitly rejected by (say) fundamentalist Chris-
tians.5 Moreover, they apply the reasonableness standard to the content of
contributions to debate, not just the motivation of speakers. Thus they are
vulnerable to criticism from difference democrats such as Young, who
accepts reasonableness as a norm for motivation but not for the content of
statements, because that involves suppression of alternative communicative
Dryzek / DELIBERATIVE DEMOCRACY 219
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
4/26
forms.6 More radical difference democrats and agonists see deliberation in
terms of theerasure of identity, a form of communication stuck in neutral that
does not recognize difference, partial in practice to well-educated white
males,especially when it prizes theunitary publicreason advanced by Rawls
and his followers. Those asserting identities for their part may feel insulted
by the very idea that questions going to their core be deliberated. What they
want is instead cathartic communication that unifies the group and
demands respect from others.7
I argue for a discursive democracy that can handle deep differences. The
key involves partiallydecouplingthe deliberative and decisionalmoments of
democracy, locating deliberation in engagement of discourses in the public
sphere at a distance from the sovereign state. I approach this argument by
examining two very different responses to divided societies. The first isagonistic, seeking robust exchange across identities. The recent history of
agonism owes much to Hannah Arendt, William Connolly, and Bonnie
Honig,8 but I focus on the work of Chantal Mouffe, because she explicitly
advocates agonism against deliberative democracy in plural societies. The
second response is consociational, seeking suppression of interchange
through agreement among well-meaning elites. I do not treat these two as
straw man extremes between which a moderate path should be sought.
Indeed, I argue that a defensible discursive democracy for divided societies
can develop elements of both.
II. AGONISM
The agonistic charge is that deliberative democracy is incapable of pro-
cessing deep difference. Mouffe argues that themain task fordemocracy is to
convert antagonism into agonism, enemies into adversaries, fighting into
critical engagement.9 Deep difference is accompanied by passions that, she
believes,cannot be resolvedby deliberation,committedas it is to rationalistic
denialof passion andthepursuit of consensus that inpractice both masks and
serves power.Her alternative is agonisticpluralism involvinga vibrant clash
of democratic politicalpositions.10 The prime task of democratic politics is
not to eliminate the passions, . . . but to mobilise these passions towards the
promotion of democratic designs.11 Acceptance of the legitimacy of the
positions of others comes not through being persuaded by argument, but
through openness to conversion as a result of a particular kind of democratic
attitude.12 The outcome is not agreement but rather relationships that com-
bine continued contestationwith deep respect for theadversaryindeed it is
220 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
5/26
noteasy to speak in terms of outcomes. Mouffe (likeGutmann andThomp-
son) is vulnerable to questions about where exactly the required attitude
should come from, especially where groups asserting identity themselves
feature hierarchy and repression.13
While accepting Mouffes identification of the need to transform antago-
nism into, if not agonism, at least more civilized engagement as the primary
task for democracy in divided societies, I differ from her on three grounds.
The first is in the content of critical interchange. Mouffe wants this inter-
change to be energized by core identities, otherwise passion is missing. Yet,
paradoxically, identities for Mouffe have to be fluid to theextent of enabling
thorough conversion inonegroups attitude to another. But if identities them-
selves are highlighted, exchange is more likely to freeze identities than con-
vert them. As Forester points out, being respectful of others is one thing;accepting at face value claims that preferences and interests are in fact basic
values is quite another, requiring a more challenging order of problem solv-
ing.14 If interchange is to move beyond confrontation and stalemate, then,
Forester argues, the focus should be on the specific needs of the parties, not
on the articulation and scrutiny of general value systems. His example con-
cerns gay activists and fundamentalist Christians meeting over HIV/AIDS
care in Colorado. The last thing that needs to be done is to reinforce mutually
hostile identities; forexample,by debating whether it is legitimate to treat the
HIV/AIDS issue in the moral terms favored by the Christians, as opposed to
the public health terms favored by the gays. But if individuals can listen to
each othersstories, they might at least acceptoneanothers specific needs
which can be reconciled, even when value systems and identities cannot.
This is a kind of reciprocal recognition, but not the kind of vibrant exchangeof passions proposed by Mouffe.
A second departurefromMouffe involves thewaydeliberative interaction
is conceptualized. Mouffe may be right that deliberation in the image of a
philosophy seminardispassionate and reasonedcannot handle deep dif-
ference. However, it is possible to formulatean account of discursivedemoc-
racy that is more contestatory than this image, so more robust in the face of
deep difference.
Third, Mouffes interpretation of themain task of democracy has no obvi-
ous place for collective decision making and resolution of social problems.
Shescornsconsensus as a coverfor power, but at least consensus implies that
decisions can get made. When agonistic pluralism does attend to collective
decisions, it is only to point to the need for them to be open to further
contestation. I explore a way to combine critical engagement and collective
decision, but this requires a differentiation of the ways politics can be con-
Dryzek / DELIBERATIVE DEMOCRACY 221
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
6/26
ducted in different sites. While Mouffe emphasizes the variety of sites (cul-
ture, workplace, home, school, etc.), for her the content of politics is
undifferentiated, everywhere agonistic.
III. ANALGESIA: CONSOCIATIONAL DEMOCRACY
A very different sort of criticism of deliberative democracys ability to
process divisive issues follows from arguments that they should be removed
from contentious democratic debate altogether. From this perspective,
Mouffes assertion that a well functioning democracy calls for a vibrant
clash of democratic political positions is naive,15 for vibrant clashes risk
disintegration. Such is the basis of Lijpharts claims for consociationaldemocracy, an agreement between the leaders of each bloc to share govern-
ment, involving grand coalition, segmental autonomy, proportionality, and
minority veto.16 Lijphart believes consociationalism is the only workable
type of democracy in deeply divided societies.17 Neither Lijphart nor his
sympathizers have taken on deliberative democracy in these termsillustra-
tion of the chasm between democratic theorists and students of real-world
democratic development. But it is not hard to deduce what they ought to say
about deliberative politics.
Lijphart points to success stories where a consociational approach has
defused religious and/or ethnic conflicts, such as his own Netherlands, Aus-
tria, Switzerland, Malaysia,South Africa in the1994 to 1996 transition from
apartheid, andIndia. (Few of thesecases actuallymeet allfour of hisdefining
criteria for consociationalism.) But conflict resolution is achieved at theexpense of several dimensions prized by democratic theorists, including the
deliberative dimension. Elections have little meaning, as the same set of
leaders willgovern irrespective of the result. Moreover, contentious delibera-
tion occurs only between the leaders of different blocs, and even then mostly
in secret (for fear of inflaming publics), ruling out much of a role for parlia-
mentary debate. The political communication of ordinary people is shep-
herdedinto within-blocchannels where it cando littledamage. This channel-
ing obstructs any kind of deliberative still less agonistic interaction across
differentblocs below theelite level,because segmental autonomyis basic.
