RffiA,$Gh{

ANM

E{HSPGT{SHffigLKTY

READINGS

IN SOh4E BASXC PROBLEMS

OF'PHILOSOPFNY

T hi rd E d i t ton

edi ted by

Joel Feinberg

T ke Rockefe l le r U n iv ers i tY

DICKENSON PUBI"ISHING COMPANY, INC., ENCINO, C,ALIFORNIA,AND BELMONT, CALIFORNIA

GFCar)

CoPYright A 1965, 1971, 1975

by Dickenson Publishing Company' Inc'

Att ighrc reserved. No part of this book may be

reproduced, stored in a retrieval systetn' or

transcribed, in any form or by any means'

electronic, mechanical, photocopying' recording'

or otherwise, without the prior written permission

of the publisher, 16561 Ventura Boulevsrd' Encino'

Californiq 91316'

Library of Congress Cotalog Card Number: 74-22624

ISBN: 0-8221-0144*X

Printed in the United States of America

1 2 3 4 5 6 7 8 9 1 0

,lPART I neaSoN AND RELrcrous BELTEF

I

CONTENTS

Preface xlll

Introduction 2

The Existence and Nature of God

SAINT ANSELMThe Ontological Argument, from proslogium 6

WILLIAM L. ROWE"The Ontological Argument" 8

SAINT THOMAS AQUINASThe Five Ways, from Summa Theologica 18

F. C. COPLESTONOn St. Thomas's Proofs, lrom Aquinas 19

, '

OeVfp HUMEDialogues concerning Natural Religion, ll-Xl 26

ANTONY FLEW, R. M. HARE, BASIL MITCHELLSymposium on Theology and Falsif icat iorr 59

Religious Experience and Mysticisnt

WILLIAM JAMESMysticism, front The I/arie rics of Raligious l".tpcrtcnct 65

Cover design: Preston J' Mitchell

5c0 SELF.LOVE AND THE CLAIMS OF MORALITY

olaced mankind under the governance oftwo sovereign mas-

i;;;.;;;;tpleasure. It iJ for them alone to point out what

i;; fi;i;i;d;;; well as well as to determine what we shall.i" " . lTrt.

"tii,ciple of utilitv recognises this subjection' and

;.";";;r^iir;il;lbil"id of t-hat svstem, the obiect of

ffiil to-i.ui trt. fabric of felicity by the hands.of reason

;il";iil. atstems rvhich attempt.t" 9l:t-tl:i^:t',9'il i"

louna. inrt.uO of sense, in caprice instead of reason' in dark-

".rt'i"J..J "r light.'-Bentiram's Principles of Morals and'triii,'ir,

J' il'sv the principle of utilitv is meant that

niiirciple.vhich approves or disapproves ofevery action what'

#;;, ;;;;;l"i io tt e tendeniv rvhich it appears to have to

ausmcnt or diminish the happiness of the party $hose.lnteresl

is irr qucsrion'-Ibid. 'Je regarde I'amour eclalre de nous-

;tr;;;;;; le principe dJtout sacrifice moral.'-['I look

un"n ."i*tt,.n.d ielf-inierest as the fundamental principle of

,ii *"irii"-ince''l-D'Alembert quoted by D' Stewart' lc-

tive and LIorol Povers, vol' i' p' 220'

2. 'Plcasure is in itself a good; nay, even setting asideimnrunitv from pain, the only good; pain is in itselfan evil'rtrJ,

"tlJ.o, rvithout exception, ihe only evil, or elsj the words

gnod and cvil have no meaning.'-Bentham's Principla ofllorals and Lcgislation, ch, x.

3. 'Good and evil are nothing but pleasure and pain' or

that which occasions or procures pleasure or pain to us' Moralgcxxl and evil thcn is orily the conformity or disagreement ofiur loluntary aclions to some larv whereby good or evil isdrrrn on us by the rvill anrl porver of the law maker' which3ood and o'il, pleasure or pain, attending our observance orbrcach of the lirv by the decree of the larv maker, is that rvecell rcsard or punishment.'-Locke's EssaH book ii. ch.rlur 'Takc away plcasures and pains, not only happiness,brrt ;urtrcc, and duty, and obligation, and virtue, all of whichbrrc bccn so claborately held up to vierv as independent ofihtm. rre so mrny empty sounds.'-Bentham's Springs of4ora ch. r. $ 15.

| ..'ll lur 6t aussi impossible d'aimer le bien pour le bien,(F tf rrnrr lc mal pour lc mal.'-[.It is just as impossible forI ro knc arxxl for irs own sake as it is to loveevil for its-'r r.Lc'l_llclrirru. Dc l,Esprit, disc. ii ch. v,

_: "lT lhc Soodnss uhich rve apprehend in God Al--...."1 3-oodn6s ro us.'_Hobbes On Human Nature,-J-t-l E ;rlctlrnd,'To lovc God is in effect the sameE-::.:,rf'ntncs(. ctcrnal happincss; and the love of' sru lhc lorc 6J 6urr.1t.s.'-Third Sernon on

or in a life to come. These may be also called the pains ofreligion.'-Ibid.

8. 'There can be no greater argument to a man ofhis ownpower, than to lind himself able not only to accomplish hi5own desires, but also to assist other men in theirs; and this isthat conception wherein consisteth charity.'-Hobbes, OnHum, Nat, ch. ix. $ 17. 'No man giveth but with intention ofgood to himself, because gift is voluntary; and of all voluntaryacts, the object to every man is his own good''-Hobbes'Leviathan, part i. ch. xv. 'Dream not that men will move theirlittle finger to serve you, unless their advantage in so doing beobvious to them. Men never did so, and never will whilehuman nature is made of its present materials.'*Bentham'sDeontology, vol. ii. p. 133.

9, 'Pity is imagination or fiction of future calamity toourselves, proceeding from the sense of another man'scalamity. But when it lighteth on such as we think have notdeserved the same, the compassion is greater, because therethen appeareth more probability that the same may happento us; for the evil that happeneth to an innocent man mayhappen to every man.'*Hobbes, Ort Hum' NaL ch. ix' $ 10.'La piti6 est souvent un sentiment de nos propres maux-dansles maux d'autrui. C'est une habile pr6voyance des malheursoi nous pouvons tomber. Nous donnons des secours auxautres pour les engager ir nous en donner en de semblablesoccasions, et ces services que nous leur rendons sont' e propre'ment parler des biens que nous nous faisons d nous'mdmes paravance.'-['Pity is often a perception of our own troubles inthe troubtes ofothers, It is an astute prescience of the misfor-tunes into which we might fall' We help others in order tocommit them to helping us under similar circurnstances; andthese services rve render them are, properly speaking' serviceswe render to ourselves in advance.']-La Rochefoucauld'Maximes, 264. Butler has remarked that if Hobbes' accountwere true, the most fearful would be the most compassionatenature; but this is perhaps not quite just, for Hobbes' notionof pity implies the union of two not absolutely identical'though nearly allied, influences, timidity and imagination.The theory of Adam Smith, though closely connected with'diff'ers totally in consequences from that of Hobbes on thispoint. He says, 'When I condole rvith you for the loss of yourion, in order to enter into your grief, I do not consider whatI, a person of such a character and profession, should sufferif I had a son, and if that son should die-I consider what Ishould suffer if I was really you. I not only change circum-stances with you, but I change persons and characters. Mygrief, therefore, is entirely upon your account. . ' ' A man mayiympathise with a woman in child-bed, though it is impossiblehe should conceive himself suffering her pains in his ownproper person and character.'-Moral Sentiments, part vii'ch. i. $ 3.

