Sentences About Believing

8/8/2019 Sentences About Believing

http://slidepdf.com/reader/full/sentences-about-believing 1/25

Sentences about Believing

Author(s): Roderick M. ChisholmSource: Proceedings of the Aristotelian Society, New Series, Vol. 56 (1955 - 1956), pp. 125-148Published by: Blackwell Publishing on behalf of The Aristotelian SocietyStable URL: http://www.jstor.org/stable/4544560

Accessed: 12/07/2010 11:23

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=aristotelian.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

The Aristotelian Society and Blackwell Publishing are collaborating with JSTOR to digitize, preserve and

extend access to Proceedings of the Aristotelian Society.

http://www.jstor.org

http://www.jstor.org/stable/4544560?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/action/showPublisher?publisherCode=aristotelian

http://www.jstor.org/action/showPublisher?publisherCode=aristotelian

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/stable/4544560?origin=JSTOR-pdf



Meetingof theAristotelianociety t 21, Bedford quare,London,W.C. ,onFebruary0th,1956,at 7.30p.m.

VII.-SENTENCES ABOUT BELIEVING

By RODERICK M. CHISHOLM

1. "I can look for him when he is not there, but nothang him when he is not there ".1 The first of theseactivities, Brentano would have said, is intentional; t maytake as its object something which does not exist. But thesecond activity is " merely physical "; it cannot be per-formed unless its object is there to work with. " Inten-tionality ", he thought, provides us with a mark of what ispsychological.

I shall try to reformulate Brentano's suggestion by

describingone of the ways in which we need to use languagewhen we talk about certain psychological states and events.I shall referto this use as the " intentional use " of language.It is a kind of use we can avoid when we talk about non-psychological states and events.

In the interests of a philosophy contrary to that ofBrentano, many philosophers and psychologists have triedto show, in effect, how we can avoid intentional language

when we wish to talk about psychology. I shall discusssome of these attempts in so far as they relate to the sortsofthings we wish to be able to say about believing. I believethat these attempts have been so far unsuccessful. And Ithink that this fact may provide some reason for saying,with Brentano, that " intentionality " is a mark of what ispsychological.

2. In order to formulate criteria by means of which wecan identify the " intentional " use of language, let usclassify sentences as simple and compound. For our

1 Wittgenstein's Philosophical nvestigations,page 133e.

N



126 RODERICK M. CHISHOLM.

purposes I think it will be enough to say that a compound

sentence is one compounded from two or more sentences bymeans of propositionalconnectives, such as " and ", " or "," if-then ", " although ", " because ", and the like. A

simple sentence is one which is not compound. Examplesofsimplesentencesare: " He is thinkingof the Dnieper Dam ",

" She is looking for a suitable husband for her daughter ",

" Their car lacks a spare wheel ", and " He believes that it

will rain ". I shall formulate three criteria for sayingthatsimple declarative sentences are intentional, or are usedintentionally.

(a) A simple declarative sentence is intentional if it uses

a substantivalexpression-a name or a description-in sucha way that neither the sentence nor its contradictory mplieseither that there is or that there isn't anything to which the

substantivalexpression ruly applies. The firsttwo examplesabove are intentional by this criterion. When we say thata man is thinking of the Dnieper Dam, we do not implyeither that there is or that there isn't such a dam; similarlywhen we deny that he is thinking of it. When we say thata lady is lookingfor a suitable husbandfor her daughter,we

do not commit ourselvesto saying that her daughter will, or-that she will not, have a suitable husband; and similarlywhen we deny that the lady is looking forone. But the nextsentence in our list of examples-" Their car lacks a sparewheel "-is not intentional. It is true that, if we affirmthis sentence, we do not commit ourselves to saying eitherthat there are or that there are not any spare wheels. Butif we deny the sentence, affirming" Their car does not lacka spare wheel ", then we imply that there is a spare wheel

somewhere.

(b) We may describea second type of intentional use byreference to simple sentences the principal verb of whichtakes as its object a phrase containing a subordinate verb.The subordinate verb may follow immediately upon theprincipal verb, as in " He is contemplating killing himself";



SENTENCES ABOUT BELIEVING. 127

it may occur in a complete clause, as in " He believes it will

rain "; it may occur in an infinitive, as in " He wishes tospeak "; or it may occur in participial form, as in " HeaccusedJohn of stealing the money " and " He askedJohn'sbrother to testify against him ". I shall say that such asimple declarative sentence is intentional if neither thesentence nor its contradictory imply either that the phrasefollowing the principal verb is true or that it is false.2 " Heis contemplating killing himself" is intentional, according

to this second criterion, because neither it nor its denialimplies either that he does or that he doesn't kill himself;similarly with our other examples. But " He preventedJohn from stealing the money " is not intentional, becauseit implies that John did not steal the money. And " Heknows how to swim" is not intentional, because its denialimplies that he isn't swimming.

