malcolm - knowledge of other minds

7/29/2019 Malcolm - Knowledge of Other Minds

1/5

ijo About Behavionsm10 . There is a dif fe rence b e tween someone gnashing his teeth and thegnashing of someone's teeth. But the di f f er ence is not a difference in

behavior: only the former is an instance of behavior; the lat ter m ay bea component of behavior.If George is gnashing his teeth then George's teeth are gnashing. But

whether or not a case in which his teeth are gnashing can rightly becharacterized as a case in which he is gnashing his teeth depends (noton whether or not the gnashing of his teeth is accompanied by "a move-me n t of the soul" but siraply) on contextual and relational matters.

11. I said that whether or not a case in which George's teeth aregnashing can rightly be characterized as a case in which George is gnash-ing his t ee th depends on contextual and relational matters. I am notsaying "Whether or not a case in which my teeth are gnashing canrightly be characterized by me as a case in which I am gnashing my teethdepends on contextual and relational matters": that would be odd. Itwould indcate that I could in general answer the following generallyodd question: "Given that your t ee th are gnashing, what entit les you tosay not merely that your teeth are gnashing but t ha t you are gnashingyour teeth, that you are doing it?" (I believe that Wittgenstein oncesaid "The first mistake is to ask the quest ion": the second is to answer it .)

What is in question here is wha t en tit les you to say that I am gnash -ing my tee th and not mere ly that my tee th are gnashing. The ques.tionwhether I am gnashing m y tee th or whether my tee th are mere ly gnash-ing is a question for you, not for me. It would general ly be odd for meto ask "Am I gnashing my teeth or are they mere ly gnashing?"

12 . Whether or not a case in which my t ee th are gnashing can rightlybe characterized by you as a case in which I am gnashing my teeth de-pends on contextual and relational matters.The teeth of a corpse may be gnashing but the corpse cannot (wi thoutoddi ty) be said to be gnashing its teeth. So I m u s t be alive, I must be-have in characteristic ways. What more is requi red? Primari ly this: my

subsequent behavior, both verbal and otherwise , must be consonant wi ththe claim that I was in fact gnashing my teeth. This is not to say thatif I assert "I was not gnashing my teeth," then I was not gnashing myteeth: I may be lying, or forgetful, or confused, e t c . But my subsequentbehavior, both ve rbal and othe rwise , is c learly re levant .There fore I deny that apart from my subsequent ve rbal behavior youca n in principie always find out whether or not I am gnashing my teeth.A nd in consequence I deny that the re is a dif fe rence be tween finding outwhether or not I am behaving in ce rta in ways and finding out whetheror not I am angry.Philosophical behaviorism is not a m etaphy sical theory: i t is the d cn i a )of a metaphysical theory . Consequent ly , it asserts not h i ng .

IX Knowledge of Other MindsNorman Malcolm

I believe that the argument from analogy fo r the existence of otherminds still enjoys more credit than it deserves, and my first aim in thispaper will be to show that it leads nowhere. J. S. Mili is one of manywho have accepted the argument and I take his statement of it as repre-sentative. He puts to himself the question, "By what evidence do Iknow, or by what considerations am I led to believe, that there existother sentient creatures; that the walking and speaking figures which Isee and hear, have sensations and thoughts, or in other words, possessMinds?" His answer is the following:

I conclude that other human beings have feelings like me, because, first,they have bodies like me , which I know, in my ow case, to be the ante-cedent condit ion of feelings; and because, secondly, they exhibit the. acts, an d other outward signs, which in my own case I know by experienceto be caused by feelings. I am conscious in myself of a series of facts con-nected by an uniform sequence, of which the beginning is modificationsof my body, the middle is feelings, the end is outward demeanor. In thecase of other human beings I have the evidence of my senses for the firstand last link of the series, but not for the intermedate link. I find,however, that the sequence between the first and last is as regular an dconstant in those other cases as it is in mine . In my own case I know thatthe first link produces the last through the intermedate link, an d couldno t produce it without. Experience, therefore, obliges me to conclude thatthere must be an intermedate link; which must either be the same in othersas in myself, or a different one: I m ust either believe them to be alive, or tobe automatons: and by believing them to be alive, that is, by supposing thelink to be of the same nature as in the case of which I have experience, an dwhich is in all other respects similar, I bring other hum an beings, as phenom-ena, under the same generalizations which I know by experience to be thetrue theory of my own existence.11J. S. Mili, /.r Sxamination of Sir William Hamilton's Philosophy, 6th ed. (London: Longman?, 1889), pp . 243-44.


