Symposium: Use, Usage and Meaning
Author(s): Gilbert Ryle and J. N. Findlay
Source: Proceedings of the Aristotelian Society, Supplementary Volumes, Vol. 35 (1961), pp.
223-242
Published by: Blackwell Publishing on behalf of The Aristotelian Society
Stable URL: http://www.jstor.org/stable/4106684
Accessed: 14/03/2010 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 at
http://www.jstor.org/action/showPublisher?publisherCode=black.
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, Supplementary Volumes.
http://www.jstor.org
USE, USAGE AND MEANING
By GILBERT RYLE AND J. N. FINDLAY
I-GILBERT RYLE
In 1932 Mr. (now Sir) Alan H. Gardinerpublished The
Theory of Speech and Language (Clarendon Press). A central
theme of his book was what, with some acknowledgedverbal
artificiality,he labelled the distinction between 'Language'
and ' Speech'. I shall draw, developand apply this distinction
in my own way.
A Language,such as the Frenchlanguage,is a stock, fund or
deposit of words, constructions,intonations,clich"phrasesand
so on. 'Speech', on the other hand, or 'discourse' can be
conscriptedto denote the activityor ratherthe clan of activities
of sayingthings,sayingthem in French,it may be, or Englishor
some otherlanguage. A stock of language-piecesis not a lot of
activities, but the fairly lasting wherewithalto conduct them;
somewhatas a stock of coins is not a momentarytransactionor
set of momentarytransactionsof buying,lending,investing,etc.,
but is the lasting wherewithalto conduct such transactions.
Roughly, as Capital stands to Trade, so Languagestands to
Speech.
A Languageis somethingto be known,and we get to know it
by learningit. We learnit partlyby being taughtit, and partly
by pickingit up. For any given part of a language,a learner
may not yet havelearnedthatpart; or he mayhavelearnedit and
not forgottenit, or he may havelearnedit and forgottenit, or he
may have half-learnedit; or he may have half-forgottenit. A
Languageis a corpusof teachablethings. It is not, of course,a
staticcorpusuntilit is a deadlanguage. Nor wouldtwo teachers
of it alwaysagreewhethersomethingshouldbe taughtas a part
of that language. Is French literary style to be taught by
teachers of the French Language or by teachers of French
Literature? Justwhendoes an acceptableturnof phrasebecome
an idiom? How old can a neologism be? What about
slang?
in a languageinvolvesbutdoesnot reduce
Sayingsomething
224
GILBERT RYLE
to knowingthe requisitepiecesof that language. The speakeris
here and now employingwhat he had previouslyacquiredand
still possesses. He is now in the act of operatingwith things of
whichhe has, perhapsfor years,beenthe possessor. The words,
constructions,intonations,etc., that he employsin saying what
he says in these words,constructions,etc., is not anotherpart of
that language. It is a momentaryoperationwith parts of that
language,just as the buyingor lendingthat I do with part of my
capital is not itself a part of that capital, but a momentary
operationwith a part of it. That, indeed,is what my capitalis
for, namely,to enableme to makepurchases,benefactions,loans,
etc., with parts of it wheneverI wish to do so. It is a set of
moderatelypermanentpossibilitiesof makingparticularmomentary transactions.
If I say somethingin French,then,eventhoughwhatI say has
never been said before, I do not thereby enlarge the French
language,i.e., increasethe amountto be learnedby a studentof
the French language. The fact that he does not know what
I said does not entail that thereis a bit of the Frenchlanguage
that he has still to learn. Dicta made in Frenchare not parts
of the Frenchlanguage. They are thingsdone with partsof the
French language. You might utilise the same parts in saying
somethingidentical with or quite differentfrom what I said.
Your act of sayingit is not mine,and neitheris a partof the fund
on whichwe both draw. But dicta can notoriouslyfossiliseinto
cliches. ' Je ne sais quoi ' can now be used as a noun; and ' Rest
and be Thankful' can be a proper name.
We are temptedto treat the relationbetweensentencesand
words as akin to the relationbetweenfaggots and sticks. But
this is entirelywrong. Words,constructions,etc., are the atoms
of a Language; sentences are the units of Speech. Words,
constructions,etc., are what we have to learn in masteringa
language; sentencesare what we producewhen we say things.
Wordshave histories; sentencesdo not, thoughtheirauthorsdo.
I must have learnedthe wordsthat I utterwhen I say something
with them. I need not, and, with reservations,cannot have
learnedthe sentencethat I come out with when I say something.
It is somethingthatI compose,not somethingthatI haveacquired.
I am its author,not its employer. Sentencesare not things of
USE, USAGE AND MEANING
225
whichI have a stock or fund. Nor are my buyingsand lendings
thingsof whichI havea hoardor purseful.
In daily life we do not often mention as such the sentences
that people produce. We speak instead of their allegations,
complaints,promises,verdicts,requests,witticisms,confessions
and commands. It is, in the main, people like grammarians,
compositors,translators,amanuensesand editors who need to
referto the thingsthat people say as ' sentences', since they are
ex officioconcernedwithsuchmattersas page-space,punctuation,
syntax,plagiarisation,and so on. None the less, what they are
interested in are instances of someone, actual or imagined,
alleging,complaining,warning,joking, etc., though their special
concern is with the punctuationof them and not with their
humourousness;withtheirlengthand not with theirtruth; with
their moods and tenses and not with their relevance or
rudeness.