Sometimes people use substantival expressionsin placeof the kind of phrases I have just been talking about.Instead of saying, " I want the striketo be called off ", theymay say, " The strike's being called off is what I want ".The latter sentence could be said to be intentional accordingto our first criterion, for neither the sentence nor its contra-dictory implies either that " there is such a thing as " thestrike's being called off, or that there isn't-that is to say,

2 This criterion must be so interpreted that it will apply to sentenceswherein the verb-phrases following the principal verb are infinitive, pre-positional, or participial phrases; hence it must make sense to speak of such-phrases as being true or false. When I say of the phrase, following the mainverb of " He accused John of stealing the money ", that it is true, I mean, ofcourse, that John stole the money. More generally, when I say of such asentence that the phrase following the principal verb is true, or that it isfalse, my statement may be interpreted as applying to that new sentence

which is like the phrase in question, except that the verb appearing in infinitiveor participial form in the phrase is the principal verb of the new sentence.I should add a qualification about tenses, but I do not believe that my failureto do so is serious. It should be noted that, in English, when the subject ofan infinitive or of a participle is the same as that of the principal verb, we donot repeat the subject; although we say " I want John to go ", we do not say"I want me to go " or "John wants himself to go ". When I say, then,that the last two words of " I want to go " are true, my statement should beinterpreted as applying to " 1 shall go ".

N2




neither implies that the strikewill be, or that it will not be,

called off.Many intentional sentences of our first type may be

rewritten in such a way that they become instances of oursecond type. Instead of saying " I would like a glass ofwater ", one may say " I would like to have a glass ofwater ". And instead of saying " He is looking for theFountain of Youth ", one may say " He is trying to find theFountain of Youth ". But some sentences of the first typeseem to resist such transformation nto the second type; for

example, " I was thinking about you yesterday".

(c) If we make use of Frege's concept of " indirectreference", which is, of course, closely related to that of"intentionality ", we can add another important class ofsentenceto ourlist of thosewhich are intentional.3 " Indirectreference" may be defined, without using the characteristic

terms of Frege'stheory of meaning, in the following way: aname (or description) of a certain thing has an indirectreference n a sentence if its replacement by a differentname(or description) of that thing results in a sentence whosetruth-value may differ from that of the original sentence.It is useful to interpret this criterion in such a way that wecan say of those names (or descriptions), such as " theFountain of Youth " and "

a building half again as tall asthe Empire State ", which don't apply to anything, thatthey are all names of the same thing. Let us add, then, thata simple declarative sentence is intentional if it contains aname (or description)which has an indirect reference n thatsentence. We can now say of certain cognitivesentences-sentences which use words such as " know ", "remember ",

" see ", " perceive ", in one familiar way-that they, too,

are intentional. I may see that Albert is here and Albert

3By adopting Frege's theory of meaning-or his terminology-we couldmake this criterion do the work of our first two. But I have made use of thefirst two in order that no one will be tempted to confuse what I want to saywith what Frege had to say about meaning. The three criteria overlap to aconsiderable extent.




may be the man who will win the prize; but I do not now

seethatthe man who will win the prize is here. And we allrememberthat although George IV knew that Scott was theauthor of Marmion'he did not know that Scottwas the authorof Waverley.

(d) With respect to the intentionality of compoundsentences-sentences constructed by means of propositionalconnectives from two or more sentences-it is enough to saythis: a compound declarative sentence is intentional if andonly if one or more of its component sentencesis intentional." I will be gratifiedif I learn that Albert wins the prize " isintentional, because the if-clause is intentional. But" The career of Ponce de Leon would have been mostremarkableif he had found the Fountain of Youth " is notintentional, because neither of its componentsis intentional.

(In order that this final criterion be applicable to sentencesin the subjunctive,we should, of course, interpretit to mean:a compound declarativesentence is intentional if and only ifone or more of the component sentences of its indicativeversion is intentional.)

3. We may now formulate a thesis resembling that ofBrentano by referringto intentional language. Let us say

(1) that we' do not need to use intentional language whenwe describenon-psychological,or " physical ", phenomena;we can express all that we know, or believe, about suchphenomena in language which is not intentional. And letus say (2) that, when we wish to describe certain psycho-logical phenomena-in particular,when we wish to describethinking, believing, perceiving, seeing, knowing, wanting,hoping and the like-either (a) we must use language whichis intentional or (b) we must use a vocabulary which we donot need to use when we describe non-psychological, or"Sphysical ", phenomena.

I shall discuss this linguistic version of Brentano's thesiswith reference to sentences about believing. I do notpretend to be able to show that it is true in its application to




believing. But I think that there are serious difficulties,

underestimated by many philosophers, which stand in theway of showingthat it is false.

I wish to emphasize that my question does not concern"subsistence " or " the being of objectswhich don't exist ".

Philosophersmay ask whether it is possible to think aboutunicorns if there are no unicorns for us to think about.They may also ask whetheryou and I can believe " the samething " if there is no Propositionor Objective toward which

each of our beliefs is directed. But I am not raising thesequestions. Possibly the feeling that the intentional use oflanguage commits us to the assumptionthat there are suchentities is one motive for seeking to avoid such use. But Iwish to ask only whether we canavoid such use and at thesame time say all that we want to be able to say aboutbelieving.