2/5

15 2 Knowledge o f Other MinasI shall pass by the possible objection that this would be very weak in-ductive reasoning, based as it is on the observation of a single instance.More interesting is the following point: suppose this reasoning couldyield a conclusin of the sort "Itis probable thatat human^jigar(pointinef at some psrson other thn oneselt^ has "itHougRts~~d feel-s."^ Then there is a question as to whether thisanythlng to the pbilosopher who drawsjt,because3here is a questionas to whether the senienre "Thaf hiiman_jurThas thoughts and feel-ings" can mean anythinp' to him. Why should this be a question? J3g:cause the assumption from which Mili starts is thatjie-ha^no criterion^or determining whether anotiler walking and speaking figure" does ordoes not nave thoughts and teelmgs.it ne naa a criterion he couldapply it, establishing with certainty that this or that human figure doesor does not have feelings (for the only plausible criterion would lie inbehavior and circumstances that are open to view), and there would beno cali to resort to tenuous analogical reasoning that yields at best aprobability. If Mili has no criterion for the existence of feelings othertrian his own then in that sense he does not understand the sentence"That human figure has feelings" and therefore does not understandthe sentence "It is probable that that human figure has feelings."There is a familiar inclination to make the following reply: "AlthoughI have no criterion of verification still I understand, for example, thesentence 'He has pain/ For I understand the meaning of 'I have apain/ and 'He has a pain' means that he has the same thing I havewhen I have a pain." But this is a fruitless maneuver. If I do not knowho w to establish that someone has a pain then I do not kriow how toestablish that he has the same as I have when I have a pain.8 Youcan-not imprve my understanding of "He has a pain" by this recourse tothe notion of "the same/' unless you give me a criterion for saying thatsomeone has the same as I have. If you can do this you will have no usefor the argument from analogy: and if you cannot then you do not un-derstand the supposed conclusin of that argument. A plosopher wh opurports to rely on the analogical argument cannot, I think, escape thisdilemma. ; ,There have been various attempts to repair ,the argument from anal-ogy. Mr. Stuart Hampshire has argued8 that its validity .as a method ofinference can be established in the following wy: others sometimes in-fer that I am feeling giddy from my behavior. Nw I have direct, non-inferential knowledge, says Hampshire, of my own feelings. So I can

*"It is no explanation to say: the supposition that he has a pain s simply the sup-position that he has the same as I. For that part of the grammar is quite clear to me: that is, that one will say that the stove has the same experience as I, if one says: it is inpain and I am in pain" (Whtgenstein, Philosophical Investigations [Oxford: Blackwell,1953]. 3 5 < > ) - '' "The Analogy of Feeling,"Mind, LXI (1952), 1-12.

Knowlcdge of Other Minascheck inferences made about me against th e facts, checking thereby theaccuracy of the "methods" of inference.

All that is required for testing the validity of any method of tactual infer-ence is that each one of us should sometimes be in a position to confrontthe conclusions of the doubtful method of inference with what is knownby him to be true independently of the method of inference in question.Each one of us is certainly in this position in respect of our commonmethods of inference about the feelings of persons other than ourselves,in virtue of the fac that each one of us is constantly able to compare theresults of this type of inference with what he knows to be true directly andnon-inferendally; each one of us is in the position to make this testing com-parison, whenever he is the designated subject of a statement about feelingsand sensations. I, H ampshire, know by what sort of signs I ma y be misled ininferring Jones' and Smith's feelings, because I have implicitly noticed(though probably not formulated) where Jones, Smith and others generallygo wrong in inferring my feelings [pp. 4-5].

