GRICE AND FLEW on Philosophy and Language

Speranza

Grice and Flew propose
to attack a miscellany of popular miscon-
ceptions, trying Incidentally to illuminate various possibly
puzzling- practices. A very typical passage from Aristotle's
Nicomachean Ethics will serve as a text:

We must also grasp the nature of deliberative excellence
vj3ov\ia and find whether it is a sort of knowledge, or of
opinion, or of skill at guessing ixjro"xia, or something different
from these in kind. Now it is not knowledge : for men do not
investigate l^rovai matters about which they know, whereas
deliberative excellence is a sort of deliberation, and deliberating
implies investigating and calculating. But deliberation is not the
same as investigation : it is the investigation of a particular sub-
ject [t.e. conduct A. FJ. Nor yet is it skill at guessing : for
this operates without conscious calculation, and rapidly, whereas
deliberating takes a long time. . . . Correctness cannot be predi-
cated of knowledge, any more than can error, and correctness of
opinion is truth (Bk. VI, ch. ix: 1142 a 32 ff.).

Objections', (i) 'But imagine that a man knew that
there was a body buried in his back garden, and neverthe-
less joined with the police in their investigations : would that
not be investigating a matter about which he already knew? J

(ii) 'But surely it is sometimes all right to speak of
erroneous knowledge: as when sarcastically I say: "He
knew the winner of the two-thirty: but he knew wrong "?'

1 This paper was originally commissioned by the Philosophical Quarterly as
a cross between a survey of work of a certain sort published since the end of the
Second German War and an apologia 'pro philosophia nostra contra murmurantes.
Hence it was to a quite exceptional degree both polemical in tone and burdened
with footnotes. For this reprinting the tone has been softened and the burden
lightened a little. But the former is considerably sharper and the latter very
much heavier than they would be if I had been writing now and for this present
purpose.

i



2 ESSAYS IN CONCEPTUAL ANALYSIS

Replies: (i) 'No, it would in his case, but not that of
the police, only be pretending to investigate, a matter of
"investigating" (in inverted commas, making the protest
that this is a bogus case of investigation). To anyone
who knows that the man knows that the body is there,
and yet sincerely persists in saying that that man is
investigating, and not pretending to investigate or
"investigating" (in snigger quotes): what else can we
say but "You just do not know the meaning of the word
1 investigate '"?'

(ii) 'You are quite right, of course: but your excep-
tion is one which, properly understood, only helps to
reinforce Aristotle's thesis. For the whole sarcastic point
of the use of the expression "knew wrong" and of saying
"he 'knew'" (in that sniggering inverted comma tone of
voice) depends absolutely on the (logical) fact that "He
knows p" entails "p"; that it is incorrect to say "He
knows p" unsarcastically if you or he to your knowledge
have reason to doubt p. 1 And, again, if anyone has reason
to doubt p (or, still better, knows not p) ; and yet sincerely
and unsarcastically insists "He (there) knows p" : what else
can we say but "Either you do not know the meaning of
the word ' know ' and are ignorantly misusing it ; or else you
have your own peculiar use for the word which I wish you
would explain and try to justify" ? 7

Notes : (i) It is appropriate to build our basic example
here upon a passage of the Nicomachean Ethics : since
most of the avant-garde of Oxford philosophy since the war
(Austin, Hart, Hare and Urmson, for instance) are soaked
in this book ; and there is a very strong analogy between
their work and it.

1 See J. L. Austin's classic 'Other Minds' in Logic and Language, Vol. II
(ed. A. Flew, Blackwell, Vol. I, 1951, Vol. II, 1953), and also in of S. E.
Toulmin's 'Probability', Chapter VIII below. I shall use LL 3 I and LL, II
as abbreviations for Logic and Language, First and Second Series, respectively.
I apologize for the frequency of these references : but those wishing to look up
some of the articles mentioned here will presumably be glad to reduce the
number of volumes with which they have to deal; while certainly no one will
wish to have repeated anything I have said before.



