You are on page 1of 30

Société québécoise de science politique

The Problem of Dirty Hands in Politics: Peace in the Vegetable Trade

Author(s): S. L. Sutherland
Source: Canadian Journal of Political Science / Revue canadienne de science politique, Vol. 28,
No. 3 (Sep., 1995), pp. 479-507
Published by: Canadian Political Science Association and the Société québécoise de science
Stable URL:
Accessed: 14/12/2009 22:15

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact

Canadian Political Science Association and Société québécoise de science politique are collaborating with
JSTOR to digitize, preserve and extend access to Canadian Journal of Political Science / Revue canadienne de
science politique.
The Problemof Dirty Handsin Politics:
Peace in the VegetableTrade*

S. L. SUTHERLAND Carleton University

The standard conception of the dirty hands problem looks at desperate

choices in the life of a political leader in terms of their consequences for
the soul: the focus is the moral condition of the autonomous moral actor.
Therefore reflection on dirty hands scenarios naturally encourages
judgments about the inherently compromising nature of politics; the
fatalistic message is that political actors will inevitably compromise
their moral being by taking part in political action. Perhaps the single
passage most cited to exemplify the supra-ethical stance of the political
leader is a short speech from Jean-Paul Sartre's Les mains sales. The
speech is taken to mean, in Coady's phrase, "that the vocation of poli-
tics somehow rightly requires its practitioners to violate important
moral standards which prevail outside politics."' The revolutionary
leader (Hoederer) apparently chides his young secretary and eventual
assassin (Hugo) for the stiffness of his beginners' principles:
Comme tu tiens a ta purete, mon petit gars! Comme tu as peur de te salir les
mains... La purete,c'est une idee de fakiret de moine. Vous autres,les intel-
lectuels, les anarchistesbourgeois,vous en tirez pretextepour ne rien faire. Ne
rien faire, resterimmobile, serrerles coudes contre le corps, porterdes gants.
Moi j'ai les mains sales. Jusqu'auxcoudes. Je les ai plongees dans la merdeet

* The author thanks David Braybrookefor his unfailing generous willingness to

comment and advise, as well as colleagues Peter Emberley, Radha Jhappanand
Denis St-Martinfor their, in equal measure,helpful and challenging comments,
and David Shugarmanfor providingthe incentive to think about the subject.The
JOURNAL'S referees also provided insightful and encouraging comments, for
which the articleis richer.
1 C. A. J. Coady, "Politics and the Problemof Dirty Hands," in Peter Singer, ed.,
A Companionto Ethics (Oxford:Blackwell, 1991), 373 (emphasisadded).

S. L. Sutherland,Departmentof Political Science, CarletonUniversity,Ottawa,Ontario

K1S 5B6

Canadian Journal of Political Science / Revue canadienne de science politique, XXVIII:3 (September/sep-
tembre 1995). Printed in Canada / Imprime au Canada

dans le sang. Et puis apres?Est-ce que tu t'imagines qu'on peut gouvernerin-


These are Sartre's words, but, I argue, they are not a distillation of
the message of his play, which is really about the inevitable sequel to a
daring "political" act, that is, retrospective political and social judg-
ment. Likewise, I hold, if there are important lessons about politics em-
bedded in dirty hands or realist problematics, they are more likely to be
found in the fifth act than in the first, that is, well after the initial execu-
tive acts, or first round of decisions.
This article is constructed in two parts. The first explores, through
the definitional features of the dirty hands perspective, whether it is pos-
sible to sustain the claim that it reveals a moral compromise at the heart
of political action. The second part argues that it is important to study
dirty hands episodes for what we can learn about the capacity of our po-
litical structures to conduct thorough-going retrospective discussions
about conduct in politics and public life, and to hold political actors to
account for the impact of their actions upon the quality of process in col-
lective life. By "process" I mean of course that public business is con-
ducted under the rule of law, but also, more generally, that it is con-
ducted with respect for deliberation and disclosure. The features of the
classical dirty hands problem that are isolated and challenged are the
following: (1) that the political actor must be pressured into acting on
consequentialist reasoning in the context of a psychologically compel-
ling narrative frame; (2) that he3 must then be judged as an individual in
the context of a deontological ethics; (3) that the exploration of the
modelled or schematic dirty hands situation captures the important as-
pects, or the heart, of politics; and (4) that the moral importance of the
political gesture is somehow directly expressed in, or is inherent to, the
act itself. This latter understanding is challenged through the dynamic
argument of Les mains sales.
2 Jean-Paul Sartre, Les mains sales (Paris: Gallimard, 1948), 104. A free translation
might be as follows: "How fond you are of your purity! How afraidyou are to
dirtyyour hands.... Purityis an idea for holy men. The rest of you, you intellectu-
als, middle-class anarchists,you use it as a pretextfor doing nothing:to do noth-
ing, to rest immobile, to hold your elbows in close to your body, to wear gloves.
Me, I have dirty hands.Up to the elbows. I plunged them into shit andblood. And
what of it? Do you imagine thatone can govern innocently?"Hoedereralso says,
more succinctly, "Tousles moyens sont bons quandils sont efficaces" (193).
3 The masculinepronounsare used because the dirty hands problem is historically
gendered,making the use of "she" anachronisticor at the least, so counterstereo-
typical as to suggest an unintendedirony. The masculine pronoun is also much
more faithful to the typical extension of ideas of masculinephysical prowess and
the aestheticof heavy physical effort to the (male) leader's oneroustask of strate-
gizing and planning-probably to lend charismafor the followers. Thus male ex-
ecutives can be portrayedas sweating while only thinking, while this would seem
quite odd in the case of a woman.
Abstract. Most treatmentsof the problem of dirty hands in politics assume that
merely holding a position of greatpolitical power will requirea political actorto violate
importantmoral standards.They assume thatthe successful political leadermust inevi-
tably be morallycorruptedby the iniquitouschoices thatmust inevitablybe made, and,
further,thatthis casts a shadowupon political life as a moralenterprise.This article ar-
gues, instead, that the conventionaldirty hands problem is not particularlysignificant
and that a much more serious test of the moralquality of public life in a given polity is
how it makesits arrangementsfor formalpublic retrospectionupon andjudgment of the
inevitable episodes of unwise, intemperateor immoral political action by leaders. In
short, it is the deliberatecorruptionof democracythat should attractour scrutiny, not
the conditionof the soul of the supra-ethicalor maverickleader.
Resume. La plupartdes analyses surle problemedes <<mains sales >>en politique pos-
tulent generalementl'exigence de la transgressiondes normes moralesparceux qui de-
tiennentd'importantespositions d'autoritdpolitique. Ces analyses pretendentque pour
reussir,l'acteurpolitiquedoit fairedes choix iniques qui corrompentin6vitablementsur
le plan moral,ce qui auraitpoureffet de donnerun caractereamorala la vie politique en
gen6ral.Cet article remet en question l'importancedu probleme des <<mains sales > et
soutient plutot qu'une fa9on beaucoupplus significative d'examiner la qualite morale
de la vie politique d'une societe donnee, est de voir commentelle organiseles processus
de revision lui permettantde discuter et de juger les actes moralementrefractaires
commis episodiquementpar ses leaderspolitiques. En bref, c'est l'effet sur la democra-
tie des actes <supra-ethiques>>du leader qui doit etre scrut6par le public, et non pas
l'effet de ses gestes sur son ame.

The second part of the article is built on three main ideas; Dennis
Thompson's concept of mediated corruption, which holds that strate-
gies to evade the deliberative phases of decision and the judgment of re-
trospective deliberation are to be seen as offences against the political
system itself;4 David Braybrooke and Charles E. Lindblom's persistent
and evolving image of policy making as an ongoing exploratory, reme-
dial and serial activity, more or less systematic;5 and some of my own
notions about how particular representative institutions are more or less
apt in the task of configuring and conducting a retrospective cycle of
concentrated deliberation to weigh the earlier acts of the executive. The
importance of the retrospective phase is that it enables the public delib-
erating in hindsight to be thoughtfully remedial (and even directly cor-
rective) in respect to the quality of process and of political conduct ex-
pected of officials as well as in respect to identifying desirable substan-
tial or policy changes. The usefulness of this perspective, which em-
phasizes deliberation, systematic retrospection and remediation, is in-
vestigated in four examples: Hollis' discussion of the sacrifice of Cov-
4 Dennis F. Thompson, "Mediated Corruption:The Case of the Keating Five,"
AmericanPolitical Science Review87 (1993), 369-81.
5 David Braybrooke,TrafficCongestionGoes Throughthe Issue Machine (London:
Routledge and Kegan Paul, 1974); David Braybrookeand CharlesE. Lindblom,
A Strategy of Decision: Policy Evaluation as a Social Process (New York:Free
Press of Glencoe, 1963); Charles E. Lindblom and David K. Cohen, Usable
Knowledge:Social Science and Social Problem Solving (New Haven: Yale Uni-
versity Press, 1979); andDavid Braybrooke,Bryson Brown and Peter K. Schotch,
Logic on the Trackof Social Change (Oxford: ClarendonLibraryof Logic and

entry;the use of the presidentialpardonin Americanpolitics; an evasion

of ministerialresponsibilityby the Conservativegovernmentin Canada
in 1991, whereit effectively triedpublic servantsby vote by constituting
a parliamentarycommitteeas a surrogatecourt;andBrecht's transposi-
tion to the Chicago vegetable trade of the story of the rise of German
fascism leadinginto the Second WorldWar.

How Can We Escape the Dirty Hands World?