Presumably I can also note when the inferences of others about my feel-ings do not go wrong. H aving ascertained the reliability of some infer-ence-procedures I can use them myself, in a guarded way, to drawconclusions about the feelings of others, with a modest but justifiedconfidence in the truth of those conclusions.My first comment is that H ampshire has apparently forgotten th epurpose of the argument from analogy, which is to provide some proba-bility that "the walking and speaking figures which I see and hear, havesensations and thoughts" (Mili). For the reasoning that he describes in-volves th e assumption that other human figures do have thoughts an dsensations: for they are assumed to make inferences about me fromobservations of my behavior. But the philosophical problem of the ex-istence of other minds is the problem of whether human figures otherthan oneself do, among other things, make observ ations, inferences, an dasserdns. H ampshire's supposed defense of the argument from analogyis an ignoratio elenchi.If we struck from the reasoning described by H ampshire all assump-tion of thoughts and sensations in others we should be left with some-thing roughly like this: "When my behavior is such-and-such there comefrom nearby human figures the sounds 'He feels giddy.' An d generallyI do feel giddy at the time. Therefore when another h uma n figure ex-hibits the same behavior and I say 'He feels giddy/ it is probable thathe does feel giddy." But the reference here to the sentence-like soundscoming from other. hum an bodies is irrelevant, since I must no t assumethat those sounds express inferences. Thus th e reasoning becomes simplythe classical argument from analogy: "When my behavior is such-and-such I feel giddy; so probably when another human figure behaves the


3/5

154 Knowledge of Other Minassame wa y he feels th e same way." This argument, again, is caught in thedilemma about the criterion of the same.

The versin of analgica! reasoning offered by Professor H. H. Price4is more interesting. He suggests that "one's evidence for the existence ofother minds is derived prmarily f rom th e understanding o f language"(p . 429). Hi s idea is that if another body gives forth noises one under-stands, like "There's the bus," and if these noises give one ne w informa-tion, this "provides some evidence that th e foreign body which utteredthe noises is animated by a mind like oneis own. . . . Suppose I amoften in its eighborhood, and i t repeatedly produces utterances whichI can u n de r s t a n d , and which I then proceed to ver i fy for myself . Andsuppose that this happens in many dif ferent kinds o f situation. I thinkt h a t m y evidence fo r believing that this body is animated by a mindl i k e m y own would then become very strong" (p . 430). Th e body fromwhich these infor mat ive sounds proceed need not be a human body. "Ifth e rustling of the leaves of an oak formed intelligible words conveyingnew information to me, and if gorse bushes made intelligible gestures,I should have evidence that the oak or the gorse bush was animated byan intelligence like m y own" (p . 436). Even if the intelligible and in-formative sounds did not proceed f rom a body they would provide evi-dence for the existence of a (disembodied) mind (p. 435).

Although differing sharply f rom the classical analogical argument, thereasoning presented by Price is still analogical in fo rm : I know by in-trospection that when certain combinations o f sounds come f rom m ethey are "symbols in acts o f spontaneous thinking"; therefore similarcombinations o f sounds, no t produced by me, "probably function asinstruments to an act of spontaneous thinking, which in this case is notm y own" (p . 446). Price says that th e reasoning also provides an explana-tion of the otherwise myster ious occurrence o f sounds which I un-derstand but did not produce. He anticipates the objection that thehypothesis is nonsensical because unver i f i ab le . "The hypothesis is a per-fectly conceivable one," he says, "in the sense that I know very wellwhat the world would have to be like if the hypothesis were truewhat sorts of entities there must be in it, and what sorts of events mustoccur in them. I know from introspection what acts of thinking andperceiving are, and I know what it is for such acts to be combined intothe unity of a single mind . . ." (pp. 446-47).

