Philosophical Review
Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty. by Stephen Engstrom;
Jennifer Whiting
Review by: David O. Brink
The Philosophical Review, Vol. 108, No. 4 (Oct., 1999), pp. 576-582
Published by: Duke University Press on behalf of Philosophical Review
Stable URL: http://www.jstor.org/stable/2998293 .
Accessed: 12/03/2014 16:36
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.
.
Duke University Press and Philosophical Review are collaborating with JSTOR to digitize, preserve and extend
access to The Philosophical Review.
http://www.jstor.org
This content downloaded from 128.54.22.253 on Wed, 12 Mar 2014 16:36:54 PM
All use subject to JSTOR Terms and Conditions
BOOK REVIEWS
ThePhilosophical
Review,Vol. 108, No. 4 (October 1999)
ARISTOTLE, KANT, AND THE STOICS: RETHINKING HAPPINESS AAD
DUTY By STEPHENENGSTROMand JENNIFER
WHITING,eds. Cambridge:
Cambridge UniversityPress, 1996. Pp ix, 310.
This collection of essayscontains revisedversionsof papers delivered at a
conference entitled "Duty,Interest,and PracticalReason: Aristotle,Kant,
and the Stoics" that was organized by Stephen Engstrom and Jennifer
Whitingat the Universityof Pittsburghin 1994. One of the main aims of
the conferencewas to bring togetherscholarson Aristotle,the Stoics,and
Kant to reevaluate the common view that Greek and Kantian ethics representfundamentallyopposed conceptions of ethical theoryand the roles
of moralityand happiness in practicalreasoning.Accordingto a common
view,the ancients are eudaimonists; they derive or justifythe virtuesby
showinghow theycontributeto the agent's own eudaimoniaor happiness.
By contrast,Kant sharplycriticizeseudaimonism for derivingor justifying
moralityin termsof happiness. This criticismapplies to eudaimonism of
all sorts,even Stoic eudaimonism,whichis perhaps closer in some respects
to Kant's own views,and of which he is somewhatless criticalthan he is
of other formsof eudaimonism. For Kant, moral dutyand respectforthe
moral law must be grounded in reason itselfand cannot be made to depend on any independent standard.These and related assumptionsabout
ancient and Kantian ethicshave helped structuremuch contemporarysystematicwork in ethical theory,as well as common conceptions of these
ethical traditions.But this common view has been under reexamination
lately;some of the most interestingworkin the historyof ethics in recent
yearshas been in Greek and Kantian ethics,and much of it challenges one
or another aspect of the received view of the ethical theoryand moral
psychologyof Kant or the Greeks.However,withsome exceptions,renewed
interestand recentworkin these two traditionshas proceeded in parallel.
The conference aimed to correct this,by bringingtogethersome of the
most distinguishedscholarsof ancient and modern ethicsto compare and
assess the role of moral dutyand happiness in the two traditions.Most of
the essays have such a comparativeassessmentas their main theme; but
even those thatfocus more exclusivelyon one of the traditionscontribute
indirectlyto thiscomparativeassessment.
Aftera useful introductionby Engstrom and Whiting,the essays are
divided into five pairs and topics.John McDowell discusses deliberation
and moral development in Aristotle'sethics,while Barbara Herman discusses related themesin Kant's ethics.Terence Irwinexplains and critically
assesses Kant's criticismsof eudaimonism,while Stephen Engstromcompares Aristotle'sconception of eudaimoniaand Kant's conception of the
576
This content downloaded from 128.54.22.253 on Wed, 12 Mar 2014 16:36:54 PM
All use subject to JSTOR Terms and Conditions
BOOK REVIEWS
highest good. Allen Wood examines the differentattitudesAristotleand
Kant have toward self-love,whereas JenniferWhitingdefends a Kantian
reading of Aristotle'sview about the relation betweenvirtueand external
goods. ChristineKorsgaard compares Kant's claims about the good will
and Aristotle'saccount of virtuousaction, whileJulia Annas contrastsAristotelianwithStoic and Kantian viewsabout the role of moralityin practical reasoning and the unityof practical reason. John Cooper examines
the metaphysicaland cosmological basis of Stoic conceptions of virtueand
happiness, while Jerome Schneewind compares featuresof Kantian and
Stoic ethics to defend a version of the received view that contrastsKant's
deontologywiththe teleologyinherentin Greek eudaimonism.