Consociation precludes any role that public deliberation construed as
social learning might play in reconciliation in divided societies. But as
Kaufman points out, ethnic hatreds are the product of symbolic politics in
particular political circumstances.18 As such, they are learned, and so can be
unlearned or transformed,19 though that can be an uphill task in the face of
222 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
7/26
persistent negative understandingsandmyths. In this light, by freezingcleav-
ages, a consociational regime may actually reinforce or, worse, create the
kind of conflict it is designed to solve.20 A deliberative democrat would hope
that reflection stimulated by interaction could contribute to less vicious sym-
bolic politics,not tied to myths of victimhood and destiny. Segmental auton-
omy precludes such a politics, because deliberation confined within seg-
ments succumbs to Sunsteins law of group polarization.21 Debate leads
only to the group positions becoming more extreme, as individuals get their
prejudices confirmed in talk with like-minded others.
IV. TOWARD A DELIBERATIVE RESPONSE
Agonists believe deliberative democracy cannot deal with divisive issues
because it is too constraining in the kind of communication it allows.
Consociationalists believe deliberative democracy cannot deal with divisive
issues because it is too open to diverse claims and claimants. Deliberative
democracy can be defended against both sides, but it has to take them seri-
ously, and beprepared to takeelementsfromeach. On the faceof itthisought
to be impossible, given their diametric opposition. The key is a differentia-
tion of political sites within a society that agonists and consociationalists
alikehave not contemplated: theformerbecause they address only politics in
the abstract rather than its institutional specifics, the latter because they see
only a politics tightly attached to the state. Deliberative democracy can pro-
cess contentious issues in a politics of engagement in thepublicsphere, even
if it has problems doing so when it comes to deliberation within theinstitutions of the state.
In this light, a conception of discursive democracy in terms of a public
sphere that is home to constellations of discourses canbe brought to bear.22 A
discourse can be understood as a shared way of making sense of the world
embedded in language.Thus any given discoursewill be defined by assump-
tions, judgments, contentions, dispositions, and capabilities. These shared
terms enable subscribers to a given discourse to recognize and convert sen-
sory inputs into coherent accounts of situations. These accounts can then be
shared in intersubjectively meaningful fashion. Thus discourses feature
storylines, involving opinions about facts and values. Familiar examples of
such discourses include market liberalism (dominant in global economic
affairs) and sustainable development (ubiquitous in environmental affairs).
Thecontent of collective decisions depends strongly (but notexclusively) on
the relative weight of competing discourses in a domain. For example, the
Dryzek / DELIBERATIVE DEMOCRACY 223
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
8/26
content of criminal justice policy varies with the weight of discourses stress-
ing, respectively, thepsychopathologyof the criminalmind,rationalcalcula-
tion of thecosts andbenefitsof criminalacts by perpetrators, and thecircum-
stances of poverty that lead individuals to a life of crime. The engagement of
discourses anditsprovisionaloutcomes aredemocratic to thedegree they are
under dispersed influence of competent actors, as opposed to manipulation
by propagandists, spin doctors, and corporate advertisers. The possibility of
contestation and engagement means discourses have to be treated as less
totalizing and constraining than some followers of Michel Foucault claim.
Discourses must be amenable to reflection, if only at themargins. The requi-
site communication is deliberation not agonism because it is oriented to
persuasion rather than conversion, and it retains some connection (however
loose) to collective decision.Some recent treatments of deliberative democracy do, then, meet theago-
nists critique.23 Agonists see deliberation as deadening and biased in the
kind of communication it allows. But the engagement of discourses can
accommodate many kinds of communication beyond reasoned argument,
including rhetoric, testimony, performance, gossip, and jokes. However,
three tests must be applied to secure the intersubjective understanding prized
by deliberative democrats. Once we move beyond ritualisticopenings, com-
munication is required to be first, capable of inducing reflection; second,
noncoercive; and third, capable of linking the particular experience of an
individual or group with some more general point or principle.24 The last of
these three criteria is crucial when it comes to identity politics gone bad. A
harrowing story of (say) rape and murder in a Bosnian village can be told in
terms of guilt of one ethnic group and violated innocence of anotherfuelforrevenge. Butthe storycanalsobe told in termsof violationof basicprinci-
ples of humanity that apply to all ethnicities, making reconciliation at least
conceivable (though not easy).
How can this discursive approach be applied to divided societies? To
begin, taking identities seriously means allowing different communicative
forms that can accompany particular identities; this is Youngs connection.
However, this recognition often helps little when it comes to deeply divided
societies, because, as Moore points out, societies deeply divided in identity
are often not divided at all in culture.25 Culturally, there are few differences
between Catholics and Protestants in Northern Ireland, and between Serbs,
Croats, and the worlds most secular Muslim community in former Yugosla-
via. It is, then, a mistake to treat identity conflictsas merelya matterof multi-
culturalism. This treatment of identity in terms of culture extends even to
Benhabibs defense of universalist deliberative democracy against cultural
224 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
9/26
relativism.26 She accepts that culture has become a ubiquitous synonym for
identity, an identity marker and differentiator,27 even as she pleads for rec-
ognition of theradical hybridityandpolyvocalityof allcultures28 thatfacili-
tates deliberation both within and across groups.29
Identities are bound up with discourses. It is in this sense that nations are
in Benedict Andersons terms imaginedcommunities,30 the product of dis-
courses, not genes, not culture. But how can reflective engagement across
discourses move beyond the vain hopes of agonists when identities are only
asserted dogmatically and so relativistically, fueled by existential resent-
ments?31
Engagement is less likely to end in hostility if the focus is on specific
needs (e.g., security, education) rather than general values. An example
comes from Turkey, where headscarves worn by young Islamic women werelong a symbolic marker that excluded them from secular Turkish universi-
ties. Beginning in 2002, a reframing of the issue in terms of the education
needs of young women and the character of education as a basic human right
gained ground, and the issue looked less intractable. Avoidance of head-on
confrontation means the other side is less easily accused of a hidden agenda
to capture the state, and ones own side cannot so easilyclaim alone to repre-
sent the people or safeguard the polity.32
Deveaux worries that the emphasis on particular needs sheadvocates will
be rejected by deliberative democrats because of its inconsistency with uni-
versalist public reason.33 But particular needs are often amenable to expres-
sion in termsof more general principles. Even thematerialadvantageof (say)
patriarchs in a cultural group can be argued in terms of stabilityand continu-
ity. And public reason itself can be plural.34
A deeper problem in emphasizing needs is that some needs can be manip-
ulated to justify hostility. Notably, advocates of ethnic cleansing in the for-
merYugoslaviaargued that it wasnecessary to ensure thebasic need of secu-
rity, at least fortheirown side.Butsucharguments could resonateonly within
their own ethnicgroup. Demagogues canmanipulate needs-talk in a destruc-
tive direction, just as they can manipulate any other kind of talk. A focus on
needs is likely to contribute to conflict resolution only in the context of an
engaged dialogue across difference, but not when communication is
segmented within groups.