10. Ce que les hommes ont nomm6 amiti6 n'est qu' unesociet6, qu'un mdnagement r6ciproque d'int6r6ts et qu'un€change de bons offiies. Ce n'est enfin qu'un commerce oirI'amour-propre se propose toujours quelque chose ir gagner.'-['What men have called "friendship" is but an "associa'tion", a reciprocal management of interests, an exchange ofgood offices. It is finally, nothing but a form of commercewhere self-interest always expects to gain'.]-La Rochefou-cauld, Mcx. 83. See this idea'developel at hrge in Helvdtius.

JOEL FEINBERG

Psychological Egoisrn*

A. THE THEORYl. "Psychological egoism" is the name given

to a theory widely held by ordinary men, and atone time almost universally accepted by politicaleconomists, philosophers, and psychologists, ac-cording to which all human actions when prop-erly understood can be seen to be motivated byselfish desires. More precisely, psychologicalegoism is the doctrine that the only thing anyoneis capable of desiring or pursuing ultimately (asan end in itsel{) is his owz self-interest. No psy-chological egoist denies that men sometimes dodesire things other than their own rvelfare-thehappiness of other people, for example; but allpsychological egoists insist that men are capableof desiring the happiness of others only whenthey take it to be a means to their own happiness.In short, purely altruistic and benevolent actionsand desires do not exist; but people sometimesappear to be acting unselfishly and disinterestedlywhen they take the interests of others to be meansto the promotion of their own self-interest.

2. This theory is called psycltological egoism toindicate that it is not a theory about what oughtto be the case, but rather about what, as amafierof fact, l3 the case. That is, the theory claims tobe a description ofpsychological facts, not a pre-scription ofethical ideals. It asserts, however, notmerely that all men do as a contingent matter of

tFrom materials composed for philosophy $tudents atBrown University, 1958. Not previously published.

fact "put their own interests first," but also thatthey are capable of nothing else, human naturebeing what it is. Universal selfishness is not justan accident or a coincidence on this view; rather,it is an unavoidable consequence ofpsychologicallaws.

The theory is to be distinguished from anotherdoctrine, so-called "ethical egoism," according towhich all men ought to pursue their own well-being. This doctrine, being a prescription of whatought to be the case, makes no claim to be apsychological theory of human motives; hencethe word "ethical" appears in its narne to distin-guish it from psychological egoism.

3. There are a number of types of motives anddesires which might reasonably be called "egois-tic" or "selfish," and corresponding to each ofthem is a possible version of psychologicalegoism. Perhaps the most common version of thetheory is that apparently held by Jeremy Ben-tham.l According to this version, all personshave only one ultimate motive in all their volun-tary behavior and that motive is a selfish one;more specifically, it is one particular kind ofselfish motive-namely, a desire for one's ownpleosure. According to this version of the theory,"the only kind of ultimate desire is the desire toget or to prolong pleasant experiences, and toavoid or to cut short unpleasant experiences foroneself."2 This form of psychological egoism isoften given the cumbersome name-psJrc hologic alegoistic hedonism.

ro thc drspleas_r.o ccrtrrn painsatUr 11 q;i'ii1i

5O2 SELF-LOVE AND THE CLAIMS OF MORALITY

ts. PRIMA FACiE R.EASONS INSUPPOR.T OF' THE TI.IEOR.Y

4. Fsychological egoism has seemed plausibleto many people for a variety of reasons, of whichthe following are typicalr

"Every action of mine is prompted by motivesor desires or impulses which are my motives andnot somebody else's. This fact might be ex-pressed by saying that whenever I act tr am al-rvays pursuing my own ends or trying to satisfymy orvn desires. And from this we might pass onto-'tr am always pursriing sornething for myselfor seeking my own satisfaction,' tr-Iere is ivhatseems like a proper description of a man actingselfishiy, and if the description applies to allactions of all men, then it lbllows that all rnenin ail their actions are selfish."8

It is a truism that r,vhen a person gets rrrha-t hewants he charaeteristicaiiy feels pleasure, Thishas suggested to many ireople that r,vhat we re-ally warrt in every case is our own pleasure, andthat we pursne other things only as a means.

Se(-freception Often we deceive ourselves intothinking that we desire something fine or noblewhen what rve really vrant is to be thought wellof by ottrers or to be able to congratulate our-selves, or to lre able to erijoy the pleasures of agood conscience. It is a well-known fact thatpeople fend to conceal their true motives fromthemselves by camouflaging them with wordslike "virtue," "duty," ete. Since we are so oftenrnisled eoncerning troth our own t'eal motivesand the real motives of others, is ii not t'eason-able to suspect that we might always be deceivedwhen rve think motives disinterested and altruis-tic? Indeed, it is a simple matter to explain awayall aiiegedly unselfish motives: "Once the con'viction that selfishness is universal finds root ina person's mind, it is very iikely to burgeon outin a thou-sand corroborating generalizations, trtwill be discovered that a friendly smile is reallyonly an attempt to win an approving nod froma more or less gullible recording angel; that acharitable deed is, for its performer, only an op-portunity to congratulate himself on the goodfortune or the cleverness that enables him to becharitable; that a public benefaction isjust plaingood business advertising. It will emerge thatgods are worshipped only because they indulgemen's selfish fears, or tastes, or hopes; that the'golden rule'is no more than an eminently soundsuccess formula; that social and political codes

are created and subscribed to only because theyserve to restrain other men's egoism as much asone's own, morality being only a speciai sort of'racket' or intrigue using weapons of persuasionin place of bombs and machine guns. Under thrsinterpretation of human nature, the ca.tegories ofcommercialisrn replace those of disinterestedservice and the spirit of the horse trader broodsover the face of the earth."a

d. iufora! Educatio,re. Morality, good manners, ds-cency, and ottier virtues must be teachable. Psy-chologieal egoists often notiee that moraleducation and the ineulcation of manners usu-ally utilize rvhat Bentham calls th.e o'sanctions ofpleasure and pain."5 Chiidren are made to ac'quire the eiviiizing virtues only by the methodof enticing rewards and painful punishments.lVluch tlie sa.me is true of the history of the lace.Feople in general have been inclined to behavewell only when it is made plain to them thatthere is'"something in it foii them." Is it not thenhighly probable that just such a meehanism ofhuma.n motivaiion as Eentharn deseribes mustbe presupposed by our nrethods of moral educa-tion?

c. c].rrTxQ{.J8, oF PSYCIJOLOGICAI,EGeIStu[: CONtr{JSIGIqS n}d TF{EA,RGUIEdENTS

5. Non-Empirical Character af the Argurnents.trf the arguments of the psychological egoisi con'sisted for the most palt of earef'ully acqriired em-pirical evidence (well-documented reports ofcontrolled experiments, surveys, interviews, labo-rataty data, aud so on), then ttre critir;al philoso-phei' would have no business carping ai thein.After all, since psychological egoism purports to

be a scientifie theory oi trruman motives, it is theeorlcern of the experimental psychologist, not thephilosopher, to accept or reject it. tsut as a matterof fact, empirical evidence of the required sort is

seldom presented in support of psychologicalegoism. Fsyehologists, on the whole, shy awayfrom generalizations about hurtan rnotives vririchare so sweeping and so vaguely fbrmulated thatthey are virtualiy incapable cfscientific testing' It

is usually the "armchair scientist" rvho holds the

theory of universal selfishness, and his ustlal ar'guments are either based siinply on his "impres-sions" or else are largely of a non-empirical sort'

The latter are often shot full of a very subtle kind

of logical confusion, and this makes their cliti'

cism a matter of special interest to the analyticphilosopher.

