Moral Coercion

volume14,no.11 may2014

Moral Coercion

Saba BazarganUniversity of California, San Diego

© 2014SabaBazarganThis work is licensed under a Creative Commons

Attribution-NonCommercial-NoDerivatives 3.0 License. <www.philosophersimprint.org/014011/>

1. Introduction1

InthispaperIwillanalyzeatypeofconductexemplifiedbythefol-lowingtwocases:

HOSTAGE

Avillaincrediblythreatenstokillfiftyinnocenthostagesshehastakenunlessyoukillthevillain’senemy—anin-nocentagainstwhomthehostage-takerholdsanirratio-nalgrudge.Ifyoudonotkillthevillain’sinnocentenemy,shewillremainunharmedbythevillain,butthehostageswilldie. Ifyoudokill thevillain’s innocentenemy, thenthehostageswillbereleased,unharmed.

Thevillainhasputyouinasituationwhereyouhaveamoralreasontocommitaprotantowrong—namely,killinganinnocent.Thevillaindoesthisbylevelingacredibleconditionalthreat:ifyoudonotaccedetoherdemandthatyoukilloneinnocent,shewillcommitamorallyworseharm.Comparethiscasetothefollowing:

SHIELD

Avillainwishes tokillan innocentenemyofhers.Sheknowsthatifshetriestodoso,youwillshoother.Soshegrabs three children and uses them as human shields;theonlywayforyoutostopthevillainfromkillingtheinnocentisbyshootingthroughthechildren.Ifyoudonot shoot the villain, her innocent enemywill die, butthethreechildrenwillbeallowedtogofree.Ifyoushootthe villain, her innocent enemywill remainunharmedbythevillain.

1. IthankCraigAguleandSamRicklessforinvaluablecriticismsofanearlierdraft.Ialsoreceivedhelpfulfeedbackinpresentingamuch-abbreviatedver-sionofthispaperattheannualmeetingoftheSocietyforAppliedPhilosophyin2013.

ImprintPhilosophers’

sababazargan Moral Coercion

philosophers’imprint –2– vol.14,no.11(may2014)

mitigateherresponsibility forwhatshedoes,andis thusnotitselfabasisfordiminishedliability.

2.The Badness Question: Intuitively,accedingtomoralcoercionallows “evil to succeed”.Howdowemake sense of thisintuition?Iwillclaimthatwecanmakesenseofthisintu-itionbyarguingthatC1’smalignintentionsaremanifestinC2’sactionswhenthelatteraccedestoC1’swishes,there-byaffectingtheweightthattheresultantharmsoughttoreceiveinourdeliberations.

Iwillbeginbydevelopingaprescriptiveaccountofmoralcoercion.Iwillthendiscusswhenandwhymoralcoercioniswrongful,beforeturningtotheLiabilityandBadnessQuestions.Butbeforebeginning,Iwillmakeseveralpreliminarypoints:

i. IwillassumethattherearethreewaysinwhichanagentA canfostersomeevente.Shecancommite,shecanenablee,orshecanallowe.Iwillleave‘commit’unanalyzed,ex-cepttosaythatifAcommitse,thenthereisnootheragentwho, subsequent toA’s act, causally contributes to e inawaynecessaryorsufficientfore’soccurrence.IfAenablese, thensheprovidesanotheragentwiththemeanstocommite,whodoesso.IfAallowse,shehasthepowertopreventanother agent from committing e but refrains from exer-cising this power. These are not full-fledged analyses of‘commit’, ‘enable’,and‘allow’;theyareinsteadsimplifyingassumptionssufficientfortheanalysisofmoralcoercionIdevelophere.

ii.Iwillassumethattheintention/foresightdistinctionismor-ally relevant in that,all thingsbeingequal, it ismorallyworsetocommitaharmintentionallythanitistodosocollaterally (i. e., foreseeably but non-intentionally). Ialsoassumethatthecommission/omissiondistinctionis

Inthiscase,thevillainhasputyouinasituationwhereyouhaveamoralreasontoallowher tocommitawrong—namely,killinganinnocent.

Though in both examples you face a dilemma, there are differ-ences inhow thedilemma is imposed. InHOSTAGE, thevillain isimposingaconditionalthreat.InSHIELD,thevillainisnot(evenim-plicitly) imposingaconditionalthreat.Rather,thevillainintention-allymakes it pragmatically impossible for you to stopherwithoutmakingthingsworse.

Despitethesedifferences,thereisacommonelementthatallowsaunifiedmoralanalysisoftheseexamples.Insuchcasesawrongdoer(C1)intentionallydeniesanagent(C2)theoptionofpreventingbothoftwodistinctsetsofharms(φandψ)frombefallingothers;C1doesthisinordertoprovideC2withanincentivetocommitorallowthelesser of the twoharms (φ), thereby achievingC1’s goal. I call this“moralcoercion”.

Itisclearmoralcoercionisworthyofconsiderationinitsownright,consideringitsprevalence,especiallyinwar.Hostage-taking,certainformsofterrorism,andtheuseofhumanshieldsarejustafewexam-plesofmorallycoercivetacticscommonlyusedinwarfare.Yetmoralcoercionremainsunder-theorizedinnormativeethics.Iwilldevelopananalysisofmoralcoercionbyaddressingtwosetsofquestionsthatwillhelpresolvehowweoughttorespondtomoralcoercion:

1.The Liability Question:Towhatextent,ifatall,isC2morallyli-ablefortheharmsshefosterswhensherespondstomoralcoercionbycommittingorenablingthelesserevil?Morespecifically,by acceding toC1’swishes and therebyocca-sioningharmtothird-partyinnocents,isC2morallyliabletodefensiveandcompensatoryharms?IwillarguethatC2canindeedbemorallyliablefortheharmsresultingfromdoingasC1wishes,evenifC2 ismorallyobligatedtoac-cede.ThefactthatC2iswrongfullycoerceddoesnotitself



C1commitsanactofmoralcoercionagainstC2ifandonlyifthefollowingconditionshold:C1 intentionallyputsC2 inaposition inwhichC2mustchooseexclusivelybetweenoneoftwoharms,φandψ,whereto“choose”aharmistocommit,enable,orallowthatharm.IfC2choosesφ,aharmwillbefallathirdparty,P1.IfC2choosesψ,aharmwillbefalladistinctthirdparty,P2.Fromanimpartialstandpoint,ψismorallyworsethanφ.C1foreclosestheconjunctiveoptionof~φand~ψ,inordertomotivateC2intochoosingthelesserharm,viz.,φ,whichC1knowsC2willhavea(perceived)moralreasontodo.

A difference betweenHOSTAGE and SHIELD is that the formerisanexampleof “active”moralcoercionwhereas the latter isanex-ampleof“passive”moralcoercion.Incasesofactivemoralcoercion,C1putsC2 in apositionwhere shemust choosebetweenφandψ,by threatening to commitψunlessC2choosesφ.This iswhathap-pensinHOSTAGE:C1crediblythreatenstocommitψ—i. e.,tokillthehostages—unlessC2commitsφ—i. e.,killsC1’s innocentenemy. Incasesofpassivemoralcoercion,C1putsC2 inapositionwhereshemust choosebetweenφandψ—butC1does thisnotby levelingaconditional threat,butbymaking itpragmatically impossible forC2tochooseneither.ThisiswhathappensinSHIELD:C1intentionallyputsC2inapositioninwhichtheonlywaytopreventC1fromkillinganinnocentisforC2tokilltheinnocentsthatC1isusingasahumanshield.Hereisanotherexampleofpassivemoralcoercion:

ALLEY

Thirtychildren,separatedfromtheirclassduringafieldtrip,findthemselvesinanalley,atadeadend.C2,whohasalsoaccidentallyturnedintothealley,isseveralme-tersfromthechildren.AndseveralmetersfromC2isaninnocentwhomC1hasbeenattemptingtokillforsometime.C1isontopofthebuildings,abovethealley,andcanseeeveryonebelow.Shehasabomb,butdoesnothaveaclearshotatherenemy.However,shereasonablyguessesthatifshethrowsittowardthechildren,C2will

morallyrelevantinthat,allthingsbeingequal,itismor-allyworsetocommitaharmthanitistoallowittooccurthroughinaction.

iii.IwilladdresstheLiabilityandBadnessQuestionsinbroadlyconsequentialistlanguage.ThisisnotbecauseIthinkthatsome version of consequentialism provides the ultimategroundsforwhatismorallyrightandwrong,butratherbe-causeconsequentialismhasexpressivepowerthatmakesspeaking in its terms especially convenient; any moral features of actions canonically emphasized in deontologi-cal accounts of morality (such as the intrinsic value of the act committed, the relevance of the commission/omissionandtheintention/foresightdistinctions,the intrinsic value of rights, agent-centered restrictions and permissions, etc.) are expressible in consequentialist terms, even if they cannot be grounded in a consequentialist theory.2 Accordingly, in de-termining whether one ought to accede to moral coercion, I will say that we ought to do a “proportionality calculation” in which we weigh the moral benefits against the moral costs, where features morally relevant from the personal standpoint are included in this proportionality calculation.