When Caesar said ' Veni; vidi; vici', he said three things,
thoughhe used only three Latinwords. Then is ' Vici' a word
or a sentence? The queernessof this disjunctivequestion is
revealing. What Caesar produced,orally or in writing, on a
certainday, was a laconicsentence,if a sentenceis an instanceof
someone saying something. In this instanceCaesarsaid somethingwhichwas true. But he said it usingonly one Latinword,
a word which had long been there for anyone to use anywhen
in sayingall sorts of considerablydifferentthings. The wordwas
not true, or, of course,false either. Caesarboasted' Vici', but
the dictionary'sexplanationof the verb ' Vici' need say nothing
about Caesar boasting. What it describeswas, perhaps,also
used by, interalios, some concussedgladiatorasking anxiously
' Vici?' The boast 'vici' was a differentsentence
from the
question ' vici?', though the authors of both used the same
Latinword, of whichneitherwas the inventor. The word' vici'
wasthere,in theircommonfund,to be employed,misemployedor
left unemployedby anyone anywhen. The boast ' vici' and the
query' vici?' weretwo momentaryspeech-actsin whichthis one
word was utilised for saying different things. Our question
" Is ' vici' a word or a sentence? " was queerbecauseits
subject
was ambiguous. Was it about a speech-episode,like a boast or
a query,or was it about an inflectedLatin verb? It was queer
P
226
GILBERT RYLE
also because'. . . a word or a sentence?'was a disjunction
betweenpredicatesof quitedifferentcategories,on a par with
'..a bat or a stroke?'
Is the interrogativesentence 'vici?' a part of the Latin
language? Well, would a studentstill have some Latinto learn
who had never met it? Surely not. What he had learnedis
enough to enable him to construeit if he should ever meet it.
Whathe construesareemploymentsof Latinwords,constructions,
etc.; what he must know in order to construe or understand
theseemployments,arethe Latinwords,inflections,constructions,
etc. He must know the word in order to understandthe oneword boast or question; but that knowing is not this understanding; what he had.long since knownis not what he has just
understoodor misunderstood. As we employ coins to make
loans, but do not employ lendings, so we employ words, etc.,
in order to say things, but we do not employ the sayings of
things-or misemploythem or leave them unemployedeither.
Dictions and dicta belong to differentcategories. So do roads
and journeys; so do gallows and executions.
Sometimesa persontries to say somethingand fails through
ignoranceof the language. Perhapshe stops short becausehe
does not know or cannot think of the required words or
constructions. Perhapshe does not stop,but producesthe wrong
word or construction,thinkingit to be the right one, and so
commitsa solecism. Perhapshis failureis of lessermagnitude;
he says somethingunidiomaticallyor ungrammatically;or he
gets the wrong intonationor he mispronounces. Such failures
show that he has not completely mastered, say, the French
language. In the extendedsense of 'rule' in which a rule is
anything against which faults are adjudged to be at fault,
solecisms, mispronunciations,malapropisms,and unidiomatic
and ungrammaticalconstructionsare breachesof the rules of,
e.g., the French language. For our purposeswe do not need
to considerthe sourcesor the statusof rulesof this kind, or the
authorities whose censures our French instructor dreads.
Solecisms are in general philosophicallyuninteresting. Nor,
for obvious reasons, do we often commit solecisms, save
when young, ill-schooled, abroad or out of our intellectual
depth.
227
USE, USAGE AND MEANING
The reproof' You cannotsay that and speakgood French' is
genericallydifferentfrom the reproof 'You cannot say that
withoutabsurdity'. The latteris not a commenton the quality
of the speaker'sFrench,sinceit could be truethoughthe speaker
had spoken in flawless French, or had not been speakingin
Frenchat all, but in Englishor Greekinstead. The comment,if
true, would be true of what was said whateverlanguageit was
said in, and whether it was said in barbarousor impeccable
French or English. A mis-pronunciationor a wrong gender
may be a bit of faulty French,but a self-contradictionis not a
fault-in-French. Cicero's non sequiturswere not lapses from
good Latin into bad Latin. His carelessnessor incompetence
was not linguisticcarelessnessor incompetence,if we tetherthe
adjective'linguistic' to the noun 'Language' as this is here
beingcontrastedwith ' Speech'.
There is an enormous variety of disparatekinds of faults
that we can find or claim to find with things that people say.
I can complain,justly or else unjustly,that what you said was
tactless, irrelevant,repetitious,false, inaccurate,insubordinate,
trite, fallacious, ill-timed, blasphemous, malicious, vapid,
uninformative,over-informative,
prejudiced,pedantic, obscure,
prudish,provocative,self-contradictory,
tautologous,circularor
nonsensical and so on indefinitely. Some of these epithets
can be appropriatealso to behaviour which is not speechbehaviour; some of them cannot. Not one of them could be
assertedor deniedof anyitemin an Englishor Frenchdictionary
or Grammar. I can stigmatizewhat you said with any one of
theseepithetswithoutevenhintingthat what you said was faulty
in its Frenchor whateverotherlanguageyou saidit in. I grumble
at your dictumbut not at your masteryof the languagethat it
was madein. Thereare countlessheterogeneousdisciplinesand
correctionswhichare meantto trainpeople not to committhese
Speech-faults. Not one of them belongs to the relatively
homogeneousdisciplineof teaching, say, the French language.
Speech-faultsare not to be equated with Language-faults.
Nothingneed be wrongwith the paints,brushesand canvaswith
which a portraitis bungled. Paintingbadly is not a pot of bad
paint.
Logicians and philosophersare, ex officio, much concerned
P2
228
GILBERT RYLE
with kinds of thingsthat people say or mightbe temptedto say.