4. The first part of our thesisstates that we do not needto use intentional language when we describe non-psycho-logical, or " physical ", phenomena. I do not believe thatthis statement presentsany seriousdifficulty. It is true thatwe do sometimes use intentional sentences in non-psycho-logical contexts. The following sentences, for example, areall intentional, according to our criteria, but none of them

describe anything we would want to call " psychological "." The patient will be immune from the effects of any newepidemics"; "A swQrdfishhas a beak like a unicorn'shorn "; and " It is difficult to assemble a prefabricatedhouse ". But these three sentences are not examplescounter to our thesis. Anyone who understands thelanguage can readilytransform hem into conditionals whichare not intentional. (A compound sentence, it should berecalled, is intentional only if it has a component which isintentional.) Instead of using intentional sentences, wecould have said: " If there should be any new epidemics,the patient would not be affected by them "; " If therewere any unicorns, their horns would be like the beak of aswordfish"; and " If anyone were to assemble a pre-




fabricated house, he would have difficulties". (Perhaps

the last sentence should be rendered as: " If anyone were totry to assemble a prefabricated house, he would havedifficulties ". In this version the sentence is intentional,once again, but since it contains the verb " to try " it can nolonger be said to be non-psychological.)

I believe that any other ostensibly non-psychological

sentence which is intentional can be transformed, in anequally obvious way, into a sentence conforming to our

version of Brentano'sthesis. That is to say, it will become asentence of one of two possible types: either (a) it will be nolonger intentional or (b) it will be explicitly psychological.

Sentences about probability may be intentional, but,

depending upon one's conception of probability, they maybe transformedeither into the first or into the second type.If I say " It is probable that there is life on Venus ", neither

my sentence noi its denial implies either that there is life onVenus or that there is not. According to one familiar

interpretationof probability,my sentencecan be transformedinto a non-intentional sentence about frequencies-sen-tences telling about places where there is life and places

where there isn't and comparing Venus with such places,etc. According to another interpretation, my sentence can

be transformed into a psychological statement about

believing-e.g., " It is reasonable for us to believe that thereis life on Venus ". Intentional sentences about tendencies

and purposes n nature may be treated similarly. If we say,non-intentionally, " The purpose of the liver is to secretebile ", we may mean, psychologically,that the Creatormadethe liver so that it would secrete bile, or we may mean,

non-intentionally,that in most live animalshaving liversthe

liver does do this work and that when it does not the animal

is unhealthy.There are people who like to ascribebeliefs, perceptions,

plans, desires, and the like, to robots and computing

machinery. A computing machine might be said to

believe, truly, that 7 and 5 are 12; when it is out of order,it may be said to make mistakes and possibly to believe,




falsely, that 7 and 5 are 11. But such sentences,once again,

are readily transformed into other sentences, usuallyconditionals, which are no longer intentional. If a mansays that the machine believes 7 and 5 to be 11, he may meanmerely that, if the keys marked " 7 " and "5 " are pressed,the machine will produce a slip on which" 11 " is marked.

Other intentional sentences about the attitudes of machines

may be more complex, but I'm sure that, if they have been

given any meaning by those who use them, they can be

readily transformed into sentences which are not inten-tional. Indeed the ease with which robot-sentencesmay be

made either intentional or non-intentional may be one

ground, or cause, for believing that sentences about the

attitudes of human beings may readily be transformedin

ways counter to our versionof Brentano's thesis.It should be noted, with respect to those universal

sentences of physics which have no"

existential import ",that they are not intentional. It is true that the sentence,"All moving bodies not acted upon by external forcescontinue in a state of uniform motion in a straight line ",

does not imply either that there are, or that there are not,such bodies. But its contradictory implies that there aresuch bodies.

5. The second part of our version of Brentano's thesisstates that, when we wish to describe anyone's believing,seeing, knowing, wanting, and the like, either (a) we mustuse language which is intentional or (b) we must use avocabulary we don't need when we talk about non-psychological facts.

Perhaps the most instructive way of looking at our thesisis to contrast it with one which is slightly different. It has

often been said, in recent years, that " the language ofphysical things " is adequate for the description of psycho-logical phenomena-this language being any languagewhose vocabulary and rules are adequate for the descriptionof non-psychological phenomena. If we do not needintentional language for describing physical things, then




this counter-thesis-the thesis that the language of physical

things is adequate for the description of psychologicalphenomena-would imply that we do not need intentionallanguage for the description of psychologicalphenomena.

The easiest way to construct a non-intentional languagefor psychology is to telescope nouns and verbs. Finding apsychological verb, say " expects ", and its grammaticalobject, say " food ", we may manufacturea technical termby combining the two. We may say that the rat is

" food-expectant " or that he " has a food-expectancy". Russellonce proposed that, instead of saying " I perceive a cat ",we say " I am cat-perceptive", and Professor Ryle hasdescribed a man seeing a thimble by saying that the man" is having a visual sensation in a thimble-seeing frame ofmind ".4 Sentences about thinking, believing, desiring,and the like could readily be transformed in similar ways.