2. Varieties of Moral Coercion

HereIexplainingreaterdetailwhatmoralcoercionis.Indoingso,Idistinguishvarioustypesofmoralcoercion.Theresultingtaxonomyisnotcomprehensive.Iwilllimitmyselftodrawingthosedistinctionsthatwillultimatelyrevealhowmoralcoercionfunctionsatthemostfundamental level, which helps answer the Liability and BadnessQuestions.Iwillnotconsider“non-central” typesofmoralcoercion,exemplifiedattheendofthissection.

2. Foradefenseofthisview,see(Portmore,2007).Formoreonthepossibilityof “consequentializing”non-consequentialistmoral theories, see especially(Dreier,1993)and(Louise,2004),butalso(Schroeder,2007).



HOSTAGE2

C1 isattemptingtorobabank.Shetakesadozen inno-cent hostages and credibly threatens to kill all of themshouldC2interferewithherplanstorobthebank.

Inthisexample,thelivesoftheinnocentsarebeingusedtomorallycoerceC2intopermittingawrongfulact.SinceC1iscoercingC2intorefrainingfrompreventingaharmfulact(ratherthanintocommittingaharmfulact),itisanexampleinwhichφiscoercer-enactedratherthan coercee-enacted.C1motivates the desired conduct—the omis-sion—bythreateningtomakethingsgomorallyworseifC2refusesto comply.This is in contrast toHOSTAGE, inwhichC2 is coercedintodoingC1’sdirtyworkbyactuallykillingC1’sinnocentenemy.Butwhether the conduct incentivized is a commission or an omissionmakesnodifferencetowhethermoralcoercionhasoccurred.

CompareHOSTAGE2withSHIELD.BecauseinSHIELDC1isco-ercingC2 into refraining from preventing themurders of her inno-cent enemy,φ is coercer-enacted, as inHOSTAGE 2. But unlike inHOSTAGE2,ψwouldbecoercee-enactedinSHIELDshouldψoccur;preventing themurder of the innocent enemy requires actually kill-ing thehumanshield,ratherthanmerelyrefrainingfrompreventingtheirdeaths.Also,unlikeHOSTAGE2,SHIELDisanexampleofpas-sivecoercion:C1deniesC2theconjunctiveoptionofpreventingthedeathsoftheinnocent’senemywithoutkillingthehumanshield—andC1doesthiswithoutconditionalizingherconducttoC2’sresponse.3

3. Notethatcoercee-enacteddeterrentsarenot limitedtopassivemoralcoer-cion.Thefollowingexampledemonstratesthis:

BOMB C2,abomb-maker,hasnegligentlylostabomb.C1knowsitslocation.Sheprovesthatitisinapopulatedarea,andthatwhenitgoesoff,manywilldie.SheclaimsthatshewillnotrevealthelocationofthebombtoC2unlessC2killsC1’senemyfirst.

Here the lesserharm,φ, is thedeathofC1’s innocent enemy.Thegreaterharm, ψ, is the death of the many innocents resulting from the bomb’s

run toward it, grab thebomb, and throw it in the onlydirectionavailable tosave thechildren: towards thevil-lain’sinnocentenemy.Thiswillachievethevillain’sendofkillinghertarget.

As inSHIELD,C1 ismakingnodemandofC2—noteven implicitly.ShedoesnotconditionallythreatentokillthechildrenasameansofmotivatingC2intokillingC1’senemy.AndC1’sactions,subsequenttothrowingthebomb,donotconditionallydependonwhatC2choosestodo.Instead,C1ensuresthatψwilloccurunlessC2choosesφ;andC1doesthisasameansofmotivatingC2intocommittingtheactthatachievesC1’saims.

Partofwhatcharacterizesallcasesofmoralcoercion,bothactiveandpassive,isthatC1intentionallydeniestoC2theconjunctiveop-tionofchoosingboth~ψand~φ—C2canonlychooseone.Theonlydifferencebetweenactiveandpassivemoralcoercionishow C1fore-closestherelevantconjunctiveoption.ThedifferencebetweenactiveandpassivemoralcoercionisadifferenceinC1’stactics.

Wecan furthercategorizemoral coercionbydistinguishingcasesinwhichC1aimstomanipulateC2intocommitting an act fromthoseinwhichC1aimstomanipulateC2intorefraining from preventing an act.Thesecasesdifferwith respect towhocommits the lesserharm, i. e.,φ.Likewise,wecandistinguishbetweencasesthatdifferwithrespecttowhetheritisC1orC2whowouldcommitthegreaterharm,i. e.,ψ,shouldC2refusetocomplywithC1’swishes.WhenC1commitsaharm,itiscoercer-enacted.WhenC2commitsaharm,itiscoercee-enacted.

Forexample,inALLEY,ifC2savesthelivesofthechildrenbyin-terceptingandthrowingthebomb,shewillbetheonewhodoesthekilling—thatis,shewillbetheonewhocommitsφ.IfC2choosesnot tointerceptandthrowthebomb,shewillhaverefrainedfromprevent-ingC1fromcommittingψ.Accordingly,ALLEYisanexampleofmoralcoercioninwhich,ifφoccurs,itiscoercee-enacted,andifψoccurs,itiscoercer-enacted;mutatismutandis forHOSTAGE.Comparethosewiththefollowingcaseofmoralcoercion:



typeof reasonoperative inC2’sdeliberationbetweenφandψ.Theoperativereasonsaremoral,asopposedtopragmatic;thisiswhatdis-tinguishesmoralcoercionfromnon-moralcoercion.ButwhatifC2’spragmaticandmoralreasonsconvergeonthesameoption?SupposeC1kidnapsC2’schild,whomC1threatenstokillunlessC2robsabankandgivesthemoneytoC1.Doesthiscountasmoral coercion?EvenifC2ismorallyobligatedtoaccedetoC1’sdemands,itisdoubtful—pre-sumingsheisapsychologicallytypicalparent—thatsheismotivatedbymoralreasons.Rather,sheismotivatedoutofwhatC.D.Broadfa-mously calleda “self-referential altruistic” concern.4 Sucha concern,despite that it isnotovertlymoral, ismorallyappropriategivenherrelationtoherchild.Thisistomaketheoft-notedpointthatspecialre-lationsgroundingmoralreasonsneednotfunctionastheagent’smoti-vatingreasons.5Inthiscase,doesC1’scoercivethreatcountasmoralornon-moral,giventhatC2hasmoralreasonstoaccedebutismotivatedby reasons of practical rationality? The account ofmoral coercion Ihaveoutlinedprovidesnodefinitiveanswer.But this is, Ibelieve,astrengthoftheaccount:thelinebetweenmoralandnon-moralcoer-cionought tobeblurrypreciselybecausethelinebetweenmoralrea-sonsandreasonsofpracticalrationalityisblurry.

3. The Wrongfulness of Moral Coercion

Nowthat Ihaveoutlinedwhatmoralcoercion is,wecanask:Whatmakeswrongfulinstancesofmoralcoercionwrongful?Inansweringthis question, it is helpful to consider non-moral coercion.6 In cases

4. (Broad,1930,pp.54–55)

5. Seeespecially(Railton,1984).

6. Thereareatleastthreetypesof(non-moral)coercion,broadlyconceived.In“act-negating”coercion,C1physicallyforcesC2tocommitφ.InsuchacaseC2literallyhasnochoicebuttobeaninstrumentinthecommissionofφ.In“autonomy-negating”coercion,C1’s threatputsC2underduress,of thesortthatmakesitpsychologicallyimpossible(inamodallyweaksense)forhertorefusetoacquiesce.See(Frankfurt,1973,p.78).In“compossibility-negating”coercion,C1providesanincentiveforC2(whoisnotunderduress)tocom-mitφbyensuringthatsomeotherevent,ψ,willoccurifC2refusestocommit

Differentiating types of moral coercion along these two dimen-sions—activevs.passiveandcoercer-enactedvs.coercee-enacted—al-lows us to appreciate precisely how the use of involuntary humanshieldsandtheuseofhostagescompare:theformerispassiveandthelatterisactive,andthedeterringactintheformeriscoercee-enactedwhereas that of the latter is coercer-enacted. Though both count asinstancesofmoralcoercioninthatC1isforeclosingthepossibilityofboth~φand~ψasameansofmorallymotivatingC2 intochoosingφ—whichisC1’sgoal—thedifferencesbetweenthemareimportant;Iwillexploretheminsections4and5.

Aspreviouslynoted,Iamsettingasidenon-centralcasesofmoralcoercion. For example, I am setting aside instances inwhichC1de-ceivesC2intojustifiablybutmistakenlybelievingthatagreaterharmwilloccurunlessC2commitsalesserharm.Isetasidesuchcases,notbecauseIdonotthinkthattheyshouldbeanalyzedundertheaegisofmoral coercion,butbecause I think that addressing instancesofsinceremoralcoercion isaprerequisite forananalysisof insinceremoralcoercion.

I am also setting aside “partial”moral coercion. These are caseswhereeitherφorψdoesnotwrongathirdparty.SupposeC1prom-isestoprovidesuppliesforfaminereliefeffortsontheconditionthatC2harmC1’sinnocentenemy.Thiscountsaspartialmoralcoercion(assumingthatrefrainingfromgivingtocharitydoesnotwrongthosewhoareinneed).Likewise,supposeC1threatenstokillaninnocentunlessC2 gives the villain a thousand dollars.Acceding toC1’s de-mand does notwrong a third party; accordingly, this is also an ex-ampleofpartialmoralcoercion.ThisisincontrasttoHOSTAGEandSHIELD,inwhichbothoptionsavailabletoC2imposeawrongonathirdparty.ThoughIwillnotconsidercasesofpartialmoralcoercionexplicitly,theaccountIdevelopwillhaveimplicationsforit.

