Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
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 support@jstor.org.
Oxford University Press and Mind Association are collaborating with JSTOR to digitize, preserve and extend
access to Mind.
http://www.jstor.org
StateoftheArtseries.
Thisarticleis thethirdofourcommissioned
I Self-determination
(i) Two featuresoffreedom.One old-fashionedway of classifyingcon-
ceptionsoffreeagencyis to distinguishbetween'libertyofindifference' and
'libertyof spontaneity'.This classificationis unobjectionableas a way of
bringingout importantlydifferentemphases in differentwriters.It is
seriouslymisleading,however,if it is meant to isolate competingcon-
ceptionsoffreedom.For whatthesetwolabels signifyare interpretations of
two different featuresof freedomthatmust be capturedin any reasonable
conception-namely,self-determination (or autonomy)and theavailability
ofalternative possibilities.Anyadequate notionoffreeagencymustprovide
forpossibilityand autonomyin somesense,and, in myview,thetraditional
conceptionsthatare stilltakenseriouslyweremeantto do so.1
Even classicalcompatibilismmade room,or triedto makeroom,forboth.
Its basic tenet,thatfreedomis dependenceofactionupon thewill,provides
a way of interpretingboth features.Alternativepossibilitiesare to be
identifiedrelativelyand counterfactually, by whatis possiblerelativeto the
subject's will. And the relevantdependencyrelationdefinesthe relevant
notionofself-determination; theselfis thewill.Hence self-determination is
determination by the will.2
1 Bernard Williamssaysthatclassicalcompatibilism opposedfreedom toconstraint, nottonecessity.
Of course,compatibilism holds thatfreedomis not opposedto necessityin the sense in which
determinism impliesnecessity. AndHobbesinparticular tookpainstostressthis.Butthatdoesnotmean
thatfreedom doesnotconflict withnecessity inanysense.Lockedidopposefreedom (butnotvoluntary
action)tonecessity: 'Wherever ... thepowertoactorforbear according tothedirection ofthought[that
is,liberty][iswanting], therenecessity takesplace'(II. 2I. I3).
I confinemyremarks in thisessayto the'moderns'.One could arguethatthesetwofeatures of
freedom werepresent in a rudimentary forminAristotle's discussion ofthevoluntary in Nicomachean
Ethics,III. 5. Therehe thinks ofthevoluntary as thatwhichhasitsorigin'in' theagent,in contrast to
actingunderconstraint 'to whichthepersoncontributes nothing'.('A humanbeingoriginates hisown
actions... as hefathers hisownchildren', I I I3bI8.) Butthevoluntary alsoinvolves a kindofdualpower:
C.. . whenactingis up to us, so is notacting'(III 3b6),thatmaybringin theelementofalternative
possibility.Thoughhe had muchto sayaboutnecessity, Aristotledid nothavea developednotionof
determinism or of freedom, perhapsbecausehe did notfacethekindof threatthatwas posed by
seventeenth-century science.(Fora libertarian seeSorabji,anda replybyFine.)For
readingofAristotle,
themedievals, generalissuesaboutfreedom wereposedbydivineforeknowledge.
To givean overview ofthemedievalandancientscenesis beyond,notthescopeofthisessay,butthe
erudition ofitsauthor.
2 Locke'sconcern toprovidefora notionofself-determination is explicit:'.. everymanis putunder
a necessity,byhisconstitution as anintelligent being,tobe determined inwilling byhisownthought and
appeal to a specialkirndofagent-causationf,
t,hatlokth
mustanadzcannot be -A
featureof the world.4The problemforlibertarianism, then,is to providea
coherentinterpretation of theseconditionstakentogether.
I wantto approachthe contemporary literatureby askinghow different
viewsoffreeagencyunitethetwofeaturesoffreedomin a singleconception.
I shall begin by discussingcontemporarycompatibilistaccounts of self-
determination.
elusivenotionofself-determination.