Deliberative rituals and indirect communication (as opposed to confron-
tation) also have roles to play in reconstructing relationships. Forester dem-
onstrates the importance of (say) small talk between erstwhile opponents
over a shared meal with no explicit connection to the issue at hand.35 Experi-
ments withn-person prisoners dilemmas show that even a period of irrele-
Dryzek / DELIBERATIVE DEMOCRACY 225
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
10/26
vant discussion can increase the incidence of subsequent cooperative behav-
ior.36 So even cheap talk canhelp moderateconflict,though by itselfsuch talk
is insufficient to produce the requisite engagement across discourses.
V. DEMOCRACY AND THE STATE IN DIVIDED SOCIETIES
I turn now from the what to the where of deliberation, beginning by
pointing to thedesirability of looseningtheconnection between thedelibera-
tion and decision moments of democracy in a divided society. Such loosen-
ing resists one strong current in deliberative democratic theory, which sees
the proper home for deliberation in the institutions of the sovereign state,
such as legislatures, courts, public inquiries, committees, and administrativetribunals. To see why a degree of separation is desirable, consider what hap-
pens when deliberation and decision are joined in the context of divisive
identity.
Mainly, decision overwhelms deliberationespecially when decision is
tied to sovereign authority. Since the peace of Westphalia in 1648, sover-
eignty has had an all-or-nothing character. Westphalia established the norm
of noninterference in internal affairs and the principle that the religion of the
princeis thereligionof thestate. At thetimein Europe, religionwasthe main,
almost sole, identity that mattered. Later, identity came also to involve
nationality, ethnicity, and class, but the idea of one identity per state per-
sisted. Identity issues could become intractable in the context of the politics
of the state: the game is all or nothing.
The very worst repression of competing identities has often come fromactorsstruggling to secure their hold over the state, and the states hold over
society. As Rae demonstrates, episodes ranging from expulsion and forced
conversion of Jews in fifteenth-century Spain to the Armenian genocide in
Turkey toethniccleansing in formerYugoslaviain the1990s canall beattrib-
utedto state-building elites.37 Theseelites pursue pathologicalhomogeniza-
tion to secure a mass identity to accompany and bolster the incipient state.
Contra Hobbes, it is leviathan under construction that creates murder and
misery, rather than curbing them.
Electoral democracy does not solve matters, and may exacerbate them.
The game becomes one of ensuring that the state is defined to ensure that
ones favoredidentitywillwinkeyvotes. Thisdefinition can involve drawing
physical boundaries or manipulating the electoralsystem or gerrymandering
or usingsuffragerestrictions(e.g.,measures taken tostopAfrican Americans
from voting in the American South, ranging overproperty qualifications, lit-
eracy tests, and exclusion of those with a criminal record).
226 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
11/26
Multicultural liberals have addressed what a multi-identity state might
look like,38 though they err by treating identity differences as mere cultural
differences.39 Such a state might involve devolution of authority to regions
dominatedby minority cultures, legal recognition and promotionof minority
languages, and group representation (e.g., parliamentary quotas for a partic-
ular group). These theorists are more compelling on the liberal aspects of
such a statethespecificationof rights for individuals and groups. They are
less compelling when itcomes to thedemocraticaspects.Proposalsfor group
representation are fraught with difficulty when it comes to specifying which
groupscount andhowmuch representation they have. If there is advantage in
being categorized as an oppressed minority, everyone will try to claim that
status, so raising divisions.
A better approach to electoral democracy in divided societies is theelec-toral engineering proposed by Horowitz and given flesh by Reilly.40 Reilly
shows that centripetal politics can be induced by systems of preferential
votingsuch as thealternative vote (AV), supplementary vote (SV), andsingle
transferable vote (STV). Leaders of ethnic parties have an incentive to seek
second, third, and fourth preferences of voters from theother side of the eth-
nicdivide, and so moderate theirpositions. Reillys clearestpositive example
is Papua New Guinea, which used the Australian-model AV before it fool-
ishly changed to first-past-the-post voting in 1975.
If there are not enough voters with moderate attitudes, preferential voting
will fail toassistreconciliationas shown byelectionsin Northern Ireland in
1973 and 1982 conducted under STV.41 Only in 1998, after a core group of
moderatesemerges from both sides of thecommunal divide does STVwork
better in Northern Ireland.42 This finding begs the question of exactly howmoderate attitudes can be promoted. Perhaps AV would work better than
STV in Northern Ireland, in keeping with Horowitzs claim that, though
majoritarian, it promotes moderation better precisely because it has a major-
ity threshold.43 So unlike under STV, under AV members cannot be elected
with thesupport only of a hardline minority. However, Reillyshowswith ref-
erence to the use of AV in Fiji that if the engineering is not done precisely
right, instability andviolence canstill result. Apparently technicalaspects of
theAV systemspecifiedas part of the1997 Fijianelectoral reform turnedout
to favor ethnic Indian parties, whose success led to an indigenist coup in
2000. Parallel problems have arisen with the use of SV in Sri Lanka. Preci-
sion electoralengineering is difficult in thecharged setting of a divided soci-
ety, especially once different sides realize that rules arenot neutral and so try
to influence their content. The deadly numbers game is transferred to the
meta level.
Dryzek / DELIBERATIVE DEMOCRACY 227
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
12/26
Despite these difficulties, Horowitzs and Reillys analyses are a step
beyond Schumpeterian minimalist accounts of the establishment of stable
electoral democracy. Przeworski argues that the stability of an electoral
democracyrestson losersaccepting defeatin theexpectationthat they might
be able towinina subsequent election,or because they fear theconsequences
of the breakdown of order more than those of defeat.44 Horowitz and Reilly
show exactly why such acceptance may be facilitated by voting systems that
draw the sting from defeat by inducing victors to moderation. Some kind of
preferential voting is clearly best for divided societies. My point is simply
that electoral engineering is not enough, because there is so much more to
politics than elections.
How else, then, might deliberative democrats respond to the challenge
posed by a deadly numbers game?At one level they could pin their hopes onthe civilizing force of deliberation to defuse conflict (and so provide one
essential precondition for preferential voting to work). But now the familiar
scale problem arises: deliberation, at least of the face-to-face variety con-
nectedtightly tostateauthority, canonlyeverbe forthe few. Perhaps thereare
a few representatives who might be so civilized, but in a politics of mass vot-
ing tightly connected to definition of the sovereign state, they canall tooeas-
ily be overwhelmed by demagogues. Thus in Northern Ireland, the Demo-
cratic Unionist Party and Sinn Fein still prosper at the expense of,
respectively, the more moderate Official Unionists and Social Democratic
LabourPartyeven at a time when compromise is in theNorthernIrishairas
never before, and the paramilitaries on both sides have laid down (most of)
their arms.
VI. LOCATING DELIBERATION IN THE PUBLIC SPHERE
A moreradical discursivedemocraticresponsewould askwhydemocracy
and deliberation must be joined to head counting and sovereign authority.