Before discussingwhat grounds thewrongfulness ofmoral coer-cion,oneissueremains.Whatpartlycharacterizesmoral coercionisthe

explosion.Thisisanexampleofactivemoralcoercioninwhichψwouldbecoercee-enactedratherthancoercer-enacted.



Iwillcallthetacticofintentionallyputtingsomeoneinapositionwherehergoalsbecomeself-undermining“hacking”heraims.8Hack-inganagent’saimsdoesnotnecessarilywrongher,inthatwedonothaveafundamentalrightagainstaim-hacking.Buttreatingsomeoneinthiswaybecomesadistinctwrong—notjustinstrumentally,butinitself—whenitisusedinfurtheranceofanendthatwrongstheagent.Ingeneral,usinganagent in furtheranceofwronginghertreatsherwronglyoverandabovethewrongfurthered.9Andusinganagent’sown aimsasameansofwrongingherisanespeciallyiniquitouswayoftreatingher.

Extant accounts of coercion are incomplete in that theymiss thenecessaryrolethataim-hackingplaysinexplaininghowwrongfulcoer-cionwrongsitsvictim.Forinstance,somearguethatwrongfulcoercionconsists in impermissibly constraining C2’s deliberative options,10 ordenyingherthestandingtolegitimatelydemandthatthecoerceraban-donherintention,11orimpermissiblyattachingacosttoanoptionshealreadyhadandtowhichshehasaright.12Buttheseaccountsunder-describehowC2iswronged.TreatingC2inthesewaysiswrongalsobecauseitinvolveshackingheraims.AdoptingthistacticinfurtheranceofwrongingC2 itselfmaltreatsher, therebycompoundinganalreadyexistingwrong.Anaccountmustappealtobothsourcesofwrongstoexplainfullywhatmakeswrongfulinstancesofcoercionwrongful.

TheaccountofwrongfulcoercionIhavepresentedis,then,open-ended.Itdoesn’tprovidenecessaryandsufficientconditionsforwhen

8. TocountashavinghackedC2’saims,C1neednotwishthatC2actuallyac-cedetoC1’sthreats.ItisenoughifC1intendsforC2tobeinapositionwhereherself-interestedcommitmentsmakeherworseoffthanshewouldbeifshedidn’thavethosecommitments.

9. Thisgeneralideaisfoundelsewhere,suchasinThomasNagel’saccountofthewrongnessofusingchemical,biological,and incendiaryweapons,andDavidSussman’saccountofwhatmakestorturewrongful(see(Nagel,1972)and(Sussman,2005)).

10. See(Shaw,2012).

11. See(Pallikkathayil,2011).

12. See(Wellman,2005,pp.132–138).

of bothmoral andnon-moral coercion,C1disincentivizesnon-com-pliancewithherwishesbyattachingacost toC2’snon-compliance.Thedifferencebetweenmoralandnon-moralcoercioniswherethosecosts lie. Incasesofnon-moralcoercion,C1 forcesC2tochoosebe-tweenperceivedcoststoher owninterests,whereasincasesofmoralcoercion,C2hastochoosebetweenperceivedcoststotheinterestsofthird-partyinnocents.7Determiningwhatthewrong-makingfeaturesofwrongfulnon-moralcoercionarecanhelpdeterminewhatmakesmoralcoercionwrongful.(Iwillrefertocasesofnon-moralcoercionsimplyas“coercion”.)

I do not here attempt to provide necessary and sufficient condi-tions forwhencoercion iswrongful. Instead, Ipresentaparticularand importantnecessary conditionofwrongful coercionwhichex-plainsthesenseinwhichsuchcoercionusesitsvictim.OntheviewIdefend,C1usesC2byturningtheteleologicalstructureofC2’sgoalsonitshead.C1putsthevictiminapositionwhereC2’sgoalsbecomeself-undermining in that she would better-achieve her own goals if she didn’t have them. That is, by having such goals,C2worse-achievesthem,asaresultofC1’sinfluence.Acoerciveact,then,usesitsvic-timbyputtingherinapositionwherehercommitmentsfurthertheachievementoftheoppositeendstowhichthosecommitmentsareteleologicallydirected.

Considertheexampleofblackmailinganadulterer.Iftheadultererdidn’thave theaimofkeepinghisadulterousconductprivate, thentheblackmailer’sattemptatblackmailwouldfindnopurchase,sincetheadultererwouldhavenoincentivetoaccedetotheblackmailer’sdemand.It istheadulterer’sowngoalofkeepinghisadulterysecretthatmakeshimworseoff.

φ,therebydenyinghertheoptionofbringingaboutboth~φand~ψ.ThisisthetypeofcoercionIamconcernedwithhere.

7. “Mixed”casesarealsopossible,inwhichC1threatenstoharmathirdpartyunlessC2 complieswith demands detrimental toC2’s interests. Though Iwillfocuson“pure”cases,wherethecoercioniswhollyprudentialorwhollymoral,theaccountIdevelopwillbeapplicabletomixedcasesaswell.



shewant togetwet,which isnecessary to retrieve thelife-jacket. So she pushes C2 into the choppy water,knowing that C2 cannot swim and that C2will conse-quentlybemotivated,onpainofdrowning,tograspanddon the life-jacket.Oncesheclimbsbackonboard,C1willhaveherlife-jacket.

In this example, C1 does not conditionally threatenC2 in anyway.There isnocoerciveproposal sheputs toC2, even implicitly.14AndC1’sactions,subsequent topushingC2,arenotsensitive tohowC2choosestorespond.C1hassetinmotionaseriesofeventsoverwhichshehasnofurthercontrol.Ofcourse,C2mightchoosetodrown.ButC1knowsthatthisisaveryunwelcomeprospectforher;thisiswhyshepushesherinthefirstplace—toforcehertochoosebetweentheexclusiveoptionsofretrievingthelife-jacketanddrowning.

Passivecoercionisimportantlysimilartoactivecoercion,whichin-volvesthepronouncementofconditionalthreats.TheybothimposeaforcedchoiceonC2,inwhichbothoptionsworse-achieveC2’saimsrel-ativetotherelevantalternative(whichisusuallythestatusquoante).15 ThechoicecontrarytoC2’spreferredoutcomeimposesagreatercostonC2 than thealternative, therebydisincentivizingnon-compliancewithC1’spreferredoutcome.PartofwhatcharacterizesbothactiveandpassivecoercionisthatC1deniestoC2theconjunctiveoptionof~ψ

14. Forthesereasons,onemightprotestthatpassivecoercionisnotaspeciesofcoercionatall.Somemightwanttorestricttheconcept‘coercion’toactsthatinvolvepronouncingconditional threatsofacertainsort. Iamhappywithsuchaview,so longas itsproponents recognizea fundamentallyunifyingfeaturebetweenthetwotypesofacts:theybothinvolveusingtheagentbyhackingheraims.

15. The problem of determining what this relevant alternative is—i. e., whatthe“baseline”isagainstwhichwedeterminewhetherC2’soptionsmakeherworseoff—haspersistedsinceNozickdiscussedit(Coercion,1969).Oneop-tion(whichAlanWertheimertakes(Wertheimer,1987))istoabandonanon-moralbaselineinfavorofamoralizedone,accordingtowhichbothφandψmakeC2worseoffthansheoughttobe.TheAim-HackingConditionisalsobuck-passinginsofarasIclaimthataim-hackingwrongsitsvictim(notjustinstrumentallybutinitself)onlyinfurtheranceofanendthatwrongsher.

coercionwrongs its victim.Rather, it says thatwhatever accountofwrongful coercionwe adopt, itmust include the claim that part ofwhatit istocoercesomeoneistoengageinaim-hacking—andthattreating someone in thiswaybecomeswrongful in itselfwhen it isusedasatacticinfurtheranceofwrongingherinsomeotherway,asspecifiedbyacandidatetheoryofwrongfulcoercion.Thisisanimpor-tantnecessaryconditionofcoercioninthisrespect:adesideratumofanyaccountofwrongfulcoercionisthatitidentifiesitswrong-makingfeaturesat leastpartly in themanner inwhich thecoerceruseshervictim.AndIclaimthatwrongfulcoerciondoessoinpartbyputtingC2inapositionsothatherowngoalsbecomeself-underminingasameansofwrongingC2.Iwillcallthisthe“Aim-HackingCondition”ofwrongfulcoercion.13

TheAim-HackingConditionofwrongful coerciongeneralizes tocasesofpassivecoercionaswell.Considerthiscase:

BOAT

C1andC2areonaboat.Agustofwindhasblownover-boardC1’sonlylife-jacket.Sheisacautiouspersonanddoesnotwant tocontinuewithout it.Butneitherdoes