The dirty hands dilemma is a circumstancein which a political actor
must strive heroically and autonomouslyto achieve the least-evil out-
come for some well-defined group,wherehe mustdo wrong to do so, in
secret,and where, in a successful test, he is accountableonly to himself.
In his beautifully crafted exposition of the Glencoe massacre, for ex-
ample, Martin Hollis endorses what he says is Machiavelli's starting
point: that "there [exist] evil persons who do not wish the good of the
people and who do not keep faith;and thatthey can only be thwartedby
marshallingthe apparatusof legitimacy against them; and that this ap-
paratushas to be used dishonestly."6The apparatusof legitimacy is
here the armand power of the state. In 1692, two companiesof Argylls
(Campbells)who, supposedlyin the context of an effort to build peace,
were billeted on the MacDonalds of Glencoe, acting on orders, rose
against their hosts in the night and slaughteredeveryone they could,
then burntwhat they could not take. The militaryhierarchywas not a
brakeon terrorbut a convenienceto its operatorswho were precipitated
into actionby the opportunityto ambush.So the tribeswere pacified-
the least evil outcome. "Politics," Hollis concludes from his examina-
tion of the dilemma of the statesmenwho organizedthe massacre, "is
an arenawhere the best is the enemy of the good, where we license our
agentsto pursuethe good andwhere they can succeed, only if they oper-
ate partlybeyond our ken and our control."7The task of the leader(and
his secret service) is, then,to act autonomouslyandcovertlyas society's
agent in areas that are beyond the prospect of being given directionor
held undercontrolby ordinarypeople, and are even beyond our under-
standing. The heroism of our agents arises from their-and our-
acceptanceof the Faustianbargainthat great crimes require,and quite
possibly even make, greatmen.
Othermuch-citeddirtyhands situationsinclude the following: the
justificationof political assassinationas an action thatcan dramatically
correctthe course of history;whetherit can be right to yield a targeted
judge to terrorists in exchange for the safety of many innocent hostages;
6 Martin Hollis, "Dirty Hands," British Journal of Political Science 12 (1982),
7 Ibid., 398.
The Problem of Dirty Hands in Politics 483

what to do if one should meet a guerrilla leader who stipulates that if one
will personally slay just one small peasant he will spare the rest of the
otherwise-doomed group; and whether, as a leader of a country in civil
war, one ought personally to order that a captured opponent be tortured
to extract information that one expects will save lives.8 In brief, the role
of the masses in a dirty hands problem is as the body count: the public is
a passive collection of individuals of whom the lone decision maker
must save the largest number possible.
Explorations of dirty hands problems almost invariably acknowl-
edge a debt to, as Rossiter puts it, "the little book that Machiavelli wrote
to wheedle a job out of Lorenzo de'Medici."9 As Albert Hirschman
points out in The Rhetoric of Reaction: Perversity, Futility, Jeopardy,
Machiavelli's The Prince gave impetus to the timeless stream of rheto-
ric that debunks any and all mechanisms that have been proposed to al-
low the masses to participate in governing.10 The basic themes of the
"realists" are those of Hirschman's title: any democratic reform will
work in the opposite way to that intended, come to nothing, or place the
whole state and society in the gravest danger. The essential message of
the critics of democracy, that "maudlin enthusiasm of humanity," is
that there exists an unbridgeable gulf between those who lead and those
who are led. Thus it is best to be led by a bold elite which will govern
with genius and fortitude. 1

8 Assassination is a subject for Albert Camus, notably The Just Assassins, in Cali-
gula and ThreeOtherPlays (New York:Knopf, 1958); thejudges problemis from
RobertNozick, Anarchy,State and Utopia (Oxford:Blackwell, 1974); the torture
example is taken up in Alan Donagan, The Theoryof Morality (Chicago:Univer-
sity of Chicago Press, 1977), 184-89; and the jungle parableis developed in Ber-
nardWilliams, "A Critiqueof Utilitarianism,"in J. J. C. Smartand BernardWil-
liams, eds., Utilitarianism:For and Against (Cambridge:CambridgeUniversity
Press, 1973), 98. But see also Williams' "Politics and Moral Character," in
S. Hampshire,ed., Public and Private Morality (Cambridge:CambridgeUniver-
sity Press, 1978), 55-74. The authors' various scenarios have taken on a life of
their own. Michael Walzer,"Political Action: The Problemof Dirty Hands," Phi-
losophy and Public Affairs 2 (1972-1973), 160-80, perhapscontains the most di-
rect discussion of the possibility of living a moral life as a public figure. In real
life, the defence presentedby the formerVichy governmentofficial, PaulTouvier,
in his 1994 trialfor crimes againsthumanityhalf a centuryearlier,was thathe had
saved 93 Jewish prisonersby designating seven for execution. See Paul Webster,
"Touvier Trial Goes Easy on Vichy," Manchester Guardian Weekly(Manches-
ter),April 10, 1994, 4.
9 Clinton Rossiter, ConstitutionalDictatorship: Crisis Governmentin the Modern
Democracies (Princeton:PrincetonUniversityPress, 1948), viii.
10 Albert O. Hirschman,The Rhetoric of Reaction: Perversity,Futility, Jeopardy
(Cambridge:The BelknapPressof HarvardUniversityPress, 1991).
11 Ibid., 93. The descriptionof democracyas a "maudlinenthusiasm"thatcould de-
stroy the institutionalcreationsof centuries of "wise heads" is from the debates
leading up to the Second ReformAct of 1867 in Britain.

How can we escape the dirtyhands world?I begin by challenging

the definitionof thatworld.

Consequentialismand Its Limits

The representativedirtyhandsscenario,as Benn notes, anticipatescosts
and benefits.12It is thereforeconsequentialistin its form. The problem
solving by the individualcaughtin the dirtyhands situationthus suffers
fromthe sametellingweaknessesas does anysystemof reasoningthatas-
sumesthatit is possibleto elaboratea rational,comprehensive,robustpre-
dictivesystemthatcan give ourrationalactordirectcontrol(agency)over
the chosen variables.BraybrookeandLindblomsummarizethe problems
in relationto utilitarianismas follows: "no interpersonalcalculus;no rule
for bringingaccounts of consequencesto an end; no rule for specifying
referencegroupsfor whom consequencescount." 3
Therefore, it seems fair to say that the dirty hands hero cannot
found his belief thathe has a "right" to take decisions for the masses in
a predictivecertaintythathis strategyis best in the circumstances.Cer-
tainly he may make probablejudgments, but where the probabilityof
success is low, the legitimacy of unilateraldecision making, outside an
emergencysituation,begins to seem very dubious.Indeed, to relax the
demandfor certaintyof a successful outcome seems to invite deception
about the reasons that may have triggereda round of expedient, secret
decision making:it is a politicalblank cheque. And certainlya lone de-
cision makermight seek to reduceuncertaintyanddiffuse responsibility
by consultingwidely, but thenthe moralresponsibilityfor the projected
action would seem to be shared,and among an arbitrarilychosen coun-
cil at that.The would-be decision makermust thereforefind legitimacy
as a political actorelsewhere thanin certaintythatpersonalrule will be
most effective and economical in terms of the sacrifice imposed on
those withoutthe positional power of office: in the inherentright to be
the decider, thatis, in elitism of birthor power (no longer legitimatein
the modern constitutionaldemocracies among which we hope to con-
tinue); or in respect for the process under which the decision maker's
role is circumscribed,thatis, in constitutionalism,whose supremacythe
dirtyhandsproblemdoes not acknowledge.

TheIncoherenceof the MoralPerspectiveof the Hero

A second common assumption is that the decider's consequentialist
outputphase, in which he tries to succeed using evil means, is suitably
12 StanleyI. Benn, "Privateand Public Morality:CleanLiving and Dirty Hands," in
S. I. Benn andG. F. Gaus,eds., Public and Private in Social Life (London:Croom
Helm, 1983).
13 BraybrookeandLindblom,A Strategyof Decision, vii.
TheProblemof DirtyHands in Politics 485

followed by a deontological phase of self-judgment.The dirty hands

leader in a wicked situation must act as a bold consequentialist.But
then, the well-intentioned expedient bad action having been accom-
plished in secret, he must turn deontological principles upon himself.
Under the set of rules by which he now judges his own conduct in his
study,he must acceptthathe has done wrong in adoptingwicked means
and heartily and sincerely condemn himself. Meanwhile, the practi-
tioner consequentialisttwin must conceal all evidence of the wrong ac-
tion by any means, keep the burdenof his guilt to himself, and must
stiffen his resolve to carryon in this vein, because he believes thatonly
by so doing can he be of real use to his followers, towardswhom he
adopts a poulterer's morality.Furthermore,he must convince himself
that at some higher level he was somehow "right" to have done wrong
and thathe has thereforebeen ennobled and temperedby the traumaof
having undertakena course of action from which ordinary persons
would shrink, and is therebyfitted to take on bigger challenges of the
same nature.The problematicthus has the heroic decideracting "politi-
cally," thatis, as a consequentialist,butit also insists thathe see himself
as what he is on universalmoral principles:a tragic failure. But is this
pose coherent(note thatone does not say "possible")? I would submit
thatit is not. The dirtyhandsleaderis always in consequentialistmode,
no matterhow unhappyat nightfall.