Consociationalists take a step in this direction on the head-counting dimen-
sion, because they suppress votings connection to collective decision. But
they do not escape the difficulties associated with construction of sovereign
authority by constitutional settlement. Consociationalism is therefore vul-
nerable to Horowitzs pessimism concerning any kind of institutional design
(including centripetal electoral systems) in divided societies: So many
forces favor the pursuit and exacerbation of conflict . . . that anything less
than a coherent package is unlikely to provide sufficient counterweight to
these forces, andyetonly partial measures that aredoomed to fall short of the
228 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
13/26
coherent package stand a real chanceof adoption most of thetime.45 Though
Horowitz recognizes no limits to the reach of this conclusion, his pessimism
actually refers only to constructionof theformal institutionsof thesovereign
state. Contemplation of the informal communicative realm might soften his
conclusion. Democratic deliberation in a public sphere at somedistancefrom
(but not completely unconnected with) the sovereign state can make a major
contribution here.
The desirability of locating deliberation in the engagement of discourses
in the public sphere in divided societies can draw on Mackies observation
that people are rarely seen to change their minds in deliberative forums.46
Even ifan individualis persuaded, itis hard for him or her toadmit it, for then
credibility is lost. Actually such changing of minds is common in what Fung
classifies as cold deliberative settingswhere participants are not parti-sans, and the forum is either unofficial or advisory.47 Citizens juries and
deliberative opinion polls exemplify this category, andit is normalto seesub-
stantial opinion shifts therein.48 In contrast, under hot deliberation, tied to
collective decision and involving partisans, participants have more strongly
formed views going into deliberation, and so cannot easily change.
Deliberation tied to sovereign authority in divided societies is about as
hot a setting as one can imagine. Most conceptions of deliberative democ-
racy require reflection and the possibility that minds can be changedin the
forum itself. This is unlikely if ones position is tied to ones identity. Locat-
ing deliberation in the engagement of discourses in the public sphere avoids
this problem because reflection is a diffuse process, taking effectover time.
With time, degree of activation of concern on particular issues can change.
Individuals can shift from partisanship to moderation to apathy and viceversa,andmayevencome toadopt differentattitudes. Nothing as dramaticas
thekind of conversion Mouffe seeks is required. This situationis less fraught
than that in hot deliberation, where reflection can take effect only in the
choices of individuals under the gaze of both opponents and those with a
shared identity. As Mackie points out, deliberation-induced reflection can
eventually lead an individual to change his or her mind. But he or she can
most easily admit that in a different setting, at another time and place, with
different participants, where face and credibility associated with having
staked out a position are no longer decisive. This consideration supports
Deveauxs specification of a revisability principle.49 A guarantee that con-
tentious issues can be revisited provides a way for those who have changed
their minds to both save face (by not admitting it for the present) and
contribute to conflict resolution (by accepting a changed position later).
Dryzek / DELIBERATIVE DEMOCRACY 229
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
14/26
VII. INSTITUTIONAL SPECIFICS
The public sphere is sometimes conceptualized as an institution-free
zone, but that is not necessarily right. Two sorts of institutions located in the
public sphere can play roles in facilitating discursive engagement: networks
and discursivedesigns, whosecontributionsI nowdiscuss. In relativelywell-
behaved political systems, the network form of organization can help estab-
lish dispersed control over the content and relative weight of discourses,
facilitating negotiation across difference. Schlosberg analyses environmen-
tal justice networks in the United States in these terms. 50 These networks
arose from a series of local actions and have no centralized leadership. They
involve individuals from very different race and class backgrounds, in some
cases from groups otherwise quite hostile to each other. Together they suc-
cessfully changed the content of public discourse on environmental affairs,
most importantly by establishing the very idea of environmental justice as a
public concern. In societies more deeply divided, the development of net-
works across divisions could be a greater challenge, given that such societies
are divided into blocs with dense within-bloc communication but little
across-bloc communication. On the other hand, even in the United States
these networks developed across groups who otherwise lived in quite
separate worlds, given the informal apartheid of American cities.
Networks are at the informal end of the institutional continuum. More
formalare the institutionsFung calls recipes forpublicspheres.51 Hehasin
mind designed forums such as citizens juries, deliberative polls, planning
cells, policy dialogues, and participatory problem-solving exercises. Any
such exercise is not in itself a public sphere, so Fungs terminology is a bitmisleading. Rather, each is a micro moment in the macro life of the pub-
lic sphere. There are many such discursive designs available. Some involve
lay citizens picked at random, some involve partisans. Some are small scale,
some try to engage larger numbers of interlinked deliberators (such as the
exercises sponsoredby theAmericaSpeaksFoundation). Some debatea spe-
cific issueor decision,othershavea broader remit.Sponsorscan include gov-
ernments, nongovernmental organizations (NGOs), academics, and founda-
tions. A few have a direct link to policy making, most lack any such direct
connection from recommendation to collective decision.
Their critics deride the general lack of direct influence on policy content,
and their sponsors often strive for such impact. But from thepoint of view of
promotingdialogue in dividedsocieties,this absenceof direct policyconnec-
tion may be positive because it provides a space for exploratory interchangeacross difference. For example, in 2001 a deliberative poll wasconducted on
the issue of relations between the indigenous and nonindigenous peoples of
230 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
15/26
Australia. Deliberating groups were made up of oversampled Aborigines
selectedin consultationwithindigenouscommunities and randomlyselected
nonindigenous others. Television coverage took the proceedings into a
broader public sphere. The results of the poll had no immediate or direct
impact on public policy, but the poll itself constituted one moment in a long
process of reconciliation across a deep divide.
VIII. BAD CIVIL SOCIETY AND ITS REMEDIES
Recognition of thecentrality of engagement across discourses in thepub-
licsphere does not mean that thestate shouldbe conceptualized as thesource
only of problems fordivided societies, andthepublicsphereonly as a benignsource of solutions. Public spheres can be segmented, the source of
interethnic conflict,52 and prone to Sunsteins law of group polarization if
individuals communicate onlywith like-mindedothers.53 Polarization canbe
exacerbated by segmented media such as right-wing talk radio in the United
States and, most notoriously, Hutu hate radio prior to the 1994 genocide in
Rwanda. (The latter was controlled by Hutu extremists associated with the
government, so not entirely a public sphere phenomenon.) The fact that sec-
tarian demagogues can flourish therein is exactly why consociationalists
seek to silence the public sphere.
Snyder and Ballentine believe that such communicative extremism is a
particular problem in societiesemerging from authoritarianism, especially if
they lack any tradition of professional journalism.54 They caution against a
liberal free-for-all in politicalcommunication,recommending bothstate reg-ulation of speech (as in Malaysia since 1969) and NGO intervention to
restrict hate speech and promote professional journalism in integrative
media.