Only wherethere can be fallaciescan there be valid inferences,
namely in arguments; and only wherethere can be absurdities
can therebe non-absurdities,namelyin dicta. We are presented
with aporiai not by the telescope or the trawling-net,but by
passagesin books or by ripostesin debates. A fallacy or an
impossibleconsequencemay indeed have to be presentedto us
in Frenchor English,etc. But it does not follow from this that
what is wrongwith it is anythingfaultyin the Frenchor English
in which it is presented. It was no part of the businessof our
Frenchor Englishinstructorsto teach us that if most men wear
coats and most men wearwaistcoatsit does not follow that most
men wearboth. This is a differentsort of lesson and one which
we cannot begin until we have alreadylearnedto use without
solecism 'most', 'and', 'if', etc. There are no French
implicationsor non-implications,so though 'p' may be said in
French and 'q' may be said in French, it is nonsense to say
' q does not follow fromp in the best French'. Similarly,what
is impossiblein 'The CheshireCat vanished,leaving only her
grin behind her' is not any piece of intolerably barbarous
English. Carroll'swording of the impossible story could not
be improved, and the impossibilityof his narratedincident
survives translationinto any language into which it can be
translated. Somethingwas amusinglywrongwith what he said,
but not with what he said it in.
I have a special reason for harkingon this point that what
someonesays may be fallaciousor absurdwithout being in any
measuresolecistic; i.e., that some Speech-faults,includingsome
of those which matter to logicians and philosophers,are not
and do not carry with them any Language-faults. Some
philosophers,obliviousof the distinctionbetweenLanguageand
Speech, or between having words, etc., to say things with and
sayingthingswith them, give to sentencesthe kind of treatment
that they give to words, and, in particular, assimilate their
accountsof what a sentencemeans to their accountsof what a
word means. Equatingthe notion of the meaning of a word
with the notion of the use of that word, they go on without
apparent qualms to talking as if the meaning of a sentence
could equallywell be spokenof as the use of that sentence. We
USE, USAGE AND MEANING
229
hear,for example,thatnonsensicalEnglishsentencesaresentences
that have no use in English; as if sentencescould be solecisms.
Should we expect to hear that a certain argumentis henceforth to contain an Undistributed Middle in B.B.C.
English?
My last sentencebut three,say, is not somethingwith which
I once learnedhow to say things. It is my saying something.
Nor is an execution something erected to hang people on.
It is the hangingof somebody. Partof whatwe learn,in learning
the wordsof a language,is indeedhow to employthem. But the
act of exercising this acquired competence, i.e., the saying
somethingwith them is not in its turn an acquiredwherewithal
to say things. It neitherhas nor lacks a use, or, therefore,a use
in English.
The famous saying: "Don't ask for the meaning; ask for
the use ", might have been and I hope was a piece of advice to
philosophers, and not to lexicographersor translators. It
advisedphilosophers,I hope, when wrestlingwith some aporia,
to switch their attentionfrom the trouble-givingwords in their
dormancyas language-piecesor dictionary-itemsto their utilisations in the actualsayingsof things; from theirgeneralpromises
whenon the shelfto theirparticularperformanceswhenat work;
from their permanentpurchasing-powerwhile in the bank to
the concretemarketingdone yesterdaymorningwith them; in
short,fromthesewordsqudunitsof a Languageto live sentences
in which they are being activelyemployed.
More than this; the famous saying, in associationwith the
idea of Rulesof Use, couldand I thinkshouldhavebeenintended
to advise philosophers,when surveyingthe kinds of live dicta
that are or might be made with these trouble-givingwords, to
considerespeciallysome of the kinds of non-solecisticSpeechfaultsagainstwhichthe producerof suchlive dicta ought to take
precautions,e.g., what sorts of dicta could not be significantly
made with them, and why; what patternsof argumentpivoting
on these live dicta would be fallacious,and why; what kinds
of verification-procedures
would be impertinent, and why;
to whatkindsof questionssuchlive dictawouldbe irrelevant,and
why; and so on. To be clear about the 'how' of the employment of somethingwe need to be clear also about its ' how not
to', and about the reasonsfor both.
230
GILBERT RYLE
Earlyin this centuryHusserland later Wittgensteinused the
illuminatingmetaphorsof'logical syntax'and'logical grammar'.
Somewhatas, say, indicativeverbs used instead of subjunctive
verbsrendersome would-beLatinsentencesbad Latin,so certain
category-skidsand logicalhowlersrenderdicta,said in no matter
which tongue, nonsensical or absurd. A so-called Rule of
Logical Syntax is what a nonsensicaldictum is in breach of.
But the analogy must not be pressed very far. The rules of
Latin syntaxare part of what we must learnif we are to be able
to produce or construe Latin dicta. They are parts of the
equipmentto be employedby someone if he is to say either
sensible or silly things in decent Latin. The Rules of Logical
Syntax, on the other hand, belong not to a Languageor to
Languages, but to Speech. A person who says something
senselessor illogicalbetraysnot ignorancebut silliness,muddleheadednessor, in some of the interestingcases, over-cleverness.
We find fault not with his schoolingin years gone by but with
his thinking here and now. He has not forgotten or misrememberedany of his lessons; he has operatedunwarilyor
over-ingeniouslyin his execution of his momemtarytask. In
retrospecthe will reproachnot his teachers,but himself; and
he will reproachhimself not for never having known something but for not having been thinking what he was saying
yesterday.
The vogue of using ' Language' and ' linguistic' ambivalently
both for dictions and for dicta, i.e., both for the words, etc.,
that we say thingsin and for whatwe say in them,helpsto blind
us to the wholesale inappropriatenessof the epithets which
fit pieces of languageto the sayingsof thingswith those pieces;
and to the wholesaleand heterogeneousinappropriatenessesof
the variegatedepithetswhichfit thingssaid to the language-pieces
that they are said in.
and language-patterns
It remainstrue that philosophersand logicians do have to
talk abouttalk or, to put it in a moreVictorianway, to discourse
about discourse. But it is not true that they are ex officio
are ex officioconcerned
concernedwith what language-teachers
with.
