Mind Association

Aristotle on Action
Author(s): John L. Ackrill
Source: Mind, New Series, Vol. 87, No. 348 (Oct., 1978), pp. 595-601
Published by: Oxford University Press on behalf of the Mind Association
Stable URL: http://www.jstor.org/stable/2253695 .
Accessed: 14/02/2011 13:28

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. 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 .
http://www.jstor.org/action/showPublisher?publisherCode=oup. .

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 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
Aristotleon Action

JOHN L. ACKRILL

Aristotle'sstatementsabout action and choice seem to involve serious

inconsistencies-and on topics centralto ethics and to his Ethics.Here
are some samples.'
(a) Aristotleholds that when we choose to do somethingwe always
choose with a view to some end, forthe sake of something;but he
also insiststhat a man who does a virtuousact is not doing it virtu-
ously-is not displayingvirtue-unless he has chosenit 'foritself'.
(b) Actions are done forthe sake of otherthings,and thingswe can do
are not themselvesthe ends with a view to which we do them; yet
action(praxis)differsfromproduction(poiesis),accordingto Aristotle,
preciselybecause it is its own end.
(c) In recommendingthe theoreticallife Aristotlesays that whereas
contemplation'aims at no end beyond itself'fineactions do 'aim at
some end and are not desirablefor theirown sake'; but in recom-
mendingthe life of action he says that doing noble and good deeds
is a thing desirable for its own sake, and that 'those activitiesare
desirable in themselvesfromwhich nothingis sought beyond the
activity'.
Passages likethesesuggesttwoproblems.First,how can actionbe good
in itselfif it is valued as a means to eudaimonia?Secondly,how can an
action be somethingdone to bring about an outcome and yet be dis-
tinguishedfroma productionbecause done for its own sake? The first
probleminvitesdiscussionof Aristotle'sview of moralityand its founda-
tion: is it valuable in itselfor only because it promotessomethingelse?
The second-with which the present note is concerned-calls for an
examinationof Aristotle'sconcept of an action, and of his distinction
betweenpraxis and poiesis.2
Commentatorsdiscussingthis distinctionoften fail to face the real
difficulty, that actions oftenor always are productionsand productions
oftenor always are actions. (The idea that some periods of the day are
occupied by action-episodesand others by production-episodeswould
obviouslybe absurd even if 'production' referredonly to the exercise
of special techniques or skills, since a period of such exercise could
certainlybe a period during which an action, of promise-keepingfor
example, was being performed.In fact however Aristotle'snotion of
I NicomacheanEthics I. I; II. 4; III. 3; VI. 2, 4, 5, I2; X. 6, 7.
a In October 1974 I delivered at the Chapel Hill Colloquium in Philosophy
a paper entitled 'Aristotleon Action'. The firstpart discussed the place
of action in eudaimonia,and it expressedviews similarto those in 'Aristotle
on Eudaimonia', Proceedingsof the British Academy, lx (I974). It is the
second part of that paper that is here published, more or less as it was
given. A number of workersin the Aristotelianvineyardhave suggested
that it would be usefulto have it in print,in spite of its evidentlimitations.
595
 596 JOHN L. ACKRILL:

production is not limited either to technical performancesor to the

makingof materialobjects.) The brave man's action is fightinguphill
to relievethe garrison,and thejust man is payingoffhis debt bymending
his neighbour'sfence. How then is one to understandthe thesis that
paying offa debt is an action but mendinga fence is a production?I
propose to examine one or two passages in which Aristotle'speaksof
choosing to do something'for itself' or of doing somethinghekousiJs
(intentionally),in order to see if theythrowany lighton the problem.
For Aristotleclosely connectsthe concept of praxis with choice; and a
man's actions,properlyspeaking,forwhichhe can be praised or blamed,
are confinedto what he does if not fromchoice at least hekousios.
In NicomacheanEthics II. 4 Aristotleconfrontsa puzzle: how can he
say-as he has said-that men become just by doing just things,when
surelymen who do just thingsare already,eo ipso,just? He firstremarks
that even in the case of skills correctperformancedoes not sufficeto
prove the performer'spossession of the relevantskill. He goes on to
make furtherpoints speciallyrelevantto virtuesas opposed to skills. It
is not enough thatthe thingdone should itselfhave a certaincharacter,
sayjustice,in orderto justifythe inferencethatit is done justlyand that
the agent is a just man; it is necessarythat he should do it knowingly,
choosingto do it foritself,and froma settleddisposition.Actual things
done (pragmata)are called just if theyare such as a just man would do.
But it is not he who does themthatis just, but he who does themin the
way in whichjust men do.
Aristotlethus draws a strongcontrastbetween what is done-which
mighthave been done fromvarious motivesor inadvertently-andwhy
it is done. If inferencesto the characterof the agentare to be made from
the characterof the thingdone, it must have been done 'foritself'.This
last,however,seems to be an unhappyformulation.For the 'actual thing
done' must be some performance-such as mending a neighbour's
fence-which is in fact (in the circumstances)just, thoughit mightbe
done by someone ignorantof or indifferent to its justice. But when it is
asked whetherthe doer chose to do it foritselfthe question is of course
whetherhe chose to do it because it was just, not whetherhe chose to
do it because it was mendinga neighbour'sfence.How can doing some-
thingbecause it is sbbe doing it foritselfor forits own sake unless the
thing done is specifiedpreciselyas sb?Only if the action is designated
not as mendinga fence but as the b act does the expression'for itself'
getthenecessarygrip.Yet the sbact is some such performance as mending
a fence,and it does not seem naturalto say in such a case thatthe agent
has done two things at the same time. It is easy to understandhow
Aristotle,not having addressed himselfto this theoreticaldifficulty,
should have said of an action both thatit is done foritselfand thatit is
done for the sake of somethingelse: the sb act is done for itself,the
mendingof a fenceis notdone forits own sake but forits sbness.
It may be thoughtthatto take mendinga fenceas one's exampleof an
'actual thingdone' is to make it unnecessarilydifficult to interpretthe
requirementthat the just agent should choose to do what he does 'for
itself'.Mendinga fenceis all too obviouslysomethingin itselfunattractive,