13. The Aim-Hacking Condition helps dissolve the paradox of blackmail. Ac-cordingtotheparadox,itiswrongforC1tothreatentorevealC2’sinfidelityasameansofcoercingmoneyfromher,eventhoughC1ispermittedtorevealtheinfidelitywithoutattachingacosttohersilence.Soitseems(contrarytoabasicaccountofcoercion(see,e. g.,(Haksar,1976))thatthewrongfulnessofactingon the threatcannotexplain thewrongfulnessof the threat itself(see (Lindgren, 1984) and (Berman, 2011)). James Shawdefends the basicaccount of coercion by arguing that the cost thatC1 attaches toC2’s non-complianceconstitutesan impermissiblesanctionsince itmanifestsamor-allyproblematic“disregard”forC2,inthattheharmC1threatenstocauseisnotoffsetbythevalueoftheendsthatC1furthersinsodoing(Shaw,2012).But Shawadmits that if revealing the infidelity absent thedemandwouldbemorallydiscretionary,thenblackmailingC2wouldnotbeawrongfulin-stanceofcoercion(185).HeinsteadappealstoC1’sproblematicdisregardforC2toexplainhowC1wrongsC2insuchcases.ButgroundingthewaythattheblackmailerwrongsC2byappealingtoakindofimpermissibleinfluenceother thancoercionleavesoutamorallyrelevantaspectoftherelationshipbetweenthem.TheaccountIpresentfillsthegapinShaw’saccount:hackingC2’saimsinfurtheranceofanendthatwrongsherimpermissiblycoercesher.



wrongC2orthird-partyinnocents.Iwillstartwiththelatter,sinceitsexplanationissimpler.

ConsidercasessuchasSHIELDorHOSTAGE.Inthesecases,C1en-dangersthird-partyinnocents.Butthisaloneisnotnecessarilywrong-ful.Itcanbepermissibletoendangerthosewhohavearightnottobeharmedifdoingsohasasufficientlyhighprobabilityofyieldingasuf-ficientlyimportantmoralgood,andiftheprobabilitythattheywillac-tuallybeharmedissufficientlylow.ButinSHIELDandHOSTAGE,C1endangersthird-partyinnocentsinfurtheranceofachievingawrong-fulend.Nomatterhowlowtheprobabilityisthattheinnocentswillactuallybeharmed,endangeringtheinnocentswrongsthemifitisinfurtheranceofawrongfulend.Sotherearetwofactorsundergirdingthe impermissibilityofendangering the innocents:1) it is impermis-siblequameanstotheachievementofsomefurtherwrongfulend,and2)itisimpermissibleinsofarasitendangerstheinnocents.Itisdueto1that2violatestherightsoftheinnocents.

ThatthecoercioninHOSTAGEandSHIELDwrongsthird-partyin-nocentsdoesnotexplain,however,howorwhyC2iswronged.Afterall,threateningtoharminnocentsdoesnotitselfviolatetherightsofC2(unless,ofcourse,shehasagent-relativeinterestsinthewelfareofthoseinnocents).Wecanappeal,then,toextantaccountsofwrongfulcoerciontoexplainhowC1wrongsC2.Atheory(suchasWellman’sorShaw’s)mighttellusthatinHOSTAGEC2hasarighttorefrainfromkillingC1’sinnocentenemywithoutsanctionsattachedtothatoption.Likewise,inSHIELD,suchtheoriessaythatC2hasarighttodisableC1withouttherebykillinginnocentbystanders.Butthisunder-describeshowC1wrongsC2inthesecases.AsNancyDavisputsit,we“thinkofourselvesasinstrumentsofevilwhenweare‘blackmailed’toinflictpainorcausedeathsatthebiddingofevilmen.Andwefindtherole‘instrumentofevil’anespeciallyrepugnantone,anassaultonourdig-nity,andathreattoourstatusasautonomousmoralagents”.18 TerrenceMcConnellmakesasimilarpoint:

18. (Davis,1980,p.202)

and~φ—C2canchooseonlyone.Theonlydifferencebetweenactiveandpassivecoercionishow C1deniesthisconjunctiveoption.

Partofwhatexplainswhatmakesparticularinstancesofpassiveco-ercionwrongfulisthattheyinvolveintentionallyputtingsomeoneinapositionwheresheworse-achievesherownaimsasaresultofhavingthoseaims—andsheisputinthispositioninordertoachieveaimsthatwrongher.InBOAT,forexample,pushingC2overboardwouldnotserveC1’saimifC2didn’thavethegoalofstayingalive.Andtheactofretrievingthelife-jacketbypushingC2overboarditselfwrongsC2.SoC1wrongsC2twiceover:byforeclosingoptionstowhichC2hasaright,andbyhackingheraimsinfurtheranceofthatend.

SotheAim-HackingConditionhelpsexplainhowboththevictimofwrongful active coercion and the victim ofwrongful passive co-ercionaremistreated; theaccount, then, treats fundamentallyalikecasesalike.Thisisnottosaythattherearenoimportantmoraldiffer-encesbetweenactiveandpassivecoercion.16Rather,thepointisthattheyshareanimportantsimilaritywhichtheAim-HackingConditionhelpsreveal.17

Wearenow in a position to better seewhatmakeswrongful in-stances of moral coercion wrongful. In general, moral coercion iswrongfulwhen it either a)wrongsC2or b)wrongs the third partywhosewell-beingisusedasleveragetocoerceC2.Ofcourse,thesub-stantivetask is toexplainhowwrongful instancesofmoralcoercion

16. BenjaminSachs,forinstanceemphasizesamorallyimportantaspectofpro-nouncingconditionalthreats—namelythatdoingsomotivatesthethreatenertoenforcethethreat(Sachs,2013).

17. TheAim-HackingConditionexplainshowC2 isusedonlygiven indepen-dent grounds for thinking that C2 is wronged. So the Aim-Hacking Con-ditionwillnotexplainhowC2 isused incasesofpassivecoercion if it iscombinedwithanaccountsuchasJapaPallikkathayil’s,sinceheraccountfo-cusesonthewrongfulnessofcoercivespeech-acts,whichareabsentincasesofpassivecoercion.ThisisnotafailureofPallikkathayil’saccount—heraimwastoprovideanaccountofactivecoercion.ButitdoesshowthatfortheAim-HackingConditiontohavetheadvantageofexplaininghowC2isused inbothactiveandpassivecoercion, itmustbecombinedwithanaccountofwrongfulcoercionthatprovidessomeindependentgroundsforhowC1wrongsC2insuchcases.



whichiswrongfulinitsownright.TheapplicationoftheAim-HackingConditiontocasesofmoralcoercionhelpsrevealtwoadditionalwaysinwhichC1wrongsC2bymorallycoercingher.

First, C1 relies on C2’s compliancewith the very norms that C1wronglyflouts(orthreatenstoflout)asameansofachievingherownends.Unilaterallyfloutingnormsisunfairtothosewhocomplywiththenorms,whentheircompliancebenefitstheflouter(asincasesoffree-riding).BycountingonC2’scompliancewiththeverynormsthatC1wronglyflouts,C1treatsC2unfairly.

Second, when C1morally coerces C2, the aims that are hackedarenotmerelyaimsthatC2isentitled tohave—i. e.,prudentialcom-mitments—butaimsthatbothC1andC2areobligated tohave—i. e.,moralcommitments.Incasesofnon-moralcoercion,itistypicallypro-crusteantoaskwhetherwearepermittedtorefusetoaccedetothecoercer’sdemands,sincethesortsofnormsoperativeinsuchacasearenormsofpracticalrationality,ratherthannormsofmorality.EveniftherearedecisivereasonsforC2toaccedetonon-moralcoercion,sheisstillmorallypermitted,allthingsconsidered,torefrainfromdoingso.ButiftherearedecisivereasonsforC2toaccedetomoralcoercion,thenbydefinition she isnotmorallypermitted to refrain.Violatingprudentialrequirementsisdiscretionaryinawaythatviolatingmoralrequirement isnot.Soeven if thepsychologicalpressureassociatedwithbothsortsofcoercionareequalinseverity,moralcoercioncantrapitsvictiminawaythatnon-moralcoercioncannot,byforeclosinganymoralpermissiontodootherthanwhatC1wants.Morally(ratherthanmerelyprudentially)foreclosinganoptiontowhichC2shouldbeentitledwrongsC2inawayoverandabovethewaysheiswrongedwhenshe isnon-morallycoerced.Consequently,whereas there isasense inwhichnon-moralcoercionismorally“escapable” (albeitatpotentiallysignificantcosttoC2),instancesofmoralcoercionmightnotbesimilarlyescapable.

Insummary,moralcoercion,whetheractiveorpassive,wrongsC2inthreeways.First,C1usesC2infurtheranceofaimsthatwrongherbyhackingheraims.Second,C1treatsC2unfairlybyrelyingonher

Forreasonsthatarenotalwayseasytoexplain,wearees-peciallyrepulsedbytheideaofonemoralagentmanipu-latingtheother.This,ofcourse,iswhathappensincasesof moral blackmail. The blackmailer attempts to get apersontodocertainactsbythreateningtodosomethingmuchworse.Totheextentthatthepersoncomplieswiththesedemands,he is surrenderinghismoralautonomy.Heis,inasense,apuppetintheblackmailer’shands.19

Partofthechallengeistomakemoreprecisehowwearebeingusedwhenwearesubjectedtomoralcoercion,andhowthisuseviolatesus.TheAim-HackingConditiondoesthis.MoralcoercionwrongsC2inthatheraim—specifically,hercommitmenttomorality—isbeingleveragedtoserveasameansinfurtheranceofanunjustend.Afterall,ifC2werenotamoralperson,C1’seffortsatcoercionwouldn’twork.ItispreciselybecauseC2hasthemoralcommitmentofpreventingtheworseoutcomesinSHIELDandHOSTAGEthatC1isinapositiontomorallycoerceC2.