II-J.
N. FINDLAY.
I AMin great agreementwith what I regardas the substantial
points in Professor Ryle's paper. His definition of language
I think ratherarbitrarilynarrow:for him it is a ' stock, fund or
deposit of words, constructions,cliche phrases and so on'.
I should have thought it would be wrong not to include in a
languagethe varioussyntacticaland otherruleswhichrestrictour
employmentof the capitalof expressionsmentionedby Professor
Ryle,thoughperhapsI am wrongin thinkinghe meantto exclude
them. That adjectives must agree with the gender of their
substantivesin certaincaseswould certainlybe held to be partof
the Frenchlanguage,as it is not part of the English. Thereis
also, I think, a furtherarbitrarinessin excludingsentencesfrom
language,and in making them the units of speech which are
produced when we say things. I think we can and should
distinguishbetweenthe sentenceJe ne said quoi as a mere possibility permittedby the Frenchlanguage,and the same sentence
as used or producedby someoneto say something. I can in fact
see no good reason why one should not have a narrowerand a
widerconceptionof a language. On the narrowerconception,a
languageincludesa vocabularyand rules, whereason the wider
conceptionit includesalso all the possible sentencesthat could
be framed out of the vocabularyin accordancewith the rules.
In this sense Frenchor Englishwould includeall the permissible
sentencesthat could be framedin it, whetheranyoneeveruttered
or wroteor thoughtthemor not. If thisconceptionof a language
makesit absurdlywide, the conceptionof it as a vocabularyplus
rules makes it unduly narrow. Certainly,however,I think we
wantto distinguishbetweena sentenceas a gramatically
permissible
word-combination,and the utteranceor writingdown or silent
thinkingof that sentenceby someoneon some occasionto make
an allegation,raisea query,expressa doubt, etc., etc., and in the
latter case I find a languageof use or employmentmore natural
than Professor Ryle's language of production. I think therefore
that Professor Ryle is legislating rather vexatiously in forbidding
us to speakof sentencesas partsof language,or to say that such
sentencescan be usedby speakers. I do not, however,thinkthat
this vexatiouspiece of legislationis in the forefrontof Professor
Ryle'sintentions.
232
J. N. FINDLAY
What ProfessorRyle is mainlyconcernedto do seemsto me
to be to distinguishbetweengrammaticalfaultsin the use of words
in constructingsentences, and faults in what may be called
' logical syntax' or 'logical grammar', whichinvolve the use of
words to constructperfectlygrammaticalsentences,but which
none the less violate a deeperset of rules,the rules of sense, the
rules of logic, the rules regulatingthe mutualrelationsof categories,etc., etc. Withall this I am deeplyin agreement,because
it involvespreciselythe recognitionthat differentsorts of words,
as it were, make differentsorts of abstractcuts in their subjectmatter,or help to executedifferentsorts of abstractcuts-some,
as Aristotlemightsay, tell us whatthingsare, othershowthey are,
others how many they are, others conjoin, others emphasize, others
bracket,etc., etc.-and that in makingsuch quite differenttypes
of cross-sectionthey become subjectto the relationsnecessarily
obtainingamong such cross-sections,so that some verbalcombinationswhich are smooth and pretty grammaticallynone the
less make hideous nonsense. ProfessorRyle, it seems to me, is
heresuggestingthatit is the relationsof differentsortsof meanings
to one another which determinethe depth-grammarof words,
and that thesemeaningsand theirrelationsare mattersthat must
consideredif we are to study logical as well as
be independently
grammaticalsyntax. If this suggestionis not implicit in his
words,perhapshe will explainwhat sort of abuse of wordsit is
as opposedto merelysurfacethatis logicalor depth-grammatical
I
abuse.
Incidentally, feel in the contextsinvoked
grammatical
it
is
that
by Ryle
doublytemptingto talk of the use and abuseof
The sentenceis there,a fully-fashioned
sentences.
grammatical
and
it is its use to express a categorially
grammaticalentity,
combination
of
meanings which is at times possible
possible
and legitimate,whereas at other times there is really only an
abuse.
Havingexpressedmy agreementand disagreementwith Ryle,
I may perhapsallow myselfto dwella little on the famousdictum
whichhe quotes and which has dominatedphilosophicaldiscussion for the past twentyyears: ' Don't ask for the meaning:ask
for the use .' I wish to make against it the not often raised
objectionthat the use for whichit bids us ask, is of all thingsthe
most obscure,the most veiled in philosophicalmists, the most
remote from detaileddeterminationor application,in the wide
USE, USAGE AND MEANING
233
rangeof philosophicalconcepts. Thereis, I think,a use of' use '
whichis humdrumand ordinary,but in whicha studyof the use
of expressionsis of only limited philosophicalimportanceand
interest. Thereis also a use of ' use' characteristicof the later
writingsof Wittgensteinwhichis utterlyremotefromthehumdrum
and ordinary,and whichhas won its way into the acceptanceof
philosopherslargelybecauseit has seemedto have the clearness
andthe straightforwardness
of the ordinaryuse. Weareall proof
the
of words like 'substance', 'being',
deceits
against
glozing
' nothingness', 'consciousness', etc., etc.: we at once see that
some occasionsof their employmentare reallyonly abuses--but
we arenot yet proofagainstthe fascinationsexertedby the singular
abuses of so ordinarya term as 'use'. When these abusesare
exposed,the whole attituderepresentedby the slogan quotedby
Ryle revealsitself as completelywithout significantbasis, which
unfortunatelyputs an end to all but a limitedemphasison' use'
and ' usage' by philosophers. Sincethe suggestionthat use and
usage-in some acceptable sense--are philosophically very
important,certainlyunderliesRyle's paper,I need not apologize
for irrelevancein proceedingto demolishthis suggestion.