McDowell defends an intuitionisticand particularistinterpretationof
Aristotle'sconception of practicalreason byattackingwhathe calls a "blueprint" model as a philistineprejudice. The blueprintmodel of practical
reasoning apparentlytakes a simple syllogisticformthatappeals to exceptionless generalizations: (1) All Fs are forbidden; (2) x is F; (3) hence,
avoid x. In defense of his own intuitionisticinterpretation,McDowell appeals to Aristotle'swell-knownremarksabout the inexactnessof political
and ethical sciences,which admit of generalizationsthathold onlyusually
or forthe mostpart (forexample, NE 1094b3-27). We can sympathizewith
McDowell's rejection of conceptions of practical reasoning that aim for
mechanical decision procedureswithoutembracingextremeparticularism
or intuitionism.For one thing,particularistswho eschew generalization
have some difficulty
explaining how the practicallywise person should deliberate about novel or especiallyperplexing situations,to which her existing sensibilityis in some sense inadequate, or how moral education,
which involvescreation of a moral sensibility,
is to proceed. Moreover,we
can be skepticalabout the possibilityof exceptionlessnormativegeneralizationswithoutembracingparticularismifwe understandnormativegeneralizationsas containingineliminable ceteris
paribusclauses. As long as it
is possible to say interestingthingsin advance (and not merelyposthoc)
about the kinds of situationsin which other thingsare not equal, we can
maintain a formof generalism.Indeed, somethinglike this is a plausible
view about the nature of generalizationsin many contexts.This sort of
generalismwould secure a kind of uncodifiabilitythatMcDowell seems to
want and provide a natural explanation of the sortof inexactnessAristotle
recognizes in ethics; it would avoid the crudeness of the blueprintmodel
thatMcDowell castigateswithoutrequiringparticularism.Indeed, itwould
allow for informativegeneralizationsthat could be used by an agent in
deliberatingabout novel or perplexingcases and could playa role in moral
education, thus avoiding some of the difficultiesthat seem to beset extremeparticularism.
Herman continues her effortsat articulatinga "kinder and gentler"
577
This content downloaded from 128.54.22.253 on Wed, 12 Mar 2014 16:36:54 PM
All use subject to JSTOR Terms and Conditions
BOOK REVIEWS
Kantian moral psychologyin which moral characterassigns an important
role to nonrationalfaculties,including emotions and desires. She focuses
on the relation between reason and desire. Desire can be what moves us
to action, withina Kantian account of moral motivation,provided the desire is reason-responsive.Desires are reason-responsivewhen theirweight,
scope, and objects are suitablyregulated by an agent's perception of the
moral law. Forms of deliberationand characterthatproceed fromreasonresponsivedesires need not be inimical to virtue.
Kant criticizesall formsof eudaimonism, including Stoic versions,for
subordinatingmoralityto inclinationand an independent standard (KpV
20-28, 35-36, 111-12). In a careful and rewardingdiscussion of Kant's
criticismof eudaimonism and its relation to Reid's assessmentof Greek
ethics,Irwin argues thatGreek eudaimonistsdid not subordinatemorality
to inclination,because they treated eudaimoniaas supplyingexternal or
categorical reasons. And because Aristotleand the Stoics viewed virtueas
a component of, rather than an instrumentalmeans to, eudaimonia,the
sense in which theysubordinate moralityto an independent standard is
not obviouslyproblematic. Indeed, Irwin seems to think that the eudaimonistmakes room, as Kant does not, foran explanation of whyan agent
should care about moral demands.' But Kant need not appeal to eudaimonism to vindicatethe authorityof moralityif these moral demands are
grounded in the capacities of rational agents as such. However,even if
eudaimonism is not necessaryto vindicatemorality'sauthority,
it mayhelp
to vindicate its supremacyinsofaras it can help reconcile categorical demands to be concerned with one's own rational agency and categorical
demands to be concerned withrational agents as such.
Engstromprovides an instructivecomparison between Aristotle'sconception of eudaimoniaand Kant's conception of the highestgood. Despite
Kant's criticismsof eudaimonism,Engstromidentifiesimportantsimilarities in the content of these two conceptions and the role they play in
Aristotle'sand Kant's ethical theories.Aristotle'saccount of eudaimoniais
an account of a final or highest good. Both view the highest good as a
complex whole consistingof virtueand other goods, in which virtueis a
condition of these other goods havingvalue. In his valuable emphasis on
similarities,however,I think Engstrom may underestimatedifferences.