I wish to arge against Price that no amount of intelligible soundscoming from an oak tree or a kitchen table could crate any probabilitythat it has sensations and thoughts. The question to be asked is: Whatwould show that a tree or table understands the sounds that come f romit? We can imagine that useful warnings, true descriptions and predic-tions, even "replies" to questions, should emanate f rom a tree, so that itcarne to be of enormous valu to its owner. How should w e establish

4 "Our Evidence for the Existence of Other Minds," Philosophy, XIII (1938), 425-56.

Knowledge of Other Minds 155that it understood those sentences? Should we "question" it? Supposethat the tree "said" that there was a vixen in the eighborhood, and w e"asked" it "What is a vixen?" and it "replied," "A vixen is a femalefox." It might go on to do as well for "female" and "fox." This per-formance might incline us to say that the tree understood the words, incontrast to the possible case in which it answered "I don't know" or didnot answer at all. But would it show that the tree understood the wordsin the same sense that a person could understand them? With a personsuch a performance would crate a presumption that he could makecorrect applications of the word in question: but not so with a tree. Tosee this point think of the normal teaching of words (e.g., "spoon,""dog," "red") to a child and how one decides whether he understandsthem. At a primitive stage of teaching one does not require or expectdefinitions, but rather that the child should pick out reds from blues,dogs from cats, spoons from forks. This involves his looking, pointing,reaching for and going to the right things and not the wrong ones. Thata child says "red" when a red thing and "blue" when a blue thing is putbefore him, is indicative of a mastery of those words only in conjunc-tion with the other activities of looking, pointing, trying to get, fetchihgand carrying. Try to suppose that he says the right words but looks atand reaches for the wrong things. Should we be tempted to say that hehas mastered the use of those words? No, indeed. The disparity betweenwords and behavior would make us say that he does not understand thewords . In the case of a tree there could be no disparity between its wordsand its "behavior" because it is logically incapable of behavior of therelevant kind.Since it has nothing like the human face and body it makes no senseto say of a tree, or an electronic computer, that it is looking or pointingat or fetching something. (Of course one can always invent a sense forthese expressions.) Therefore it would make no sense to say that it didor did not understand the above words. Trees and computers cannoteither pass or fail the tests that a child is put through. They cannoteven take them. That an object was a source of intelligible sounds orother signs (no matter how sequential) would not be enough by itself toestablish that it had thoughts or sensations. How informative sentencesand valuable predictions could emanate from a gorse bush might be agrave scientific problem, but the explanation could never be that thegorse bush has a mind. Better no explanation than nonsenselIt might be thought that the above difficulty holds only for wordswhose meaning has a "perceptual conten" and that if we imagined, forexample, that our gorse bush produced nothing but pur mathematicalpropositions we should be justified in attributing thought to it, althoughnot sensation. But suppose there was a remarkable "calculating boy" whocould give right answers to arithmetical problems but could not applynumeris to reality in empirical propositions, i.e., he could not count


4/5

156 Knowledge o/ Other Mindsany objects. I believe that everyone would be reluctant to say tha t heunderstood the mathematical signs and truths that he produced. If hecould count in the normal way there would not be this reluctance. An d"counting in the normal way" involves looking, pointing, reaching,fetching, and so on. That is, it requires the human face and body, andhuma n behavioror something similar. Things which do not have thehuman form, or anything like it, not merely do not but cannot satisfythe criteria for thinking. I am trying to brng ut part of what Wittgen-stein meant when he said, "We only say of a human being and what islike on that it thinks" (Investigations, 360), and "The human body isthe best picture of the human soul" (ibid., p. 178).I have not yet gone into the most fundamental error of the argumentfrom analogy. It is present whether the argument is the dassical one(the aualogy between my body and other bodies) or Price's versin (theanalogy between my language and the noises an d signs produced byother things). It is the mistaken assumption that one learns from one'sow n case what thinking, feeling, sensation are. Price gives expressio n tothis assumption when he says: "I know from introspection what acts ofthinking and perceiving are . . ." (op. cit., p. 447). It is the most naturalassumption for a philosopher to make and indeed seems at first to be theonly possibility. Yet Wittgenstein has made us see that it leads first tosolipsism and then to nonsense. I shall try to state as briefly as possiblehow it produces those results.A philosopher wh o believes that one must learn what thinking, fear,or pain is "from one's ow n case," does not believe that the thing to beobserved is one's behavior, but rather something "inward." He considersbehavior to be related to the inward states an d occurrences merely as anaccompaniment or possibly an effect. He cannot regard behavior as acriterion of psychological phenomena: for if he did he would have nouse for the analogical argument (as was said before) and also the pri-ority given to "one's own case" would be pointless. He believes that henotes something in himself that he calis "thinking" or "fear" or "pain,"and then he tries to infer the presence of the same in others. He shouldthen deal with the question of what his criterion of the same in othersis. This he cannot do because it s of the essence of his viewpoint to re-ject circumstances an d behavior as a criterion of mental phenomena inothers. An d what else could serve as a criterion? He ought, therefore,to draw the conclusin that the notion of thinking, fear, or pain inothers is in an important sense meaningless. He has no idea of wha twould count for or against it.5 "That there should be thinking or painother than my own is unintelligible," he ought to hold. This would bea rigorous solipsism, and a corred outcome of the assumption that on e