6. The psychological egoist's first argument(4a, above) is a good example of logical confu-sion. It begins with a truism*namely, that all ofmy motives and desires are ftty motives and de-sires and not someone else's. (Who rvould denythis?) Eut from this simple tautology nothingwhatever concerning the nature of my motives orthe objective of my desires can possibly follow.The fallacy of this argument consists in its viola-tion ofthe general logical rule that analytic state-ments (tautologies) cannot entail synthetic(factual) ones.8 That every voluntary act isprompted by the agent's own motives is a tautol-ogy; hence, it cannot be equivalent to ".d personis always seeking something for himself, or,.Allof a person's motives are selfish," rvhich are syn-thetic. What the egoist must prove is not merely:

(i) Every voluntary action is prompted by a motiveof the agent's own.

but rather:

(iD Every voluntary action is prompted by a mo-tive ofa quite particular kind, viz. a selfish one.

Statement (i) is obviously true, but it cannot allby itself give any logical support to statement (ii).

The source of the confusion in this argument isreadily apparent. trt is not the genesis ofan actionor the origin or its motives which makes it a"selfish" one, but rather the "purpose" ofthe actor the objectiye of its motives; nat where the mo-tive comes frorn (in voluntary actions it alwayscomes from the agent) but what it aims at deter-mines whether or not it is selfish. There is surelya valid distinction between voluntary behavior, inwhich the agent's action is motivated by purposesof his own, and selfsh behavior in which theagent's motives are of one exclusive sort. Theegoist's argument assimilates all voluntary actioninto the class of selfish action, by 4equiring, inefect, that an unselfish action be one which is notreally motivated at all. In the words of LuciusGarvin, "to say that an act proceeds from ourown . . . desire is only to say that the act is ourown. To demand that we should aet an motivesthat are not our own is to ask us to make our-selves living contradictions in terms.,,?

JOEL FETNBERG 5o3

7. But if argument 4a fails to prove its point,argument 4b does no better. Frorn the fact that allour successful actions (those in which we getwhat we were after) are accompanied or followedby pleasure it does not follow, as the egoistclaims, that the objective of every action is to getpleasure for oneself. To begin with, the premiseof the argument is not, strictly speaking, eventrue. Fulfillment of desire (simply getting rvhatone was after) is no guarantee of satisfaction(pleasant feelings of gratification in the mind ofthe agent). Somefimes when we get rvhat tve rvanttve also get, as a kind of extra dividend, a tvarm,glowing feeling of contentment; but often, far toooften, we get no dividend at all, or, even lvorse,the bitter taste of ashes. Indeed, it has been saidthat the characteristic psychological problem ofour time is the dissatisfaction that attends thefulfillment of our very most powerful desires.

Even ifwe grant, however, for the sake ofargu-ment, that getting what one wants usually yieldssatisfaction, the egoist's conclusion does not fol-lorv. We can concede that we normally get plea-sure (in the sense of satisfaction) when our desiresare satisfied, no matter whot our desires are for;but it does not follow from this roughly accurategeneralization that the only thing we ever desireis our own saiisfaction. Fleasure may rvell be theusual accompaniment of all actions in which theagent gets rvhat he wants; but to infer from thisthat what the agent always rvants is his own plea-sure is like arguing, in William James's example,sthat because an ocean liner constantly consumescoal on its trans-Atlantic passage that thereforethe puryose of its voyage is to consume coal. Theimmediate inference from even constant accom-panirnent to purpose (or motive) is always a nonsequitur.

Perhaps there is a sense of "satisfaction" (de-sire fulfillment) such that it is certainly and uni-versally true that we get satisfaction whenever rveget what rve want. But satisfaction in this senseis simply the "coming into existence of that whichis desired." I{ence, to say that desire fulfillmentalways yields "satisfaction" in this sense is to sayno more than that rve always get what we tvantwhen we get what lve want, which is to utter atau-tology like "a rose is a rose." It can no moreentail a synthetic truth in psychology (like theegoistic thesis) than"a rose is a rose" can entailsignificant information in botany.

b.

__.....,-_

5c1 SELF-LOVE AND THE CLAIMS OF MORALITY

cient evidence for his universal generalization.His argument is not fallacious, but inconclusive.

No one but the agent himself can ever be cer-tain what conscious motives really prompted hisaction, and where motives are disreputable, eventhe agent may not admit to himself the true na-ture ofhis desires. Thus, for every apparent caseof altruistic behavior, the psychological egoistcan argue, with some plausibility, that the truemotivation might be selfish, appearance to thecontrary. Philanthropic acts are really motivatedby the desire to receive gratitude; acts of self-sacrifice, when truly understood, are seen to bemotivated by the desire to feel self-esteem; and soon. We must concede to the egoist that all appar-ent altruism might be deceptive in this way; butsuch a sweeping generalization requires consider-able empirical evidence, and such evidence is notpresently available.

ll. The "Paradox of Hedonism', ond lts Con-sequences for Education. The psychologicalegoistic Hedonist (e.g., Jeremy tsentham) has thesimplest possible theory of human motivation.According to this variety of egoistic theory, allhuman motives without exception can be reducedto one-namely, the desire for one's own plea-sure. But this theory, despite its attractive sim-plicity, or perhaps because of it, involves oneimmediately in a paradox. Astute observers ofhuman affairs from the time of the ancient Greekshave often noticed that pleasure, happiness, andsatisfaction are states of mind which stand in avery peculiar relation to desire. An exclusive de-sire for happiness is the surest way to preventhappiness from coming into being. F{appiness hasa way of"sneaking up" on persons when they arepreoccupied with other things; but when personsdeliberately and single-mindedly set offin pursuitof happiness, it vanishes utterly from sight andcannot be captured. This is the famous ,.paradoxof hedonism": the single-minded pursuit of hap-piness is necessarily self-defeating, for the woy toget hoppiness is to forget it; then perhaps it willcome to you. If you aim exclusively at pleasureitself, with no concern for the things that bringpleasure, then pleasure will never come. To de-rive satisfaction, one must ordinarily first desiresomething other than satisfaction, and then findthe means to get what one desires.

To feel the full force ofthe paradox ofhedon-ism the reader should conducf an experiment in

JOEL FETNBERC 5o5

his imagination. Imagine a person (let's call him"Jones") who is, first of all, devoid of intellectualcuriosity. He has no desire to acquire any kind ofknowledge for its own sake, and thus is utterlyindifferent to questions of science, mathematici,and philosophy. Imagine further that the beautiesof nature leaves Jones cold: he is unimpressed bythe autumn foliage, the snow-capped mountains,and the rolling oceans. Long walks in the countryon spring mornings and skiing forages in the win-ter are to him equally a bore. Moreover, let ussuppose that Jones can find no appeal in art. Nov_els are dull, poetry apain, paintings nonsense andmusic just noise. Suppose further that Jones hasneither the participant's nor the spectator's pas-sion for basebail, football, tennis, or any othersport. Swimming to him is a cruel aquatic formof calisthenics, the sun only a cause of sunburn.Dancing is coeducational idiocy, conversationa waste of time, the other sex an unappealingmystery. Folitics is a fraud, religion mere su-perstition; and the misery of millions of under-privileged human beings is nothing to be con-cerned with or excited about. Suppose finally thatJones has no talent for any kind of handicraft,industry, or commerce, and that he does not re-gret thet fact.