The reason why it is absurdto tell us not to attend to the
meaningof expressionsbut to concentrateon theiruse, is perfectly
simple:it is that the notion of use, as it ordinarilyexists and is
used, presupposesthe notion of meaning(in its centraland paradigmaticsense),and that it cannotthereforebe used to elucidate
the latter, and much less to replaceor to do duty for it. The
notion of use is a widernotion than the paradigmaticnotion of
meaning:it covers many otherthings beside the meaningof an
in its paradigmaticsense is
expression,but the meaning-function
it is not possibleto give
one
of
the
and
it
covers,
certainly
things
a clearaccountof use withoutpresupposingthis function. What
I am sayingis simplythat we cannot fully say, in a great many
cases, how an expressionis used, without saying what sort of
thingsit is intendedto referto, or to bringto mind,andjust how,
or in what angleor light, it purportsto referto them, or to bring
themto mind. And in caseswhereit wouldbe wrongand absurd
to say that an expressionindependently
brought somethingto
none the less be
it
would
or
it
in
a
certain
mind, presented
light,
do
such
to
it
to
that
thingsin some
uncontestablyright say
helped
definite matter, so that what was brought to mind would be
234
J. N. FINDLAY
different, or differentlypresented, if the expression were not part
of our total utterance. Thus if I make use of the word ' dragon '
in a large number of contexts, I use it to refer to a human being
or beings, generally mature and female, and I use it also to
represent such a human being or beings as being restrictive,
uncompromising and somewhat terrifying. And if I apply the
term in a certain context I see that to which I apply it in the light
connoted by my words. And if I use the words ' such a' before
uttering the word ' dragon ', these words certainly help to suggest
that what I am describing is very restrictive, very uncompromising
and very terrifying, i.e., they contribute to the force of my description without playing an independent part of it. In saying what
the use of my expressions is, I therefore have to say what, in the
ordinary diction of logicians, they denote and connote, what their
precise reference is or what their general scope, or how they
contribute to connotation or denotation, and it is not thought
possible to say how many expressions are used, without bringing
in such connotative and denotative particulars
The notion of use of course goes far beyondthat of connotation
and denotation, and it is one of the extremely important discoveries of modern semantics that there are some expressions
whose use, in certain contexts, is not to connote or denote anything,
nor even to help to do either, but to do such things as give voice
to feelings and wishes, evoke certain attitudes in others, or perform
certain formal social acts, e.g., promises, which have certain
definite social consequences, etc., etc. That not all expressions,
on all occasions of their use, perform the functions of reference
or characterization, or assist in such performance, is certainly a
discovery not to be underestimated, which has cleared the deck
of much tangled tackle and many stumbling-blocks. But this
kind of non-referential, non-connotative use is parasitic upon a
connotative, referentialone, and could hardly exist without it. It
is one of Wittgenstein's more irresponsible fancies that there could
be a language composed only of commands, or only of curses,
or only of greetings. The concept of use also certainly covers all
the hidden implicationsand suggestionswhich attach to the writing
or utterance of a word or a sentence, but which are not strictly
part of what it means or says: thus when I say' He did not commit
this murder' I may use this sentence to imply that he committed
certain other murders, that I absolutely believe him to be no
USE, USAGE AND MEANING
235
murderer,that we live underlaws forbiddingthe taking of life,
etc., etc. But all such implicationsand suggestionsaie likewise
dependentupon the function of directlyconnotingor denoting
something,and are in fact an extensionof the same. Use also
obviouslycoversthe mererequirementsof accidenceand syntax,
in the
thoughthese,as Ryle has shown,are mereinstrumentalities
task of significantdiction.
What is implicit,however,in the slogan 'Don't ask for the
meaning:ask for the use' is not that use coversmuchmorethan
the connotativeand denotativefunctionsof language,but that it
somehowresumesand completelyexplainsthe latter,that we can
completelysee aroundand see throughtalk about the reference
and connotationof expressionsby takingnote of the way people
operatewith suchexpressions,how they combinethemwith other
in
expressionsto form sentences,and the varyingcircumstances
whichproducingsuch sentencesis reckonedappropriateor fully
justifiable. This studyof verbalmanoeuvres,and of appropriate
and justifyingcircumstances,must not, however,be confinedto
the single instant of utterance:it must point backwardsto the
all-importantsituationsin whichuse was learntor taught,and it
must pointforwardsto the innumerablesituationsin which the
utterancein questionwill again be found appropriate,or will be
found to be more and more abundantlyjustified. The study of
use thereforeincludesa genealogyand a prognosisof the most
indefiniteand complexkind, much more extensivethan any that
occursin a merelygrammaticalor philologicalstudy. In another
respect, however, the slogan gives 'use' an extraordinarily
restrictedinterpretation. The operationsinvolvedin use are not
to be operationsconductedprivatelyin anyone'shead, or at least
such operationscan only be broughtinto considerationin so far
as theycan be narrowlytied up with othernon-privateoperations,
and the circumstancesin which such operationsare conducted
must all be circumstancesbelongingto what may be called the
common public environment,circumstancesin which bricksare
being assembledinto buildings,apples taken from drawersand
handed over to customers,railway-signalsaltered, or hunting
expeditionsconducted. Thesortof changewhichis a merechange
in perspectiveor in conscious'light' is not among the circumstancesmentionablein describinguse.