10The pictureofidentification
as some
kindof bruteself-assertion
seems totallyunsatisfactory,
but I have no idea
whatan illuminatingaccountmightbe.11
II Alternative
possibilities
... judgethenwhata pretty kindoflibertyitiswhichismaintained
byT. H., sucha
liberty as is in littlechildrenbeforetheyhavetheuse ofreason,beforetheycan
consultordeliberate ofanything.Is thisnota childish andsucha liberty
liberty; as
is in brutebeasts?(BishopBramhall, p. 43.)
Hobbes,Locke,Hume,Moore,Schlick,Ayer,Stevenson,anda hostofothershave
donewhatcanbe done,oroughtevertohavebeenneeded,toremovetheconfusions
thatcanmakedeterminism seemtofrustrate
freedom.
(Davidson(2), p. 63.)
(v) Theambiguity of'can'. The discussionofthelastsectionwas devotedto
the followingreasoning:our viewsabout Brave New World cases, and the
like, commitus to criteriafor freedomthat could not be satisfiedin a
deterministic world.Because it focusesupon ourintuitionsabout particular
cases, this mightbe called a 'bottom-up'argumentforincompatibilism.
Whatevertheplausibilityofthisbottom-upargument, top-downarguments
whichdeduce incompatibilism fromgeneralprinciplespresentedas self-
evident-have seemed to many verycompelling.These argumentshave
centredon the interpretation of the conditionof alternativepossibilities.
In thispart,I shall reviewthesearguments.
As I notedearlier,the compatibilisttreatment ofalternativepossibilities
usuallyis a consequenceof its view of self-determination.Once autonomy
is characterizedin termsofthenotionoftherelevantnotionoftheself(and
this is refinedto meet the Brave New World cases), then alternative
possibilitiesare definedby whatis dependenton the self,thatis, upon the
will. The resultis some kindof conditionalaccountof 'can'. 16
Two points about such accounts are obvious fromthe outset. First,
compatibilismwill insistthatthereare different notionsof possibility(or
differentsenses of 'possibility'),and that the notion picked out by a
conditionalaccount is the one relevantto freeagency.The second point
is that a conditionalanalysiscannot (withoutvicious regress)be applied
to the relevantnotion of 'will', so that in an importantsense, no com-
patibilistaccountof 'freewill can be given.I will turnto the second point
lateron.
If determinism is true,thenclearly,in somesense,thereareno alternative
possibilities.Relativeto the laws of natureand antecedentconditions,it is
notpossiblethatone does anythingbut whatone does. Compatibilistsmust
dismiss this sense as irrelevantto freeagency,and oftentheyclaim that
incompatibilist argumentstradeon ambiguitiesof 'can' or 'possible'.
Some philosophersregardtheseclaimsto ambiguityas ad hoc,as having
no forceindependentof some programmeto make determinismsafe for
human freedom.But this claim should be acknowledgedfromthe start.
Conditionaliststake themselvesto be givingan account of such ordinary
contextsas 'S is able to walkout oftheroom'.Such a sentencemightindeed
be trueeven thoughthereis no possibilitythatS will do so. For example,
supposeS has no interest, concern,orreasonto walkoutoftheroom.In that
case, it may well be truethatthereis no possibilitythatS will do so, even
thoughS is perfectly able to do so. This pointdoes not ofcoursesupporta
conditionalanalysis,but it does permitus to ask whetherthesensein which
determinism entailstheabsence of alternativepossibilitiesis a sense thatis
relevantto the interpretation of such sentences.
Furthermore, the exampleshowsthatthe ordinarynotionofabilityis in
some way relativeto attributionsof desireor will. It is simplya misuse of
'able to' to say: 'S is unable to leave theroombecause he doesn'twantto (or
has no reason to).' This gives some initial plausibilityto the idea that
attributions ofabilitiesand powersofactionareindependentofattributions
of 'will'.17
One can be drawn to a conditionalanalysis,not out of adherenceto a
programme,but because it can seemto accommodatethesepoints.Further,
the notionpicked out by a conditionalanalysisis arguablyof considerable
practicaland ethicalinterest.If to learnabout whatsomeoneis able to do is
to learn about what depends upon the person's will, then to learn that
someone was able to do what was leftundone will be to learn something
about the person'sactual will.