What then is Jones interested in? He must de-sire something. To be sure, he does. Jones has anoverwhelming passion for, a complete preoccupa-tion with, his own happiness. The one exclusivedesire of his life is to be happy. trt takes littleimagination at this point to see that Jones,s onedesire is bound to be frustrated. People who-likeJones-most hotly pursue their own happinessare the least likely to find it. Flappy people arethose who successfully pursue such things asaesthetic or religious experience, self-expression,service to others, victory in competitions, know!-edge, power, and so on. Ifnone ofthese thines inthemselves and for their own sakes mean iny-thing to a person, if they are valued at all thenonly as a means to one's own pleasant states ofmind-then that pleasure can never come. Thervay to achieve happiness is to pursue somethingelse.

Almost all people at one tirne or another intheir lives feel pleasure. Some people (though per-haps not many) realiy do live lives which are onthe whole happy. But if pleasure and happinesspresuppose desires for something other than plea-

8. Disinterested Benevolence. Thefallacy in ar'

gument 4b then consists, as Garvin puts it' "in

Ihe supposition that the apparently unselfish- de-

sire tobenefit others is transformed into a selfish

one by the fact that we derive pleasure from-car-

rying it out."s Not only is this argument falla-

cioui; it also provides us with a suggestion of a

counter-argument to show that its conclusion(psychological egoistic hedonism) is false' Not

only is the presence ofpleasure (satisfaction) as a

by-product of an action no proof that the action

wai selfish; in some special cases it provides

rather conclusive proof that the action was un'

selfsh. For in those special cases the fact that we

get pleasure from a particular action presupposes

lnoi ,, desired something e/se-something other

than our own pleasure-as an end in itself and

not merely as a means to our own pleasant state

of mind.This way of turning the egoistic hedonist's ar-

gument back on him can be illustrated by taking

i typical egoist argument, one attributed (per'

haps apouyphally) to Abraham Lincoln, and

then examining it closelY:

Mr. Lincoln once remarked to a fellow-passenger onan old-time mud-coach that all men were prompted by

selfishness in doing good. His fellow-passenger was an-

tagonizing this position when they rvere passing over a

"oidotoy bridge that spanned a slough' As they

crossed this bridge they espied an old razor-backed sow

on the bank making a terrible noise because her pigs

had got into the slough and rvere in danger ofdrown-ing. es the old coach began to climb the. hill' Mr'

Lincoln called out' "Driver, can't you stop Just a mo'

ment?" Then Mr. Lincoln jumped out, ran back andlifted the little pigs out of the mud and water andplaced them on the bank. When he returned, his com-panion remarked: "Now Abe, where does selfishnesscom. in on this little episode?" "Wh!, bless your soul

Ed, that was the very essence of selfishness' I shouldhave had no peace ofmind all day had I gone on and

left that suffeiing old sow worrying over those pigs' I

did it to get peace of mind, don't you see?"lo

If Lincoln had cared not a whit for the welfare

of the little pigs and their "sufering" rnother, but

only for hii own "peace of mind," it would be

didcult to explain how he could have derived

pleasure from helping them' The very fact that he

iiO feet satisfaction as a result ofhelping the pigs

presupposes that he had a preexisting desire for

something other than his own happiness' Then

when that desire was satisfied, Lincoln of course

derived pleasure. The obiect of Lincoln's desire

was not pleasure; rather pleasure was the conse'quence of his preexisting desire for- something

ilse. [f Lincoln had been wholly indifferent to theplight of the little pigs as he claimed, how could

6e possibly have derived any pleasure from help'

ingthem? Ftre could not have achieved peace of

mind from rescuing the pigs, had he not a prior

concern-on which his peace of mind depended-for the welfare of the pigs for its own sake'

trn general, the psychological hedonist analyzes

appaient benevolence into a desire for "benevo'

lent pleasure." No doubt the benevolent man

does get pleasure from his benevolence, but in

most casei, this is only because he has previously

desired the good of some person, or animal, or

mankind atlarge' Where there is no such desire,

benevolent conduct is not generally found to give

pleasure to the agent.9. Malevolence. Difficult cases for the psycho-

logical egoist include not only instances of disin'

teiested benevolence, but also cases of

"disinterested malevolence." Indeed, malice and

hatred are generally no more "selfish" than be-

nevolence. Both are motives likely to cause an

agent to sacrifice his own interests-in the case of

binevolence, in order to help someone else, in the

case of malevolence in order to harm someone

else. The selfish man is concerned ultimately only

with his own pleasure; happiness, or power; the

benevolent man is often equally concerned with

the happiness of others; to the malevolent man'

the injuiy of another is often an end in itself-an

end to bi pursued sometimes with no thought for

his orvn interests. There is reason to think that

men have as often sacriflced themselves to injure

or kill others as to help or to save others, and with

as much "heroism" in the one case as in the other'

The unselfish nature of malevolence was first no'

ticed by the,A'nglican Bishop and moral philoso-

pher Joseph Butler (16g2-i752), who regretted

that men are no moie selfish than they are'll

10. Lack of Evidence for universol Self-Decep'

tion. The rnore cynical sort of psychological

egoist who is impressed by the widespread phe-

n"o*.non of self-deception (see 4c above) cannot

be so quickly disposed of, for he has commited no

Iogicai mistakes. We can only argue tha.t the ac'

kriowledged frequency of self-deception is insuffi-

5c/6 SELF-LOVE AND THE CLAIMS OF MORAI-ITY

ilI

,irl

,[

sure and happiness, then the existence ofpleasureand happiness in the experience of some peopleproves that those people have strong desires forsomething other than their own happiness-egoistic hedonism to the contrary.

The implications of the "pafadox of hedonism"for educational theory should be obvious. Theparents least likely to raise a happy child arethose who, even with the best intentions, traintheir child to seek happiness directly. E{ow oftenhave we heard parents say:

I don't care if my child does not become an intellec-tual, or a football stat, or a great artist. I just want himto be a plain average sort ofperson. Happiness does notrequire great ambitions and great frustrations; it's notworth it to suffer and become neurotic for the sake ofscience, art, or do-goodism. I just want my child to behappy.

This can be a dangerous mistake, for it is the child(and the adult for that matter) without "outer'directed" interests who is the most likely to beunhappy. The pure egoist would be the mostwretched of persons.

The educator might well beware of "life adjust'ment" as the conscious goal of the educationalprocess for similar reasons. "Life adjustment"can be achieved only as a by-product of otherpursuits. A whole curriculum of "life adjustmentcourses" unsupplemented by couses designed toincite an interest in things other than life adjust'ment would be tragically self-defeating.

As for moral education, it is probably true thatpunishment and reward are indispensable meansof inculcation. But if the child comes to believethat the sole reasons for being motal ate that hewill escape the pain of punishment thereby andlor that he will gain the pleasure of a good reputa-tion, then what is to prevent him from doing theimmoral thing whenever he is sure that he willnot be found out? While punishment and rewardthen are important tools for the moral educator,they obviously have their limitations' Beware ofthe man who does the moral thing only out of fearof pain or love of pleasure. He is not likely to bewholly trustworthy. Moral education is truly suc-cessful when it produces persons who are willingto do the right thing simply because it is right, andnot merely because it is popular or safe.