ARISTOTLE ON ACTION 597
nor is it by any means alwaysthe just thingto do. However,Aristotle's
positionwould hardlybe easier if an example like repayinga debt were
used. It is not alwaysjust to repay a debt either.In any case, even if
what is done could be given a descriptionsuch thatany such act would
be just, yet such an act would inevitablyhave othercharacteristics too.
What 'foritself'points to will be clear only if the act is brought,before
us preciselyas having the relevantcharacteristic, e.g. as the just act: it
is not enoughthatit should actuallyor even necessarilyhave it.
One way of bringingout the point at issue is to distinguishtwo ways
of understandingthe expression'do somethingthat is O'. It may mean
(thatis
'do something n)',wherethereis no implication
thatthe doer
necessarilyknows or supposes that what he does is sb.Or it may mean
'do something-that-is-sb, where it is implied that the doer knows or
supposes it to be sb(whetheror not he does it becauseit is sb,'foritself').
Aristotlecomes close to this kind of formulationin NicomacheanEthics
V. 8. Here, before contrastingthe characterof what is done with the
characterof the agent (along the same lines as II. 4), he raises a pre-
liminaryquestion: before asking whethersomeone did a just act 'for
itself' (or forulteriormotives)we must ask whetherhe did a just act at
all, properlyspeaking.
Aristotlefirstdistinguishesbetween 'doing a thing that is in fact
wrong'and 'doing-wrong'(adikein,a singleword). To do-wrongis to do
somethingwrongknowinglyand intentionally.If one does what is as a
matterof factwrongbut does not know thatwhat one is doing is wrong
one cannotbe said to do-wrong(save per accidens)(V. 8. i). Later in the
chapter(v. 8.4) Aristotleapplies the same principleto expressionslike
'doing what is right'and 'doing thingsthatare wrong'. A man who has
been compelledto returna deposit cannot be said to have done-rightor
even to have donewhatis right,save per accidens.So it seems thatwhat
a man can be said to have done strictly,withoutqualification,not per
accidens,is what he has done unforcedand knowingly.
The contrastbetween doing something,properlyspeaking,and only
doing somethingper accidens,differsfromthe earliercontrastbetween
doing somethingforitselfand doing it foran ulteriormotive.But here
again what is involvedis a contextthatdoes not permitfreesubstitution
of alternativedescriptionsof the agent's performance.In the strictuse
a man 'does -' only if he 'does - knowingly'.An action of his, then,
is not somethingsome of whose featuresor circumstanceshe may be
ignorantof. Rather it must be definedby featureshe is aware of, since
it is onlyas so definedthathe can be said to have done it knowingly and
hence to have doneit at all (strictlyspeaking).
Aristotleimplies then that nothing a man does unknowinglycan
count as an action of his. Does he recognisethat,since thereare on any
occasion a greatnumberof factsan agentknowsabout what he is doing,
therewill be a greatnumbe' of different ways of characterising what he
is doing knowingly?Does he see that what is done may be subject to
praiseunderone descriptionand blame underanother,or may constitute
one offenceunder one descriptionand a different one under another,or
may invitemoralappraisalunder one descriptionand technicalappraisal
under another?
598 JOHN L. ACKRILL:

In NicomacheanEthicsV. 8, betweenthe sectionsalreadysummarised,

Aristotleexplains what counts as hekousion:'whateverof the thingsin
his power a man does in knowledge and not ignorance of either the
person, the instrument,or the result-e.g. whom he strikes,what he
strikeswith, and with what result-and <knowing> each of them not
per accidens'.' This last requirementis explained by an exarmple:you
may knowthatyou are strikinga man but not knowthatthe man is your
father; so, Aristotleimplies, you do not know whomyou are striking
(your father)save per accidens.Similarly,he adds, as regardsthe result
and the whole action.
Here, then, Aristotletouches on some of the various factorsor cir-
cumstancesof any practicalsituation-whose number and diversityhe
often,of course,stresses;and he uses the notionof knowingper accidens,
a notionthatis essentiallyconnectedwiththe idea thatfreesubstitution
of extensionallyequivalent expressionsis not always permissible.Yet
he conspicuouslyfailsto remarkthatthoughon his account you do not
strikeyour fatherhekousios,you do strikea man hekousios;or that, in
virtue of different known factorsin a given situation,a man may be
accused of-and offerdiverseexcuses for-differentoffences.Aristotle's
mind is clearlyon giving conditionsfor ascribingresponsibilityfor an
act as already specifiedin the accusation: 'he struckhis fathera fatal
blow witha sword'.
We must,however,examine Aristotle'sfulleraccount of actions and
excuses in NicomacheanEthicsIII. i. He startswiththe privativeterm,
akousion:a man can deny responsibility forsomethingdone-claim that
it was akousion-if he can plead forceor ignorance.By 'force' is meant
real physicalforce,where it would in factbe misleadingto say that the
man had done anything-'the arche [originatingprinciple] is outside
and nothingis contributedby the person who acts or ratheris acted
on'. By 'ignorance' is meant ignorance of facts, circumstances,and
consequences,not ignoranceof 'the universal',of whatis good or lawful.
Correspondingto these negativetests forakousion-not due to an arche
in the person,not known-is the positiveformula:the hekousion is 'that
whose originatingprincipleis in the agenthimself,he being aware of the
circumstancesof the action'.
-particular
of the ignorancetest,
Various questions arise as to the interpretation
and Aristotlediscussessome of them. But the point of concernto us he
does not bringout, and indeed his way of speakingservesto conceal it.
I give someone a drinknot knowingit to be poison-I think it will
refreshbut in fact it will kill. Ignorance makes my act akousion.What
act? Clearlywhat I did throughignorancewas to poison my friend,not
to givehima drink.The ignorancethatmakesmyact akousionis ignorance
of a featurethat goes to definethat act and not ignoranceof a feature
that simply characterisesit. Now some of Aristotle's formulations
could perhapsbe construedin such a way as to accommodatethispoint.
When, afterreferringto the various circumstancesof action, he says
I The flow of the sentence is in favour of understanding'knowing' rather
than 'doing' before 'each of them'. (For 'knowing per accidens' see for
example PosteriorAnalytics76a I, 93a 25, 93b 25.)
ARISTOTLE ON ACTION 599
that 'the man ignorantof any of these acts akousios' (III. I.I5), we
mighttake him to mean that correspondingto ignoranceof any factor
therewill be some act the man can be said to have done akousids.And
when he says that that is hekousionwhich a man does in knowledgeof
person,instrument,and result(V. 8.3), we mighttake him to be using
only by way of example a case where the performancein question is
specifiedas the bringingabout of a certain result by using a certain
instrumenton a certainperson. This would then be consistentwith his
allowing that that is also hekousionwhich a man does in knowledgeof
person and instrument(but in ignoranceof result): he struckhis father
hekousios(thoughhe struckhis fathera fatalblow akousids).
These would, however,be veryforcedways of interpreting Aristotle's
words.His own approachis indicatedby the factthat,aftergoingthrough
a numberofthingsofwhichone mightbe ignorant,he saysthatone who
was ignorantof any of these is thoughtto have acted akousios-and
especiallyif he was ignoranton the most importantpoints (III. i.I8).
It is clear that Aristotleis not associatingknowledgeor ignorance of
this,that,or the otherwithvariousact-descriptions involvingthis,that,
or the other,with respect to each of which the question 'did he do it
hekousios?'could be asked. Rather he is asking simplywhethera man
'acted hekousios'on some occasion, and sayingthat he did so only if he
knew all the importantcircumstances.
It is easy to understandwhyAristotleshould have spoken as he does.
In a simple expositionhe considers simple and strikingcases. We all
know what Oedipus did, and we are quite willingto say simplythat he
'acted akousios'.The enormityof the chargeof strikinghis fathera fatal
blow pushes aside any minorinfelicitiesof whichhe may simultaneously
have been guilty,and even submergesthe quite serious charge (which
he mightwell findit harderto evade) of havingstrucka man. In such a
dramaticcase one can ask simplywhethera man 'acted hekousios',or
whetherhe 'did it hekousios',withoutenteringintoor even noticingtheo-
reticalquestionsabout the identification of actions.
It is difficult,however,to see how closer considerationcould have
leftAristotlesatisfiedwithhis way of speaking.For whetherhe identified
the thingdone ('it') with the person's bodily movementM or with the
total package M (a, b, c . .. )-where the lettersin bracketsstand for
various circumstancesetc.-he would find it impossible to raise the
questionsthatwe (and the courts)want to raise. But if he treatedM(a),
M(b), etc. as different things done (perhaps different offences),about
each of which separatelythe question whetherit was done hekousios
could be asked, he could not say that the knowledge required for an
affirmative answerwas knowledgeof all or of the mostimportantfactors
in the situation.The knowledgerequiredforan affirmative answerto the
question about M(a) would be simplythe knowledgethat M would be
M(a).
It mightbe said that,thoughwhata man does on a particularoccasion
mustbe (as it were) takenapartin thisway-the questionabout intention
or 'voluntariness'being directed not at the whole package but at the
elementsin it, M(a), M(b), etc.-, yet what a man does hekousioson a
6oo JOHN L. ACKRILL:

particularoccasion can be treatedas a single action (the action he per-

formed)-say, M(a, g, m ... .), wherethe lettersin bracketsstand forthe
circumstancesetc. known to the agent. Certainly,however much he
disliked some of the circumstances,however much he regrettedthat
doing M(a) would be doing M(g), he did know thatit was preciselythis
package-M(a, g, mi ...) that he was taking,and he took it because on
the whole he wantedto do so ratherthannot.
There are, nevertheless,stillreasonsforpickingM(a, g, m .. .) apart.
Firstly,he may well have to go to differentcourts to meet different
charges in respect of M(a), M(g) etc. In one court M(a) will be the
action complained of, and that it was also M(g) will be, perhaps, a
mitigatingcircumstance.Secondly, even if our knowledgethat he took
the package because on the whole he wanted to make it superfluousto
ask separatelywhetherM(a) was hekousion, whetherM(g) was hekousion,
etc., we may well want to ask with respectto each whetherhe was glad
or sorry(or indifferent) thathe was doing that.Was thatwhatmade him
take the whole package, or was it perhaps an elementhe regrettedbut
had to accept in orderto get some other?He wanted M(a, g, m .. .) on
the whole. Was it perhapsonly (or precisely)M(a) thathe reallywanted?
This takesus back to thefirstpartofAristotle'saccountofthehekousion
-'that whose arche [originatingprinciple] is in the agent himself,he
being aware of the particularcircumstances'.The arche relevantfor
action is no doubt desire,orexis.(For, as Aristotlerecognises,not every
internalarcheleads to performances classifiable-evengivenknowledge-
as hekousia.Many processes of a biological kind are not influencedby
our wishes and desires; they are not hekousiaand they are not akousia
either.V. 8.3. I I35a 33-b 2.) But ofwhatexactlyis desirethe originating
principle?Is it, to use the above crude symbolism,M or M(a) or M(a, g,
m .. .) or M(a, b, c ... .)? Does Aristotle'sgeneralaccountof human and
animal movementthrowany lighton this?
The centralfeaturesofthisaccountare familiar.If an objectofthought
or imaginationbecomes an object of desire a man's facultyof desire is
stimulatedand moves him towards realising or achieving it. Three
'causes'-or explanatoryfactors-are mentionedhere: the final cause,
the object of desire; the efficientcause, the man's actual desire; and the
formalcause, the essence or definitionof the movementproduced. In a
certainway these three 'causes' coincide, as Aristotlesays, for example
in PhysicsB. 3, wherehe takeshis illustrationsfromproductivecrafts.
It would appear thenthatwhatactionpreciselyhas been performed-
what action is genuinelyexplained by the arche in the agent-depends
on what the object of thoughtand desirewas. Unfortunately difficulties
at once arise.Aristotleoftengivesas the object of desire(or of its species,
appetiteand wish) a characteristic (like the pleasant,the noble), and not
somethingthat could strictlybe done. When he does speak of what we
may want to do he is naturallyoften concerned with cases in which
deliberationis involved,where one thingis done as a means to another
or wherethe pros and cons of a course of actionhave to be weighedup.
So an immediatedistinctionpresentsitselfbetweenwhat one primarily
wantsto do and what one wantsto do derivatively, insofaras one thinks
ARISTOTLE ON ACTION 6oi
it necessaryto achieve one's real aim. Should we then say thatwhat we
reallywant to do, or want to do withoutqualification,is only what we
wantnon-derivatively to do? Aristotlecomes close to thisin Nicomachean
EthicsVII. 9.I: 'if a person chooses or pursues this forthe sake of that,
per se it is that that he pursues and chooses, but per accidensit is this.
But when we speak withoutqualificationwe mean what is per se'. This
suggestsa seriesor hierarchyof descriptionsof what a man does because
he desires to, each successive descriptioncoming nearer to revealing
exactlywhat he aims at. In our simple case, M(a, g. m . . ) comes first,
and is followedby M(a): his wantingthe package was derivativefrom
his wantingM(a). But ifit is alwaysforsome desirablecharacteristic that
a possible line of action appeals, there will be M(?b) afterM(a) in the
series. Desire is then the archeof M(a, g, m ... .), M(a), and M(?b)-but
primarilyof the last and only derivativelyof the others.
It is clear,I think,thatwhatAristotlesaysaboutdesireas theoriginating
principleof action does not provide an answerto the sortsof question
about actions and action-descriptions that were leftunansweredby his
discussionsof responsibility.Moreover his account of the physiology of
animal movement,which shows how desire operatesas a physical(non-
intentional)process leading to muscular and limb movements (how
desireis in a waythearcheofM), givesno clue as to how thephysiological
storyis connectedto the psychologicalone, or how questions about the
individuationof movementsare relatedto questionsabout the individua-
tion of actions.
I conclude thatwhile Aristotlehas much to tell us about the responsi-
bilityforactions,the motivesof actions,and the physiologyof actions,
he does not direct his gaze steadily upon the questions 'What is an
action?' and 'What is an action?'. It is not thatsuch questionswould be
beyondhim.He revelsin questionsofthiskind,and he has the conceptual
and linguisticequipmentneeded to tackle them. Whateverthe reasons
why he did not tacklethese questions head-on, it seems likelythat this
failureis itselfthe reason for many of the 'incoherences'and 'contra-
dictions'to be foundin passages such as those I quoted at the beginning.

OXFORD UNIVERSITY