Generalizationto All Politics of a Flawed Model

The thirdassumption,that the typical dirty hands scenarios reveal the
true natureof political action in general, is also flawed. It involves a
generalizationfrom the particularscenario(which can also be described
as a modelled political problem)to some universalizedcharacteristicof
"political action." But a dirtyhands scenario can shed light only upon
the variablesit explicitly takesinto accountin its specificationof the im-
portantinternalcausal conditions. And, further,the feature of internal
validity, assuming it is attained-that such a set of identified circum-
stances would plausibly interact for the predicted result-does not
Externalvalidity, the appropriatenessof attributinga result or dy-
namic observed in the model to the world, is to be judged by the ade-
quacy with which the "experiment" or problem that has apparently
been plumbedwas appropriatelymodelled upon each of the telling con-
ditions of the domainto which it is hoped to generalize.Unless each of
the causal featuresof the settinghas been seized in the schematicprob-
lem, one must not generalizefrom the model or experimentback to the
settingfrom which the model was schematized.The dirtyhandsprotag-
onist is a one-personexecutive,unencumberedby eitherassemblyorju-
diciary. As Anthony King tells us, executives are the primitivecore of

governmentactivity,being "alreadythere" at the startof organizedpo-

litical life.14Dirty handsscenariosthereforeexclude the prospectof the
reciprocitybetween citizens and leaders which arises throughthe citi-
zens' role in choosing the governors,andis regulatedby the structureof
justice that predatesthem both: there is no prior structureof justice in
the raw "politics" of a dirtyhandsproblembecause thereare no courts
or legislatures, only what was alreadythere-power and domination.
This is essentially a feudal world, in which the noble has rights of life
and death over the inhabitantsof his territories,and deals as an equal
only with othernobles.
Because the dirtyhandsexecutive acts outside any constitution,all
action in a dirty hands problem takes place within the paradigmof the
emergency situation.The protagonistis therebyin a game with the fol-
lowing rules: he is the only actor in his round;if he refrainsfrom bold
action, the innocent will pay; the only way he can save at least some in-
nocents is with the blood of otherinnocents;and time presses. His mo-
ralityis engaged only in the utilitarianor "lesser evil" rule thathe will
seek to minimizesufferingas he engages the problem.
Thus the feature thatis seen to be missing here is thatpolitics are
social, andthuspolitical activityis inevitablypublic (sooneror later).

Sartre'sPlay as a Perspectiveon Dirty Hands

The final assumptionof dirtyhands scenariosexaminedhere is one that
Sartrehimself challenges: that the distinctivemeaning of the political
act is to be found in the act itself, and thus can be neatly exploredin just
such scenarios. Les mains sales repays examination in this respect,
because it creates a reflection on this issue as its action progresses.Its
protagonistacts, judges himself, is judged and is made to suffer from
the brash "innocence" of his earlieracts as they take their own trajec-
Sartre'sstory takes place as the war is coming to an end and vari-
ous factions arejockeying for a favourablepeacetimerole. "Politics" is
moving away from wartimestrugglefor absolutestowarddialogue and
compromisewith yesterday'senemies. The play is essentiallya psycho-
drama,focused as it is upon threestruggles:first on the long momentof
choice of the idealistic and thus uncompromisingHugo as to whetherto
kill the realist and compromisingHoederer;secondly on Hugo's moral
and philosophical exploration of the meaning of his own act; then,
finally, on Hugo's struggleto control,in a social context, the meaningof
his original choice by exercising his freedom to reaffirmor to reinter-
14 "The [contemporary]executive... is neither more nor less than what is left of
government... when legislatures and the courts are removed" (Anthony King,
"Executives," in F. I. Greensteinand N. W. Polsby, eds., Handbookof Political
Science, Vol. 5 [Reading,Mass.:Addison-Wesley,1975], 181).
The Problem of Dirty Hands in Politics 487

pret it through his subsequent actions. (Authenticity, as the existential-

ists tell us, is not won forever in one act, but in each act, every day.)
Soon after its appearance, Les mains sales was virulently attacked
by the Communist press, which not unnaturally saw its story line as an
attack.15 It was then appropriated as a propaganda weapon in the Cold
War. To give one example of how the play was treated: the 1948 Ameri-
can production was renamed "Red Gloves," cast the elegant Charles
Boyer in the role of the virile man-of-the-people, Hoederer, invented for
him a new speech about Abraham Lincoln, and saved Hugo from his in-
exorable fate.16
Another hijacking of Les mains sales is the constant invoking of
the Hoederer speech, cited earlier, as though it were a final judgment on
the essential nature of politics. Sartre himself did not think he had come
down on either the side of realism or on that of idealism in the play:
Je donne raison a tous: au vieux chef realiste du PartiProletarien,qui, parce
qu'il transige provisoirementavec la reaction, se voit qualifie de "social-
traitre"par pur opportunisme[of the partyfaction that wants to take over the
fruitsof Hoederer'slong struggle].Et aussi a sonjeune disciple, eperdud'idea-
lisme, que les "durs"ont charged'executercelui qui fut son idole.17
When Hoederer tells Hugo that he has dealt in both merde et sang, one
must come to terms with the merde because Hoederer mentions it first,
thereby qualifying his use of the word sang. Merde has a weaker scato-
logical focus than the English "shit"; it comprises nonsense, doubt, pet-
tiness, confusion and futility. Putting it first places blood as the other,
fated, pole of this sad, muddled, human spectrum. Thus also the injus-
tice of the standard finicky translation of Sartre's merde as "filth,"
which implies the speaker's distance from, and scorn for, the disorder
and confusion of human struggle. Hoederer, on the contrary, is saying
that he, unlike the bourgeois anarchists, is unconditionally willing to
lose himself in the merdier of mass political action.
Therefore, the "filth and blood" translation obscures Hoederer's
condemnation of those who pursue revolution as a series of poses. This

15 See Marc Buffat, Les mains sales de Jean-Paul Sartre (Paris:Gallimard, 1991),
16 At first Sartretried washing his hands of the whole business, declaring that the
play's fate in the world representedan "objective" path that had nothing to do
with him. But by 1952, a declared critical sympathizerof Communism, he was
refusing to authorizethe play's productionin certain cities, "des points nevral-
giques." This semi-banlasteduntil the mid-1960s (ibid., 167-71).
17 Ibid., 166, quotingMichel ContatandMichel Rybalka,Les Ecrits de Sartre(Paris:
Gallimard, 1970), 178-79: "I make everyone right:the old 'realist' leader of the
proletarianparty, who, because he makes some provisional concessions with the
reactionaryparty,sees himself portrayedby opportunistsas a traitorto society, is
correct.And so also is the young disciple, lost in idealism, whom the hard-liners
have sent to execute he [Hoederer]who was once his idol."

is the error-the bad faith-of which Hoedereris hoping to cure Hugo

by telling him to abandonhis naively intransigentidealism. To no avail,
for Hugo peevishly replies, "On s'apercevrapeut-etreun jour que je
n'ai pas peur du sang." Hoederer's answer to this is: "Parbleu:des
gants rouges, c'est elegant. C'est le reste qui te fait peur. C'est ce qui
pue a ton petit nez d'aristocrate."'8To think of proletarianblood as
"red gloves," as a layer of protectionfrom ce qui pue, is the fatuous
(but still fatal) and narcissisticstupidityof the bourgeoisanarchist,who
is willing to wear the peoples' blood as a symbol of his own commit-
mentto theiremancipation.
Hoederer rejects Hugo's argumentthat intransigenceis owed to
the slain martyrsof the party. He rejects the idea that only still more
blood can redeem the blood alreadyspilled: "Je me fous des morts.Ils
sont mortspour le Partiet le Partipeut deciderce qu'il veut. Je fais une
politiquede vivant,pourles vivants."19Living people are to come first.
Even knowing that Hugo is a danger to him, Hoederer so much
prefers dialogue that he thinks he can save his own life by disarming
Hugo with words.20Hoedereris not afraid to die, only of being killed
before his work is done. But Hugo is afraidto kill, and, alas for Hoed-
erer,he is even more afraidof discoveringthathe is incapableof killing:
Hugo apparentlyinterpretsthe seductivenessof Hoederer's arguments
as a sign thathe, Hugo, lacks faith.
And here the characterswould be hung for eternity like those in
Huis Clos, had Sartrenot providedHugo with a young and attractivefe-
male companionat headquarters,who is presentedas his wife, Jessica.
Hoederer'srationalityand virility work somewhatmore quickly on Jes-
sica. Still hesitating as to whether to obey the orders to kill Hoederer,
Hugo stumblesupon Jessica and Hoedererin each other's arms.He ac-
quiresthe will to pull the triggerbecause of the shock of the doublefail-
ure:Jessica's physical treasonto theircause, made possible by his own
temporizing.21But with his last breath,Hoederer,infused with his char-
acteristiccharitytowardshumanmuddle,cheats Hugo of his momentof
false decision by describinghis own assassinationas a crime of passion,
exculpatingHugo. The Sartreandilemmais thatHugo has not yet killed
18 Sartre,Les mainssales, 194 (emphasisadded).Hugo: "Youwill see one daythatI am
not afraidof blood." Hoederer:"redgloves, that'selegant.It is the restof it [i.e., the
merde]thatfrightensyou. That'swhatstinksto yourlittlearistocraticnose."
19 Ibid., 191. Minutes laterhe says, "Pour moi fa compte un homme de plus ou de
moins dansle monde."
20 Buffat,Les mains sales de Jean-PaulSartre,69.
21 The apparentlychauvinistconceptionof Jessica's role, while it may date the play,
is unimportant.There are serious women in the party, such as the one who is
chargedwith the responsibilityof deciding whetherHugo must be killed, a task to
which she seems equal.
TheProblemof Dirty Hands in Politics 489