In cases of non-moral coercion, the self-interestedmotivationsC2has areused against her; if she lacked amotivation to prevent per-ceivedharmtoherself,thenparadigmexamplesofcoercionwouldfailtoincentivizecompliance.Incasesofmoralcoercion,C1’smoralmoti-vationsareusedasameansofunderminingtheirownpurpose.Atoneremove, then,wrongfulmoralcoercionwrongsC2 in thesamewaythatwrongfulnon-moralcoerciondoes—bothinvolveusingC2asameanstotheachievementofC1’sendsbyintentionallyputtingC2inapositionwheresheworse-achievesherownlegitimateaimsasaresultofhavingthoseaims.SowecanappealtotheAim-HackingConditiontoexplainamannerinwhichwrongfulmoralcoercionwrongfullyusesC2:sheiswrongednotonlyinthewayspecifiedbystandardtheoriesofcoercionbutmoreoverbyhavingheraimshackedinfurtheranceofthosewrongs.Doingsousesherownaimsasameanstowrongingher,

19. (McConnell,1981,p.562)



isobligatedorevenpermittedtoabideby.ThusC1doesnottreatC2unfairlybyrefrainingfromabidingbythemherself.

Third, thoughC2 ismorallyobligated todowhatC1wants—i. e.,tosavetheinnocent—thisisnotbecauseC1hasforeclosedanoptiontodootherwise.Rather(asInoted),C2wasantecedentlyrequiredtosavetheinnocent(thoughC2mistakenlythinksotherwise).

Onemightraiseanotherchallenge.Ihaveclaimedthatmoralcoer-cionwrongsC2byputtingherinapositionwheresheworse-achieveshermoralaimsasaresultofhavingthoseaims.AtthetimethatC2iscoerced,theworldwouldbebetteroffifC2didn’thavehermoralcommitment—which is precisely theopposite effect thathermoralcommitmentsaresupposedtohave.Onemightarguethatthisanaly-sisproblematicallypresumesthatC2mustbeaconsequentialist.Butconsideracommitteddeontologistwhobelieves,forinstance,thattheDoctrineofDoing andAllowingprovides absolutist agent-centeredconstraints against committingharms.Now suppose that this deon-tologistisC2inALLEY,orSHIELD,orHOSTAGE.HowwouldC2re-spond?Shecertainlywouldn’tcomplywithC1’swishes—eventhoughdoing sowouldmake theworldbetter.This is because compliancewouldrequireviolatingabsolutistagent-centeredconstraintsagainstcommittingharms.Thissuggeststhathackingmoralaimsworksasatacticonlyagainst thosewhosemoral reasoning isat leastpartly te-leological.Thisshouldn’tbeasurprise,sincemoralcoercionfunctionspreciselybythreateningtomakethingsgoworse.Suchathreatwillhavelittlepurchaseonanabsolutistdeontologist.TheupshotisthatthesortofconsequentialistreasoningItacitlyimputetoC2isappro-priate.Otherwise,C2couldn’tbecoerced.20

20.ButsupposeC1saystoC2,anabsolutedeontologist,“Commitharmxnow,orIwillputyouinapositionwhereyouwillbeforcedtocommittensuchharmsinthefuture.”WhetherC2shouldcomplydependsonwhetheragent-centeredconstraintsagainstdoingharmarenotonlyagent-relative,buttime-relativeaswell.Adoubly relativedeontologistwouldn’tcommitx,even ifitmeant that shewouldhave to commitworseharms in the future.But atime-neutral deontologist would indeed be coercible. This is precisely be-causesuchadeontologistiscommittedtopromoting herownnon-violationof agent-relative constraints over time. So it is because there is a residual

compliancewiththeverynormsthatC1wrongfullyflouts.Andfinally,insofarasthesearecommitmentsthatC2isnotmerelyentitledbutobligatedtohave,moralcoercioncanwrongC2byillicitlyforeclosinganymoralpermissiontodootherthanwhatC1wants.

Onemight argue that the account I havepresentedover-gener-alizes, in thatmoral coerciondoes not always involve a pro tantowrong.Considerthefollowingcaseofwhatmightbecalled“reversemoralcoercion”.

RACIST

A flash flood is endangering an innocent. C1 is not inapositiontosaveher,butC2candoso,at littlecosttoherself.However,C2isaracistwhobelievesthatmoral-ityrequirestheexterminationoftheinnocent’srace.Sheconsequently believes that the innocent should not besaved.Knowingallthis,C1conditionallythreatenstode-votehimselftoalifetimeofcharitydirectedtomembersoftheinnocent’srace,unlessC2savesthatinnocent.

Here,C1iscoercingC2intocommittingagoodactwhichC2mistak-enlybelievestobewrongfulbythreateningtodosomethingwhichC2mistakenlybelievesisevenmorewrongful.OnemightarguethatC1’sactisnotevenprotantowrongful.ButtheaccountofmoralcoercionthatIhavepresentedisconsistentwiththisresult,inthatnoneofthethreeexplanationsofmoralcoercion’swrongfulnessapplyinRACIST.

First,hackingC2’saimsinfurtheranceofC1’sgoalswrongsC2onlyifC1’sgoalswrongC2.InRACIST,C2hasanenforceablepositivedutyto save the innocent—consequently, it does not wrong her to usehermistakencommitmentsinfurtheranceofenforcingthatduty.So,thoughC1usesC2,thisdoesnotwrongC2.

Second,C1’sconductdoesnottreatC2unfairly.ThoughC1isrely-ingonC2’scompliancewithracistnormsthatC1isherselffloutingasameansofincentivizingthedesiredact,thesearenotnormsthatC1



thatshecan impose themonC2.ThisseemsunfairnotbecauseC2wasmorallyobligatedtocommitφ—onecanhaveagent-relativerea-sonstopreventwhatanotherpersonhasanagent-neutralobligationtodo.Rather,imposingtheharmsonC2mightseemunfairbecauseC2is,afterall,avictimofcoercion.Incasesofmoralcoercion,thereisaharmwhichmustfallsomewhere,andC2isinapositiontochoosewheretheharmwillfallamongalimitedmenuofoptions—butsheisnotresponsiblefor the factthattheharmmustfallsomewhere.Rather,theagentwhoismorallycoercingheristheoneresponsibleforthepredicament.ItseemsunfairtoholdC2responsiblefortheharmshecommits,whenC1 is responsible forputtingC2 in thepredicamentinthefirstplace.ThoughC2istheproximatecauseoftheharm,andC1iscausally“upstream”,C1bearsmoreresponsibilityforthatharm,because the degree of responsibility depends on the options avail-able—andC1hastheoptionofchoosingboth~φand ~ψ,whereasC2hasonlytheoptionofchoosing~φor~ψ,exclusively.

BecauseC2islessresponsiblethanC1,thevictimofφ(orheres-tate)shouldseekredressfromC1ratherthanfromC2,giventhatshecanchooseonlyoneortheother.Likewise,ifthepotentialvictimofφcanpreventherowndeatheitherbykillingC1orbykillingC2,itisclearthatC1isthemoremorallyappropriatetarget.

TheclaimthatC1ismoreresponsiblethanC2forφis,however,compatiblewith the claim thatC2 is no less responsible than shewouldbe if she committed theharm in response to a functionallyequivalent adventitiously imposed dilemma (i. e., a dilemma im-posed by happenstance), rather than in response to C1’s coercion.Thispointcanbeputdifferently.C2’sdiminishedresponsibilityforφisgroundedsolelyinthefactthatshehasalimitedmenuofoptionsfromwhichtochoose.Thathermenuofoptionsislimitedas a result of being coerced by C1doesnotitselfdiminishherresponsibilityoverandabovethedegreetowhichitismitigatedasaresultofhavinghermenuofoptionsreduced.Toseethis,considerthefollowingmodifi-cationofALLEY:

Nowthatwehaveabetterpictureofwhatmakesmoralcoercionwrongful—andspecificallyofhowitinvolvesusingC2—wecanturntotheLiabilityandBadnessQuestions.

4. The Liability Question

SupposeC2 ismorallyobligated toaccede toC1’scoercivedemand.Thatis,supposeC2ismorallyobligatedtochooseφoverψ.Eveninsuchacasethethird-partyinnocentvictimofφmightstillbemorallyentitledtodefendherselfagainsttheharmthatC2iscommitting,ortocompensation. That is, C2 is morally liable to defensive or compen-satory harm (where a person is liable to be harmed just in case she has done something to forfeit her right not to be harmed in that way). ButhowcanitbethatC2’svictimsareentitledtoprevent(orseekcom-pensationfor)whatC2ismorallyobligatedtodo?