16 Not all compatibilists
areconditionalists. changeshismind
See Lehrer(i) and(3): he apparently
in (4); andDavidson(2).
17 See Chapteri ofHampshire (2).
totheconditional
(vii) Objections analysis:(a) Lehrer'sargument.
Let us turn
to criticismsof the conditional analyses themselves. In the current
one findstwogeneraltypesofcriticism.The firsttypeis, as faras
literature,
I know,new. It is due to Keith Lehrer (i). It seems to me unsound. The
second type captures objectionsthat have long been made against such
analyses,and, thoughit is inconclusive,it succeeds in raisingdeep issues
about the notionof freewill. I beginwiththe first.
A conditionalanalysishas the form:
(i) If C, thenS willM.
Lehrer claims that,for any such analysis,the followingwill be logically
possible:
(2) If not-C,thenS cannotM.
This claim is based on the considerationthat (i) is meant to be a causal
conditional,and causes are logicallyindependentof effects.But suppose:
(3) C doesnotobtain.
Then:(4) S cannotM.
So (i) cannotentail:(5) S can M.
G. E. M. Anscombe agrees with Lehrer. Alvin Goldman and Donald
Davidson disagree. Goldman (i) argues that if Lehrer's reasoningwere
sound,no dispositionalanalysisofsolubilitycould be correct.Since it is not
obviousthatsuch analysesare correct,thispointis inconclusive.Davidson
says:
responsetoLehreris simplythatifoneanalysessolubility
The correct bya causal
21 See Slote(2).
totheconditional
(viii) Objections analysis:(b) Chisholm's
argument.One gets
a differentargumentwhenitis supposedthatI am somehowpreventedfrom
willing. This supposition is pivotal to Roderick Chisholm's argument
against conditionalanalyses. At least since Bishop Bramhall replied to
Hobbes, conditionalanalyseshave seemed inadequate because theyseem
andfreewill
III Libertarianism
(ix) 'Free will? Both in ordinarylanguage and in currentphilosophical
literature,the use of the phrase 'free will' is a curious one. In common
parlance,the phraseis eitherencumberedwitha lot of nonsenselearnedat
Sunday School, or irrelevantto traditionalconcerns.As an exampleofthe
former,I happened to share a table some time ago with a high-ranking
officialof the Orange County (California)Republican Party.During the
conversation,his companionbemoaned her unsuccessfulcampaignto get
the City Council of Santa Ana to take action to amelioratethe plightof
homelesspeople in thatcity.With a sense of bewildermentand exaspera-
tion,she notedthatthe Cityhad been willingto fundsheltersfordogs and
catswhileit did nothingforpeople. The partyofficial was untroubled.'The
difference',he said, 'is thatpeople have freewill.' If freewillcan explainand
justifythe myriadplagues and miseriesthatmarkhuman history,thereby
justifying God's waysto mortals,surelyit can handle a bit ofdestitutionin
Santa Ana, California.
Ordinarylanguagehas moreinnocentusages.We speakof'actingofone's
own free will' in connectionwith coercion. If someone shows up for
inductionintothearmyonlyto avoid a jail term,thatis a paradigminstance
ofnotactingofone's own freewill.But thisusage has littleifanyconnection
withtraditionalconcernsfor freewill. What happens in these paradigm
cases is notthatthepersonhas no options,butthatcertainoptionsare made
prohibitivelyexpensive.But we do not speak of restrictionson freewill
whensuchcostsareconsequencesofnatural,ratherthanhuman,forces.It is
doubtfulthatthesourceofthecostshouldmatterto thequestionoffreewill.
So it is doubtfulthatone can findin theseordinarylocutionsa serioususe
forthe phrase'freewill'.25
25 (3) on coercion.