Now things become a little complicated. On his release from the

shortjail sentence imposed for crimes of passion, Hugo still cannotde-
cide whetherhe is a simple murdereror a principledpolitical assassin.
Again, externalevents provide a resolution. He finds that the group on
whose ordershe had killed Hoedererhas now embracedHoederer's leg-
acy of compromise,endorsingthe very policies for which he was killed.
The outstandingquestion for the partyis whetheror not Hugo, its tool
for the killing, is "recuperable."ShouldHugo agreeto change his iden-
tity andremainsilent, the partywill continueto give him shelterof sorts
and make use of him. But if he chooses to claim the murderas his own
political testament,the party will arrangeto inter that testamentwith
Hugo under one and the same tombstone. In short, the only way for
Hugo to stay alive is to acceptthathe is a deeply compromisedandfool-
ish young man who has now lost almost any marginof freedom or hope
of authorshipof his own life. Thus Hugo finds himself in a replay of his
first set of decisions, the pressure in this retrospectiveround coming
from the party's conditionthathe must now interprethis past action in
conformitywith the party'scurrentrequirements,therebyrelinquishing
any ideals thatmay remainto him, or die.
But Hugo is revoltedat the idea thatHoederer'sdeathshould stand
as a kind of accident. He wants to honour Hoedererby affirminghis
murderas a politically motivatedact; as an assassinationand not as a
blunderthat became the cynical partisantactical opportunityin which
he would collude. Thus, in gallant irony and existential good faith, he
standshis groundas, at one and the same time, his own man at last and
the party's past dupe, and has his last and finest moment in crying out
the words "non recuperable,"triggeringhis own assassination.Hugo's
more educatedchoice, his second chance at political action, says to the
party that too much opportunismis a bad thing. (Hoederer's murder,
like his arguments,show the opposite, thattoo muchprinciplein politics
is a bad thing.) In his death,Hugo yields to his fate as one of the party's
unnamedsoldiers,the meaningof whose actionis open to interpretation
and appropriationby the living. In so doing, he takes his place with
Hoederer in the confused egalitarianismof the merdier of action and
death.Hoederer'sdeathof course always belonged quite squarelyto his
partyjust as he had said:he died for the partyand the partymade of his
deathwhatit wanted.
Thus while Sartreforces one to strugglewith the politics in his play
("Je donne raison a tous . ."22), the central theme is clear: events may
be enacted by individuals,but meaning (political morality)is found in
the process of social negotiation,to which the actorhas no choice but to
give up his act and perhapshimself, somewhatas Sartrelost his play to
the constructionsfitted upon it by others. "First-round"events may
22 Ibid., note 17.

well be behavioursqueezed out of the equivocatingprotagonistby the

relentless pressure of the situation. But later, as the protagonistcon-
tinues to try to negotiate the meaning of the deed in society, one sees
more clearly the political meaningof the action. Thus, regardlessof an
actor's experience and beliefs as to his own motivationat the time of an
action, if indeed they were clear, the author'ssignaturedoes not remain
with the action to become the authoritativesocial interpretation.Instead,
the action as it has been witnessed at second remove is subjectedto a
public reconstructionthatmoralizesit from the exterior.In the vocabu-
lary of this argument, the social process of retrospective delibera-
tion-a wickedly manipulatedprocess in Sartre'splay-appropriates
andconfers layers of meaningupon the first-roundbehaviour.23
Therefore Sartre'splay, I argue, is a complex extended examina-
tion and, ultimately,a fatalisticmockeryof the bold adventurerin poli-
tics who feels fit to decide for othersas thoughthe futurewere a game of
chess in an empty room: the play insists that one cannot predict what
will happen as a result of one's act from the basis of what one might
wantto thinkone had intendedto happen.24

Political Action as Constituted by Rules, Deliberated

and Retrospectively Evaluated
The dirty hands world and the emergency situation are inadequateto
model political life, because the entire prospect of constitutionalde-
mocracyis missing. Constitutionalismcan be definedas the belief in the
rule of law and the belief thatideals andrules abouthow deliberationis
realized and how the mechanismsof democraticprocedurewill operate
are importantenough to be summarizedand expressed in normative
documents that should and usually do guide the processes of govern-
ment, and, secondly, thatthese rules should try to ensurethatthe power
of government will be linked and subjugated in some way to the
23 As Gadamerinsists, the fissure that time creates means that meaning must be
reconstructed, and, being reconstructed, is different from what went before
("alienated"). The past is remade socially in a mediation with the present. See
Joel C. Weinsheimer,Gadamer'sHermeneutics:A Reading of Truthand Method
(New Haven:YaleUniversityPress, 1985), 131-34.
24 On self-authorshipand authenticitysee David Carr,Time,Narrative and History
(Bloomington: Indiana University Press, 1986), 80-99. In Carr's terms, Hugo's
choice of the timing of his own deathcan be seen as his clinging to authorship:he
makes the issue of what social meaning is to be given Hoederer'sdeath the sum-
mationor, perhaps,roundingout, of his own "project."Carrnotes thatHeidegger
and Sartre"sharethe same ideal of authenticityas self-authorship:it is simply that
Sartre,with tragicpathos, believes the ideal unattainable"(84). In the sameplace,
Carrsays thatto suppose thatone can take complete chargeof the stories in which
one figures, or to work oneself up into a sense of tragic regretthat one cannot do
so, "is to succumbto the illusion of being or desiringto be God."
TheProblemof Dirty Hands in Politics 491

people.25This does not mean thatmoralandpolitical safety areguaran-

teed by constitutionalismthroughruleby law andthe respectand adher-
ence to democraticparticipationin decision makingthatit underwrites,
because, as noted, emergencies and exceptional situationswill always
arise in which leaders may behave in unforeseenways. What constitu-
tionalismadds is the morningafter:a public process of retrospectionin
which leaders' actions are reviewed. As its main purpose,political ac-
tivity in representativeinstitutionsmust channeland control,assess and
sometimes condemn the use of power. This is so whether the political
action is to be, or was, exercised within the state or externally,whether
the action is routineor, on the other hand, involves the exercise of dis-
cretionin poorly understoodcontexts.
The decision phase of the executive is thus inevitablyfollowed by
phases of judgment. These subsequent rounds of discussion, whose
form and impact are dictated by the contingent configurationsof the
representativeinstitutions,operate to moralize and, indeed, politicize
the events of previous rounds:there will be reassessment,redefinition,
approbationor disapprobation,and formulationof new issues for the
agendas of all parts of the institutionalframework.Public judgment
takes the place of the autonomousphase of self-judgmentin the dirty
hands scenario. And retrospectivedeliberationalso takes place, it must
be remembered, outside the representativeinstitutions in the "big
loose" systems of social judgment (the latter now their own kind of
merdier, because they are largely media-drivenor media-appropriated
for profit) and political influence that run through the whole society.
Sometimes retrospectivedeliberationwill lead to new policy, and occa-
sionally to destitutionor censuring of particularactors. These cycles
will be more or less open, timely, exploratory,inexorable,just or reme-
dial dependingon actualstructuresof the representativeinstitutionsin a
given political system andthe issues in question.
This perspective on democracy as both a policy delivery system
and a conversationalsystem had its origin in David Braybrooke'sbook
about public debate over traffic congestion, and his earlier book with
CharlesLindblom,Strategyof Decision. Braybrookeand Lindblomsay
that public policy is "remedial, serial and exploratory."26What is ar-
gued here is that particular political institutions provide particular
structuresand routesfor ongoing public conversationsaboutpublic life
25 See,fora recentexploration of democracies
as communitiesof inquirycapableof
intelligentcollectiveadaptation,CharlesW. Anderson,"Pragmatic Liberalism,
theRuleof Law,andthePluralistRegime,"in StephenL. ElkinandKarolEdward
Soltan,eds., A New Constitutionalism: DesigningPoliticalInstitutionsfor a
GoodSociety(Chicago:Universityof ChicagoPress,1993), 106. An important
sourcefor suchtreatments is JohnDewey,Liberalismand SocialAction(New
26 Braybrooke andLindblom, A Strategyof Decision,vii.

that are inexorably serial and differentiallyexploratoryand remedial

(and perhapsdifferentin deterrenteffect). My goal is to emphasizethat
political conversationdoes take place, not only in regardto the content
and form of policy, but also, importantly,in regardto political action
overall. In particular,I want to stress that these public conversations
(withina particularlystructuredset of rules in the communityof conver-
sation) retrospectivelyexplore and pass judgment on the moralityand
general aptness of previous rounds of political action. Thus society
strugglestowardsubstantialpublic policy and also towardpublicmoral-
ity throughits exercise of the deliberativeopportunitiesprovidedunder
the so-called proceduralrules of its public life. Both phases of political
life count mightily:for new policy, the first-rounddecision phase; and,
for maturingpolicies, results of decisions, the style and quality of con-
ductof the leadership,andthe phases of retrospectivedeliberation.
David Carr'sdiscussion of the narrativecharacterof everydayex-
perience illuminatessuch a perspective.Individualstake a prospective-
retrospectivestance towardtheirown projects,he says, in orderto, like
a storyteller,select andthen attendto whatcounts, so they are betteren-
abled to negotiatewith the future.Real hindsightis a luxuryindividuals
poised for action do not have, yet they do ruminateon probableconse-
quences of possible courses of actionon an "as if" basis. But real hind-
sight is essential to the storyteller'sposition:it is not unjust,but is nec-
essary to select the frameof the storyitself, actively constitutinga point
of view. All narrationis essentially intersubjectiveand thus social.
Building on Carr,it is interestingto thinkof the democraticcommunity
of inquiry as constructingits own narrativeon the meaning of events
and actions as they flow past. It can decide upon the organizationalfea-
tures of the narrativeand, in a sense, negotiate with the future on the
basis of what is gleaned from the past:it collectively can make the deci-
sions about "beginning-middle-end,suspension-resolution,departure-
return,repetition,and the like."27In such a perspective,the representa-
tive machineryof governmentcan be seen as creatingdifferentkinds of
fora and schedules for political conversationand the craftingof the nar-
rative on public life: the House of Commons, "the greatinquest of the
nation," has long been so conceived. More concretely, one can hypo-
thesize that a constitutionaldesign that ensures retrospectivedelibera-
tion within the representativeinstitutionsmight have the following im-
pact on the public policy process: allow it to be more effectively educa-
tional and exploratorybecause it can be made more systematic;make it
more remedialbecausejudgmentcan be more timely andmore concen-
trated, bringing closer the possibility of corrective action; and more
calm and civil because the agendafor discussion and grievancewill be