Though virtueis a complete good, forAristotle,it is not an unconditionally
complete good; it is chosen forthe sake of eudaimonia.Moreover,Aristotle
specifiesvirtuein termsof the human function.But then even ifvirtueis
the controllingingredientin Aristotle'sconception of eudaimonia,virtueis
'Also see Terence Irwin,"Moralityand Personality:Kant and Green," in Selfand
Naturein Kant's Philosophy,ed. Allen W. Wood (Ithaca: Cornell UniversityPress,
1984).
578
This content downloaded from 128.54.22.253 on Wed, 12 Mar 2014 16:36:54 PM
All use subject to JSTOR Terms and Conditions
BOOK REVIEWS
posteriorin justificationto eudaimonia,and this appears to violate Kant's
strictureson the relation between virtueand the good. Moreover,I think
one may wonder whetherthe value of external goods for Aristotleis, as
happiness (Gliickseligkeit)
is for Kant, entirelyconditional on virtue.Some
external goods seem to contributeconstitutively
to a complete good independentlyof their role in virtue (1099b1-7, 1100a6-8, 1100b27), and
it's not clear thatAristotlerestrictstheirvalue to those who lead virtuous
lives. (However,Whiting'sclaims about the kaloskagathos
go some way toward defendinga Kantian reading of Aristotlehere.)
Aristotlecontrastsvulgarand proper self-loveand insiststhatthevirtuous
person is the true self-lover(1 168b28-34). By contrast,Kant thinksthata
will that is determinedby the principle of self-loveis heteronomous and
cannot provide a foundationformorality(KpV 34). Wood's essayis a nice
examination of the claims about self-worth
and human nature on which
Kant's reservationsabout self-lovedepend.
Whitingexplores Aristotle'ssuggestionin the EudemianEthics(8.3) that
so-called natural goods are intrinsicallyneither good nor bad, insofaras
theycan be misused by,and prove harmfulto, the nonvirtuousagent.This
thesisechoes Socratic claims in Plato's Euthydemus
(280a-81e) and prefigures Kant's reasons in the Groundwork
for claiming that a good will is the
only thinggood withoutqualification(G 393-94). In the course of exploring this thesis,Whitingargues thatAristotle'sdistinctionbetween the agathos(the good person) and the kaloskagathos
(the fine-and-goodperson)
vircorrespondsto his distinctionbetween naturalvirtueand authoritative
tue, in which the authoritatively
virtuousagent is responsible for his own
fate in a wayin which the naturallyvirtuousagent is not.
Korsgaardaddresses apparent differencesbetweenAristotle'sand Kant's
conceptions of moral motivationand virtuousaction that are consequent
upon apparent differencesin theirassessmentsof the naturallysympathetic
person. Aristotlemayseem to thinkthe naturallysympatheticperson could
be virtuous,whereas Kant may seem to deny thispossibility.But Korsgaard
argues that these apparent differencesare illusory,because both Aristotle
and Kant insistthatvirtuousactions must be chosen for theirown sakes.
Kant famouslyclaims that a good will requires not just thatan agent conform to dutybut that she do so froma sense of duty (G 390). An action
in conformity
withinclinationcan nonethelessexpressa good willbut only
if the agent's choice of actions is suitablyregulated by her beliefs about
to dutyacts
what's morallyrequired. In such cases, it's not thatconformity
as the agent's ultimateaim; it's ratherthatit servesas a constraintor filter
on eligible motivations.Kant focuses on the case of the person who acts
from a sense of duty in the absence of supportinginclinationor in the
presence of recalcitrantinclination,not because she alone has a good will,
but because the operation of the sense of dutyis especially clear in her
579
This content downloaded from 128.54.22.253 on Wed, 12 Mar 2014 16:36:54 PM
All use subject to JSTOR Terms and Conditions
BOOK REVIEWS
case. According to Korsgaard,Aristotle'sposition is essentiallythe same as
Kant's. For, according to Aristotle,the virtuousperson must choose virtuous actions because they are fine (kalon). So there is a sense in which
Aristotle,as well as Kant, must say that the person who is beneficentor
generous merelyfrominclination,in a way not regulated by the requirements of virtue,fails to be virtuous.If there is a residual sense in which
Aristotleis more inclined to regard the person who is reliablybeneficent
as generous, it is because, Korsgaard thinks,Aristotle,unlike Kant, treats
as perceptionsof normativereasons. But this
emotions,such as sympathy,
is more of a psychologicalthan a normativedisagreement.While there is
much in Korsgaard's account to agree with,I wonder if she overstatesthe
similaritiesbetween Aristotelianand Kantian conceptions of virtueby underestimatingtheirdisagreementabout the need foraffectand inclination
withinan account of virtue.For whereas Kant maynot thinkthatthe presence of inclinationdisqualifiesan agent fromdisplayinga good willor that
a good willrequirescontramoralinclination,nonethelesshe seems to think
that the person in whom a sense of dutymust overcome indifferenceor
contraryinclination (the grudgingmoralist) can and does displaya good
will (G 398). But Aristotlewould certainlythinkthatthe grudgingmoralist
displayscontinence,ratherthan virtue,inasmuch as he thinksthatit is the
markof the virtuousperson thathe does not experience a conflictbetween
the rational and nonrationalparts of his soul and that appetite and emotion harmonize with rightjudgment (1102bl4-28). Korsgaard discusses
thisissue brieflybut, I think,underestimatesits significance.