Butsee Slote(i) andFrankfurt
26 It wasAlbritton's
paperthatfirst
mademesee thispoint.
electivepower of the rationalwill' (p. 42); that is, the power of rational
beings to will one way ratherthan another,as alternativespresentthem-
selves.27No subtle discussionof internalobstacles will help us with this
question (thoughit will help us to understandthe relevantnotionof the
will). Even Frankfurt'sdiscussion,whichis unusual among compatibilist
writingin itsinsistenceon distinguishing freedomofactionfromfreedomof
will, is unhelpfulhere. For if we set aside his technicalnotionof will as
effective desire (which is afterall thoroughlyHobbesean), and
first-order
thinkof the will as one's higher-ordermotivations,then we can say that
Frankfurtgives us a betterunderstandingthan Hobbes of the efficacy of
will,but not of its freedom.
At the end of Part II, I suggested that it is a mistake for the in-
compatibiliststo deny that the conditionalistssucceed in identifyinga
notionthat is importantin our practice,and does not depend upon this
distinctnotionoffreedom.That denialdetractsattentionfromwhatto them
mustbe themostimportantissue: whatthispoweris, and whetherwe have
it. They can granta distinctnotionoffreedomof actionand stillinsistthat
that is not all the freedomwe want, or the most importantfreedom.
Freedomofactionis importantenough,theymaygrant,buttheywantto tie
what is special about human beings to this furtherpower; that is what
underwritesthe sense of human dignity.(Note Bramhall'sumbrageat the
Hobbesean picturein the passage quoted at the beginningofthispart;that
pictureis not merelywrongbut demeaning.)
It is far fromobvious what this furtherpower is. What is obvious,
however,is that,whateverit is, it is not somethingwhichcan be captured
by an impedimentmodel. It is this fact about it, I suspect, that leads
such an otherwiseuncartesianphilosopheras Rogers Albrittonto endorse
Descartes's view thatwe enjoyperfectfreedomofwill.28This breathtaking
opinionis widespreadamongdefendersoffreewill,eventhoughitis hardto
27
Cp. Reid:'Bytheliberty ofa moralagent,I understand a poweroverthedeterminations ofhisown
will'(p. 323). See Kane fortheinsistence on separating questionsoffreewillfromquestionsoffree
action.
28 This thought restson a consideration we havealreadyseen:that,whilethereare all too many
obstaclestoimplementing oreffecting ourwills,therecanbe no obstacletowillingitself.In thissense,
thewillcannotbe blocked.(CompareO'Shaughnessy: 'thewillitself
cannotbe paralysed',vol.2, p. 42.)
The truth behindthisthought is I thinkthis.The conceptofwilling is suchthatthereis nosuchthingas
failing towill;willingis necessarily successful.Butan obstacleis suchthatitblocksthepathtosuccess.
In thecaseofwilling, thereisnopath.Ratherwilling is 'goingforsomething', onthepath.There
starting
is roomforan obstaclebetweenwillinganditsobject,butno obstacletowillingitself.
Whether onethinks thistobetrue,trivially, becausewilling justistrying(andAlbritton saysthatthere
mightbe nothing to it),it does notfollowthatone cannotbe prevented fromwilling, notbyhaving
obstaclesplacedinthepath,butbyhavingone'swillpushedas itweretowardonepathoranother. So it
mightseemin casesofbrainwashing or hypnotism. Albritton findstheidea of unfreedom ofthewill
'inconceivable',and suspectshe couldhandleanyexamplestothecontrary; butitis significantthathe
doesnotconsider theseexamples.Perhapshewouldsayinsuchcasesthewillis notforcedbutbypassed
(as in hisexamplesofautomatism). But I don'tsee why.
ThomasReiddisagrees withthepurists. He findsthatfreewillis 'impaired orlost'byvicioushabits,
melancholy, madness,divineintervention ... (p. 326).
29
See Broad,and also Bonjour,Goldman(2), Nagel,and Thalberg(i). For a partialdefenseof
'immanent causation',
see vanInwagen(3).
References
Albritton, R., 'Freedomof Will and Freedomof Action',Presidential Address,
Proceedings oftheAmerican Philosophical
Association,
Novemberi985.
Anscombe, G. E. M. (i), 'CausalityandDeterminism',in Anscombe(3).
(2), 'SoftDeterminism', in Anscombe(3).