27 Carr, Time, 60-65.

TheProblemof DirtyHands in Politics 493

to a largeextent in the handsof the assembly, thus the need for strategic
overstatementis less.
In relation to constitutionalismand the dirty hands problem, it is
useful to study the "politicalmorality"of individualactions in termsof
theircapacityto subvertthe intentionsof actualpoliticalinstitutions,un-
derstoodas rules for guidingpublicbehaviour.One can maintainthatthe
choices political actorsmay make underpressureshouldbe seen as raw
events to be studiedand moralizedin the course of the orderly,constitu-
tionally appropriateprocesses of retrospectivedeliberationand evalua-
tion.The politicalmechanismsof each constitutioncan thenbe examined
for theircapacityto structureretrospectiveroundsof publicdiscussionin
a timelyandtransparent way, andso remedy,lay to rest,or systematically
andcontinuouslyreformulatethe issues embeddedin the choices.
In regard to a certain class of actions that effectively subvert the
exploratoryphases of politicallife butwhich arenot usually regardedas
crimes, Dennis Thompson's concept, that of mediated corruption,is
helpful. For the purposesof this article,the most importantaspect of his
definitionstates that acts that subvertthe democraticprocess by under-
mining deliberationand thus competitionare morally corrupt.Thomp-
son uses the word "mediated" to describe the form of corruptionbe-
cause the actions at issue, undertakenby politiciansin a kind of "loop-
hole" thinking in the context of the system rules, are filtered through
politicalpracticesthatmay in othercontextsbe legitimate:
mediatedcorruptiondiffers from conventionalcorruptionwith respect to each
of these three elements [thatnormallydefine corruption]:(1) the gain that the
politicianreceives is political, not personal,and is not illegitimatein itself, as in
conventionalcorruption;(2) how the public official provides the benefit is im-
proper,not necessarilythe benefititself or the fact thatthe particularcitizen re-
ceives the benefit; (3) the connection between the gain and the benefit is im-
properbecause it damagesthe democraticprocess, not because the public offi-
cial provides the benefit with a corruptmotive. In each of these elements, the
concept of mediatedcorruptionlinks the acts of individualofficials to qualities
of the democraticprocess. In this way, the concept provides a partialsynthesis
of conventional corruption(familiar in contemporarypolitical science) and
systemiccorruption(foundin traditionalpolitical theory).28

Thompson stresses that mediatedcorruption"precludesdelibera-

tion about whether to deliberate." Actors become corruptingagents
when they undertaketo manipulatethe rules for their covert political
gain, "farming" the system to appropriateresources to consolidate
their hold on power. Of course, many analysts before Thompson have
woven moraljudgments into their descriptionsof the founding princi-
ples of particularsystems of democraticrepresentationandresponsibil-
28 Thompson,"Mediated 378.

ity. HermanFiner, for example, says, in regardto British institutions,

thatthe process of deliberationhas to be complementedby a process of
politicalresponsibilityfor the decisions taken,and that,further,this re-
sponsibility for leadershiphas to be a public arrangementthat all can
see, a process externalto the executive that operateswithin the power
nexus of the representativemachinery of government.29Finer comes
within a hair of equatingdemocracywith the mechanismsof responsi-
ble government.Nonetheless, the principlesthat are importantto both
Thompsonand Finer are thatdemocracyis somethingthatis public and
deliberativein its characterand nature.And Thompson has defined a
clear standardof judgment:respectfor deliberation.
Thus, although his conception of mediated corruptionis broadly
consistent with other theories of democracy, Thompson recommends
thatwe might try foundingrespect for our system squarelyupon the de-
liberativeconception, "which prescribesthat officials act on consider-
ations of moral principle,ratherthan only on calculations of political
power." The moralprinciplethatis appealedto is thatit is deeply wrong
to bypass or otherwisecorruptthe democraticprocesses of deliberation
and competition.In this conception, neitherends nor means are legiti-
mate unless they have been addressedin legitimatedeliberativeproce-
dures.One can add thatthis is not the same thingas saying thatall things
thatare chosen throughlegitimateproceduresarejust or necessary,only
thatthe "warrantof the democraticprocess" must be obtainedfor two
reasons:to legitimatea substantialcourse of action-a policy-and to
supportthe health of the process itself. (In this regard,one must again
insist upon the importanceof the retrospectivephase of a political pro-
cess. "Loophole man" can be seen for what he is: a criminal of pro-
cess.) The very idea of constitutionalismcan thusbe seen to incorporate
the idea of deliberativedemocracy,and further,of deliberativedemoc-
racy as exploratory,serial and remedial(thatis, as explicitly retrospec-
tive as well as active or positive).30
Thompson'sidea suggests a standardby which to assess the effec-
tiveness of the machineryof governmentof an actual democraticsys-
tem. The moralityof an individualpolitical actioncan be judged (along-
side other standards)31by whetheror not its agents anticipatedthatre-
29 HermanFiner, "AdministrativeResponsibilityin DemocraticGovernment,"Pub-
lic AdministrationReview 1 (1940/1941), 335-50.
30 Thompson, "MediatedCorruption,"377.
31 We are familiarwith the standardof majorityrule. In his 1987 book, Thompson
develops an argumentto the effect that morallyjustifiable decisions can be as-
sessed againstthreeadditionalsets of standards:generality,autonomyand public-
ity. Verybriefly, "Standardsof generalityrequirethatlegislative actionsbe justifi-
able in termsthatapply to all citizens equally ... Standardsof autonomyprescribe
thatrepresentativesact on relevantreasons ... Standardsof publicity requirethat
an interventiontake place in ways thatcould be justified publicly." See Dennis F.
TheProblemof DirtyHands in Politics 495

trospectiveexplorationof the meaning and effects of the act would oc-

cur, took this appropriatelyinto account, and did not try to evade it or
corruptits meaning. The integratinginstrumentof the individualistic
and structuralistviews of responsibleaction is our priorconception, in-
deed our moraljudgment,of what kind of democraticprocess we want
to encourage.The next challenge is to think about whole political sys-
tems as distinctivein the way they provide and ensure the realizationof
deliberativeopportunities.In what follows, we extend Thompson's at-
tentiveness to the deliberativeprocesses to explore some examples of
political advantage-takingthatinvolve broadstrategiesto hold on to of-
fice (andafflict one's enemies).

Hollis' Treatmentof the CoventrySacrifice

The deliberationstandardas a test of political morality suggests that
contemporary political systems-configurations of institutions and
rules-differ in their capacityfor engaging focused public discussion.
They would thereforealso differ in their characteristicvulnerabilityto
dirty hands decision makers,and might even provide distinctiveincen-
tives to leadersto slip into secretive,consequentialistand self-interested
modes of thinkingand acting. Certainlythe requirementsfor delibera-
tion are different,or at least differentlyconfigured,in differentpolitical
The article by Hollis that was cited earlier makes an almost-
offhand observationabout the British cabinet system which, I submit,
errs in an interestingand useful way in thatit illustratesthe importance
of contingentconfigurationsof rules. The "storygoes," he tells us,
that Churchillwas told duringthe afternoonof 14 November 1940 that there
was to be a massive air raidon Coventrythat night. A word of warningwould
have saved hundredsof lives. Yet no warning was sent. For, if the Luftwaffe
found themselves expected, they would know that their Enigma cypher had
been brokenandthatwas a consequenceworthmore in the wareffort thanhun-
dredsof lives in Coventry.It might seem thatChurchillhad sacrificedhis integ-
rity to the wareffort. But, on reflection,would it not have been self-indulgence
to give the word?Did integritynot demandthe keepingof a widerfaith?32

But, in a note at page bottom, Hollis tersely informs us that "Churchill

was given no such information."The note then dismisses the impor-
tanceof Churchill'sfactuallack of involvementin the episode, in which
his own integrity was nonetheless allegedly at stake. Why? Because,

Thompson, Political Ethics and Public Office (Cambridge:HarvardUniversity

Press, 1987), chap. 4. The concept of mediatedcorruptionis distinguishablefrom
the publicity standardin thatit specifically considersthe form and requirementsof
thesystem, ratherthansimple exposureof a deed.
32 Hollis, "DirtyHands," 391.