But this way of avoiding intentional language has oneserious limitation. If we wish to tell anyone what ourtechnical terms mean, we must use intentional languageagain. Russell did not propose a definition of his technicalterm " cat-perceptive" in familiar non-intentional terms;he told us, in effect, that we should call a person " cat-perceptive" whenever the person takessomething to be acat. Our version of Brentano's thesis implies that, if we

dispense with intentional language in talking about per-ceiving, believing, and expecting, we must use a vocabularywe don't need to use when we talk about non-psychologicalfacts. The terms " food-expectancy ", " thimble-seeingframe of mind", and " cat-perceptive" illustrate such avocabulary.

I shall comment upon four general methodsphilosophersand psychologists have used in their attempts to provide"physical " translations of belief sentences. The first ofthese methods makes use of the concept of " specific

4 See Russell's Inquiry ntoMeaning andTrutfi(American edition), page 142and Ryle's Concept f Mind, page 230.




response "; references to this concept appeared in the

writings of the American " New Realists" and can still befound in the works of " behaviouristic" psychologists. Thesecond method is one which refers to " appropriatebehaviour "; ProfessorBraithwaite has provided us with aclearstatementof it. The third refersto "verbal behaviour";its clearest statement is to be found in Professor Ayer'sThinking ndMeaning. The fourth refers to a peculiar typeof " fulfilment" or " satisfaction"; its classic statement isWilliam James' so-called " pragmatic theory of truth ". Ishall try to show that, if we interpret these various methodsas attempts to show that our version of Brentano's thesis isfalse, then we can say that they are inadequate. I believethat the last of these methods-the one which refers to" fulfilment" or " satisfaction"-is the one which has thebest chance of success.

6. When psychologists talk about the behaviour ofanimals, they sometimes find it convenient to describecertain types of response in terms of the stimuli with whichsuch responses are usually associated. The bird's " nestingresponses," for example, might be defined by reference towhat the bird does in the presence of its nest and on noother occasions. Rats might be said to have "bell

responses" and " obstacle responses"; dogs may evenhave " automobile responses" and " newspaperresponses".In the writings of some of the " New Realists " we findsuch expressionsas " rain response" and " fire response"

applied to complex human behaviour and we find principlesasserted which, if true, would enable us to introduce andapply more determinate expressions-say, " dictionaryresponse" and " Christmas-treeresponse." These philo-sophers apparently assumed that, for every object of whicha man can be said ever to be conscious, there is somereaction which he exhibits when and only when he is inthe presenceof that object.5 And they felt that the specific-

5 See Chapter Nine of the Concept f Consciousness,y E. B. Holt.




response vocabulary-" rain response", "fire response",

etc.-provided a means of formulating sentences aboutbelieving and other phenomena of the sort Brentano called"international ".

A man who smells smoke may believe, in consequence,that there is a fire nearby. If we use the specific responseterminology, we may be tempted to say, therefore, that thesmoke-or the smell of smoke-has called up the man's" fire responses ". And then we may be tempted to say

furtherthat the man's belief that there is a fire near by maybe described by reference merely to the " activation " ofthese responses. (It will not do to say, merely, that his fireresponses have been put in a state of readiness-i.e., thatthe man is ready o make them, or preparedo make them-for sentences using " ready "-words in this way are inten-tional by our criteria.)

Our example could have been put in terms of " signs ",or " sign behaviour ". The smoke signifies ire to theman-or, he takes it to be a sign of fire. If we can describethe man's believing-his believing that there is a firenearby-by reference merely to the activation of his fireresponses, then we have at hand a general method fordescribing "sign behaviour". We can say that so-and-sois a sign of such-and-such for a given person if and only

if so-and-so calls up his such-and-such responses. But letus look at these responses more closely.

If smoke is sufficient to call up a man's fire responsesthen we can no longer say that man's fire responsesare theways he acts when and only when he is in the presence offire. For we are now saying that he may make theseresponses in the presence merely of smoke. What, then,

could his fire responses be ? Should we say that they areresponses which are like those he makes in the presence offires ? This would be saying too much, for in some espectsany response he ever makes is like those he makes in thepresence of fire. All of his responses are alike in beingthe result of various neural and physiological events. Butwe don't want to say that every one of his responsesis a




fire response. We can't say that a man's fire responsesare

all those things he does when he is in the presence of afire, for there will be some hings he does in the presence offire that he doesn't do when he just smells smoke.