(3), Metaphysics and thePhilosophy ofMind: CollectedPhilosophical
Papers,
vol.III, University ofMinnesotaPress,I98I.
Aristotle,Nicomachean Ethics.
Aune,B., 'Hypotheticals and "Can"', inWatson(3).
3 I amindebtedtoLorraineRappforbibliographical
assistance
inthepreparation
ofthispaper.I am
alsoverygrateful,
indifferent
ways,toGerrySantas,andtoSaraLundquist,forproviding
mewithsome
freedom ofopportunity
(thoughnotfreedom ofwill)to writethispaper.
inWatson(3).
Ayer,A. J.,'FreedomandNecessity',
Bonjour,L., 'Determinism, and AgentCausation',Southern
Libertarianism,
JournalofPhilosophy,1976.
Bishop,'LibertyandNecessity',
Bramhall, in MorgenbesserandWalsh.
Reidel,1976.
Brand,M. andD. Walton(eds.),ActionTheory,
Bratman,M. (i), 'Davidson'sTheoryof Intention',in B. Vermazenand M.
Hintikka(eds.), Essays on Davidson: Actionsand Events,OxfordUniversity
Press,I985.
(2), 'Intention andEvaluation', Studiesin
in P. Frenchetal. (eds.),Midwest
Philosophy, vol.X, UniversityofMinnesotaPress,I986.
Broad,C. D., 'Determinism, Indeterminism, in Morgen-
and Libertarianism',
besserandWalsh.
Chisholm, R. (i), 'HumanFreedomandtheSelf',inWatson(3).
(2), 'The Agent as Cause',in BrandandWalton.
Davidson,D. (i), 'How is WeaknessoftheWillPossible?',inDavidson(4).
(2), 'Freedomto Act',in Davidson(4).
(3), 'Intending',in Davidson (4).
(4), Essays on Actionsand Events,OxfordUniversityPress, I980.
Dennett,D., ElbowRoom,MIT PressandOxfordUniversity Press,I984.
Dworkin,G. (i) 'ActingFreely',Nous, 1970.
(2) 'Autonomy andBehaviorControl',Hastings CenterReport, 1976.
Fine,G., 'Aristotle's Determinism', ThePhilosophicalReview,I98I.
Fischer,J. (i), 'Responsibilityand Control', The ournalofPhilosophy,1982.
(2), 'Incompatibilism',PhilosophicalStudies,I983.
'Van Inwagenon FreeWill',Philosophical
(3) Quarterly, I986.
Foley,R., 'Compatibilism andControloverthePast',Analysis,1979.
Frankfurt, H. (i), 'Alternate andMoralResponsibility',
Possibilities TheJournalof
Philosophy,I969.
(2), 'Freedomof the Will and the Conceptof a Person',TheJournalof
Philosophy, 1971.
in Honderich.
(3), 'CoercionandMoralResponsibility',
(4), 'Three Concepts of Free Action',Proceedingsof theAristotelianSociety,
Supplementary Volume,1975.
in AmelieRorty(ed.), TheIdentities
and Externality',
(5), 'Identification of
Persons,University ofCalifornia
Press,1976.
(6), 'The Problemof Action',AmericanPhilosophicalQuarterly,1978.
Gert,B., andT. Duggin,'FreeWillas AbilitytoWill',Nous,1979.
Ginet,C. (i), 'MightWe Have No Choice?',in Lehrer(i).
(2), 'The Conditional
AnalysisofFreedom',in vanInwagen(4).
(3), 'A DefenseofIncompatibilism',
PhilosophicalStudies,I983.
Glover,J.,Responsibility,
Routledgeand Kegan Paul, 1970.
Goldman, A. (i), A TheoryofHuman Action,Prentice-Hall,1970.
(2), 'Chisholm's TheoryofAction',Philosophia,
1978.
Greenspan, P. S., 'BehaviorControland Freedomof Action',ThePhilosophical
Review,1978.
Hampshire,S. (i), Thoughtand Action,VikingPress, 1959.
(2), 'FreedomoftheIndividual',HarperandRow,I965.
in Hampshire
(3), 'Two KindsofMorality', (5).