Hollis says, "certainly those who decipheredthe German signals did

know; so the case can stand, even if Churchillis not the focus of it."33
The anonymousofficers who did contingentlypossess the information
about the raid could have warnedtheir fellow citizens in Coventry,but
chose not to do so for wartimeemergencyreasons (with which we have
no quarrel here), sacrificed their moral integrity, Hollis says, even
thoughthey had no choice in strategicterms.
But how does this involve Churchill?Hollis even picks up the Cov-
entrydecision a second time. His implicationis clearly thatthe "integ-
rity" dilemma does hold vicariously for Churchill,that the cleanliness
of his hands in the Coventrydeathsis seriously in question.But, again,
why? It seems most likely thatHollis can let his referencesto Churchill
standbecause his own thinkingis immuredin the Britishframeworkof
responsible government:the minister is politically responsible (must
answerin public and must providefor redressor a rationalefor not act-
ing) for the actions of subordinatesin the portfolio.WereHollis Ameri-
can, the concept of "deniability"would certainlyhave enteredhis cal-
culations, because deniability,or the contingent ignorance defence, is
an explicit repudiationof the possibility of vicariousmoralresponsibil-
ity, and thus of political responsibility,at least in an expedient version
of the Americansystem of government.
Certainly,in the British constitution,Churchillwas politically re-
sponsible for the suppressionof a warning by the officers who deci-
phered the German signals-just as though he himself had been the
officer who had contingentlydecided to let the citizens of Coventrybe
struck.But politicalresponsibilityhere means retrospectiveanswerabil-
ity to the House of Commonsand before the public, followed by any le-
gal or electoralconsequences.It is not the same thingas vicariousmoral
responsibility.Churchill'sformalpolitical duty would be to explain the
reasons that the anonymous decision makers could offer for having
overriddenthe rights to reciprocityof the citizens of Coventry,and to
attemptto reconcile the public to such reasoning.(If Coventrywere dis-
cussed under a new war minister, that person would interpretand de-
fend the government's past action in the public discussion.) But
Churchill could only be morally responsible for the blood spilled in
Coventry were it to be shown retrospectivelythat he had personally
known of the raid before it happened,and had personallychosen not to
warnthe city for reasons thatcould not survive scrutiny;thatis, for rea-
sons thatwere self-serving,quitepossibly followed by attemptsto cover
up his bungling. In short, unless his decision making was later seen to
have been irregularandflawed, Churchill'spoliticaltestamentwouldbe
found in the formalpolitical relationsof the time. One can note thatthe
concept of mediatedcorruptionadds the standardthatChurchillwould
33 Ibid., note 5.
TheProblemof Dirty Hands in Politics 497

have been at fault had he in any way evaded or shirkeda duty to make
himself available to consult with his officers-had he acted in such a
way as to keep himself "out of the loop" of information.Thus the tactic
of providingpolitical chiefs with "deniability"can be given a little dig-
nity as the systemic offence of a conspiracyto commitdeliberate negli-

MediatedCorruptionand theAmericanPresidency
The idea of mediated corruptionas action that defeats the established
lines of the system as they are understoodby the main players and by
citizens-action that precludes public deliberation-provides help to
those wishing to grasp some of the dilemmas of the modernAmerican
presidency. One can offer as examples the Watergateand Iran-Contra
scandals and their sequels. PresidentNixon apparentlythoughthe had
the right to order any action that he believed might be instrumentalin
maintaining the safety and stability of American life. This included
breakinginto the offices of the DemocraticNational Committeein the
Watergatebuilding.One reconstructionof a motive is thatNixon andhis
officials honestly thoughtthatinformationwould come into theirhands
thatcould help preservethe stabilityof the country.Anotherinterpreta-
tion is thatthey only wantedto collect partisaninformationto hurttheir
opponentsin the forthcomingelection. (Orboth motives could have fig-
ured.) Nixon himself told David Frost on television that no law is
violated if a president's staff members are simply doing whatever it
takes to implementa presidentialdecision.34The opinion of the Ameri-
can courts was different,for Nixon's officials were tried and sentenced
for criminalactions. Yet Nixon was not impeachedbecause he resigned,
and,further,he was nevertriedfor any crime, because PresidentGerald
Fordpardonedhim prospectively,as it were.
Likewise, both the Reagan and the Bush presidencies were
hauntedby the fear of the revelationof the scope of secret weapon sales
to Iranin exchange for the release of Americanhostages, as well as the
money that was used to make undeclaredwar on the Nicaraguangov-
ernment.PresidentReagan duckedthe question of his own political re-
sponsibilityfor the acts of severalofficials who reporteddirectlyto him,
claiming thathe had never been personallyinvolved in these particular
matters.PresidentBush subsequently,and likewise, long claimed that
he had been "out of the loop" of informationduringthe same events. In
December 1992, at the end of his term of office, with proof of his per-
34 Michael Foley, TheSilence of Constitutions(London:Routledge, 1989), 68. Benn
notes thatto lie for the government'ssurvivalcan be acceptable,if embarrassingto
its supporterswhen found out. Nixon's errorwas to keep on lying when therewere
no grounds for believing that lying was in the public interest (Benn, Public and
Private Morality, 166).

sonal involvementappearingever more likely, Bush short-circuitedthe

possibility of a resolutionof the questionof both Reagan's and his own
involvement,by using the power of presidentialpardonto preventthe
impendingcriminal trial of his associates. In his announcementof the
ChristmasEve pardons, President Bush claimed that the officials he
pardonedhad been motivatedby patriotism,which made their deeds
pure.The motives of those who were prosecutingthe cases, he said, was
opportunisticallyto seize the law in order to criminalize what Bush
called mere policy differences. "Thus did he define law-breakingas a
publicduty, andlaw enforcementas an extensionof partypolitics."35
For a similar dynamic but higher stakes, one can read C. A. J.
Coady's discussion of the Cubanmissile crisis. Coady notes thatPresi-
dent Kennedythoughtthatthe probablegains for his political career-
his need to regainprestige,demonstratecourage,eliminatethe prospect
of impeachmentand avoid Democratic defeats in upcoming congres-
sional elections-justified what he assessed as a risk of between a third
andone half of the situationendingin nuclearholocaust.36
What one can say about this, following Thompson,is that the use
of the presidentialpardonhas forestalleddeliberationand moralization
of the chief executive's actions. The very existence of the power to ex-
tend a full pardonbefore a case is tried may well have incited Reagan
and Bush to use it in this way, and thus begs to be remedied, even
though the actions which its liberal use covered up might eventually
have been judged to have been acceptablein the circumstances.(Mali-
cious prosecutionis a lesser evil, because it at least invites deliberation.)
One might say that the presidentialpardonso-used amountsto a short-
circuitingof the deliberativeevaluationphase of the qualityandcorrect-
ness of the original round of decision making and executive action.
Fromthis perspective,it looks like a system flaw. And if Kennedyreally
felt free to use a nuclearthreatto improvehis chancesfor re-election,as
Coady charges, the lack of a tighter,more formal and much more re-
sourcefulcycle of accountabilityagain looks like a flaw. One sees that
the power to pardonprospectively,combinedwith the solitarynatureof
presidentialdecision making in foreign policy in general, and in the
emergency situationin particular,can effectively trumpconstitutional-
ism. They combine to forestallformalpublic processes of retrospective
deliberation.But the power to pardoncannot put an end to media and
35 Hugo Young, "If This Is the New EthicalOrder,It Stinks,"ManchesterGuardian
Weekly(Manchester),January17, 1993, 5.
36 Coady, "DirtyHands," 373. Lewis Laphamhas identifieda kind of social pathol-
ogy in his nation's life, in thatin manyAmericannarrativesthe violent man is the
hero, and he unerringlydiscovers evil in "even the most rudimentaryattemptsat
civilization." Complementarily,the villains always form part of "the system"
which even the little childrenknow to be corrupt(Lewis Lapham, "BurntOffer-
ings," Harpers,April 1994, 11-15).
The Problem of Dirty Hands in Politics 499

public speculation and thus it conduces to the erosion of the public's be-
lief in the legitimacy of the system.

Mediated Corruption in Canadian Parliamentary Life

But parliamentary systems are no better if not appropriately run. One
can find a rich example of mediated corruption in Canada in the episode
in 1991 when an Iraqi diplomat who had served as his country's ambas-
sador to the United States in the lead-up to the Gulf War was, with un-
common dispatch, admitted to Canada as a landed immigrant within
weeks of his first enquiries. The ensuing media coverage and opposition
interest made the admission a political event of considerable impor-
tance. But the ministers of the Conservative government whose depart-
ments were involved flatly refused to give reasons for the speedy admis-
sion, on the grounds that they had not been personally engaged in the
decision, thus could know nothing of it and could not provide answera-
bility for it-perhaps nothing more than a case of the deniability strain
of the idea flu virus picked up from their American homologues. The
ministers, consulting with the clerk of the privy council (then Paul Tel-
lier) and the other most senior officials in the executive's main support
agency, the Privy Council Office, arranged for an inquiry by a House of
Commons committee, one that would naturally operate as a creature of
the House of Commons, dominated by the large government majority.37
Thus, by allowing the problem to be framed in a such a manner, the
government constituted itself as a dirty hands actor to debase the politi-
cal processes for partisan gain in the following ways: it evaded its duty
(under responsible government) to provide political answerability to the
House of Commons for the rationales behind the original decision; it
sacrificed the rights to naturaljustice of the individuals who were desig-