The problem of what Chambers and Kopstein (2001) call bad civil soci-
ety is not confined to postauthoritarian societies.55 Focusing on racist hate
groupsin theUnited States, Chambers andKopstein advocategreater income
security and social justice, which would mean fewer insecure individuals to
be tempted by sectarian extremism. Chambers and Kopstein also guardedly
endorse intervention to shape group life through (for example) subsidies to
relatively benign organizations that provide services.56 They approve of the
role played by Ford, the Eurasia Foundation, and Soros in promoting benign
group life in the postcommunist world, while recognizing that such efforts
may hinder homegrown groups.
Calling thestate to therescue of badcivilsociety is problematic if thestate
itselfis the instrument ofone group ina divided society, or ifit isengaged ina
Dryzek / DELIBERATIVE DEMOCRACY 231
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
16/26
homogenization project to bolster its own support. A consociational state is
not much better if it seeks to suppress engagement in the public sphere. But
only a power-sharing state (or a majoritarian government with incentives to
appeal to minorities) is in a position to contribute to deliberationacross divi-
sion in the public sphere. The state need not be the exclusive source of solu-
tions here; NGOs and foundations can play similar parts. And there might
even be a role for political theorists when it comes to exposing the false
necessities pushed by sectarian groups.
IX. LOCATING THE PUBLIC SPHERE TRANSNATIONALLY
Engagement across division can be further promoted by transnationalaspects of deliberation in the public sphere. Channels of political influence
can be extended to and from intergovernmental bodies such as the European
Union, international NGOs, transnational corporations, and other states.
Some groups in divided societies have already succeeded in making such
links.Forexample, in response to governmental repression andenvironmen-
tal destruction associated with oil production, the Ogoni people in Nigeria
sought help from NGOs based mainly in developed countries. These NGOs
in turn pressured their own governments and corporations such as Shell
which operate in Nigeria. In Mexico, the Zapatistas in Chiapas have devel-
oped an Internet-based network of sympathizers. This sort of outreach comes
with an obligation to behave according to emerging transnational norms of
civility. Snyder and Ballentine recommend transnational intervention to curb
the contribution of partisan journalism to hostility in divided societies.57
Appropriate measures might include professional journalism education,
press codes, sponsorship of nonpartisan media, and subsidies conditional on
accurate and balanced coverage. They point to the success in Cambodia of a
UN media program.
Of course, more negative forms of transnational linkage are possible too,
especially by nationalists reaching out to a diaspora. The Irish Republican
Army long depended on financial support from Irish Americans, and much
of the Serb diaspora was in the 1990s vocal in supporting nationalism and
excusing ethnic cleansing. The opening of channels to a neighboring state of
sharedethnicity by a minorityis also dangerousandhashistorically provided
a justification for invasion of the Sudetenland by Nazi Germany, of Cyprus
by Turkey, of Croatia by Serbia. So only outreach beyond shared national
identity has a civilizing force. This caveat aside, strengthening of trans-
national sources of political authority would be conducive to the weakening
232 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
17/26
of the connection between engagement in the public sphere and the deadly
contest for sovereign authority.
Any associated weakening of the sovereign state might be especially at-
tractive to those on the receiving end of oppression in countries like Sudan
and Rwanda, for whom a centralized state has always brought misery be-
cause it has only ever been experienced as the instrument of one segment.
Such weakening is also consistent with the increasing conditionality of sov-
ereignty in the international system. NATO intervention in Kosovo in 1999
helped reinforcethe idea that sovereignty is nota barrier behindwhich a state
can terrorize sections of its people. Theconditionality of sovereignty and trans-
nationalization of authority will not please consociationalists, whose plans
require conflict tobecentralizedandthen resolved ina fully sovereignstate.
X. PUBLIC SPHERE INFLUENCE ON
THE STATE: LOOSE CONNECTIONS
Emphasizing the public sphere (and its transnational connections) as the
focus for discursive engagement does not have to mean banning public
sphere influence over state actions. This influence is central to Habermass
model of deliberative democracy.58 Habermas endorses diffuse subjectless
communication in thepublicsphere, producingpublicopinion whose influ-
ence can then be turned into communicative power through elections, then
into administrative power through legislation. This sequence is insufficient
for divided societies for two reasons. The first reason is that subjectless
communication is too amorphous when the identity of subjects themselvesis the key issue and public opinion is deeply plural. It is better under such cir-
cumstances to think of engagement across discourses (discourse itself
being defined in non-Habermasian terms,because forHabermas discourse is
completely unconstrained communication). The second reason is that elec-
tions are highly problematic transmission mechanisms. In any society, com-
petitive elections are largely strategic and symbolically manipulated exer-
cises. In divided societies, the results they register when it comes to the
weight of competing groups and of their extremists and moderates depend
crucially on the design of the electoral system. And as discussed earlier, a
deadlynumbers game at themeta levelcan resultonce all sides recognize the
importance of electoral engineering.
Electionsare not theonly sourceof democratic legitimacy, which canalso
be secured through responsiveness of public policy to the relative weight of
discourses in the public sphere, which does not have to involve the direct
Dryzek / DELIBERATIVE DEMOCRACY 233
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
18/26
counting of heads.59 However, some electoral systems are better than others
when itcomes topromotingdiscursiveengagement ina divided society. Pref-
erential voting has the merits of promoting communication across divides
involving voters and leaders.
Electoral or otherwise, the link from public sphere to state ought not be
too tight, because then the deadly contest for sovereign authority resumes.
But if influence is absent entirely, there is a danger the public sphere may
decay into inconsequentiality. Such decay would undermine the legitimacy
of the state itself.60 Between these two extremes one can think of state and
public sphere as being loosely connected, or semidetached. Discursive
engagement in the public sphere can influence state action in many informal
ways. These ways include changing the terms of discourse in ways that even-
tually come to pervade the understandings of governmental actors. AsHabermas puts it in a moment of expansiveness beyond his stress on elec-
tions, Communicative power is exercised in the manner of a siege. It influ-
ences the premises of judgment and decision making in the political system
without intending to conquer the system itself.61 Much of the success of
environmentalism and feminism in the late twentieth century can be inter-
preted in these terms. These two movements provided a new vocabulary
including, forexample, the term environment, which did not exist prior to the
1960s. Individuals versed in thesediscourseseventually occupied influential
positions in government.
Social movements have at times achieved more formal integration into
policy making, though sometimes this has proved to be a bad bargain if the
movement has received mostly symbolic rewards. Genuine inclusion as
opposed to symbolic inclusionis facilitated to thedegree a movement canes-tablish a link between its defining interest and a core function in the states
system of priorities. For example, the alignment of environmentalism with
the core economic priority has recently been facilitated in Northern Europe
by theidea of ecological modernization.62 In these terms,a group that defines
one side in a divided society has the capacity once included to connect to the
core interest of the state in securing internal order, or at least its leadership
does, as is clear from the historical success of consociational settlements in
Europe. But as this example makes clear, inclusion of group leadership begs
some larger questions about adversary politics versus consensual politics in
the institutions of government, and howthisaffectssocial learningacross dif-
ference in the public sphere. Other unresolved questions include the charac-
terof theleaders included (radicalsor moderates) andincentives fordifferent
sorts of behavior once included.