Annas explores the distinctivefeaturesof Aristotle'srelation to Kant in
part by seeing how he (Aristotle)differsfrom the Stoics. Whereas Kant
and the Stoics distinguishsharplybetween moral and nonmoral goods,
puttingthem in effecton separate scales, Aristotleincludes both moral
and nonmoral goods withineudaimonia.Annas suggeststhatAristotle'sunified conception of value and practical reasoning is less mysteriousand
closer to common sense. But, she thinks,Aristotle'sattemptto preserve
endoxacomes at a price; once we concede that happiness includes nonmoral goods, it becomes hard to explain the primacythat he assigns to
moral virtuewithinhappiness. One mightwonder about Annas's contrast
between Kant and Aristotlehere. If one pays attentionto Kant's claims
about the role of virtueand happiness withinthe highestgood (KpV 11019) and argues, as Whitingand Engstromdo, thatforAristotlevirtueis a
condition of the value of external goods, one might suppose that both
Aristotleand Kant recognize nonmoral goods as well as moral goods but
only such nonmoral goods as are conditioned by moral goods. Moreover,
one might wonder whether Aristotle'sconstraintson eudaimonia,introduced in NE 1, don't providesome resourcesfordefendinghis claim about
the controllingrole of virtuewithineudaimonia.The functionargument
580
This content downloaded from 128.54.22.253 on Wed, 12 Mar 2014 16:36:54 PM
All use subject to JSTOR Terms and Conditions
BOOK REVIEWS
(NE 1.7) implies thathappiness consistslargelyin a lifeof activityexpressing one's rational capacities; the life of virtueis such a life,whereas a life
of nonmoral goods is not (1099b25-28). Moreover,it seems clear thatAristotlethinksthathis requirementsthateudaimoniabe comparativelystable
and withinour controlimplythatvirtuemustbe the controllingingredient
in happiness (NE 1099bl8-20, 1100bl-7, 12-15; EE 1215a8-18). Indeed,
insofaras Aristotleidentifiesand individuatesvirtuesin termsof theircontributionto the agent's eudaimonia,the most pressingquestion is not so
much whyvirtueis the controllingingredientin eudaimonia,as whyfamiliar
and justice, are genuine
traits,such as courage, generosity,
other-regarding
virtues.
Cooper argues thatwe cannot reach a fairunderstandingand assessment
of what is distinctiveabout Stoic ethics withoutappreciating the way in
which Stoic ethical viewsdepend upon theirviews about nature and cosmology.The Stoics identifyeudaimoniawitha life lived in accordance with
nature,which theyidentifywith a life lived in accordance withreason or
virtue. However,Cooper thinksthat this is not just an endorsement of
Aristotle'sfunctionargument.Stoic beliefsabout the rational and benevolent characterof natureprovideadditional assurance thatthe lifeofvirtue
is beneficial.Indeed, the Stoic belief thatthe world is ordered forthe best
view thatvirtueand happihelps explain theirotherwisecounterintuitive
ness are identical. For if virtueexhausts happiness, it is hard to see how
the demands of virtuecould call on us to sacrificegenuine goods. This is
a paradoxical view.But it becomes more intelligibleif we remember that
the Stoics thinkthatnature is ordered for the best and thatvirtueis a life
in accord with nature. For then what might appear to be costs of virtue
can be seen to be essentialpartsof the verybest whole. However,I wonder
if Stoic theodicyjustifiesStoic claims about happiness. For even compensated sacrificesappear stillto be sacrifices;yetthe Stoics cannot admitthat
virtueexacts a price, even (as Plato and Aristotleclaim) one worthpaying.