Whatjustifieschoosingφisthatitisthelesserevil;thismeansthattheoccurrenceofφwillstill infringe(thoughnotviolate)therightsofan innocent.Since thisvictimhasdonenothing to loseher rightnot tobeharmed, toharmherwrongsher,even ifwrongingher istherightthingtodo,allthingsconsidered.21Thisiswhythevictim(orherestate)canbeowedcompensationfortheharmshesuffers,evenifC2permissiblyimposesthisharm.22Andthefactthattheinnocentiswrongedcanalsogroundanagent-relativepermissionfortheinno-centtoengageinproportionatedefensiveviolencenecessarytopre-venttheharmsheisthreatenedwith—eventhoughC2isobligated,fromanagent-neutralstandpoint,toimposethatharm.

So, at least inprinciple, the third-party victim canbe entitled toimposedefensiveandcompensatoryharms.Buton whomisshemor-allypermittedto imposesuchharms?Atfirst, itseemsunfair tosay

teleology toC2’s non-consequentialism—i. e., its time-neutrality—that sheis coercible.So imputingsomesortof teleological thinking toC2 isappro-priateifwearetoassumethatsheiscoercible.(Formoreonthedistinctionbetweentime-relativeandtime-neutralviews,see(Louise,2004)).

21. Formoreonthedistinctionbetweeninfringingandviolatingrights,see(Mc-Mahan,2009,p.10).

22. Formoreonthis,see(Rodin,2012).



liabletobedefensivelykilled—dependsonwhetherthevictimisabletotargetC1instead.Ifshecan,thenC2isnotliabletobekilled.23

The upshot is this: C2’s liability to defensive and compensatoryharmsdependson a) thedegreeof responsibility shebears for theharmsheimposesonthevictim,b)whetherthereisanyoneelsewhoismore responsible for that harm, and c)whetherC2’s victims areabletoimposedefensiveandcompensatoryharmsonamorerespon-sibleparty(viz.,C1).WhatisimportanttorecognizehereisthatbeingwrongedbybeingcoerceddoesnotitselfaffectC2’sliabilityforwhatshedoesinresponsetobeingcoerced.

Notethateverythingthathasbeensaidalsoappliesincaseswhereφiscoercer-enacted.IncasessuchasSHIELDorHOSTAGE2,C2mustdecidebetweenallowingalesserharmandcommittingagreaterharm.Supposethatshechoosestheformer;sheisclearly lessresponsiblethanC1fortheharmssheallows,bothbecauseC1isresponsibleforputtingC2inthispredicamentandbecauseC1istheonewhoactuallycommits theharm.Butshe isno less responsible for theharmssheallowsthanshewouldbeifsheallowedthesameharmsinresponseto facinga functionallyequivalent adventitiously imposeddilemma.Again,whetherC2 is liable todefensiveorcompensatoryharmsde-pendsonwhethersuchharmscanbeimposedonC1instead.

ThoughbeingwrongeddoesnotdiminishC2’s responsibility forwhatshedoes,itcan(asweshallsee)serveasaprotantoreasonforrefusingtoaccedetomoralcoercion.

23. Itmightseemstrangethatone’sliabilityiscontingentinthisway—thatitcanappearanddisappeardependingonwhetherthemoreresponsiblepartycanbetargeted.Liabilityisinstrumentalinthiswayduetotherolethatnegativerightsplayinourmoraleconomy.Negativerightsprotectusfrombeingusedwithoutourconsentasameanstotheachievementofanother’sends.Whenapersoninfringesanother’srightnottobeusedasameans,theinfringerher-selfforfeitsherrightnottobeusedasameanstopreventingorrectifyingtheharmswhichshewasthreateningorimposing.Arights-infringercanbecomeliable to themeans required to prevent or rectify that rights-infringement.Thusliabilityisnecessarilyinstrumental,becauseitsfunctionistopreventarights-infringementortorestorearight.

ALLEY2

Thirty children, separated from their class during a field trip, find themselves in an alley, at a dead end. C2, who has also accidentally turned into the alley, is several me-ters from the children. And several meters from C2 is an-other innocent. At thatmomentabomb, leftover fromthepreviouswar,happenstorollofftheroofofthetallbuildingaboveC2.Asitbouncesdowntowardthechil-dren, it arms.Theonlyway forC2 to save the livesofthechildrenistointerceptthebombandthrowitintheonlydirectionavailabletoher,whichisdownthealley,toward the innocent several meters away, foreseeablykillingher.C2doesso.

The only difference between ALLEY and ALLEY 2 is that in the former case anagent intentionally foreclosesoptions thatwouldotherwisebeopentoC2,whereas inthe lattercasetheoptionsare foreclosedasamatterofhappenstance.WecanevenimaginethatbothcasesareidenticalfromC2’sstandpoint:fromherperspectiveinbothcasesthebombsimplyfallsfromthesky.ItwouldbestrangetothinkthatC2islessresponsiblefortheprotantoharmshecauses inALLEYthat inALLEY2,eventhoughsheiswrongedintheformercasebutnotthelatter.Thissuggeststhatcoercionper sedoesnotdiminishC2’srespon-sibilityforwhatshedoes.Butthefactthatheroptionsaredelimited,incombinationwiththefactthatshedoesnotintendthedeathoftheinnocent,makesherlessresponsiblethanC1forφ.

ThoughC2’sresponsibilityforwhatshedoesinALLEYandALLEY2isthesame,thedegreeofliabilityshebearscandifferbetweenthetwocases.Thisisbecause,inALLEYbutnotinALLEY2,thereissome-oneelsewhoismoreresponsiblethanC2.SoiftheinnocentcouldsaveherownlifebykillingC2inALLEY2,shewouldbepermittedtodoso.ButwhethersheissopermittedinALLEY—thatis,whetherC2is



Oneconstraintuponimposingaharmistheconstraintofpropor-tionality,accordingtowhichtheaversionoftherelevantevilsmustbeworththeseverityoftheprotantowrongfulharmsimposed.Andonefactor relevant to determiningwhether a harm satisfies the propor-tionality constraint iswhether thatharm is committed intentionally.Committingaharmintentionallyratherthancollaterallyisrelevantinthattheformerreceivesgreaternegativeweightinthecalculationofproportionalitythanthelatter.24Itmightseem,then,thatinstancesofcoercioninwhichC1coercesC2intointentionallycommittingaharm(suchasHOSTAGE) areworse than instancesof coercion inwhichC1coercesC2intocommittingaharmmerely foreseeably(suchasinALLEY).Butthisisanoverlysimplisticpictureofhowintentionfunc-tionsincasesofcoercion.InallsuchcasesC2isfulfillingC1’saims.Byforeclosingtheconjunctiveoptioncomposedof~ψand~φ,C1inten-tionallyincentivizesthecommissionofφ,therebyeffectivelyenlistingC2’s assistance—albeitwithouther consent—in furtheranceofC1’saims. Under these conditions, themoralmeasure of C1’s action de-pends,inter alia,ontheintentionalstatusofC1’saction.

Toseethis,consideragainALLEY.SupposeC2throwsthebomb awayfromthechildrenwhomitwouldotherwisekill.Indoingso,sheforeseeablekillsan innocentbystander—which isprecisely theout-comeatwhichC1wasaiming.C1’sroleaffectsthemoralityofwhatC2does in the followingway: thoughC2didnot intendthe innocent’sdeath, it shouldbeweighed asheavily as an intentional killing, be-causea)C1aimedatthedeathoftheinnocent,andb)C1furtheredthataimbycontributingsubstantiallytothatdeath.Indeed,itisnotin-felicitoustodescribetheinnocent’sdeathasintentional,eventhoughC2didnotkillherintentionally.

Thissuggeststhat,evenpresumingthemoralrelevanceofthein-tention/foresightdistinction,theprotantowrongwhichC2commitsinHOSTAGE isnoworse than thepro tantowrongwhichC2 com-mitsinALLEY—even though in the latter case C2 kills merely foreseeably

24. See(McMahan,2009),Ch.1.

5. The Badness Questions

SofarIhavearguedthatbeingmorallycoerceddoesnotitselfaffectC2’sliabilityfortheharmsshefosterswhenshechoosesφ.ButhowdowedeterminewhetherC2oughttoaccedetoC1’swishes,whenC2isbeingmorally coerced? Iobviously cannotaddressall the factorsrelevanttodeterminingwhetherC2oughttoaccede;butIwilladdressone factor thatplaysaspecial role incasesofmoralcoercion—andthatistheroleofintention.

Inaddressingthisissue,Iwillargueasfollows:

1.Moralcoercioncaneffacetherelevanceoftheintention/fore-sightdistinction,inthatwhenC2isweighingφagainstψ,theformershouldbeweighedasheavilyasitwouldbeifitwerecommittedintentionally,evenifC2isactuallycommit-tingitcollaterally(i. e.,foreseeablybutnon-intentionally).

2.Indefending1,Iwillarguethatweneednotthinkthatinten-tionhasonlyfirst-personal(andnotthird-personal)relevanceinourmoraldeliberations.Evenifwhatgroundsthemoralrelevance of the intention/foresight distinction iswhollyfirst-personal,itstillhasthird-personalrelevance.

3.Aconsequenceof1isthatwehaveabasisforthinkingthatac-cedingtomoralcoercionisworsethanaccedingtoafunc-tionallyequivalentadventitiousmoraldilemma.