So far as I can see, no matter how we try to define aman's " fire responses", we will find-unless we make useof intentional language-either that they are responseswhich only fire can call up or that they are responseswhichhe

will sometimes make when he doesn't believe thatanything is on fire.I think that some of the people who used. the specific

responseterminology were inclined to say these two things:(1) that a man's fire responses arethose responseshe makeswhen and only when there is a fire in his environment; and(2) that when, as we say, he believes something is on firethere is a fire there-for him. And then they could say

that if some stimulus is a fire for him it is quite capable ofcalling up those responses he makes only in the presenceof fires. But this unfortunate way of talking accomplishesvery little. Instead of saying, "The man believes there isa fire nearby ", we now say, "Something is on fire forhim". But the new sentence is no less intentional than theold; for neither the new sentence nor its contradictory mplythe unqualified sentence " Something is on fire ", and

neither imply its denial.6

7. It is easy to pass from the " specific response"terminology to the " appropriate action " terminology.The second way of defining believing want to consider is

6 Instead of saying " Something is on fire for him", some psychologistswould adopt a more serious-sounding terminology and say, " There is a firein his geographical nvironment. It is then necessary to contrast the man'sgeographical environment with that other place-his " behavioural environ-ment "-in which there may be smoke but no fire. (See Kurt Koffka'sPrinciplesof GestaltPsychology,Chapter Two.) Brentano could now put hispoint by saying that the objects we talk about in psychology, unlike those wetalk about in the non-psychological sciences, have things in their geographicalenvironments which they don't have in their behavioural environments.




one which says that to believe a proposition p is to be

disposedto act in a way which is " appropriateto p's beingtrue ". Psychologists sometimes say that, if a dog believesthere is food in the room, then he acts in a way which is

appropriateto food being there-" he displays those of hisresponses which are appropriate to the presence of food ".

But unless we can find a way of defining " appropriate",

this way of talking is intentional by-ourcriteria. When we

affirm, or when we deny, " The knight is acting in a waythat is appropriate to the presence of dragons ", we do notimply either -that there are, or that there are not, anydragons in his presence. Professor Braithwaite has triedto define " believing " in terms of " appropriate action "

and has proposedan independent definition of " appropriateaction ". I wish now to consider this definition briefly.

To say that a man believes a proposition p, according

to Professor Braithwaite, is to say that, under suitablecircumstances, the man would act " in a manner appro-priate to p's being true ". And to say that he would actin a manner appropriate to p's being true is to say this:" If at a time when an occasion arises relevant to p, his

springs of action are s, he will perform an action which issuch that, if p is true, it will tend to fulfil s, and which issuch that, if p is false, it will not tend to satisfy s. His

belief in the truth of p, if it is a sincerebelief,will lead himto act in a differentway from that in which, given the sameoccasion and the same springs of action, he would act hadhe believed p to be false ". Professor Braithwaite notesthat " a person with normal self-preservative nstincts, whocannot swim, will not jump into a pool which he believesto be deep: a person possessedof a suicidal impulse having

exactly the same belief will jump ".IIt should be noted that, in all probability, the conceptof " springs of action ", as it must be interpreted in Pro-fessor Braithwaite'scriterion, is one which would normally

7" Belief and Action ", AristotelianSociety,Supplementary Volume XX, p.10.




be defined in intentional terms. To say that a man is

" possessedof a suicidal impulse ", as this expression s usedabove, is to say " He wants to kill himself"-or "One of

his goals, or purposes, is that of killing himself". The

latter two sentencesare intentional. Hence if the definitions

are to provide us with a means of avoiding intentional

language, we must have a non-intentional definition of

" desire" or of " purpose". (This point is not a criticism

of Professor Braithwaite, for he did not set out to avoidall intentional language.) And if we adopt Professor

Braithwaite's definitions, we cannot define " desire " and

"purpose ", as some philosophers have, in terms of

"believing ".A somewhat more serious difficulty lies in the fact that

our true beliefs often contribute to the frustration of our

purposes,or springsof action. And I believe that this fact

indicates that ProfessorBraithwaite'scriteria do not supplyus with necessaryconditions for saying that a man believes

a propositionp. Let us suppose that a driver, wishing tokeep an appointmentwith a friend and believing, truly, that

the friend is waiting along a certain road, acts on his belief

by taking that road; and let us suppose further that he

has an accident, with the consequence that many of his

springs of action-including his desire to meet his friend-

are frustrated. (We may say, if we choose, that his falsebelief that the road would be safe was the real source ofhis trouble; for, had he not believed, falsely, that the roadwas safe, he would not have had the accident. But thisfact does not absolve his true belief; for the fact remains

that, had he not believed, truly, that his friend was waitingon the road, he would not have had the accident.) Here,then, we have a case where a man believesa true proposition,but where, contrary to ProfessorBraithwaite's criteria, theman's action on that true belief fails to fulfil his purposes.Hence the criteria do not supply us with necessary condi-tions for saying that a man believes a propositionp.

Nor do they supply us with sufficient conditions. Letus suppose that someone, en route to the bank where he




wants to get some money, finds a pocket-bookwhich is full

and that this find satisfies the relevant springs of action.If I interpret ProfessorBraithwaite correctly we must say,in this case, that the man believed he would find thepocket-book; for he acted in a manner which, because ofhis find, satisfied the relevant springs of action. (Thesituation becomes more difficult if we add the assumptionthat the man was believing, falsely, that the road he wason would take him to the bank.) I

think,then, that

Professor Braithwaite's criteria for saying that a manbelieves a proposition p may be satisfied by the behaviourof a man who does not believe the propositionp.8

I do not wish to minimize the relevance of our " springsof action " to our beliefs, nor do I wish to deny that thereis a close relation between them. It may be that we canmodify Professor Braithwaite'sdefinition of the expression.

acting in a manner appropriate to p's being true", in thisway "A man may be said to act in a manner appropriateto p's being true provided that: if at a time when anoccasion arises relevant to p, his springs of action are s, hewill perform an action such that, if p is true-and if all ofthe other relevant propositions which he believes are alsotrue-then his action will tend to fulfil s ". But this

definition of " appropriate action" makes use of the term

" belief" and is thus no longer available for defining"belief".