12. Pleasure as Sensation. One final argumentagainst psychological hedonism should suffice to

put that form of the egoistic psychology to restonce and for all. The egoistic hedonist claims thatall desires can be reduced to the single desire forone's own pleasure, Now the word "pleasure" isambiguous. On the one hand, it can stand for acertain indefinable, but very familiar and specifickind of sensation, or more accurately, a propertyof sensations; and it is generally, if not exclu-sively, associated with the senses. For example,certain taste sensations such as sweetness, ther'mal sensations of the sort derived from a hot bathor the feel of the August sun while one lies on asandy beach, erotic sensations, olfactory sensa-tions (say) of the fragrance of flowers or perfume,and tactual and kinesthetic sensations from agood massage, are all pleasant in this sense. Letus call this sense of "pleasure," which is the con'verse of "physical pain," pleasurel.

On the other hand, the word "pleasure" is of'ten used simply as a synonym for "satisfaction"(in the sense of gratification, not mere desirefulflllment.) In this sense, the existence of plea'sure presupposes the prior existence of desire.Knowledge, religious experience, aesthetic ex'pression, and other so-called "spiritual activities"often give pleasure in this sense. In fact, as wehave seen, we tend to get pleasure in this sensewhenever we get what we desire, no matter whatwe desire. The masochist even derives pleasure(in the sense of "satisfaction") from his orvnphysically painful sensations. I-et us call the senseof "pleasure" which means "satisfaction"-plea-SUf€2.

Now we can evaluate the psychological hedon-ist's claim that the sole human motive is a desirefor one's own pleasure, bearing in mind (as heoften does not) the ambiguity of the word "plea-sure." First, let us take the hedonist to be sayingthat it is the desire for pleasurel (pleasant sensa'tion) which is the sole ultimate desire of all peopleand the sole desire capable of providing a motivefor action. Now tr have little doubt that all (or

most) people desire their own pleasure, some'times. Eut even this familiar kind of desire oc'curs, I think, rather rarely. When I am veryhungry, I often desire to eat, or, more specifically'to eat this piece of steak and these potatoes. Muchless often do I desire to eat certain morsels simplyfor the sake of the pleasant gustatory sensationsthey might cause. I have, on the other hand, beenmotivated in the latter way when I have gone to

especially exotic (and expensive) French or Chi'

nese restaurants; but normally, pleasant gastro-nomic sensations are simply ahappy

"on..qu.n..or by-product of my eating, not the antecedentlydesired objective of my eating. There are, ofcourse, others who take gustatory sensations farmoye seriously: the gourmet wiro eats only tosavor the textures and flavors of fine foods. andthe wine fancier who ,,collects,' the exquisitelysubtle and very pleasant tastes of rare old wines.Such men are truly absorbed in their taste sensa-tions when they eat ancl drink, and there mayeven be some (rich) persons whose clesire for suchsensations is the sole motive for eatingand drink-ing. It should take little argument, however, toconvince the reader that such persons are ex*tremely rare.

Similarly, I usually derive pleasure from takinga hot bath, and on occasion (though not veryoften) I even decide to bathe simply for the sakeof such sensations. Even if this is equally true ofeveryone, however, it hardly provides groundsfor inferring that no one ever bathes from anyother motive. It should be empirically obviousthat we sometimes bathe simply in order to gecclean, or to please others, or simply from habit.

The view then that wr; are never after anythingin our actions but our own pleasure*that ail menare complete "gourmets,, of one sort or another-is not only moraily cynical; it is also contraryto common sense and everyda;r experience. In.r%ct, the view that pleasairt sensations pla.y sueiran enormous role in human affair:s is so patenilvfalse, on the availa.bie evidence, that we mrrsiconclude that the psyehologica,l hedonisf has theother sense of "pleasure,,-satisfaction--in mindwhen he states his thesis. If, on the other hand,he really does try to reduce the appar"ent multi-tude of human motives to the one desire for pleas-ant sensations, fhen the abundance of historicalcounter-examples justifies our rejection out ofhand of his thesis. It surely seems incredible thatthe Christian martyrs were ardenfly pursuingtheir own pleasure when they marcheel off to fur'.the lions, or that what the Russian solcjiers atStalingrad "teally" wanted when thev douseclthemselves with gasoline, ignited thernselves, andthen threw the flaming torches of their own bod-tes on German tanks, was simply the experienceof pleasant physical sensations.

13. Pleasure as Satisfoction. tr-et us considernow the other interpretation ofthe hedonist,s the-sis, that according to rvhich it is one's own plea-

JOEL FETNBERG 5o7

sure2 (satisfaction) and not merely pleasurel(pleasant sensation) which is the sole ultimateobjective of all voluntary behavior. In one re-spect, the "satisfaction thesis', is even less plausi_ble than the "physical sensation thesis',; for thelatter, at least is a genuine empirical hypothesis,testable in experience, though contrary to thefacts which experience discjoses. The former,however, is so confused that it cannot even becompletely stated without paradox. It is, so tospeak, defeated in its own formulation. Anv at_tempted explication of the theory that all men atall times desire only their own satisfaction leadsto an infrnite regress in the following nay:

"AIl men desire only satisfaction.""Satisfaction of what?',"Satisfaction of their desires.""Their desires for what?""Their desires for satisfaction",,"Satisfaction of what?""Their desires.""For rvhat?""For satisfaction"-etc., ad infnitum.

In short, psychological hedonism interpreted inthis way attributes to all peopie as iheir sole inr:-tive a vrholly vacuorls and infiniteiy sel&defeatingdesire. T'he source oi this absurdity is in the nn-tion that satisfaction eail, so to speak, feed onitseif; and perfbrrn the nriracle of perpetual seii-regeneration irr t?re atrsence of desires for any_thing other th.an itselfl'Uo

summarize the a_rgument of sections l l ancl12: 'I'he

wol'd "pleasure,, is ambiguous. Fiea-sDr€l r?€3_t1s a certain indefinable chaizcteristic ofphysical sensation. Fleasure2 refers to the feelingof satisfaetion that often comes when one getswhat one desires whatever be the nalure of thatwhich one desires. Now, if the heclonist meailspleasurcl vuhen he says that on€'s own pleasurc isthe ultirnate objective of all of oire's beha,lor,then his view is not supported by the iacts. Cn theother hand, if he means pleasure2, then his theorycannot even be clearly formulated, sinee it leaclsto the following infinite regress: ,,I clesire onlysatisfaction of my desire for satisfaetioir of rnydesire for satisfaction . . . etc., ad iryInitum.,, [,conclude then that psychological hedonism (thernost common form of psychological egoism),however interpreted, is untenable.

i?F' :

t'IE1

Ei,

FH:

i

508 SELF-LOVE AND THE CLAIMS OF MORALITY

D. CRITIQUE OF PSYCHOLOGICALEGOISM: UNCLEAR LOGICAI, STATUSOF THE THEORY

14. There remain, however, other possibleforms of the egoistic psychology. The egoistmight admit that not all human motives can bereduced to the one ultimate desire for one's ownpleasure, or happiness, and yet still maintain thatour ultimate motives, whether they be desire forhappiness (J. S. Mill), self-fulfillment (Aristotle),power (Hobbes) or whatever, arc always self'regarding motives. He might still maintain that,given our common human nature, wholly disin-terested action impelled by exclusively other-regarding motives is psychologically impossible,and that therefore there is a profoundly impor-tant sense in which it is true that, whether theybe hedonists or not, all men are selfish.