37 See, for a descriptionof the inquiry,S. L. Sutherland,"The Al-MashatAffair:Ad-

ministrativeAccountabilityin ParliamentaryInstitutions,"Canadian Public Ad-
ministration 34 (1991), 573-603. Governmentdocuments since released to Ken
Rubin, an access investigator,suggest the hypothesis that the Privy Council Of-
fice, having perhapsinitially misinterpretedsome of the technical aspects of the
case and of the modalities available to bring a foreign national into the country,
carrieda battle, which it may have seen as one for its writ in nationalsecuritymat-
ters, into ExternalAffairs, where it easily became focused upon the person of the
associate undersecretary,RaymondChr6tien,because such a focus was attractive
to the Conservativepartisans.But even if PCO did play a role in directingthe parti-
san dynamicin the inquiryandthe subsequentdemoralizationof one of the impor-
tant ministries of government,the political responsibility belongs to the prime
ministerand the other relevantministers.The damage that this episode did to the
Canadianpublic service ethos lost its profile when the Conservativepartywas left
with only two MPs after the subsequentnational election. Its legacy is a public
service convictionthatministerswill henceforthclaim a rightto victimize arbitrar-
ily public servantsratherthan to test theirpolitical skills in providing answerabil-
ity in the representativeinstitutions.

natedto take the blame (and therebyto protectministers)by pretending

that the ministerswere only cooperatingwith "Parliament"in making
theirofficials availablefor questioning,whereas de facto the committee
could not have become seized of the issue withoutexplicit government
managementof the House of Commons' committee agendas;it know-
ingly allowed its MPs to cross-examine aggressively the accused offi-
cials, thus play-actingfor the public in show-trialfashion in orderto le-
gitimatethe "verdict," a verdictwhich was eventuallyand shamelessly
arrivedat by a partisanvote directedby the government,which amounts
to gross manipulationof the system of legitimation;and it knowingly
used "nationalsecurity" rationalesas a shield for itself, and to protect
from scrutiny the government's reasons for which the individual in
questionhad been admitted,and the issues of whether,and how, blame
should be fixed for having mishandledthe case (assuming that it had
been mishandled),actuallysilencing all attemptsto provide a fuller ex-
planationof the event and its administration,one that might have sup-
portedboth the working-level bureaucratsand the merits of the appli-
cant, which latterhad been put in grave question by the government's
steadfastdenialof answerability.
In effect, the government strategy pretendedthat the admission
(past,presentandfuture)of the diplomatto Canadahad in fact been, not
a case that could be discussed on its merits,but a dirtyhands dilemma
(between, perhaps, giving shelter to an enemy national or sentencing
him to death at the hands of his enemies) that belonged completely to
officials, somethingthatis not possible in our system of government.As
has been discussed in relationto the Hollis example,the workingframe-
work of the constitutionprovides that ministersare politically answer-
able in retrospect-are the spokespersons-for everythingthatgoes on
in theirministries,presentand past, whetherthey knew of it at the time
or not. Theirduty is to know of it now. The new theory said, simply, in
an imitationof American-styledeniability,but withoutceding an iota of
Westminster-stylegovernmentpower to manipulateand impose, that
ministers simply could not be expected to speak and to argue in the
House (held politically responsible)for anythingwithin theirportfolios
of which they had not personallyknownat the time.
Then the governmentwent still further.It moved from self-protec-
tion to prosecutionof its enemies. It used the apparatusof coercion to
hide its own handand force a certainline of what it pretendedwas a dis-
covery. In the event, two individuals,one civil servantand one political
officer, were selected for blame by partisanvote. That one of the civil
servants-a career bureaucratof 25 years-who was chosen to be
guilty of havingfailed to inform his minister was the nephew of the
then-leaderof the official oppositionpartywas presentedalternatively,
and conveniently, as the merest of coincidences, or as only to be ex-
TheProblemof Dirty Hands in Politics 501

pected, given the government'spure presumptionas to his loyalties.38

The crime againstthe democraticprocess consists in disarmingand sub-
verting the structureof justice of the formal retrospectivephase of re-
Parenthetically,one can notice thatThompson's concept of medi-
ated corruptionis helpful in thinkingabout the systemic immoralityof
politicians'refusalto accountfor nationalsecurityfunctions.The politi-
cal leadershipis allowed to seal itself off from its own currentsecurity
operationsso thatthese operationsbecome self-managing.In effect, the
kind, scope and amountof activitybecomes essentially what the secu-
rity agentsdecide to supply.

MediatedCorruptionas Fascism (throughCaesarism)in Brecht

The purpose of Brecht's play, The Resistible Rise of Arturo Ui,39 is to
strip the political adventurer,here Ui, of glamour by casting him as a
common gangster.The play shows Ui's rise in the murderousstep-by-
step transpositionto the Chicago vegetable tradeof the chronology and
strategyof Hitler's takeoverbeginning in 1929, culminatingin the 98
per cent popularvote in 1939 for his party.Brecht's play startsin peace-
time, in a world of democraticpolitical structuresoperatingunderlaw,
mostly honourabledecision makers,a free press, mostly good citizens
and business and political elites who know enough to distrust Ui.
Nonetheless, political gangsterism triumphs.The play is therefore a
demonstrationof the concept that political actors who are personally
wholly amoral can commit political crimes that are nonetheless prop-
erly understoodas being in themselves moraloffences againstthe prior
structureof justice. It shows mediatedcorruptionin action.
The result being a foregone conclusion, the interest of the play
comes from the relentless peeling away of the spectator's illusions
aboutthe solidity of the normativestructureof political life. In Ui, those
entrustedwith the highest positions in governmentand society make
their choices on a calculus of personal expediency, rarely thinking
aboutthe honourof the system as a factorin itself. Even when the vege-
table wholesalers and the elite businessmen begin to wake up, the
closest they can come to wisdom is the bromide, "Moralsgo overboard
38 Actually, two memorandahad gone to the current minister of external affairs
through the careeradvice streamof Chr6tien'sshop, both from Tony Vincent, a
senior official, dated March8 and April 2, the lattera reminderalong with press
lines. These memorandawere releasedto Ken Rubinas partof a largerinformation
request.No tracewas found of any warningof the ministerof immigrationbefore
the fact of Al-Mashat's entry,otherthana draftmemorandumthatwas turnedback
to its authorandneverrevised.
39 Bertolt Brecht, The Resistible Rise of Arturo Ui (London:Eyre Methuen, 1976).
The play was only producedin Britainin 1967, and publishedtherein 1976. Later
the BBC madean excellent film with Nicol Williamsonas Ui.

in times of crisis."40Every individual looks for a strong man, rather

than to any process availableunderlaw, to offer resistance to Ui: "We
only hope with Gods help that the bastard/ Some day comes across
some guys thatshow / Theirteeth."41
Before each stage of his rise, Ui retreatsto brood until his plan for
the next strategicphase is ready, and until the situationis ripe. He is a
caricatureof the torturedand sweating strategist,in fact, of the Weber-
ian statesman.As in 1932, when Hitler's party was out of money and
Hitler had to hold back and wait for the ripening of a scandal which
would open the way, Ui, rebuffedby the king-makersof his own world,
sets the simple trapsof greed and complicity that will compromisethe
smug powerbrokers.Even the most noble statesmenturnout to be quite
affordable-the Hindenburgcounterpartdelivers himself body and
soul because he had thought he deserved the gift of a country estate
some time before. He is led to suppressthe capacity of political chan-
nels to deal with Ui, bothbecause he wantsto save his own personalleg-
acy of respect and because he wants to show the heavens morejust than
could be believed were he to allow himself to be seen as the pawn that
Ui has madeof him.
Ui, althoughhis initial conception of the masses had been as con-
sumers and producers,begins to see the usefulness of the mass as a
source of legitimacy-as a Caeasaristelectorateratherthanas citizens,
for citizenship entails a relationof reciprocityand respect between the
leader and the led, somethingthat does not enter into Ui's calculus. So
he courts the vote, and when the electorate comes to its moment of
choice, it is as individualswithouta system andthereforewithouta rem-
edy, and the mass legitimatesUi ("we," "oui"), the beast in itself. Ui
promisesto fulfil his partof the bargain:
PrideI acceptyour thanks.Some fifteen years
Ago, when I was only a humble,unemployed
Son of the Bronx;when following the call
Of destinyI sallied forthwith only
Seven staunchmen to bracethe WindyCity
I was inspiredby an iron will
To createpeace in the vegetabletrade.
We were a handfulthen, who humblybut
Fanaticallystrovefor this ideal
Of peace! Todaywe area multitude.
Peace in Chicago's vegetabletrade
Has ceased to be a dream.Todayit is
Unvarnishedreality.And to secure

40 Ibid., 12.
41 Ibid., 91.
TheProblemof DirtyHands in Politics 503

This peace I haveput in an order

Are not alone in clamouringfor protection!
Thereareothercities: WashingtonandMilwaukee!
And othertowns wherevegetablesaretraded!
Philadelphia!Columbus!Charleston!And New York!
They all demandprotection.And no "phooey!"
No "That'snot nice!" will stop ArturoUi !42

In Brecht's play, democracy(as due process, deliberationandparticipa-

tion) is Ui's victim. Brechtgave the task of guardingthe qualityof polit-
ical life to the collectivity, which here failed because it allowed the insti-
tutions that would ordinarilyhave provided organization-and thus
mobilizationand potency-to public discussion to become ineffective
and corrupted.In other words, Brecht invests the collectivity with the
moralresponsibilityfor the failureof its constitutionalism,precisely be-
cause those who believe thatthey are fated to answerthe call of destiny
(the lone hero leaders) cannot be trustedto restraintheir impulses to
hasten and autonomouslyinterpretthatcall. There is no difference be-
tween Capone-style "protection"in cauliflower and the Nazi regime,
Brechttells us: the apparatusof legitimacy, as Hollis calls the power of
the state, is always vulnerableto being appropriatedby crass adven-
turers,who will, if allowed, turnit into an apparatusof exploitationfor
their own ends, and even demandthat the public see their intemperate
and ignorantgestures as sacrifices. Politics is not about blind trustbut
Thereforelearnhow to see andnot to gape.
To act insteadof talkingall day long.
The worldwas almostwon by such an ape!
The nationsputhim wherehis kindbelong.
But don't rejoicetoo soon at yourescape-
The womb he crawledfrom still is going strong.43