8. In the third type of definition we refer to theverbal behaviour" which we would ordinarilytake to be

symptomatic of belief. This time we try to describe aman's belief-his believing-in terms of his actual uses ofwords or of his dispositionsto use words in various ways.

Let us consider a man who believes that the MissouriRiver has its source in the northern part of Montana. In

8 In suggesting these counter-examples, I am assuming that, if an actionsatisfies s, it " tends " to do so, and that, if it does not satisfy s, it does not"tend " to do so.




saying that he believes this, we do not mean to imply that

he is actually doing anything; we mean to say that, if theoccasion arose, he would do certain things which he would

not do if he did not believe that the Missourihad its source

in northern Montana. This fact may be put briefly bysaying that when we ascribe a belief to a man we are

ascribinga certain set of dispositionsto him. What, then,are these dispositions? Accordingto the presentsuggestion,the man is disposed to use language in ways in which he

wouldn't use it if he didn't have the belief. In its simplestform, the suggestion is this: if someone were to ask the

man " Where is the source of the Missouri River ? " the

man would reply by uttering the words, " In the Northern

part of Montana "; if someone were to ask him to name

the rivers having their sources in the northern part of

Montana, he would utter, among other things, the word"

Missouri";if

someone were to ask" Does the

Missouriarise in northern Montana ? " he would say " Yes "; and

so on.

Using this type of definition, we no longer have to say

that the man, en route to the bank and about to find apocket-book,believed he was going to find the money; for,had he been questioned, he would not have displayed theappropriate verbal behaviour. And we can say that the

man who was headed for an accident did believe his friendwas waiting for him, for he would have displayed theappropriateverbal behaviour.

We should note that this type of definition, unlike the

others, is not obviously applicable to the beliefs of animals.Sometimes we like to say such things as " The dog believeshe's going to be punished" and " Now the rat thinks he'sgoing to be fed." But if we accept the present type of

definition, we cannot say these things (unless we are pre-pared to countenance such conditionals as " If the ratcould speak English,he'd now say' I am about to be fed ' ").I do not know whether this limitation-the fact that thedefinition does not seem to allow us to ascribe beliefs toanimals- should be counted as an advantage, or as a




disadvantage, of the " verbal behaviour" definition. In

any case, the definition involves a number of difficultiesofdetail and a general difficultyof principle.

The if-then sentences I have used as illustrationsdescribe the ways in which our believer would answercertain questions. But surely we must qualify thesesentences by adding that the believer has no desire todeceive the man who is questioning him. To the question" Where is the source of the Missouri ? " he will reply by

saying " In northern Montana "-provided he wants totell the truth. But this provisobringsus back to statementswhich are intentional. If we say " The man wants to tellthe truth " we do not imply, of course, either that he doesor that he does not tell the truth; similarly, if we assertthecontradictory. And when we say " He wants to tell thetruth"-or, what comes to the same thing, " He doesn'twant to lie "-we

mean,I

suppose,he

doesn't want to sayanything he believes o be false. Perhaps we should alsoadd that he has no objection to his questionerknowingwhatit is that he believes about the Missouri.

We should also add that the man speaks English andthat he does not misunderstand the questions that are putto him. This means, among other things, that he shouldnot takethe other man to be saying something other than

what he is saying. If he took the other man to be saying"Where is the source of the Mississippi? " instead of" Where is the source of the Missouri ? " he might replyby saying " In Minnesota " and not by saying " In Mon-tana ". It would seem essential to add, then, that he mustnot believe he other man to be asking anything other than"Where is the source of the Missouri" ?

Again, if the man does not speak English, it may be

that he will not reply by uttering any of the wordsdiscussedabove. To accommodate this possibility, we might qualifyour if-then statements in some such way as this. " Ifsomeone were to ask the man a question which, for him,had the same meaning as 'Where is the source of theMissouri ?' has for us, then he would reply by uttering an

Q




expression which, for him, has the same meaning as 'In

the northern part of Montana' has for us ".9 Or wemight qualify our original if-then statementsby adding thisprovision to the antecedents: " and if the man speaksEnglish." When this qualification is spelled out, then,like the previous one, it will contain some reference to the

meanings of words-some reference to the ways in whichthe man uses, applies, or interprets words and sentences.These referencesto the meanings of words and sentences-

to their use, application, or interpretation-take us to thedifficultyof principleinvolved in this linguisticinterpretationof believing.