Movement impact from the public sphere by means of changes in the
terms of discourse can occur before, during, and after any such inclusion. In
234 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
19/26
divided societies, it is easy to identify rapid change in the terms of discourse
that createdivisions rather than heal them. Forexample, Hutu andTutsi iden-
titieshardly existed inRwandabeforeBelgiancolonial rule. Butmorebenign
shifts are possible, as indicated by the rethinking of identity on all sides but
especially on the part of formerly dominant whites in South Africa in the
1990s.
Changes in the terms of discourse can be brought about by the power of
rhetoric,which canalso reach from thepublicsphere into thestate.Such was
the achievement of Dr. Martin Luther King Jr. in the 1960s. Because King
appealed to the emotional commitment of white Americans to symbols such
as the Declaration of Independence and the Constitution, he could not easily
be dismissed, and eventually the rhetoric forced redefinition of the ways in
which dominant liberal discourse was understood. When Nelson Mandelaemerged from prison he could have espoused a rhetoric of victimhood and
revenge; instead, he developed a rhetoric of reconciliation that looked for-
ward rather than backward, with telling effect on the state structure. Argu-
ments honed in the public sphere may be noticed and heeded by state actors,
andrhetoricianssuch as King andMandela didof courseaccompanyrhetoric
with argument. This sort of influence is what Habermas (following Arendt)
meansby communicative power (though Habermas accepts onlyargument
and rejects rhetoric).
XI. POSITIVE EXAMPLES
No polity that I know of exemplifies the sort of discursive democraticengagement in a semidetached public sphere that I am endorsing here. But
elements can be discerned in some systems. Consider Canada (classified by
Lijphart as semiconsociational, even though not one of his four defining
features truly applies).63 Canada features occasional attempts to rewrite the
constitution to accommodate thecompetingaspirations of Francophones and
Anglophones, as well as episodes where Quebec looks as though it might
secede and then draws back. Attempts to rewrite the constitution normally
end in deadlock, frustration, and failureeven if elites manage to bargain a
resolution,as in theMeechLakeaccords of 1987, which failed to attainratifi-
cation because of opposition from Anglophones and indigenous peoples.
Failure is generally followed by a period of inaction at the constitutional
level. In these periods of inaction, Canada is at its best, because individuals
on the various sides can then get back to engaging one another in the public
sphere where struggle over sovereignty is not at stake. Political leadership
can get back to the modus vivendi that makes Canada such a generally suc-
Dryzek / DELIBERATIVE DEMOCRACY 235
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
20/26
cessful society. The peace is disturbed only by political philosophers who
believe a constitutional solution is required. This is exactly what is not
requiredas should be clear from the lessons of what happens when it is
tried.
A secondpositiveexample canbe found in South Africas transition in the
mid-1990s. Thoughclaimed by Lijphart for consociationalism, thatdesigna-
tion applies mainly in terms of the grand coalition that oversaw transition.
There was no suppression of engagement across racial and ethnic lines as
required by consociationalisms segmental autonomy. Engagement and
reflection were promoted by Archbishop Desmond Tutus 1995-98 Truth
and Reconciliation Commissionwhich operated at arms length from the
coercive authority of the sovereign state (and withstood legal challenges
from both formerapartheid PresidentF.W. deClerkandtheAfricanNationalCongress). The commission was a deliberative institution whose terms of
reference were themselves the product of broad public debate (though the
commission was established under the new constitution). It could offer
amnesty and recommend reparations, though the implementation of its rec-
ommendations in public policy were haphazard, so its influence on the state
mayhavefallen short of theoptimum in thetermsI have developed.Perpetra-
tors and victims of apartheid-era political crimes told their stories, and there
were some very public episodes of reconciliation between perpetrators and
survivors. South Africa also featured mixed-race discussion groups, and
efforts to rethink identity in the media, educational institutions, and else-
where in the public sphere.
Deep division in South Africadidnot endwith thedeparture of apartheid.
In 1996, a liberal constitutionwasadopted that specifiedequal rights formenand women, clashing with the institution of customary marriage in some
African communities. Deveaux discusses a series of consultations initiated
by the South Africa Law Commission to resolve this conflict, which threat-
ened the authority of traditional leaders.64 These consultations produced a
compromise, which was reflected in legislation, that was acceptable to both
womens groups and traditional leaders. This compromise entailed some
reform to traditional practices, while retaining the nonliberal bridewealth
payment practice, and avoided confronting the authority of traditional
leaders. Deveaux does not address the issue, but this avoidance of challenge
to the sovereignty of traditional leaders may have facilitated deliberative
resolution. So although there was a (rare) tight connection between delibera-
tive forum and legislative outcome, this was possible because the
sovereignty issue was not confronted.
236 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
21/26
XII. THREE KINDS OF FAILURE
To further strengthen the case for emphasizing the engagement of dis-
courses in a semidetached public sphere in divided societies, consider three
kinds of failure in these terms.
Thefirstconsists of tootight a connection between publicsphereandsov-
ereignauthority. The tighter this connection, thegreater is the likelihood of a
deadly contest over thecontent of sovereignauthority. Northern Irelandsince
the 1990s illustrates this difficulty. Northern Ireland is a highly politicized
society, so there is plenty of public debate in the media, clubs, bars, commu-
nity groups, and so forth. However, the organizations active in this debate
have close links to the political leadership negotiating with British and Irish
governments over how government in Northern Ireland shall be organized.
Community groups, paramilitaries, and politicians are tightly connected.
There is great difficulty in maintaining a public sphere at any distance from
the sovereignty contest. Heroic attempts have been made by activists to
develop networks concerned with issues such as health care, employment,
and welfare across the communal divide, but such networks remain precari-
ous in the face of sectarian public spheres joined to each other mainly in the
sovereignty contest. Perhaps the most successful antisectarian institutions in
Northern Ireland todayareCommunityPolice Boards withrepresentativesof
both communities.Theseboards deal with some of themostdivisive andcon-
tentious issues in day-to-day life in Northern Ireland but stay away from the
sovereignty question. As such they are elements of a semidetached public
sphere.
A second, very different, kind of failing exists when a public sphere con-fronts a completely unresponsive state. Indeed, this kind of polity comes
close to failing to bea deliberative democracy by definition (unless collective
outcomes sensitive to public opinion can be produced in nonstate or trans-
state locations). Northern Ireland at the commencement of the Troubles in
the late 1960s may illustrate this condition. At the time, the province had
been governed for decades by the Ulster Unionist Party, whose leadership
was upper-middle class. The Troubles began as a civil rights movement on
theCatholicside. Butunresponsiveness andrepression on thepartof thestate
played into thehands of theIrishRepublican Army, andthesocialmovement
gave way to paramilitary action and terror. The struggle stopped being about
civil rights and started being about sovereignty. On the Unionist side, work-
ing-class activists denied access by the traditional unionist eliteorganized in
paramilitary fashion.