4.C2hasa(defeasible)agent-centeredprerogativeagainstbe-ing morally coerced, which means that in weighing thedisvalueofaccedingtoC1’smoralcoercionagainstthedis-valueofresistingit,C2canaugmentthelatter.

TodeterminewhetherC2ismorallypermittedtoaccedetomoralcoer-cion,wehavetoweighthemorallyrelevantcostsofdoingsoagainstthemorallyrelevantbenefitsrelativetotherelevantalternative—whichistorefusetoaccede.Thatis,C2hastochooseamongevils,andindoingso,C2willbecommitting,enabling,orallowingaharm.



one person from being intentionallymurdered or another from be-ingkilledcollaterally,Ihaveastrongerreasontodotheformer.26Onthisview,theagent-relativereasontoabidebytheconstraintagainstintentionallycommittingprotantowrongsisnotgroundedinthepro-motion of anyvalue that aggregates across instances of compliancewith theconstraint.Aworld inwhich innocentsaremerely foresee-ablykilledratherthanintentionallykilledisindeedabetterworld,butthis isnotwhy there isastrongeragent-relativedutyagainstkilling.Rather,aworldinwhichinnocentsaremerelycollaterallyratherthanintentionallykilled isabetterworldbecausesuchaworld isone inwhichpeopleareabidingbytheagent-relativeconstraintagainst in-tentionallykilling.Thisreflectsatraditionalpictureofdeontology,inwhichtherightprecedesthegood.

Butevenifweaccepttheagent-relativeview,thereare,Ibelieve,stillagent-neutral reasons tosee to it thatpro tantowrongsarenotcommitted intentionally. Onmy view, all agent-relative constraintsgenerate a corresponding agent-neutral reason to promote compli-ance with the agent-relative constraint.27 That is, the agent-relativeviewentailstheagent-neutralview.Accordingly,aworldinwhichin-nocentsarenotintentionallykilledisabetterworld—butthisisnotwhatgroundstheagent-relativeconstraint.Rather,suchaworldisbet-ter precisely because agent-relative constraints are beingmet. Andthoughthisdoesnotgroundtheagent-relativereasontorefrainfromkillingintentionally,itdoesgenerateanagent-neutralreasontoseetoitthatinnocentsarenotkilledintentionally.

Against this,onemightdenynotonly thatmaking theworldgobetter iswhatgroundsagent-relative constraintsagainst intentionalharming,butalsothatabidingbyagent-relativeconstraintshasagent-neutralvalue.Butsuchaviewcomesatasignificantcost.Comparetwoworldswhicharethesameexceptthatinthefirsteveryonevio-lates agent-relative constraints, and in the second everyone abides

26. (Nagel,1986,p.178)

27. Theviewclosesttothispicturebelongsto(McNaughton&Rawling,1995).

rather than intentionally.Accordingly,whenwedo the calculationofproportionalitydeterminingwhethertocommitφ,thatactshouldbeweighedasheavily as an intentional harm in the calculation. In thisrespect,moralcoercioncaneffacetherelevanceoftheintention/fore-sightdistinction.

Onemightargueinresponsethatintentionhasonlyfirst-personal andnotthird-personalrelevanceinthecalculationofproportionality.Itismorallyworseformetointentionallycommitawrongfulharmthatit is todosocollaterally, in thatsuchaharmshouldreceivegreaterdisvalueintheproportionalitycalculationdeterminingwhatIshoulddo.ButthefactthataharmIbringaboutwasintentionallysoughtbysomeoneelse—whocoercedmeintocommittingit—doesnotentailthat it should similarly receivegreaterdisvalue in thecalculationofproportionality.Onthisview, theconstraintagainst intentionallykill-inginnocentsisagent-relative.Iwillcallthis“theagent-relativeview”.

Ontheagent-relativeview, theaugmenteddisvalue thataharmreceivesforbeingintentionalisagent-relativeinthatitappearsonlyin the intender’s calculation determining the permissibility of com-mitting theact. It doesnot appear in anyoneelse’s, including caus-ally“upstream”agentswhoenabletheintendertocommittheharmaswellascausally“downstream”agentswhocommittheintendedharm.25Indoingthecalculationofproportionality,thecausallydown-streamagentsshouldnotaugmentthedisvalueoftheharmtheyarecommitting,eventhoughitwasintended—sinceitwasnottheywhodidtheintending.Thisisincontrasttoaviewaccordingtowhichthemoralrelevanceoftheintention/foresightdistinctionisgroundedinagent-neutralreasonstopreferthelattertotheformer.Iwillcallthisthe“agent-neutralview”.

Theagent-relativeviewhasbeenexplicitlyendorsedbysome,in-cluding,notably,ThomasNagel.Hesays,in“TheViewFromNowhere”,thateventhoughit’smorallymoreobjectionableformetoharmsome-oneintentionallythancollaterally,it’snottruethatifIcaneitherstop

25. Anexceptionisiftheyareintentionallycooperatinginfurtheranceofajointgoal.Iaddressthisin(Bazargan,2013).



intention/foresightdistinction isgrounded inagent-relative reasonsnottointentionallycommitprotantoharms.

ThefactthatC1’sintentionsarerelevanttothemoralmeasureofwhatC2doeswhenC2accedesdoesnotmeanthatC2oughtnevertoaccedewhendoingsorequiresenactingφ.Butthefactthatφre-ceivesaugmenteddisvalueinthecalculationofproportionalitymakesitlesslikelythattheproportionalitycalculationwillworkoutinfavorofaccedingunderthesecircumstances.Inaddition,thefactthatC1’sintentionsaffects themoralityofwhatC2doesmeansthatwehavegroundsforthinkingthat it isworseforC2toaccedebyenactingφthanitwouldbeifsherespondedinananalogousfashiontoamor-allyadventitiousdilemma(suchasALLEY2).Wehave,then,groundsfortheintuitionthatbyaccedingtomoralcoercionweallow“eviltosucceed”:themanipulator’sevilintentionsaremanifestinouractionswhenweenactheraims;choosingφoughttobeweighedaccordingly.

Doesthesameargumentapplytocasesofmoralcoercionwhereφ is coercer-enacted, as in SHIELD or HOSTAGE 2? In these cases,C2 is instructed todonothing—to refrain frompreventingC1 fromachievingherwrongfulaims(whichC2wouldotherwiseprevent).In-tentions canbe relevant to themoral assessment of omissions. ButC2’somissioncannotproperlybesaidtobeacauseofφ;asaresult,we lackanecessarybasis for thinkingthatC1’s intentionsaffect themoralmeasureofC2’somission.For thisreason, Idonot thinkthattheagent-neutralviewprovidesabasisforaugmentingthedisvalueof choosingφwhenφ is coercer-enacted.A consequence is that itshouldbemoredifficulttojustifyaccedingtomoralcoercionincaseswheredoingsoinvolvesactuallycommittingφ,notmerelybecauseofthemoralrelevanceofthedifferencebetweendoingandallowing,butbecauseφisawrongintendedbyC1.Thatis,theaction/omissiondis-tinctioncombineswiththeintention/foresightdistinctiontoexplainhowandwhyaccedingtocoercee-enactedmoralcoercionisespeciallyegregiouswhenφisamorallywrongfulaim.

SofarIhavediscussedtherelevanceoftheintention/foresightdis-tinctionasitappliestodeterminingwhenandwhetherweoughtto

bythem—andwecanchoosebetweenthesetwoworlds. Ifabidingbyagent-relativeconstraintshasnoagent-neutralvalue,thenthefactthat in the first world everyone violates agent-relative constraintswoulditselfprovideliterallynoreasontopreferthealternativeworld.Andthisseemslikeabigbullettobite.

Of course, the claim that agent-relative constraints generate cor-respondingagent-neutralvaluesleavesopenthepossibilityofconflictbetweenanagent-relativereasontorefrainfromintentionallykillinganinnocent,andanagent-neutralreasontoseetoitthatintentionalkillingsofinnocentsareminimized.Howweresolvethisconflictde-pendsonhowweweighthesecompetingreasons—whichisbeyondthescopeofthispaper.28

So the agent-relative reason toprefer committing collateral overintentionalharmsgeneratesanagent-neutralreasontopreferthatoth-erscommitcollateralover intentionalharms.Wheredoesthis leaveus?IarguedthatevenifC2doesnotintendφ,itshouldbeweighedasheavilyasanintentionalharm,becausea)C1aimedatφ’soccurrence,andb)C1furtheredthataimbycontributingsubstantiallytoφ.Onemightattempttoforestallthisargumentbyclaimingthatsomeone else’s intentionsregardingwhatI dogenerallycannotaffectthemoralityofwhatIdo.ButIrespondedtothisbyarguingthattheintention/fore-sightdistinctioncanhavethird-personalrelevance:bycommittingφ,C2bringstofruitionC1’swrongfulintentions,whichiswhyitshouldreceiveasmuchweight inC2’sdeliberationsasan intentionally com-mittedharm.Hencemoral coercion can efface the relevance of theintention/foresightdistinction,even thoughthemoralrelevanceofthe

28.Onemightnotethatifthemoralrelevanceofthedoing/allowingdistinctionisgroundedinanagent-relativereasonagainstcommittingharms,andiftheagent-relative view entails the agent-neutral view, thenwehave an agent-neutralreasontochooseallowingsovercommittingsingeneral(forharmsofequalseverity).Itmightseemthatchoosinginthiswayisnotpossible,sinceallowinganactentailsthatsomeone elsecommitsthatact.Butthisoverlooksthe fact that not all harms are the result of acts. There is, then, a straightfor-wardwaytocharacterizetheagent-neutralcounterpartoftheagent-relativegroundsfortherelevanceofthedoing/allowingdistinction:itisareasontochoosenon-committedharmsovercommittedharms(ofequalseverity).



oftheveryendsthattheyaretaskedwithavoidingimper-missiblyusestheagentwhenit isdoneinfurtheranceofendsthatwrongtheagent.