And thereis yet anothermost extraordinary
restrictionplaced
236
J. N. FINDLAY
in whicha wordis correctly
uponouraccountof the circumstances
used: we must not employthe word or its equivalentto explain
those circumstances. We mustnot, e.g., say, that whena man is
confrontedby threeapplesin a drawer,or by an appleandanother
appleand yet anotherapple, he is thenjustifiedin employingthe
word ' three' in connexionwith such apples. The word ' three'
may be employed in describingthe circumstancesjustifying
countless other sorts of utterance,but not the circumstances
justifyingits ownemployment. In the same way we must never
say that it is when a man is confrontedby a red object, or has
learnt to discriminateits colour, that he is justifiedin callingit
'red'. Such accountsare held to be wholly trivialand unilluminating,and are moreoverheld to suggestvariousdeep philosophicalfallacies:the beliefthatmeaningsexist' out there' in the
thingswe deal with beforewe findthe appropriatewordsto ' pick
them out ', or that they exist ' in the mind' or the understanding
before we find wordsto expressthem. Whateverwe suggestby
our accounts of use, we must never suggest that there arepreexistentmeanings. Wordsenjoy meaning and referencein and
by our use of them, and our use cannot be explainedin terms
of any meaningthat antedatesthe use of words. And since
understandingand thinking are defined in terms of the
operationwith signs, we must neverspeak as if we could understand or think anythingbeforewe dispose of appropriateverbal
expressionsand have been taught to employ them. The programmeof this extreme'utilitarianism'-as one may perhaps
call the use-doctrine-is impressive:it resemblesnothingso much
as the braveempiricistprogrammeof Lockeand Hume,in which
no idea was to be admittedinto the charmedcircleof thoughtand
knowledge without producing a genealogy purer than any
requiredby the Nuremberglaws, exhibitinga proper origin in
sensationand reflection,and a derivationfromtheseby approved
processes. But, like that braveprogramme,it faces the crucial
objectionthat it cannot be carriedout completely,and that no
comprehensiveaccountof use and usagecan be givenwhichdoes
not contain some membersof impure origin. That the brave
programmewas hopelessWittgensteinhimselfperhapsobscurely
realized, when he wrongly said of the BrownBook, the most
profoundand wonderfulof his writings,that it was nichtswert.
But if success, rather than stimulus and provocation, is the
USE, USAGE AND MEANING
237
criterion of philosophical value, his judgment was entirely
justified.
I neednot rangefar nor cite manyinstancesto makeplainthe
totally unilluminating,indeed deeply obfuscatingcharacterof
attemptsto give a completeaccountof the use of expressionsin
termsof merelypublic operationsand circumstances. The very
conceptionof a rule, centralto the 'utilitarianism'in question,
aboundsin difficulty. For we are expresslytold that to follow a
ruleis not necessarilyto be guidedby a spokenor writtenformula,
since each such formula admits of interpretationin terms of
anotherformula,and this in termsof another,and so on indefinitely. Nor is the followingof a ruleto be identifiedwithany sort
of innerpersonalunderstanding
of that rulewhichcan guideone's
subsequentperformance,since to hold this would be to accept
pre-existentmeaningsresidentin the queermediumof the mind.
Nor can the following of a rule be identifiedwith one's actual
performanceup to a point, sincethis is alwayscompatiblewithan
infinityof rules. In the end it would seem that followinga rule
must be an ineffablesort of affair:it must be somethingthat can
be accomplishedin one's doing(in this case, speaking),but not
effectivelyspokenabout. It is somethingthat one can knowhow
to dowithoutbeingableto knowhow whatone doesis done. The
conceptionof a linguisticrule has, in fact, all the irretrievable
obscurityof the structuralresemblanceconstitutiveof meaningin
the Tractatus,whichcannotbe expressedbut only shown. If it is
at least possiblethat a rule should at times be understood or
graspedin thought,we can understandwhatit is like to follow it
without thought, but if graspingis a function of following,the
whole activityof followingdissolvesin mystery. I do not myself
know how it differsfrom the most arbitraryirregularityexcept
that it mysteriouslyfeels right at every stage, and that others,
standingat my side,mysteriouslyagreein suchfeelings. And if it
is hardto throwlighton the followingof rulesin termsof outward
circumstancesand performances,how muchharderit is to say in
what lies conformityto an openrule, one whichis to be applied
over and over indefinitely. While the thoughtexpressedby the
phrase' and so on indefinitely' is most absolutelysimpleand easy
to entertain,it is a thought logicallyimpossibleto evince adequatelyin one's performance. Much has been written,from the
standpointof the use-doctrine,aboutthe differencebetweenclosed
and open games,but the discussionendsup with verymuchwhat
238
J. N. FINDLAY
it startedfrom, that it is a differencein the spiritwith whichthe
respectivegames are played. A man, e.g., using an open arithmetic simply has a system or general rule for constructing
numerals indefinitely. That a spirit is operative in this case
I should not care to deny, but that it consorts well with
the use-doctrine, or establishes its superiority, I cannot
conceive.
Similardifficultiesconfrontus if we considerthe use-accountof
the use of descriptiveadjectiveslike those of colour. We are
forbiddento talk of prior colour-differences
in objects,or prior
in persons,as this wouldinvolve the grave
colour-discriminations
error of positing pre-existentmeanings. We are introduced
to imaginarytribalactivitieswhichinvolvethepicturesquecarrying
about of chartsof colour samplesand their comparisonwith, or
impositionon objects,but these it would seem explainlittle or
nothing,sincethe chartsare dispensableand admitmoreoverof a
wrong use. From the use of chartsthe tribe progressesto the
use of colour samplescarriedsomehowin the mind's eye, and
ultimatelyto the mereunhesitantpronouncement,aftersufficient
training,of certaincolour-wordsin the presenceof certainobjects.