I have arguedthat the problemof dirty hands, as focused on the inten-
tions of the lone politicalactortrappedin a vicious scenario,distortsour
understandingof the natureof politics.
Most obviously, it does not model the most difficultandinteresting
aspectsof politics:those thataspireto democraticactionunderconstitu-
42 Ibid., 95.
43 Ibid.,epilogue, 96.

tionalism. The dirty hands problematicis deceptive in its assumptions

thata sole political actorcan routinelyand somehow properlyengage in
"lesser-evils" choices on behalf of the whole of society, takingon, as a
personalmoralburden,society's role in publicly deliberatinguponhard
choices, at least retrospectively,and, further,dispensing with public
answerabilityfor such choices. This is in partbecause, as Braybrooke
andLindblompoint out, the necessary calculationsto supportsuch hard
choices are in real life subjective,andinformationis scarce, andin large
partonly availableafterthe fact.
Indeed, the illusion of heroic, statesmanlikewisdom successfully
executed in dirty hands situations, is probablypreserved only by the
suppressionof deliberativepolitics in both the decision makingandret-
rospection cycles, as was shown in the examples of side-tracked
answerabilityin Americanand Canadianpolitics. While it may be nec-
essary to bypass deliberativepolitics in emergencies,in principle,such
substantialdebates as may have been avoided in the decision-making
phases, with justificationor not, can still be undertakenafterthe event.
One mighteven say thatpracticalwisdom would suggest thatit is trulya
false economy to suppress the retrospective-deliberativeprocess, be-
cause it representsthe only hope that society has to become educated
aboutits challenges. In short,the purposeof political institutionsis not
exhaustedby the simple fact thatthey can sometimeslimit the arbitrary
use of political power in the first-roundor executive phase, or even that
they might be used to restrainthe scope of government action: they
should facilitate intelligent social problem solving and help form the
characterof citizens by conductingformalretrospectivedeliberationin
a frameworkof orderlinessand civility thatcannotbe appropriatedfor
partisan purposes.44
In systems of responsible government,the apportioningof blame
for previous activity (or inactivity)is the defining political ritualof the
adversarialassembly. Thus, the cycle in which retrospectivedelibera-
tion is undertakenis comparativelyshort or timely, and "the system"
can ideally see the executive forced to respond quite thoroughlyto the
political opposition and the signals from the electorate it may convey
within very short time periods-provided of course that secrecy is
either not imposed or fails. The American system of separatedinstitu-
tions, on the other hand, concentratesits energy on substantialdebate
about the content of the first or outputphase of new political decision
making,and, further,on politically and legally regulatingthe amountof
influence thatthe separatedinstitutions,sharingpower, can have on the
substanceof new policy in each instance. Thus the institutionalplan of
Americangovernmentis to force intenseprospectivediscussion of posi-
44 Stephen L. Elkin, "Constitutionalism'sSuccessor," in Elkin and Soltan, eds.,
A New Constitutionalism,124.
TheProblemof Dirty Hands in Politics 505

tive political options by the device of distributingvetoes widely. In

some areas, such as health care, to mobilize the separateinstitutionsof
the political system to act in concert to undertakea substantiallynew
direction,virtuallyrequiresthatthe whole politically active layer of so-
ciety speak with one will. The first or substantialphase of decision be-
ing so deliberatelyundertaken,with so many proceduralrequirements,
and drawing on all institutions,it is perhapsthe case that formal, sys-
tematic retrospectivedeliberationinside the representativeinstitutions
is less importantthan in cabinet governmentsystems where planning
takes place in closed fora and the executive controls the legislative pro-
gramme.But in areas such as foreign policy and national security, the
US presidentcan act almostunilaterally,andexempthimself andhis of-
ficials from investigation.Nevertheless,retrospectivejudgmentstill oc-
curs in the "big loose" accountabilitysystem outside the frameworkof
representativeinstitutions,anda numberof the intractabledifficultiesof
the modernpresidencyare due to a lack of orderlinessand fair-minded
procedurein such ad hoc evaluation,often pursued equally as "info-
tainment"or for low andeven hystericalpartisanreasons. In parliamen-
tarysystems, on the otherhand,the executivegets power to act unilater-
ally by takingresponsibilityfor anythingthatmay occur in an areathat
is within the scope of a ministerialdepartment:at any given time the
"disaster-bring-forward"mechanism,from which the House of Com-
mons takes its energy, can lay responsibilityfor any state of affairs at
the feet of the government,which in turncan effectively acquirethe ju-
risdictionand legitimacyto act remediallyor in expiation,by accepting
retrospective,currentand prospectiveresponsibilityas a government.It
is perhapsbecause the acceptanceof responsibilitycreates political ca-
pacity that adherentsof the small state battle against the Westminster
design of institutions,and in particularmisrepresentthe doctrineof in-
dividualministerialresponsibilityas a duty to resign ratherthanas what
it is, a dutyto converse.
Perhapsthe most harmfulaspect of dirtyhandspolitics is its bland
acceptance that political leadership, individual or collective, inciden-
tally or deliberately,has a kind of naturalright to deny reciprocity to
membersin good standingof a duly constitutedpolity: to sacrificethem.
It is obvious to anyone thatthe executive's status and positional power
can put into its hands the capacity to sacrifice strategicallymembersof
the groupin "firstround"politics (firstroundat least for the survivors).
As Evan Simpson points out, dirtyhandsproblemsflatly pit expediency
(or sympathy)against rights. In utilitarianthinking, as it is standardly
represented,the welfare of the whole (assuming it can be calculated)
can legitimatelytrumpthe rights of individuals.45That conflicts some-
45 "Willingness to get one's hands dirty in politics expresses an orientationtoward
thinking in termsof welfare at the expense of principlesof right.It is permanently

times exist between the rights of individualsand the needs or expedien-

cies of the group, or between the needs and expediencies of individuals
andthe rightsof a group,is a fact of life. But one can suggest thathere is
wherepolitical institutionscome into theirown. The task,again,is to in-
sist thatthe retrospectivephases of deliberativedemocracyare meticu-
lously executed. Because the power to trumprights of those who are
less well-positioned is conferredon leaders by their control of process
and mechanisms,political actorsmust neverbe allowed to be judge and
jury in theirown case. No system worthits name could be satisfiedwith
the ad hoc privatemoralblame of individualsouls: rather,political re-
sponsibility, located in the structuresand processes of the political
framework,must be engaged, and the citizens must accept their duty,
not to realise substantialjustice in the aftermathof events (for which it
will often be too late), butto ensurethatthe answerabilityphase of polit-
ical life can operate deliberately and calmly, and that the public can
learnfromits sacrifices.
JonathanWolff, in his reflection on Robert Nozick's work, pro-
poses thattheremustbe two elements for a meaningfullife. Nozick, fol-
lowing Kant, is correct, he says, that a meaningful life must be self-
shaping.Dirty hands leadersdeny to othersthe right to shape theirown
lives. Wolff then adds his own condition:a meaningfullife must include
the feature of attachmentsto others, which involves seeing oneself as
partof a networkof humanrelations.46The dirtyhands leader,by defi-
nition, operates quite alone in a supra-ethicalrealm of elitist thinking.
Even if the leadershipcan be excluded from the normalnetworksof hu-
manrelationsthatwould bind it into society, which seems doubtful,and
which Sartrecontests, the networksof reciprocitybetween ordinarycit-
izens that constitute normal peacetime society require active tending.
Deliberationwithin and aboutthe priorstructureof justice is an impor-
tant way of tending these mutualrestraints.Any serious discussion of
how processes of answerabilityare to be arrangedwill also help to clar-
ify and rank (if not to formulate)at least some of the social goals that
such processes are designedto protectandfulfil.
The article also found in Sartre'splay, Les mains sales, a demon-
strationthatthe will of one actor(individualor corporate)cannotunilat-
erally impose political meaningfor a society. The meaningof a particu-
lar politically intended action, rather,is not a pure philosophicalcon-
cept, butcan be puzzling even to the actors,who may havehad confused
motives, and is ultimatelylost to the protagonist,being negotiatedin so-

subject to question from the alternativeorientationbut not to refutation"(Evan

Simpson, "Justice,Expediency and Forms of Thinking," abstractof a paperde-
liveredatthe "DirtyHands"Workshopat YorkUniversity,December12, 1993).
46 JonathanWolff, RobertNozick:Property,Justice and the MinimalState (Oxford:
Polity Press, 1991), 28-30.
TheProblemof Dirty Hands in Politics 507

ciety. The escape from the charismaof arbitrarypower and elitism-

from the Faustiandramain which those who believe they are destinedto
write history will willingly undertaketo sacrifice others-is the (only)
relative safety of constitutionalpolitics. As Brecht's almost stridently
anti-romanticplay demonstrates,a democraticpublic has in any case no
alternativebut to tend to its political processes and its claims to reci-
procity and respectful public conversation,including formal, patient,
thorough,unsentimentaland intelligentlycharitableconversationabout
past deeds many think would be better forgotten:Ui is the alternative.
Thus "the guys" thatshow theirteeth must do so in a timely and deter-
mined fashion, while the living still matter,and within formal, public,
political structures.