PHILOSOPHY AND LANGUAGE 3

When someone like Ryle says 'We don't say* or 'We
can't say' or uses any of the semi-equivalent expressions of
the material mode of speech ; and we can think of occasions
on which we might and do intelligibly and not incorrectly
say precisely what he says we cannot say : it is a good rule
to consider whether these exceptions do not in fact actually
reinforce the point he is really concerned to make, or whether,
if not, they are really relevant to it, involving the same use
of the word. No one is infallible, and certainly not Ryle in
this matter, but we should allow for the fact that a self-
contradictory or otherwise logically improper expression may
get a piquancy precisely as such ; and can thus acquire a
use, a point, which depends entirely on the fact that it is a
misuse, and is thus parasitical on the logico-linguistic rule
to which it is an exception. 'He knew but he knew wrong',
'bachelor husband', and 'the evidence of my own eyes' all
get their piquancy in this way.

(A) ' But Aristotle was not concerned with mere words :
whereas your replies to objections involve nothing else.' A
closer look at the example will show that and how this anti-
thesis is here crucially misleading. The replies are not about
words in the way in which protests at the replacement of
'men (and women)' or 'people' by '(male and female) per-
sonnel' are about words. 1 Nor do they concern English
words to the exclusion of equivalents in Greek or Chocktaw.
Nor do they even concern words as opposed to non-verbal
signs doing the same jobs. (Consider the camp-fire version
of 'Underneath the spreading chestnut tree', of which our
late King was so fond, in which gestures replace some of the
words.) Rather they are about the uses of certain words,
the jobs they do, the point of employing them : their -mean-
ing^ and the implications which they carry.

Thus it would be no more necessary to mention the
particular Ehglish words 'investigate' and 'know' in trans-
lating the replies into another language than it is to mention
ts- and emar^^ in rendering Aristotle's argument from

1 'God created personnel in his own image' (Sir Alan Herbert)



4 ESSAYS IN CONCEPTUAL ANALYSIS

the Greek. Though English-speaking philosophers some-
times speak of correct or standard English this must not be
mistaken to imply that they are concerned with English as
opposed to other languages (usually: but see (B) below). 1
The replies, like Aristotle's theses and the objections to them,
are all equally concerned with logic as much as with lan-
guage. The whole enquiry is logical rather than philological,
an examination of the 'informal logic' 2 of two workaday
concepts. Hence the fashion for expressions such as 'the
logic of (our) language \ 'logic and language ', 'the logic of
" probable '", 'the logical behaviour of " God "-sentences',
and even 'logical geography' is not necessarily just a point-
less irritating fad ; though nothing we have to say will do
anything to justify 'The Logic of British and American
Industry' or 'The Logic of Liberty' when used of enquiries
neither in the linguistic idiom nor even conceptual.

(B) This suggests why philosophers given to talking about
correct English 'seem to take little account of the existence
of other languages whose structure and idiom are very
different from English ... but which seem equally if not
more capable of engendering metaphysical confusion'. 3
Being, like their colleagues, concerned with conceptual
matters, their protests against the misuse of English are not
primarily motivated by a concern for correct English as
opposed to faultless Eskimo. But the matter should not be
allowed to rest there. The existence of other natural lan-
guages whose structure, idiom, and vocabulary are not
completely congruent with those of our own is philosophically
relevant in at least three ways.

(i) They provide concepts not available in the stock of
our language group. Notoriously there are in all languages
words untranslatable into English : no English words, that
is, have precisely the same use. And many of the concepts

1 Cf. L. J. Cohen, * Are Philosophical Theses relative to Language ? ' in
Analysis, 1949.

z G. Ryle, 'Ordinary Language', Philosophical Review, 1953.
3 PQ t 1952, p. 2 (top),



PHILOSOPHY AND LANGUAGE 5

concerned are of philosophic interest ; eitfaeT directly in
themselves ; or indirectly because it is necessary to master
them in order to understand some philosopher who used or
discussed the concept in question. Perhaps the best examples
are ethical, such as vflpis, dpcTij, or tabu.