With this pronouncementothersas unhesitatinglyagree. From
the Scylla of holding that 'blue' stands for a discriminable
bluenessin objects,or expressesan awarenessof bluenessin one's
mind, one proceedsto the Charybdisof sayingthat those things
are blue whichwe and othersagree,and havebeentrained,to call
so. It is plain,of course,thatone musthaveultimatessomewhere,
and it is plain also that thereare differentpossibilitiesof colourdiscriminationcorrespondingto differentpossibilitiesof usage:
whatis notplainis whyone shouldprefersucha strange,secondary
ultimateas a use to the more obvious,understandableultimates
of discriminatingthoughtsin the mind, or discriminablefeatures
in things.
The most superbexampleof the problem-increasing
character
of the use-semanticsis, however,to be foundin its treatmentof
caseswheremen use expressionswithoutobviousreferenceto any
palpablefeatureof the publicenvironment,whenthey give voice,
e.g., to recollectionsor anticipations,or describetheir personal
feelingsor impressions,or reporttheirfantasiesor their dreams.
Here the course is followed of attemptingto account for such
uses by supposingmen to be suchas spontaneously
to wantto use
in
certain
in
contexts contexts where their
expressionstaught
USE, USAGE AND MEANING
239
normaljustificationis absent,andthat thesenon-normalneeds,so
strangelyuniversal among us, constitute the basis for a new
secondaryset of linguisticusages,wherethe sole fact that we agree
in feeling certain linguisticurges is the sole criterionof their
correctness. Thus childrenperhapsspontaneouslyrun over the
namesof objectsrecentlypresentedto them,or can be encouraged
to do so withoutdifficulty:meaningcan then be givento thepast
tense,andtheycanlearnto saythattheyhada ball,a stick,a rattle,
etc. To ' referto the past' is merelyto learnto employthe past
tense in such circumstances,an account as amusinglyfree in
presupposingpastnessand temporalpassagein the circumstances
of thelearning,as it is firmin denyinganynon-verbalunderstanding
of them. Men then spontaneouslybegin to use the past tense
wherethereis no such recentprovocation:we then give a use to
talk about 'remembering', particularlyif others agree in such
spontaneousinclinations. The referenceto the pastin memoryis
thereforenot the ultimate,mysteriousthing that Husserl,Broad
and othershave supposedit to be: it merelyreflectsthe strange
tendencyof men to talk preteritivelybeyondthe limitsof recency,
andthefurtherlinkageof thisfactwiththe readingsof instruments,
the reportsof others,and manyotherobservedmatters. It may
now happenthatmenwakingfromsleepspontaneouslytalkin the
pasttenseas if recallinghappeningswhichno one remembers,and
which do not fit in with the observablecontemporarystate, or
with the memory-inclinations
of others. The conceptof' dreaming' now makes its ddbutto take care of these extraordinary
performances. Malcolm,the admirableexponentof a preposterous analysis,admitsthat on it dream-language
is veryodd: it is as
if one is faithfullyrecallingsomething,but one cannot explain
this fact by sayingthat one didexperiencewhatone is disposedto
report,since this would involvean unintelligiblehypothesis,one
excludedby the guidingassumptionsof the doctrineof use. What
these queernessessurely show is the profound mistakenness
somewhereof these guiding assumptions. To make use of a
gnostic principleused by Moore in other contexts: we know
certainfactsaboutmeaningmuchmoreabsolutelythanwe can be
sureof the premisses,or the inferentialrules,of semanticarguments
designed either to establish them, or to explain them away.
Obviouslywe cannotmakestraightsenseof manylinguisticusages
without postulatingjust those pre-existentunderstandings(not
confined to matters in the public forefront) and the possibility of
240
J. N. FINDLAY
communicatingsuch understandingsto others, which it is the
whole aim of the use-doctrineto exclude.
The use-doctrinemay furtherbe objectedto for its profoundly
circular,question-beggingcharacter. This is a point ably made
by Mr. Gellner in a book where some of the most profound
criticismsof the use-doctrineand its consequenceslie hidden
undera somewhatpopularexterior. To haveseenanunacceptable,
unarguednaturalismbehind Wittgenstein'sbrilliant fagade of
exposition,is no meaninsight. By describingthe functioningof
linguisticexpressionsexclusivelyin publicand social terms,we at
once go toofar in assumingsuchapproachesto be whollyjustified
and clear,and we also do notgofar enoughin refusingto recognise
aspects of languagenot fitting an approachof this sort, or in
' proving' themto be misguidedor senseless. Thesetwo lines of
objectionreallycoincide,sinceit is by turningawayfrom aspects
of languageit cannotreadilyaccommodatethat the use-doctrine
is unable to see its own difficultiesand obscurities. The usetheoristshave dwelt much on the profoundsubtletyof ordinary
language,but they have been far from recognizinghow subtleit
actuallyis. For it not only uses expressionsto point to, or to
throwlight on, ordinaryobjects,but it also uses them reflexly, in
the mannerstudiedin Husserl'sLogischeUntersuchungen,
to point
to or throw light on its own meanings,thereby setting up
an orderof objectsas clear-edgedandpartialas its normalobjects
are fuzzyand full, and as delicatein theirabstractionas they are
indispensablefor the higherflights of thought. That a phrase
like 'the third door on the right' can be used both straightforwardlyto refer to a door, and reflexlyto refer to its own
meaning,is a truthplain to babes, but occasioningheadachesto
the semanticallyover-wiseandprudent. Ordinaryspeech,further,
providesus with an instrumentfor communicatingwith others
aboutmatterspublicand common,whichis also an instrumentfor
purely personal use, in which differentobservations,different
views,differentjudgmentsprovidemuchthe samecomplementary
parallax,and the same correctiveor confirmatorytesting as in
the interpersonalcase. But not only is it thus double in its use,
it also managesto incorporatethe personalin the publicuse, and
the publicin the personal,in a regresspursuableas far as even
we choose. Thus we all understandother people's first-person
talk by analogywith our own, and its imperfectpublicintelligibility is also perfectlyand publicly intelligible,since everyone
USE, USAGE AND MEANING
241
makes just such first-person statements in his own case. The
manner in which we smoothly swing over from another man's
perfectly understood use of the first-person pronoun 'I ', and
replace it with ' he ' in reporting the content of his statement, and
expect the other man to do the same in regard to us, as well as
the children's games in which these proprieties are amusingly
violated: all these show an understanding of the antithesis of
contrasted privacies, and of their overcoming in a wider
publicity, of which the use-semantics betrays no inkling. In the
same manner, ordinary speech has in it the germs of what may be
called the Cartesian or the Lockean inversion, the reversal of the
ordinary approach from outward things to the mind, into an
approach to outer things from the facts of our subjective life.