Now it seems to me that this highly paradoxi-

cal claim cannot be flnally evaluated until it isproperly understood, and that it cannot be prop'

erly understood until one knows what the psy-

chological egoist is willing to accept as evidenceeither for or against it. In short, there are twothings that must be decided: (a) whether the the-ory is true or false and (b) whether its truth orfalsity (its truth value) depends entirely on themeanings of the words in which it is expressed orwhether it is made true or false by certain facts,in this case the facts of psychology.

15. Anatytic Statements. Statements whosetruth is determined solely by the meanings of thewords in which they are expressed, and thus canbe held immune from empirical evidence, are of'ten called analytic statements or tautologies. Thefollowing are examples of tautologies:

(l) All spinsters are unmarried.(2) All effects have causes'(3) Either Providence is the capital of Rhode Is'

land or it is not.

The truth of (1) is derived solely from the mean-ing of the word "spinster," which is defined (inpart) as "unmarried woman." To find outwhether (1) is true or false we need not conductinterviews, compile statistics. or perform experi'ments. .dll empirical evidence is superfluous andirrelevant; for if we know the meanings of "spin-ster" and "unmarried," then we know not only

that (1) is true, but that is is necessarily true-i.e.,that it cannot possibly be false, that no futureexperiences or observations could possibly upsetit, that to deny it would be to assert a logicalcontradiction. But notice that what a tautologygains in certainty ("necessary truth") it loses indescriptive content. Statement (1) imparts no in-formation whatever about any matter of fact; itsimply records our determination to use certainwords in a certain way. As we say' "It is true bydefinition."

Similarly, (2) is (necessarily) true solely in vir-

tue of the meanings of the words "cause" and

"effect" and thus requires no further observationsto confirm it. And ofcourse, no possible observa'tions could falsify it, since it asserts no matter offact. And finally, statement (3) is (necessarily)

true solely in virtue of the meaning of the Englishexpression "either . . . or". Such terms as "either. .. orrt' "If . , . thenr" "andr" and "not" arecalled by logicians "logical constants." The def'nitions of logical constants are made explicit in

the so-called "laws of thought"-the law of con-tradiction, the law of the excluded middle, and

the law of identity. These "laws" are not laws in

the same sense as are (say) the laws of physics.

Rather, they are merely consequences of the def'

nitions oflogical constants, and as such, thoughthey are necessarily true, they impart no infor'

mation about the world. "Either Providenceis the capital of R.hode trsland or it is not" tells

us nothing about geography; and "Either it is

now raining or else it is not" tells us nothingabout the weather. You don't have to look at a

map or look out the window to know that

they are true. Rather, they are known to be

trve a priori (independently of experience);and, like all (or many)lz a prioti statements'they are vacltolts, i.e', devoid of informativecontent.

The denial of an analytic statement is called a

contradiction. The following are typical examplesof contradictions: "Some spinsters or€ married"'

"Some causes have no effects," "Providence both

is and is not the capital of Rhode Island." As m

the case oftautologies, the truth value ofcontra-

dictions (their falsehood) is logically necessary'not contingent on any facts of experience, and

uninformative. Their ialsity is derived from the

meanings (definitions) of the words in which theY

are expressed.

16. Synthetic Statements. On the other hand,statements whose truth or falsity is derived notfrom the meanings of worcJs but rather from theflctl ole](perience (observations) are called syn_tltetic,ts Prior to experience, there can be no good

, reason to think either that they are true or thatthey are false. That is to say, their truth value iscontingent; and they can be confirmed or dis_confirmed only by empirical evidence,ra i.e., con_trolled observations of the world. Unlike analyticstatements, they do impart information aboutmatters of fact. Obviously, ,,It is raining in New-port now," if true, is more informative than ,.Ei-ther it is raining in Newport now or it is not.,,even though the former could be false, while thelatter is necessarily true. I take the following tobe examples of synthetic (contingent) statements:

(1') All spinsters are frustrated.(2') All events have causes.(3) Providence is the capital of Rhode Island.(3") Newport is the capital of Rhode Island,

Statement (3') is true; (3") is false; and (1') is amatter for a psychologist (not for a philosopher)to decide; and the psychologist himself canoniydecide empirically. i.e., by making many observa_tions. The status of (2') is very Oifficult and itstruth value is a matter of great controversy. Thatis because its truth or falsity depends on sll thefa,cts ("all events"); and, needless to say, not allof the evidence is in.

17. Empirical Nypotheses. perhaps the mosrinteresting subclass of synthetic sta-ternents arethose generalizations of experience of the sorreh_aracteristically made by scientists; e.g., ,,Allreleased objeets heavier than air fall,,' ,.A11 swansare white," "A.ll men have Oedipus compiexes.,,I shall call such statements ,.empirical hypothe-ses" to indicate that their function is to ium uppast experience a_nd enable us successfuilv to ore-dict or anticipate future experience.rs T.hry"uranever logically certain, since it is alwavs at ieasrconceivable that future experience will dis-conlirm them. For example, zoologists once be_lieved that all swans are white, until black swanswere discovered in Australia" The most impor_tant cha,racteristic of empir ical hypotheses forour present purposes is their relation to evidence.,4. person can be said to unclerstand. an empiricalhypothesis only if he knows how to recognize

roEL FErNtsERc 5Og

evidence against it. If o person osserts or believesa generol stotement in such a way that he csnnotconceive of ony possible experience which ke wouldcount os evidence agoinst it, then he cannot be saidto be asserting or believing an empirical hypothe-sls. We can refer to this important characteristicof empirical hypotheses as fatsifability in princi-ple.

Some statements only appear to be empiricalhypotheses but are in fact disguised tautologiesreflecting the speaker,s determination to usewords in certain (often eccentric) ways. For ex_ample, a zoologist might refuse to allow the exis_tence of "Australian swans', to count as evidenceagainst the generali zation that all swans arewhite, on the grounds that the black Australianswans are not "really" swans at all. This wouldindicate that he is holding wLtiteness to be part ofthe definition of "swan,', and that therefoie, thestatement "All srvans are white,, is, for him, ,,trueby definition"-4pd thus just as immune fromcounterevidence as the statement ,,All spinstersare unmarried." Similarly, most of us wouldrefuse to allow any possible experience to counras evidence against ,,2 + 2 : 4', or,,Either uni_corns exist or they do not,', indicating that thepropositions of arithmetic and logic are not em-pirical hypotheses.

18. Ordinary Languoge and Equivocation.Fhilosophers, even more than ordinaiy men, areprone to make startlin g and paradoxical claimsthat take the form of universal generalizationsand hence resemble empiricai hypotheses. Forexample, "All things are rnental (there are nophysical objeets),', .,All things are good (there isno evil)," ".4.11 voluntary behavior is selfish,', etc.Let us confine our attention for the moment tothe latter which is a rough statement of psycho_Iogical egoism. At fi.rst sight, the statemeni,,Alivoluntary behavior is selfish,' seems obviouslyfalse. One might reply to the psychological egoistin some $uch manner as this:

\ know some behavior, at least, is unselfish, becauseI saw my Aunt Emma yesterday give her last cenr roabeggar.I{orv she wiil have to go a whole rveek rvithnothing to eat. Surely, that was not selfish of her.