(ii) Different languages offer different temptations. J. S.
Mill must have been beguiled into his disastrous argument
from what is in fact desired to what is in morals desirable
by the 'grammatical' analogy between English words like
'audible 1 and c visible ' and the English word 'desirable'. 1
(There might be a language in which there was no such
morphological analogy between a class of words meaning
' able as a matter of fact to be somethinged ' and one meaning
'ought as a matter of value to be somethinged'.) The mis-
construction of 'infinity' as being the word for a gigantic
number is made attractive by the morphological analogy
between the expression 'to infinity' and such as 'to one
hundred'. If we always said 'for ever' or 'indefinitely'
instead of 'to infinity', and if 'aleph-nought' did not happen
to sound like the word for a colossal number, then this
temptation would disappear. 2 It has been said that it is
hard to make Hegel's dialectic plausible or even intelligible
in English for the lack of any word with ambiguities parallel
to those of the German aufheben* Kant, in a significantly
phrased passage, noted:

The German language has the good fortune to possess expres-
sions which do not allow this difference [between the opposites of
das Ubel and das Bose A. F.] to be overlooked. It possesses two
very distinct concepts, and especially different expressions, for that
which the Latins express by a single word bonum.*

While the Greek way of forming abstract noun substitutes

1 Utilitarianism (Everyman), p. 32 (bottom) : Mill argues from this morpho-
logical analogy, explicitly. Though even here it is doubtful if this was more
than the occasion for a mistake, the true cause of which was the quest for a
'scientific ethics'.

2 See P.A.S., Supp. Vol. XXVII, pp. 42-3 and 47-8, for a recent example
of this howler and its criticism.

3 T. D. Weld on, The Vocabulary of Politics (Pelican, 1953), p. 107.
* Critique of Practical Reason, trans. T. K, Abbott, p. 150.



6 ESSAYS IN CONCEPTUAL ANALYSIS

from the neuter of the definite article and the adjective does
something, though not of course very much, to explain the
attractions for Plato of the Theory of Forms. 1

(iii) The existence of natural languages with radically
different logical characteristics gives the opportunity for
logical explorations of ways of thinking far more diverse
than those embraced in most of these singly : for, as it were,
logico-linguistic travel, which can broaden the mind and
stimulate the imagination and so provide benefits of the sort
which alert people are able to get from physical travel.

Consider, for example, the analogy between the recog-
nition of the legitimate existence of non-Euclidean geometries
which helps to undermine rationalist hopes of a quasi-
geometrical deductive system of knowledge about the world
based on self-evident necessary premises; and the realiza-
tion that there actually are natural languages to which the
subject-predicate distinction can scarcely be applied, which
are not saturated with the concept of cause, and which pro-
vide words to pick out different differences and likenesses
from those which English, and indeed most European lan-
guages, are equipped to mark. To realize this is to discredit
ideas that the subject-predicate distinction must be inextri-
cably rooted in the non-linguistic world, 2 that the notion of
cause is an indispensable category of thought, 3 and that
language must reflect the ultimate nature of reality. 4 Of
course, it is theoretically possible to imagine other conceptual
systems and categories of concept. 5 But this is excessively

1 For some of the many more worthy attractions see D. F. Pears, 'Uni-
versals', in LL, II. For Aristotle's battle against the temptations of this idiom,
in which he had to express his definition of goodness, see the early chapters of
Me. Eth., Bk. I. r

2 This point was originally made by Sayce; and reiterated by Russell,
Analysis of Mind, p. 212.

3 See articles by D. D. Lee mentioned below, though her interpretation of
Trobriand thought is disputed.

* See Republic, 596 A 6-8 for a suggestive admission : 'We have been in the
habit, if you remember, of positing a Form, wherever we use the same name in
many instances, one Form for each many'.