s One reason why philosophers have not commonly drawn this conclusin may be,as Wittgenstein acutely suggests, that they assume that they have "an infallibleparadigm of identity in the identity of a thing with itself" (Investigations, 215).

Knowlcd'c oj Other Minds 157can know only from one's own case wha t th e mental phenomena are.An equivalent way of putting it would be: "When 1 say 1 am in pain,'by 'pain' I mean a certain inward state. When I say 'He is in pain,' by'pain' I mean behavior. I cannot attribute pain to others in the samesense that I attribute it to myself."Some philosophers before Wittgenstein may have seen the solipsisticresult of starting from "one's ow n case." But I believe he is the first tohave shown how that starting point destroys itself. This may be pre-sen ed as follows: on e supposes that on e inwardly picks out somethingas th inking or pain an d thereafter identifies it whenever it presents it-self in the soul. But the question to be pressed is, Does on e make correctidentifications? The proponent of these "prvate" identifications hasnothing to say here. He feels sure that he identifies correctly the occur-rences in his soul; but feeling sure is no guarantee of being right. Indeedhe has no idea of what being right could mean. He does not know howto distinguish between actually makin g correct identifications and beingunder the impression that he does. (See Investigations, 258-59.) Sup-pose that he identified the emotion of anxiety as the sensation of pain?Neither he or anyone else could know about this "mistake." Perhapshe makes a mistake every time! Perhaps all of us do! We ought to seenow that we are talking nonsense. We do not know what a mistakewould be. We have no standard, no examples, no customary practice,with which to compare ou r inner recognitions. Th e inward identifica-tion cannot hit the bull's-eye, or miss it either, because there is no bull's-eye. When we see that the ideas of correct an d incorrect have no ap-plication to the supposed inner identification, the latter notion loses itsappearance of sense. Its collapse brings down both solipsism and theargument from analogy.

77This destruction of the argument from analogy also destroys theproblem fo r which it was supposed to provide a solution. A philosopher

feels himself in a difficulty about other minds because he assumes tha tfirst of all he is acquainted with menta l phenomena "from his owncase." What troubles him is how to ma ke the transit ion from his owncase to the case of others. When his th ink ing is freed of the Ilusin ofthe priority of his own case, then he is able to look at the familiar factsand to acknowledge tha t th e circumstances, behavior, an d utterances ofothers actually are his criteria (not m erely his evidence) for the existenceof their menta l states. Previously this ha d seemed impossible.But now he is in danger of flying to the opposite extreme of behav-iorism, which errs by believing that through observation of one's ow ncircumstances, behavior, an d utterances one can find out tha t one isthinking or angry. The philosophy of "from one's own case" and be-


5/5

15 8 Knowledge o f Other Mindshaviorism, though in a sense opposites, make the common assumptionthat the first-person, present-tense psychological statements are verifiedby self-observation. According to the "one's ow n case" philosophy theself-observation cannot be checked by others; according to behaviorismth e self-observation would be by means of outward criteria that ar eavailable to all. The first position becomes unintelligible; the second isfalse for at least many kinds of psychologicals ta tements . We are forcedto conclude that the first-person psychological statements are not (orhardly ever) verified b y self-observation. It follows that they have noverification at all; for if they had a verification it would have to be byself-observation.