Dryzek / DELIBERATIVE DEMOCRACY 237
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
22/26
In divided societies, a state that is completely obtuse in the face of move-
ment activismmayplay into thehands of warlords whopreferviolence to the
traditional social movement repertoire, exacerbating a sectarian politics that
is both irresponsible and violent. Of course, Northern Ireland was already a
sectarian statethough beginning in the early 1970s, direct rule from Lon-
don began to ameliorate this aspect. But even a consociational state that is
completely unresponsive to events in the public sphere may be vulnerable.
Many factors conspired to drive Lebanons consociational system into civil
war in the 1970s, but one factor was the complete lack of responsiveness of a
system dominatedby traditional elites to emerging social forces, particularly
on the Muslim side. Warlords could then harness these forces.
A third kind of failure exists when there is no autonomous public sphere
worth speaking of.Again this failure is oneof deliberative democracy almostby definition, if (as I have argued) deliberative democracy depends crucially
on the engagement of discourses in the public sphere. But there may alsobe a
threatto politicalstability. In thecaseof Austria, decades of a noncontentious
party politics and consensus government eventually provided fertile ground
for the rise of right-wing populism in the form of Jrg Haiders Freedom
Party in the late 1990s. In a very different setting, Yugoslaviaunder Tito sup-
pressed any kind of contestatory politics, be it within the state or the public
sphere, partly for fear of ethnic nationalist mobilization. While the story of
the breakdown of Yugoslavia is complex, there were no substantial political
forces to stand in theway of powerfulfigures from theold regime reinventing
themselves as murderous ethnic nationalists.
XIII. CONCLUSION
Agonism may feature plenty in the way of authentic democratic commu-
nication, but is hard to apply to any divided society in the real world. States
with consociational aspects for their part can sometimes preserve political
stability in real-world divided societies, but they undermine the ability of
groups to live together through deliberative and democratic social learning.
Many positionsmay existin thelarge territorybetween these twomodels, but
my aim has not been merely to stake out a moderate position, for there is
validity in aspects of both these extremes, and I have tried to show that
these aspects can be redeemed and developed in a discursive democracy in
divided societies that emphasizes engagement in the public sphere only
loosely connected to the state. Contributions to its development could come
from the following:
238 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
23/26
deliberative institutions at a distance from sovereign authority,
deliberative forums in the public sphere that focus on particular needs rather than general
values,
issue-specific networks,
centripetal electoral systems,
a power-sharing state that does not reach too far into the public sphere,
the conditionality of sovereignty, and
the transnationalization of political influence.
NOTES
1. Benjamin R. Barber,Jihad vs. McWorld: How Globalism and Tribalism Are Reshaping
the World(New York: Ballantine, 1995).
2. Joseph M. Bessette,The Mild Voice of Reason: Deliberative Democracy and American
National Government(Chicago: University of Chicago Press, 1994).
3. Seyla Benhabib,TheClaims of Culture: Equality andDiversity inthe GlobalEra (Prince-
ton, NJ: Princeton University Press, 2002), 108-12.
4. Amy Gutmann and Dennis Thompson,Democracy and Disagreement(Cambridge, MA:
Harvard University Press, 1996).
5. Stanley Fish, Mutual Respect as a Deviceof Exclusion, inDeliberative Politics:Essays
on Democracy and Disagreement,ed. Stephen Macedo (New York: Oxford University Press,
1999), 88-102, at 92-93.
6. Iris Marion Young,Inclusion and Democracy(Oxford, UK: Oxford University Press,
2000), 16-51. Youngidentifies herself withagonism(pp. 49-51),butthe reasonableness motiva-
tionalnormwould sether apart from many agonists.She does push deliberative democracyin an
agonistic direction.
7. WilliamH. Simon, ThreeLimitationsof Deliberative Democracy: IdentityPolitics, Bad
Faith, and Indeterminacy, inDeliberative Politics,49-57, at 50-52.
8. William E. Connolly,Identity/Difference: Democratic Negotiations of Political Paradox
(Ithaca, NY: Cornell University Press, 1991); and Bonnie Honig,Political Theory and the Dis-
placement of Politics(Ithaca, NY: Cornell University Press, 1993).
9. Chantal Mouffe, Deliberative Democracy or Agonistic Pluralism?Social Research66
(1999): 745-58; Chantal Mouffe,The Democratic Paradox(London: Verso, 2000); and Chantal
Mouffe,Deliberative Democracy or Agonistic Pluralism,Political Science Series 72 (Vienna:
Institute for Advanced Studies, 2000).
10. Mouffe,Deliberative Democracy or Agonistic Pluralism,16.
11. Mouffe, Deliberative Democracy or Agonistic Pluralism? 755-56.
12. Ibid., 755.
13. Ilan Kapoor, Deliberative Democracy or Agonistic Pluralism? The Relevance of the
Habermas-Mouffe DebateforThirdWorldPolitics,Alternatives 27(2002):459-87,at 472-73.
14.John Forester, Dealingwith DeepValue Differences, inTheConsensusBuildingHand-
book,ed. Lawrence Susskind (Thousand Oaks, CA: Sage, 1999), 463-93, at 470-72.
15. Mouffe,Deliberative Democracy or Agonistic Pluralism,16.16. Arend Lijphart, Varieties of Nonmajoritarian Democracy, inDemocracy and Institu-
tions:The LifeWork of ArendLijphart, ed.MarkusM. L. Crepaz, ThomasA. Koelble,and David
Dryzek / DELIBERATIVE DEMOCRACY 239
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
24/26
Wilsford (Ann Arbor: University of Michigan Press, 2000), 225-46, at 228. See also Arend
Lijphart,Democracy in Plural Societies: A Comparative Exploration (New Haven, CT: Yale
University Press, 1977).
17. Arend Lijphart, Prospects for Power Sharing in the New South Africa, inElection 94
SouthAfrica:An Analysis of the Results, Campaignand Future Prospects, ed.AndrewReynolds
(New York: St. Martins, 1994), 222.
18.Stuart Kaufman,Modern Hatreds:The Symbolic Politics of Ethnic War(Ithaca,NY:Cor-
nell University Press, 2001).
19. See also Jorge M. Valadez, Deliberative Democracy, Political Legitimacy, and Self-
Determination in Multicultural Societies(Boulder, CO: Westview, 2001), 36-38.
20. Andrew Reynolds, Majoritarian or Power-Sharing Government, in Democracy and
Institutions: The Life Work of ArendLijphart, ed.MarkusM. L. Crepaz, ThomasA. Koelble,and
David Wilsford (Ann Arbor: University of Michigan Press, 2000), 155-96, at 169-70.