5 Cf. the 'language games' of Wittgenstein, imaginary truncated languages
used as diagrams in Philosophical Investigations : and the Newspeak of George
Orwell's 1984, Appendix.



PHILOSOPHY AND LANGUAGE y

difficult, as witness the calibre of some of the philosophers
who have assumed or even asserted contingent, though per-
haps admirable, characteristics of their particular languages
to be necessities of thought. In any case there is actual
material waiting to be studied, 1 and there is much to be
said for the use of real, as opposed to imaginary, examples
in philosophy. It can add vitality to discussion and help
to break down the idea that philosophical training and
philosophical enquiry can have no relevance or value in the
world outside our cloistered classrooms.

(C) The use of a word is not the same as, though it is
subtly connected with, the usage of that word. The former
(see above) is language-neutral : if we enquire about the
use of 'table' then we are simultaneously and equally con-
cerned with the use of 'tavola' and other equivalents in
other languages ; with, if you like, the concept of table.
The latter is language-specific : if we enquire about the
usage of 'table' then we are concerned with how that par-
ticular English word is (or ought to be) employed by those
who employ that word, and not 'tavola*.

But the two are crucially related. No word could be said
to have a use except in so far as some language group or
sub-group gives it a use and recognizes as correct the usage
appropriate to that use : for the sounds we use as words are
all, intrinsically and prior to the emergence of any linguistic
conventions about them, almost equally suitable to do any
linguistic job whatever. Whereas a knife, say, could not be
used, or even misused, as a tent, ' glory' might have been
given the use we have in fact given to 'a nice knock-down
argument ' .

The uses of words depend subtly on the correct usages
of words. Humpty Dumpty can only be accused of misusing
' glory 1 because the accepted, standard, correct usage of
Lewis Carroll's language group was radically different from

1 See LL, II, p. 3 : to the references given there in the second note can be
added D. D. Lee in Psychosomatic Medicine, Vol. XIII, 1950, and in The Journal
of Philosophy y 1949.



8 ESSAYS IN CONCEPTUAL ANALYSIS

his private usage. It was perverse, ill-mannered, misleading,
and endangered the possibility of linguistic communication,
thus wantonly and without explanation to flout the linguistic
conventions. (No doubt, like contemporary 'prophets of a
new linguistic dispensation', 1 he regarded such linguistic
conventions as 'preposterous restrictions upon free speech*. 2 )
Furthermore, as academic philologists 3 and people concerned
with maintaining and increasing the efficiency of the Eng-
lish language 4 (and others) have often urged, what is cor-
rect usage of any language group depends ultimately upon
actual usage. It is because use depends on correct usage
while this in turn depends ultimately upon actual usage
that changes in actual usage can enrich or impoverish the
conceptual equipment provided by a language. If a new
usage Is established by which a new use is given to a word,
a use not previously provided for, then to that extent the
language concerned is enriched. 5 Whereas if an old usage
whereby two words had two different uses is replaced by a
new one in which one of them loses its job to become a mere
synonym of the other, then similarly there is a proportionate
impoverishment. Since the actual usage of any language
group or sub-group is never in fact completely static, both
processes are usually going on, and together constitute a
considerable part of the history of any language. ('The
history of language ... is little other than the history of
corruptions': Lounsbury was writing as a grammarian,
but the same is true from a logical point of view ; though
'corruption' must be taken as value-neutral here.)

To come at the matter from a new angle : consider how
the historical theologian studies the concept of nephesh in
Israel. He has and can have no other method but the

1 PQ, 1952, p. 12. 2 PQ f 1952, p. 2.

3 To the point here would be references given by P. L. Heath, PQ, 1952,
p. 2, n.

See Sir Alan Herbert's What a Word! Sir Ernest Gowers' Plain Words
and ABC of Plain Words, etc.