But if sentences like "My head aches" or "I wonder where she is" dono t express observations then wha t d o they do ? What is the relation be-tween my declaration that m y head aches and the fact that m y headaches, if the former is not the report of an observation? The perplexityabout th e existence of other minds has, as the result of crit icism, turnedinto a perplexity about th e meaning of one's ow n psychological sen-tences about oneself. At our starting point it was the sentence "His headaches" that posed a problem; but now it is the sentence "M y head aches"t h a t puzzles us. *

One way in which this problem can be put is by the question, "Howdoes one know when to say the words 'M y head aches?" The inclinationto ask this question can be made acute by imagining a fantastic but notimpossible case of a person who has survived to adut years without everexperiencing pain. He is given various sorts of injections to correct thiscondition, and on receiving one of these one day, he j u m p s and exclaims,"Now I feel pain!" One wants to ask, "How did he recognize the newsensation as a pain?"

Let us note that if the man gives an answer (e.g., "I knew it mustbe pain because of the way I jumped") then he proves by that very factthat he has not mastered the correct use of the words "I feel pain." Theycannot be used to state a conclusin. In telling us ho w he did it he willconvict hmself of a misuse. Therefore the question "How did he recog-nize his sensation?" requests the impossible. The inclination to ask it isevidence of our inability to grasp th e fact that the use of this psychologi-cal sentence has nothing to do with recognizing or identifying or observ-ing a state of oneself.

The fact that this imagined case produces an especially strong tempta-tion to ask the "How?" question shows that we have the idea that it mustbe more difficult to give the right ame of one's sensation the first time.Th e implication would be that it is not so diff icult after the first time.Why should this be? Are we thinking that then the man would havea paradigm of pain with which he could compare his sensations and sobe in a position to know right of whether a certain sensation was or wasnot a pain? But the paradigm would be e i ther someth ing "outer" (be-

Knowledge of Other Minds 159havior) or something "inner" (perhaps a memory impression of thesensation). If the former then he is misusing the first-person sentence.If the latter then the question of whether he compared correctly thepresent sensation with the inner paradigm of pain would be withoutsense. Thus the idea that the use of the first-person sentences can begoverned by paradigms must be abandoned. It is another form of ourinsistent misconception of the first-person sentence as resting somehowon the identification of a psychological state.These absurdities prove that we must conceive of the first-person psy-chological sentences in some entirely different light. Wittgenstein pre-sents us with the suggestion (to which philosophers have not beensufficiently attentive) that the first-person sentences are to be thoughtof as similar to the natural nonverbal, behavioral expressions of psycho-logical states. "My leg hurts," for example, is to be assimilated to crying,limping, holding one's leg. This is a bewildering comparison and one'sfirst thought is that two sorts of things could not be more unlike. Bysaying the sentence one can make a statement; it has a contradictory;it is true or false; in saying it one lies or tells the truth; and so on. Noneof these things, exactly, can be said of crying, limping, holding one'sleg. So how can there be any resemblance? But Wittgenstein knew thiswhen he deliberately likened such a sentence to "the primitive, thenatural, expressions" of pain, and said that it is "new pain-behavior"(ibid., 244). Although my limits prevent my attempting it here, I thinkthis analogy ought to be explored. For it has at least two importantmerits: first, it breaks the hold on us of the question "How dos oneknow when to say 'M y leg hurts'?" for in the light of the analogy thiswill be as nonsensical as the question "How does one know when to cry,limp, or hold one's leg?"; second, it explains how the utterance of afirst-person psychological sentence by another person can have impor-tance for us, although not as an identificationfor in the light of theanalogy it will have the same importance as the natural behavior whichserves as our preverbal criterion of the psychological states of others.

malcolm - knowledge of other minds

Documents