Nevertheless, the psychological egoist is likelynot to be convincerl, and insist that, in this case,if we knew enough about z\unt Emma, we rvouldlearn that her primary rnotive in helping the beg_

__.....|-F

5ao SELF-LOVE AND THE CLAIMS OF MORAI-ITY

gar was to promote her own happiness or assuage

f,., o*n conscience, or increase her own self'

esteem, etc. We might then present the egoist

with even more difrcult cases for his theory-

saints, martyrs, military heroes, patriots, and oth'

ers who have sacrificed themselves for a cause' If

the psychological egoist nevertheless refuses to

a"..!t any of these as examples of unselfish be-

havior, then we have a right to be puzzled about

what he is saying. Until we know what he would

count as unselfsh behavior, we can't very well

know what he means when he says that all volun-

tary behaviot is selfsh. And at this point we may

suspect that he is holding his theory in a "privi-

leged position"-that of immunity to evidence'

that he would allow no conceivable behavior to

count as evidence against it. What he says then,

if true, must be true in virtue of the way he defines-or redefines-the word "selfish'" And in that

case, it cannot be an empirical hypothesis'

Ifwhat the psychological egoist says is "true by

redefinition," then I can "agree" with him and

say "It is true that in your sense of the word'selfish' my Aunt Emma's behavior was selfish;

but in the ordinary sense of 'selfish,' which im-

plies blameworthiness, she surely was not

selfish." There is no point of course in arguing

about a mere word. The important thing is not

what particular words a man uses, but rather

whether what he wishes to say in those words is

true. Departures from ordinary language can of'

ten be justified by their utility for certain pur'

poses; but they are dangerous when they invite

equivocation. The psychological egoist may be

saying something which is true when he says that

Emma is selfish in ir sense, but if he doesn't

realize that his sense of "selfish" differs from the

ordinary one, he may be tempted to infer that

Emma is selfish in the ordinary sense which im-

plies blameworthiness; and this of course would

be unfair and illegitimate' It is indeed an extraor'

dinary extension of the meaning of the word

"self-indulgent" (as G. C. Chesterton remarks

somewhere) which allows a philosopher to say

that a man is self-indulgent when he wants to be

burned at the stake.19. The Fallacy of the Suppressed Correlative'

Certain words in the English language operate in

pairs-e.g., "selfish-unselfish," "good-bad,";'large-smal1," "mental-physical." To assert that

a thing has one of the above characteristics is to

contrist it with the opposite in the pair. To know

the meaning of one term in the pair, we must

know the meaning of the correlative term with

which it is contrasted. trf we could not conceive

of what it would be like for a thing to be bad, for

example, then we could not possibly understand

what is being said of a thing when it is called

"good." similarly, unless we had a notion of what

iiwould be like for an action tobe unselfsh, we

could hardly understand the sentence "So'and-so

acted selfishly"; for we would have nothing to

contrast "selfishly" with. The so-called "fallacy

of the suppressed correlative"16 is committed by

a person-who consciously or unconsciously re'

defines one of the terms in a contrasting pair in

such a rvay that its new meaning incorporates the

sense of its correlative.Webster's Collegiate Dictionary defines

"selfish" (in part) as "regarding one's own com-

fort, advantage, etc. in disregard of, or at the

expense of that of others." In this ordinary and

pt'opaa sense of "selfish," Aunt Emma's action in

giving trer last cent to the beggar certainly was

ior selfish. Emma disregarded her own comfort

(it is not "comfortable" to go a week without

eating) and advantage (there is no "advantage" in

malnutrition\ for the sake of (not "at the expense

of') another. -similarly,

the martyr marching off

to the stake is foregoing (not indulging) his "com'

fort" and indeed his very life for the sake of(not

at the expense of) a cause. If Emma and the mar'

tyr then are "selfish," they must be so in a strange

new sense of the word'A careful examination of the egoist's argu'

ments (see especially 4b above) reveals what new

sense he givei to the word 'nselfish'" fle redefines

the word so that it means (roughly) "66fiv2fsd"'or perhaps "intentional." "After all," says the

egoist, "Aunt Emrna had some purpose in giving

tte beggar all her money, and this purpose (de'

sire, inilntion, motive, aim) was &er purpose and

no one else's' She was out to further some airn of

her own, wasn't she? Therefore, she was pursuing

her own ends (aeting from her own motives); she

was after something for herself in so acting' and

that's what I mean by calling her action selfish'

Moreover, all intentional action-action done'on

purpose,' deliberately from the agent's-own rno'

iiu.i-i. selfish in the same sense"' We can see

now, from this reply, that since the egoist appar'

ently means by .,selfish" simply .,motivated,,,when he says that all motivated action is selfishhe. is no-t asserting o synthetic empirical hypotltesisabout human mofives; rothen iis stateient is atautology roughly equivalent to ,,all motivated oc-tions are motivated.,' And if that is the case, thenwhat he says is true enough; but, Iike all tautolo-gie,s, it is empty, uninteresting, and trivial.

Moreover, in redefining ,,selfish" in this way,the psychological egoist has committed the fal-Iacy of the suppressed correlative. For what canwe now contrast ,,selfish voluntary action', with?Not only are there no octual caies of unselfishvoluntary actions on the new definition; there arenot even any theoretically possible or conceivablecases of unselfish voluntary actions. And if wecannot even conceive of what an unselfish volun-tary action would be like, how can we give anys:nr-:- t9_ the expression ,,selfish voluntary ac-tion"? The egoist, so to speak, has so blown upthe sense of "selfish,, that, like inflated currency,it will no longer buy anything.

20. Psychological Egoism as a Linguistic pro-posol. There is still one way out for the egoist. Hemight admit that his theory is not really a psycho-Iogical hypothesis about human natuie desisneOto account for the facts and enable us to pr;dictor anticipate future events. He may even willinglyconcede that his theory is really a disguisedredefinition of a word. Still, he might argue, hehas made no claim to be giving un u".u.-ut. d._scription of actual linguistic usage. R.ather, he ismaking a proposal to revise our usage in the inter-est ofeconomy and convenience,just as the biolo-gists once proposed that we change the ordinarymeaning of "insect'o in such away thatspiders areno longer called insects, and the ordinary mean_ing of ,,fish,' so that whales and seals are nolonger called fish.

What are we to say to this suggestion? First ofall, stipulative deflnitions (proposals to revise us-age) are never true or false. They are simply use-ful or not useful. Would it be useful to ieiefine"selfish" in the way the egoist recommends? It isdifficult to see what would be gained thereby. Theegoist has noticed some respects in which actionsnormally called "selfish', and actions normallycalled o'unselfi.sh" are alike, namely they are bothmotivated and they both can give satisfaction_etther in prospect or in retrospect-to the agent.Because of these likenesses, the egoist feelsj-usti-

fi ed in attach,"r rr.;; ;:ilT; ..r:"::Thus one word-"selfish,'-must for him do thework of trvo words (,,selfish" and ,.unselfish', intheir old meanings); and, as a result, a very realdistinction, that between actions for the sake ofothers and actions at the expense of others, canno longer be expressed in the language. Becausethe egoist has noticed some respects in which twotypes of actions are alike, he wishes to make itimpossible to describe the respects in which theydiffer. It is difficult to see any utility in this stateof affairs.