Though the language in which we talk of and to ourselves-the
best subject-matter and audience-may have had its source in
contexts of public ostensibility, it can, by a profitable ingratitude,
use the personal language thus painfully acquired to cast doubt
upon, or to throw light on, its own origin. We may illuminate
our understanding and knowledge of public matters in terms of
just those personal experiences and pre-existent understandings
which talk about public matters first renders possible. And
this personal Cartesian or Lockean story can then achieve the
widest publicity, since to have back rooms continuous with those
opening on the public square is the most universal, most
inescapable of predicaments. It is no doubt by a creative transformation that the rumour of the square penetrates backwards,
and is re-echoed in the small back rooms, and it is likewise by a
creative transformation that these transformed echoes rejoin the
rumour of the square. All this, however, unquestionably
happens, and it is the task of a philosophical semantics to make
sense of it, and not to declare it unintelligible.
Nothing that has been said in the foregoing is meant to reflect
on the painstaking, detailed study of linguistic usage, or the actual
manner of its teaching, if used to show how we actually come to
mean what we undoubtedly do mean, or to throw light on the
complexity and subtlety of our meanings, or to show how we
come to be misled into supposing we mean what really conflicts
with the ' depth-grammar' of our meanings. Our criticisms are
only of a radical use-theory carried to extremes, which constructs
fables as to how we might have been taught the meanings of words
in order to buttress a priori doctrines as to what we must or cannot
Q
242
J. N. FINDLAY
mean. If anyone thinks such doctrines archaic and superseded,
and so not requiring rebuttal, he is wide of the truth. Wittgenstein's accounts of language-games are so arresting, so novel, so
subtle in their detailed development, so daring in their frank
embrace of the unplausible, so imbued with intellectual seriousness
and earnestness, and so great, finally, in their aesthetic appeal, that
it is hard to see through them or around them. They fascinate
the philosopher in the same way that Wittgenstein claimed that
philosophers were fascinated by the forms of ordinary language,
and against such fascination determined steps are necessary.
The steps I have taken in this paper may not have been sufficiently
subtle, and may have involved certain misunderstandings
of detail: I shall hope, at least, to have incited others to do
better.
All this should not, of course, be taken as reflecting on the
philosophical greatness of Wittgenstein. Wittgenstein is the
author of three wholly differing accounts of meaning, all of which
merit entire rejection: meaning is not reduplication of structure,
it is not verification or verifiability, it is plainly not what he meant
by ' use '. It is not these things, though it is of course intimately
connected with them all, but it will be best illuminated by construing further the old humdrum notions of connotation and denotation, and by seeking painfully to throw light on the 'thought
behind our words ', for which, on account of the peculiar categories it involves, it would seem that no adequate surrogate has
been, or can be, offered. It is, I surmise, in the 'intentional
nature of thought' that the true solution of the problems of
meaning is to be found. But by formulating these three inadequate accounts, Wittgenstein has given the semantic problem the
central place it deserves in philosophy, and has contributed vastly
to its solution. Through his inability to account satisfactorily
for certain linguistic performances, he has indicated the precise
nodes where language makes its various creative leaps and
has thereby given philosophical semantics its opportunity and its
task. Moreover, each of Wittgenstein's frequent rhetorical
questions is such that, if answered in the sense not intended by the
question, it will lead to an illuminating result: they are practically
all arrows which, if read in the reverse direction, point unerringly
to some truth. A philosophy of meaning so valuably wrong
does not differ profoundly from one that is systematically
right.
ARISTOTELIAN
ARISTOTELIAN SOCIETY
SOCIETY
CORRIGENDA
CORRIGENDA
for
ume
335
5
for
Supplementary
Supplementary VolVolume
lline
ine
44 ffrom
rom
tthe
he
ffoot.
o t.
7,
PagePage 1117,
FFor
or
rread
ead
incompatible
ease
incompatibleplplease
compatible
compatible
llines
ines
44 tto
o
55.
.
2224,
4,
Page
Page
FFor
or
The
read
His
The
words,
read
His
words,
constructions,
intonations,
constructions,
intonations,plplease
ease
o
f
t
h
e
employment
constructions,
employment of the words,
words, constructions,
intonations,
intonations,
ine
1.
8,
, lline
PagePage 2228,
paragraph
paragraph 22,
1.
FFor
or
r
e
a
d
read
harki
ng
plplease
ease
harking
harpi
ng
harping