The sentences we use to describe the meanings and uses

of wordsare ordinarily ntentional. If I say, " The German

word Riesemeans giant ", I don't mean to imply, of course,either that there are giants or that there aren't any giants;

similarly, if I deny the sentence. If we think of a wordas a class of sounds or of designs, we may be tempted to

say, at first consideration, that intentional sentences aboutthe meanings and uses of words are examples which runcounter to our general thesis about intentional sentences.For here we have sentences which seem to be concerned,not with anyone's thought, beliefs, or desires, but ratherwith the propertiesof certain patterns of marksand noises.

But we must remind ourselves that such sentences areelliptical.

If I say, of the noises and marks which constitute the

German wordRiese, hat they mean giant, I mean somethinglike this: " When people in Germany talk about giants,they use the word Riese to stand for giants, or to refer to

giants ". To avoid talking about things which don't exist,we might use the expression" gigantic " (interpretingit inits literal sense) and say: " People in Germany would calla thing ein Riese if and only if the thing were gigantic ".

nBSeeAlonzo Church's

"1On Carnap's Analysis of Statements of Assertionand Belief ", Analysis,Vol. 10 (1950).




And to make sure that the expression " to call a thing ein

Riese" does not suggest anything mentalistic, we mightreplace it by a, more complex expression about noises and

marks. " To say 'A man calls a thing einRiese' is to saythat, in the presence of the thing, he would make the noise,or the mark, ein Riese ".

Let us ignore all of the difficulties of detail listed aboveand let us assume, for simplicity, that our speakershave achild-like desire to call things as frequently as possible bytheir conventional names. Let us even assume that every-thing having a name is at hand waiting to be called. Isit true that people in Germany would call a thing einRiese-in the present sense of " to call "-if and only if the thingwere gigantic ?

If a German were in the presence of a giant and tookit to be something elsev-say, a tower or a monument-he

would not call it ein Riese. Hence we cannot say that,if a thing were a giant, he would call it ein Riese. If hewere in the presence of a tower or a monument and tookthe thing to be a giant, then he would call the tower orthe monument ein Riese. And therefore we cannot sayhe would call a thing ein Riese only if the thing were a

giant.Our sentence "The German word Riese means giant"

does not mean merely that people in Germany-howeverwe may qualify them with respect to their desires-wouldcall a thing ein Riese f and only if the thing were gigantic.It means at least this much more-that they would call athing by this name if and only if they took he thing to begigantic or believedt to be gigantic or knew t to be gigantic.And, in general, when we use the intentional locution," People use such-and-such a word to mean so-and-so",part of what we mean to say is that people use that wordwhen they wish to express or convey something they knowor believe-or perceiveor take-with respect to so-and-so.

I think we can say, then, that, even if we can describea man's believing in terms of language, his actual use oflanguage or his dispositions to use language in certain




ways, we cannot describe his use of language, or his dis-

positions to use language in those ways, unless we referto what he believes, or knows, or perceives.

9. One of the basic points in the grammar of our talkabout states of consciousness, as Professor Findlay hasobserved, is that such states always stand opposed to otherstates which will " carry them out " or " fulfil " them.'0The final approach to belief-sentences I would like to

discuss is one based upon this conception of fulfilment.I believe that, if we are to succeed in showing that Brentanowas wrong, our hope lies here.

Let us consider a lady who reaches for the tea-kettle,expectingo find it full. We can say of her that she has a" motor set " which would be disruptedr frustratedf thetea-kettle turns out to be empty and which would be

fulfilled or satisfied f the tea-kettle turns out to be full.In saying that the empty tea-kettle would disrupt orfrustrate a " motor-set," I am thinking of the disequilibra-tion which might result from her lifting it; at the veryleast, she would be startled or surprised. But in sayingthat her set would be fulfilled or satisfied if the tea-kettleturns out to be full, I am not thinking of a positive statewhich serves as the contrary of disruption or frustration.

Russell has introducedthe terms " yes-feeling" and " quite-so feeling " in this context and would say, I think, that ifthe tea-kettle were full the lady would have a quite-sofeeling."- Perhaps she would have such a feeling if herexpectation had just been challenged-if someone hadsaid, just before she lifted the tea-kettle, 'I think you'remistaken in thinking there's water in that thing." Andperhaps expectation alwaysinvolves a kind of tension,whichis relieved, or consummated, by the presence of its object.But we will be on surer ground if we describe the requisitefulfilment or satisfaction, in negative terms. To say that

10 " The Logic of Bewusstseinslagen , PhilosophicalQuarterly,Vol. 5 (1955)."ISee HumanKnowledge (American edition), pp. 148, 125; compare the

Analysisof Matter, p. 184.



9ENthNCVS ABOUT BELIEVING. 145

a full tea-kettle would cause fulfilment, or satisfaction, is

merely to say that, unlike an empty tea-kettle, it wouldnot cause disruption or frustration. The kind of " satisfac-tion " we can attribute to successful expectation then, isquite different from the kind we can attribute to successfulstrivings or " springs of action ".

Our example suggests the possibility of this kind ofdefinition: " S expects hat E will occur within a certainperiod " means that S is in a bodily state which would befrustrated,or disrupted, if and only if E were not to occur

within. that period. Or, if we prefer the term " fulfil ",we may say that S is in a bodily state which would befulfilled if and only if E were to occur within that period.And then we could define " believes " in a similar way, orperhaps define " believes " in terms of " being-disposed-to-expect ".