21. Cass R. Sunstein, The Law of Group Polarization,Journal of Political Philosophy10
(2002): 175-95.
22. John S. Dryzek,Deliberative Democracy and Beyond: Liberals, Critics, Contestations(Oxford,UK: OxfordUniversityPress,2000);and JohnS. Dryzek,Legitimacyand Economyin
Deliberative Democracy,Political Theory29 (2001): 651-69.
23. Dryzek, Deliberative Democracy and Beyond, chap. 3; Young, Inclusion and
Democracy.
24.Dryzek,Deliberative Democracy and Beyond,68. Young,Inclusion and Democracy, 77-
79, proposes a complementary set of standards: to ask if an intervention is respectful, publicly
assertable, and does it stand up to public challenge?
25. Margaret Moore, Beyond the Cultural Argument for Liberal Nationalism, Critical Re-
view of International Social and Political Philosophy2 (1999): 26-47.
26. Benhabib,The Claims of Culture.
27. Ibid., 1.
28. Ibid., 25.
29.See alsoMoniqueDeveaux,A Deliberative Approachto Conflicts of Culture, Political
Theory31 (2003): 780-807, at 781.
30. Benedict Anderson,Imagined Communities: Reflections on the Origin and Spread ofNationalism(London: Verso, 1983).
31. As lamented by Connolly,Identity/Difference.
32. See the guidelines proposed by Meindert Fennema and Marcel Maussen, Dealing with
Extremistsin PoliticalDiscussion: Front Nationaland Front Republicanin France,Journalof
Political Philosophy8 (2000): 379-400.
33. Deveaux, A Deliberative Approach, 788.
34. James Bohman,Public Deliberation: Pluralism, Complexity, and Democracy (Cam-
bridge, MA: MIT Press, 1996), 83-85.
35. John Forester,The Deliberative Practitioner(Cambridge, MA: MIT Press, 1999), 115-
53.
36. JohnM. Orbell,Alphons J. C. van de Kragt, and Robyn M. Dawes, Explaining Discus-
sion-Induced Cooperation in Social Dilemmas,Journal of Personality and Social Psychology
54 (1988): 811-19.
37.HeatherRae,StateIdentitiesand the Homogenisationof Peoples(Cambridge, UK:Cam-
bridge University Press, 2002).38.Forexample,Will Kymlicka,Multicultural Citizenship (Oxford,UK: Oxford University
Press, 1995).
39. Moore, Beyond the Cultural Argument.
240 POLITICAL THEORY / April 2005
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
25/26
40. Donald Horowitz,Ethnic Groups in Conflict(Berkeley: University of California Press,
1985); and Benjamin Reilly,Democracy in Divided Societies: Electoral Engineering for Con-
flict Management(Cambridge, UK: Cambridge University Press, 2001).
41.STV combinespreferentialvotingwith proportionalrepresentation in multimember con-
stituencies. The ballot requires voters to rankall candidates. Candidates achievinga quota are
declared elected; theirsurplus votes are then redistributed,along with the secondpreferences of
candidates eliminated on the basis of their low number of first preferences. If there are (say) six
seatsperconstituency, a quota couldbe 16.67percentof votes. AV usessingle-memberconstitu-
encies, again requiring voters to rank all candidates. Candidates are eliminated beginning with
those with the fewest first preferences, the votes for eliminated candidates being reallocated
according to the next preference on the ballot. Under SV, voters identify only their first and sec-
ond preferences for candidates in a single-member constituency. If no candidate receives a
majorityof first preferences, all butthe toptwo candidates(based on first preferences) are elimi-
nated, and second choices of votes for all the other candidates are reallocated to determine the
winner.
42. Reilly,Democracy in Divided Societies,136-37.43. Donald Horowitz,A Democratic South Africa? Constitutional Engineering in a Divided
Society(Berkeley: University of California Press, 1991), 189.
44. Adam Przeworski,Democracy and the Market: Political and Economic Reforms in
Europe and Latin America(Cambridge, UK: Cambridge University Press, 1991), 24.
45. Donald Horowitz, Constitutional Design: An Oxymoron? inDesigning Democratic
Institutions (Nomos XLII),ed. Ian Shapiro and Stephen Macedo (New York: New York Univer-
sity Press, 2000), 253-84, at 262.
46. GerryMackie, Does DemocraticDeliberation Change Minds? (paper presented at the
annual meeting of the American Political Science Association, Boston, 2002).
47. Archon Fung, Recipes for Public Spheres,Journal of Political Philosophy11 (2003):
338-67.
48.Graham Smith andCorinneWales,CitizensJuriesandDeliberative Democracy,Polit-
ical Studies48 (2000): 51-65; and James Fishkin,The Voice of the People: Public Opinion and
Democracy(New Haven, CT: Yale University Press, 1995).
49. Deveaux, A Deliberative Approach, 792.50. David Schlosberg,Environmental Justice and the New Pluralism:The Challenge of Dif-
ference for Environmentalism(Oxford, UK: Oxford University Press, 1999).
51. Fung, Recipes for Public Spheres.
52. Claire Jean Kim,Bitter Fruit: The Politics of Black-Korean Conflict in New York City
(New Haven, CT: Yale University Press, 2000).
53. Sunstein, The Law of Group Polarization.
54. Jack Snyder and Karen Ballentine, Nationalism and the Marketplace of Ideas,Inter-
national Security21 (1996): 5-40.
55.SimoneChambersandJeffreyKopstein,Bad Civil Society,PoliticalTheory 29 (2001):
837-65.
56. Ibid., 855.
57. Snyder and Ballentine, Nationalism and the Marketplace of Ideas, 38-39.
58. Jrgen Habermas,Between Facts and Norms: Contributions to a Discourse Theory of
Law and Democracy(Cambridge, MA: MIT Press, 1996).
59. Dryzek, Legitimacy and Economy.60. Ibid.
61. Habermas,Between Facts and Norms,486.
Dryzek / DELIBERATIVE DEMOCRACY 241
by Nenad Zakosek on October 14, 2008http://ptx.sagepub.comDownloaded from
http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/http://ptx.sagepub.com/8/13/2019 Dryzek Deliberative Democracy in Divided Societies 2005
26/26
62. John S. Dryzek, David Downes, Christian Hunold, and David Schlosberg, with Hans-
Kristian Hernes,Green States and Social Movements: Environmentalism in the United States,
United Kingdom, Germany, and Norway(Oxford, UK: Oxford University Press, 2003).
63. Lijphart,Democracy in Plural Societies.
64. Deveaux, A Deliberative Approach, 795-800.
JohnS. Dryzek is professorin the Social and Political Theory Program, ResearchSchool
of Social Sciences, Australian National University. Recent books include Deliberative
Democracy and Beyond(Oxford University Press, 2000),Post-Communist Democrati-
zation(coauthored, Cambridge University Press, 2002), andGreen States and Social
Movements(coauthored, Oxford University Press, 2003).
242 POLITICAL THEORY / April 2005