But suppose we adopt the egoist's ,.proposal,'nevertheless. Now we would have to say tirat allactions are selfish; but, in addition, we wouldwant to say that there are two different kinds ofselfish actions, those which regard the interests ofothers and those which disregard the interests ofothers, and, furthermore, that only the latter areblameworthy. After a time our ear would adjustto the new uses of the word .,selfish,,' and wewould find nothing at all strange in such srate-ments as "Some selfish actions are morally praise-worthy." After a while, we rnight even invent twonew words, perhaps ,,selfitic" and ,,unselfitic,,, todistinguish the two important classes of ..selfish"actions. Then we would be right back where westarted, with new linguistic tools (,.selfish,' for"motivated," "selfitic" for "selfish,,' and,,unselfi-!ig" for "unselfish") to do the same old necessaryjobs. That is, until some new egoistic philosopheiarose to announce with an air of discovery that"All selfish behavior is really selfltic-theie areno truly unselfitic selfish actions.,' Then, Godhelp us!

NOTES

_ l. See his Introduction to the principles of Morals andL.egislation .(1789), Chap. I, first paragiaptr,-,.Xatuie trasplaced manklnd under the governance of trvo sovereign mas_ters, poin and pleasure. It is for them alone to point oirt whatwe ought to do, as well as to determine what we shal do. . . ..They govern us in all we do, in all we say, in a[ we?ink:'every

effort rve can make to throw offour su6:..ti"" *itt-r..u.but to demonstrate and confirm it."

.. -2. C. D. Broad, Ethics and the History of philosophlt (NewYork: The Humanities press, 1952), Essay"lG_..Egoi'., us uTheory of Human Motives," p. 2ta. fiis essay"is iilghlyrecommended,

, 3. *Austin Duncan-Jones, Butler's Moral philosophy (Lon-don:.Penguin Books, t952), p. 96. Duncan-Jon., go.s on toreJecr rnrs argument. See p. 5 l2 f.

542 SELF.LOVE AND THE CLAIMS OF MORALITY

4, I-ucius Garvin, A Modern Introduction to Ethics (Bos'

ton: Houghton Miffiin, 1953), p. 37' Quoted here by permis-

sion of the author and Publisher.5. Op. cit, Chap. III.

6. See Part D, 15 and 16, below.

7. Op. cit., P.39.8. The Principles of Psychology, (New York: Henry Holt'

1890), vol. II, p. 558.

9. Op. cit., P,39.10. Quoted from the Springlfeld (Illinois) Moniton by F'

C. Sharp inhis Ethics (Nerv York: Appleton-Century' 1928)'p . 7 5 .

ll. Seehis Fifteen Sermons on Human Nature Preachedat the Rolls Chapel (1726), especially the first and eleventh'

12, Whether or not there are some a priori statementsthat are not merely analytic, and hence rot vacuous, is stilla highly controversial question among philosophers'

MICT{AEL A. SLOT'E

An Empirical Basis for Fsychological Egoisnan

13. Some philosophers (those called "rationalists") be-Iieve that there are some synthetic statements whose truth canbe knorvn a priori (see footnote I 2)' If they are right, then thestatement above is not entirely accvtate,

14. Again, subject to the qualification in footnotes 12 and13.

15. The three examples given above all have the generic

character there indicated, but they also differ from one an-other in various other ways, some of which are quite impor'tant. For our present purposes however, we can ignore thervays in which they differ from one another and concentrateon their common character as generalizations of experience("inductive generalizations"). As such they are.sharply con'irasted withiuch a generalization as "All puppies are young

dogs," which is analYtic.

16. The phrase was coined by .1. Lowenberg' Seehis -arti-cle "What is Empirical?" in the Jaurnal of Philosophy, May1940.

(like hunger, thirst, sleep, eliminaexplain all other, higher-order, drias derived genetically from the pricertain "laws of reinforcement.',deny the "functional autonomy" of torder drives or motives.4 Now it isbated issue in contemporary learwhether any theory such as we habriefly above could adequately explzman behavior. I shall, horvever, arga theory of the above kind may welithat from such a theory, foriifieOadditional psychological premise,egoism (non-altruism) Iogically follshow, thereby, that the question ofpegoism is still an open empirical issfallacious be the philosophicol argun

But what is "functional autonomvril!6r,3 .

I

ilr,'ii.Fi

does the lack ofit help to show ourselfish? According to behavioristicchologists a higher-order (acquifunctionally autonomous when itally independent of primary motivesof those motives association with whiit to be acquired in the first place) inthat one will indefinitely keep actin

It is commonly believed in the philosophical

world today that the age-old problem of psycho-

logical egoism is merely a pseudo-problem and

that this is true just because the a priori philo-

sophical arguments that have traditionally beengiven in favor of egoism depend in the main upon

confusions about the logic of our ordinary lan'guage. It has been claimed, for example, that the

well-known argument that we act selfishly even

when we want fo help others because in such

cases we are stiil attempting to satisfy our owt!

desire to help others, is fallaciously generated by

misunderstandings of the proper use of terms like'want', 'satisfY', and 'desire'"r

\n Butler's Moral Philosophy, Austin Duncan-

Jones, expressing Butler's view, and, it seems

from the eontext, his own as well, states that if

there is son,ething wrong with aii the a priori

philosophical arguments that have traditionally

*Michael A. Slote, "An Empirical Basis for FsychologicalEgoism," Journal of Philosophy' V9l' 6l' No. 18 (October i'1i64), 530-531. This slightly revised version reprinted by-permission of the authoi and the editors of the Journal ofFhilosoph2.

been given in favor of egoism (which he has ear'

lier identified with the doctrine that all human

acts are selfish),z then there is little else to recom-

mend the theory, since "the appearance of things'

undistorted by theory," is that men sometimes do

act unselfishly, disinterestedly. Only one who al'

ready believeC in tfre validity of the philosophical

argurnents for egoism would have any reason to

interpret the facts of human behavior in a way

comfatible with the doctrine of egoisrn'3 Thus

fluncan-Jones seems clearly to be ruling out the

possil',ility that the (empirical) facts as they stand

coulcl, wittr any semb'lance of objectivity, be used

to support egoism. -{nd rnany other eontempo-

rary philosophers would, tr think, tend to agree

r,vith him.(n the present paper I wish to argue that psy'

ohological egoism may rvell have a basis-in the

empirical facts of human psychology',-Certatn

contemporary lea.rning theoriits, e.g', Flull and

Skinnei have put foriard behavioristic theories

of the origin aird functioning of human motiv€s

which posit a certain nrimber of basically

"seifish,;' unlearned primary drives or motives

motive, even if rewards for those othimotives are no longer in general asso,such actiorr.s We have reason to belgiven higher-order drive or motive itionally autonomous, i.e., is functident, if when we cut offall reinforcempnmary rewards (rewards of primarythere are, in addition, both a sufficient

It is necessarv for a motive derivedflrom "selfish" (-or at least not unselfs

complete absencg during those extincof primary rervards for any similar himodves (to eliminate the possibiliry oftion of primary rewards from motivesthat belng extinguished), the higher<

"extinction trials" (occurrences ofthat higher-order motive whichciated even indirectly, i.e., throughorder motives, with primary rewar

or motive actually does extinguish; ison whose higher-order motive is brguished eventually, even if perhapsgradually, ceases to act from that himotive.

drives also to be functionally depen,