I would like to remark, in passing, that in this type ofdefinition we have what I am sure are the essentials ofWilliam James' so-called "pragmatic theory of truth "-

a conception which has been seriouslymisunderstood, bothin Great Britain and in America. Although James usedthe terms "fulfil" and "fulfilment," he preferred" satisfy" and " satisfaction". In his terms, our suggesteddefinition of " believing " would read: " S believes that

E will occur within a certain period" means that S is ina bodily state which would be satisfiedf and only if E wereto occur within that period. If we say that S's belief istrue, hat he is correct in thinking that E will occur withinthat period, then we imply, as James well knew, that Eis going to occur in that period-and hence that S's beliefwill be satisfied. If we say that S's belief is false, we implythat E is not going to occur-and hence that S's belief willnot be satisfied. And all of this implies that the man'sbelief is true if and only if he is in a state which is goingto be satisfied. But unfortunately James' readers inter-preted " satisfy" in its more usual sense, in which it isapplicable to strivings and desirings rather than tobelievings.




Our definitions, as they stand, are much too simple;

they cannot be applied, in any plausible way, to thosesituations for which we ordinarilyuse the words " believe ",

take ', and " expect " Let us consider, briefly, thedifficultiesinvolved in applying our definition of " believe"to one ofJames' own examples.

How should we re-express the statement "Jamesbelieves there are tigers in India ?" Obviously it wouldnot be enough to say merely:

"James isin a state which

would be satisfied if and only if there are tigers in India,or which would be disrupted if and only if there are notigers in India ". We should say at least this much more:"James is in a state such that, if he were to go to India, thestate would be satisfied if and only if there are tigersthere ".What if James went to India with no thought of tigers andwith no desire to look for any? If his visit were brief and

he happened not to run across any tigers, then the satisfac-tion, or disruption,would not occur in the mannerrequiredby the definition. More important, what if he came upontigers and took them to be lions? Or if he were to go toAfrica, believing imself to be in India-or to India, believinghimself to be in Africa?

I think it is apparent that the definition cannot beapplied to the example unless we introduce a number of

intentional qualifications into the definiens. Comparabledifficulties seem to stand in the way of applying the termsof this type of definition in any of those cases we wouldordinarily call instances of believing. Yet this type ofdefinition may have an advantage the others do not have.It may be that there are simple situations, ordinarilydescribed as " beliefs" or " expectations," which can be

adequately described, non-intentionally, by reference tofulfilment, or satisfaction, and disruption, or surprise.Perhaps the entire meaning of such a statement as " Thedog expects to be beaten " or " The baby expects to befed" can be conveyed in this manner. And perhapscsatisfaction " or " surprise" can be so interpreted thatour ordinary beliefs can be defined in terms of " being




disposed to have " a kind of expectation which is definable

by reference to " satisfaction" or " surprise". And if allof these suppositions are true then we may yet be able tointerpret belief-sentencesin a way which is contrary to thepresent version of Brentano's thesis. But, I believe, wearen't able to do so now.

10. The philosophers and psychologists I have beentalking about seem to have felt that

theywere

tryingto

dosomething important-that it would be philosophicallysignificant if they could show that belief sentences can berewritten in an adequate language which is not inten-tional, or at least that it would be significantto show thatBrentano was wrong. Let us suppose for a moment thatwe cannotrewrite belief sentences in a way which is contraryto our linguistic version of Brentano'sthesis. What would

be the significanceof this fact? I feel that this question isitself philosophically significant, but I am not prepared toanswer it. I do want to suggest, however, that the twoanswerswhich are most likely to suggest themselvesare notsatisfactory.

I think that, if our linguistic thesis about intentionalityis true, then the followers of Brentano would have a rightto take some comfort in this fact. But if someone were to

say that this linguistic fact indicates that there is a ghostin the machine I would feel sure that his answer to ourquestion is mistaken. (And it would be important toremind him that belief-sentences, as well as other inten-tional sentences, seem to be applicable to animals.)

What if someone were to tell us, on the other hand,that intentional sentences about believing and the like

don't really say anythingand

that,in

consequence, thehypothetical fact we are considering may have no philo-sophical significance? He might say something like thisto us: " The intentional sentences of ordinary languagehave many important tasks; we may use the ones aboutbelieving and the like to give vent to our feelings, toinfluence the behaviour of other people, and to perform




many other functions which psychiatristscan tell us about.

But such sentencesare not factual; they are not descriptive;they don't say things about the world in the way in whichcertain non-psychological sentences say things about theworld ". I do not feel that this answer, as it stands, wouldbe very helpful. For we would not be able to evaluate itunless the man also (1) gave some meaning to his technicalphilosophical expressions, " factual ", " descriptive ", and"

they don't say things about the world ", and (2) hadsome way of showing that, although these expressionscanbe applied to the use of certain non-psychological sentencesthey cannot be applied to the use of those psychologicalsentences which are intentional.

Documents

Sentences About Believing