B)ShouldC2accedetoC1’sdemands,shecanbeliabletode-fensive or compensatory harms, even if C2wasmorallyobligated to accede. Whether she is liable depends onwhetherthedefensiveorcompensatoryharmscanbeim-posedonC1 instead, sinceC1, unlikeC2, is responsiblefor imposing the dilemma in the first place. Still, forC1towrongC2bymorallycoercingherdoesnotinitselfdi-minish thedegreeof responsibility thatC2bears for theharminwhichφconsists.Thoughbeingmorallycoercedcanserveasgroundsforanagent-centeredprerogativetorefrainfromacceding,C2isnolessresponsiblethanshewouldbe if she respondedanalogously toa functionallyequivalentadventitiouslyimposeddilemma.Still,weneedto account for the robust intuition that aharm resultingfromamorallycoercivedilemmaisworsethanaharmre-sultingfromananalogousmorallyadventitiousdilemma.Incasesofmoralcoercion,C1’sintentionsaremanifestinC2’s actions, even though the latter is not intentionallycooperatingwith the former.The result is that theharmoughttobeweighedasheavilyasanintentionalharminC2’sproportionalitycalculation,evenifshecommittedtheharmcollaterallyratherthanintentionally.

C) One of the goals of this paper is to articulate the intuition that byacceding tomoralmanipulationwe “allowevil to suc-ceed”.IfwhatIhaveadumbratedinAandBiscorrect,therearetwowaysaccedingallowseviltosucceed.ForC2toac-cede tomoral coercionallowsher tobeused in thewayIhavedescribedinsection3.ButthiscannotexhaustthesenseinwhichcompliancewithC1’swishesallowseviltosucceed.Thisisbecausethedutywehavenottobeused

accedetomoralcoercion.Thereisanothercharacteristicendemictomoralcoercionandrelevant to theproportionalitycalculationthat Iwillbrieflymention.

AsIarguedinsection3,C1usesC2asameremeansinfurtheranceofC1’sevilplansbyhackingC2’saims.C2hasa(defeasible)agent-cen-teredprerogativeagainstbeingusedasameremeansingeneral,andbeingmorallycoercedspecifically.Havingthisprerogativemeansthat,inweighingthedisvalueofaccedingtoC1’smoralcoercionagainstthedisvalueofresistingit,C2canaugmentthelatter—withinlimits.Ifthedisvalueofthetwoissufficientlyclose,theagent-centeredprerogativecantipthebalance,therebypermittingC2torefrainfromdoingwhatwouldmakethingsgoimpersonallybest.

Note that ifan individual is facedwithamorallyadventitiousdi-lemma,asinALLEY2,therewouldbenosimilaragent-centeredpre-rogativetorefrainfromcommittingthelesserharm,sinceindoingsotheagentisnotbeingmorallycoerced.

6. Conclusions

There are three morals I wish to draw from my discussion of thewrongnessofmoralcoercion,theLiabilityQuestion,andtheBadnessQuestion.

A)Ifmycharacterizationofmoralcoercioniscorrect,thenater-roristwhouseshostagesascoinforconcessionsisengagedinfundamentallythesameactivityasawarcriminalwhostrapsinnocentstothesideofhistank.Theyarebothen-gagedinmoralcoercion.Wearerepulsedbythewayourmoralcommitmentsarepliedinthesesortsofcases.TheaccountIhavelaidoutprovidesabasis forthis intuition.C1hacksC2’saims—sheintentionallyputsC2inapositionwherehermoralcommitmentsbecomeself-undermining,inthatC2worse-achievestheobjectofthosecommitmentsbecause shehas thosecommitments.PuttingC2 inaposi-tionwherehermoralcommitmentsareusedintheservice



Dreier, J. (1993). Structures of Normative Theories. The Monist, 76 (1), 22—40.

Epstein, R. A. (1983). Blackmail, Inc. The University of Chicago Law Re-view, 50, 553–566.

Feinberg, J. (1986). Harm to Self. New York: Oxford University Press.Frankfurt, H. (1973). Coercion and Moral Responsibility. In T. Hond-

erich (Ed.), Essays on Freedom of Action (pp. 65—86). London: Rout-ledge & Kegan Paul.

Haksar, V. (1976). Coercive Proposals. Political Theory, 4 (1), 65—79.Lindgren, J. (1984). Unraveling the Paradox of Blackmail. Columbia Law

Review, 84 (3), 670—717.Louise, J. (2004). Relativity of Value and the Consequentialist Umbrella.

The Philosophical Quarterly, 54 (217), 518—536.McConnell, T. C. (1981). Moral Blackmail. Ethics, 91 (4), 544—567.McMahan, J. (2009). Killing in War. New York: Oxford University Press.McNaughton, D., & Rawling, P. (1995). Value and Agent-Relative Rea-

sons. Utilitas, 7 (1), 31—47.Nagel, T. (1986). The View From Nowhere. New York: Oxford University

Press.Nagel, T. (1972). War and Massacre. Philosophy & Public Affairs, 1 (2),

123—144.Nozick, R. (1974). Anarchy, State, and Utopia. New York: Basic Books.Nozick, R. (1969). Coercion. In S. Morgenbesser, P. Suppes, & M. White

(Eds.), Philosophy, Science, and Method: Essays in Honor of Ernest Nagel (pp. 440—472). New York: St. Martin’s Press.

Pallikkathayil, J. (2011). The Possibility of Choice: Three Accounts of the Problem with Coercion. Philosophers’ Imprint, 11 (16), 1—20.

Portmore, D. W. (2007). Consequentializing Moral Theories. Pacific Phil-osophical Quarterly, 88 (1), 39–73.

Railton, P. (1984). Alienation, Consequentialism, and the Demands of Morality. Philosophy & Public Affairs, 13 (2), 134–171.

Rodin, D. (2012). Justifying Harm. Ethics, 122 (1), 74–110.Sachs, B. (2013). Why Coercion is Wrong When it’s Wrong. Australasian

Journal of Philosophy, 91 (1), 63–82.

asameremeansisfundamentallyaself-regardingduty.C1,though,wrongsnotjustC2,butthethird-partyinnocentsaswell.Anycharacterizationofhowcomplianceallowseviltosucceedshouldcapturethisfactaswell.SupposethatanuninvolvedthirdpartyconfrontsC2andremindsherthatifshedoesasC1wishes,C2willhavetherebyallowedevilto succeed. And suppose C2 responds by saying, “Don’tworry—Idon’tmindbeingsoused.”Sucha response in-tuitivelymissesanadditionalsense inwhichcomplianceallows evil to succeed.And the argument I presented insection5explainsthis:complianceallowseviltosucceednotonlyinthatitwrongfullyusesC2,butalsointhatital-lowsC1’swrongfulintentionstocometofruition—viz.,thegoalofharmingthethird-partyinnocents.Sotherearetwo(instrumentallyrelated)evilsherethatweallowtosucceedwhenwecomplywithC1’swishes:oneevilisself-directed,andtheotheristhird-partydirected.

Thisaccountisbutafirststepinthemoralanalysisofmoralcoercion.Ihopeheretohaveprovidedausefulfoundationforfurtherdiscussion.

Works Cited

Bazargan, S. (2013). Complicitous Liability in War. Philosophical Studies, 165 (1), 177–195.

Berman, M. N. (2011). Blackmail. In J. Deigh & D. Dolinko (Eds.), The Oxford Handbook of Philosophy of Criminal Law (pp. 37—106). New York: Oxford University Press.

Broad, C. D. (1930). The Philosophy of C.D. Broad. Open Court Publishing Co.

Davis, N. (1980). The Priority of Avoiding Harm. In B. Steinbock (Ed.), Killing and Letting Die (pp. 172—214). Englewood Cliffs, New Jersey: Prentice Hall, Inc.



Schopp, R. F. (1998). Justification Defenses and Just Convictions. Cam-bridge: Cambridge University Press.

Schroeder, M. (2007). Teleology, Agent-Relative Value, and ‘Good’. Eth-ics, 117 (2), 265–295.

Shaw, J. R. (2012). The Morality of Blackmail. Philosophy & Public Affairs, 40 (3), 165–196.

Sussman, D. (2005). What’s Wrong with Torture? Philosophy & Public Affairs, 33 (1), 1–33.

Wellman, C. H. (2005). A Theory of Secession: The Case for Political Self-Determination. Cambridge: Cambridge University Press.

Wertheimer, A. (1987). Coercion. Princeton, New Jersey: Princeton Uni-versity Press.

Moral Coercion

Documents