Also, one may wonder whether the agent is alwayscompensated for her
sacrificeseven if theyare elements of the verybest whole. Others or even
nature as a whole may be betterfor the agent's sacrifices,but thatdoesn't
implythat theymake herbetteroffunless we define the good of the part
in termsof the good of the whole. I am not sure thatStoic theodicyitself
justifiesthis organic conception of individualhappiness.
Schneewind attemptsto preserve the contrastivereading of Kant as deontologistand the ancient Greeks as teleologistsby criticizingtwo recent
attempts(by Guyerand Herman) to read Kant as appealing to the value
of a rationalwill and by contrastingKant's appeal to respectforthe moral
law and Stoic appeal to the perceptionof naturalgoods. I am not sure that
the distinctionbetween doing somethingout of respect for the moral law
and doing it because it is perceived to be good fora rationalagent is clear
581
This content downloaded from 128.54.22.253 on Wed, 12 Mar 2014 16:36:54 PM
All use subject to JSTOR Terms and Conditions
BOOK REVIEWS
or strongenough tojustifySchneewind'sversionof the traditionalcontrast
between the Greeks and Kant.
This collection of essaysrepresentsan importantand stimulatingcontributionto the historyof ethics. The qualityof the essaysis consistently
high; as a rule, theydefend interestinginterpretiveand systematicclaims
in resourcefulways,marshal evidence fromthe textspersuasively,and significantlyadvance the level of discussion on the topics they address.
Though each of the essays repays careful reading, some of them-especially those by Irwin, Engstrom,Wood, Whiting,Korsgaard,Annas, and
Cooper-strike me as especiallyrewarding.StudentsofAristotle,the Stoics,
and Kant will certainlywant to studythese essaysnot onlyfortheirinsights
about these figuresbut also to achieve a bettercomparativeunderstanding
and assessmentof these figures.Nor should readership for this book be
confined to specialistsin the historyof ethics; those whose primaryinterests are in systematicethicsshould also read these essayswithconsiderable
interestand profit.
DAVID
University
ofCalifornia,San Diego
0.
BRINK
ThePhilosophical
Review,Vol. 108, No. 4 (October 1999)
POLITICS: BOOKS VII AND VIII. Clarendon AristotleSeries. ByARISTOTLE.
Trans. RICHARD KRAUT. New York: Oxford UniversityPress, Clarendon
Press, 1997. Pp. x, 229.
Despite itsimportanceboth historicallyand withinthe Aristoteliancorpus,
scholarshipon Aristotlein the past
and despite the outpouringof first-rate
thirtyyears,the Politicshas received much less attentionthan it deserves.
This situationis, however,beginning to be rectified.The magisterialfourvolume nineteenth-century
commentarybyW. L. Newman has been joined
in recent years by numerous new translationsas well as commentariesby
Richard Robinson with supplementarymaterial by David Keyt on Politics
III-IV, TrevorSaunders on PoliticsI and II, P. Simpson on the whole of the
on PoliticsI-VI.' Richard Kraut's
Politics,and in German by E. Schfitrumpf
4 vols. (Oxford: Oxford UniversityPress,
'W. L. Newman, ThePoliticsofAristotle.
PoliticsBooksIII and IV 2d ed., with
Clarendon Press, 1887-92); R. Robinson,Aristotle
a supplementaryessay by D. Keyt (Oxford: Oxford UniversityPress, Clarendon
PolitikBuchI (Berlin:AkademieVerlag,1991);
Aristotles:
Press,1995); E. Schfitrumpf,
E. Schvitrumpf,
Aristotles:
PolitikBuchII-III (Berlin:Akademie Verlag,1991); E. SchilPolitikBuch IV-VI (Berlin: Akademie
trumpfand Hans-JoachimGehrke, Aristotles:
Politics:A Philosophical
Commentary
(Chapel Hill:
Verlag,1997); P. Simpson, Aristotle's
PoliticsBooksI and
Universityof North Carolina Press, 1998); T. Saunders, Aristotle's
II (Oxford: Oxford UniversityPress,Clarendon Press, 1995).
582
This content downloaded from 128.54.22.253 on Wed, 12 Mar 2014 16:36:54 PM
All use subject to JSTOR Terms and Conditions