5 This point is developed by F. Waismann in his 'Analytic-Synthetic', and
stressed to a point at which some might complain that it encouraged anarchic
Humpty-Dumptyism (Analysis, 1950, Vol. X ff.).



PHILOSOPHY AND LANGUAGE 9

examination of the occurrences of the word ' nephesh ' In his
texts : the attempt to discover from a survey of usage what
was its use, what job this word did in the vocabulary of the
people who employed it. Or, again, consider how Professor
H. J. Paton objects decisively to the translation oiabgeleitet as
' deduced' because 'an examination of Kant's usage will show
that it seldom or never means this' (The Categorical Impera-
tive, p. 134, .). Or consider how the cryptographer tries to
discover the meaning of an unknown element in a code. He
has and can have no other method but a similar examination
of its occurrences, hoping by a study of usage to hit upon
its use, its meaning. Appeals to use and usage in creative
philosophy can be regarded as a belatedly explicit application
of the tried and necessary methods of the historians of ideas.
Before passing to section (D), various minor points:
First, 'linguistic conventions' here means those by which
we use ' pod * rather than ' pid ' or ' nup * to mean pod ; and
so forth. Second, 'language group or sub-group' is not
here a precise expression. It is intended to cover the users
of recognized languages, of their dialects, of jargons and
private languages of all kinds, down to and including in-
dividuals who develop terminologies private to themselves
and their readers and interpreters, if any. Our point is one
about the presuppositions of linguistic communication.
Third, not all features of the usage of a word will be relevant
to questions about its use : that the personal pronouns ' I ',
'he', and 'she* are subject to radical morphological trans-
formation in other cases is of concern to Fowler, but not to
the philosopher ; for their use would be unaffected if usage
were to send these transformations the way of other unneces-
sary case-indications. But this is a matter for caution, for
it is hard to be sure without examination what will turn out
to be relevant : Fowler would be concerned with the spread of
the usage which makes * contact ' a transitive verb ; but per-
haps this change also subtly affects the notion of contact. 1

1 On the analogous difficulty of knowing in advance the 'logical breaking
strains of concepts ' see Kyle's Inaugural, Philosophical Arguments (Oxford, 1945).



io ESSAYS IN CONCEPTUAL ANALYSIS

Fourth, it is possible for people to communicate, in a way
which depends partly on words (or other conventional signs),
in spite of misusing many of the words (or other conventional
signs) they employ : for the intelligent appreciation of con-
text (in the widest sense) can do much to compensate for
such deficiencies. But to the precise extent to which it needs
to, communication is thereby not depending upon words (or
other conventional signs). Fifth, this stress on use derives
mainly from Wittgenstein : the idea is present unexploited
in the Tractatus Logico-Philosophicus : 'In philosophy the
question " For what purpose do we really use that word . . . ? "
constantly leads to valuable results' (6.211, cf. also 3.328,
3.326, and 5.47321); and it became the slogan ' Don't ask for
the meaning^ ask for the use' in the early thirties after his
return to Cambridge. 1 The explicit concern with correct usage
as the determinant of use seems to derive mainly from J. L.
Austin. 2

(D) Notoriously there is often a gap between actual and
correct usage. It is possible for some usage which is (even
much) more honoured in the breach than the observance to
be one which defaulters are prepared to acknowledge as
correct, mainly because certain people and reference books
are recognized as generally authoritative : there is still, in
Britain at any rate, no question as to what is the correct
usage of such non-technical Jogical terms as * refute', 'imply',
and 'infer', but it seems most unlikely that the actual usage
of the majority even of first-year university students con-
forms with it. This gap is of the greatest importance to
anyone who wishes to understand 'what is at the bottom of
all this terminological hyperaesthesia, and all the whistle-
blowing and knuckle-rapping and scolding that goes along
with it '.3

(i) It enables a piece of 'logical geographizing', telling

1 See the Philosophical Investigations > especially ad init, for his own account
of the reasons for this maxim.

2 See M. Weitz, 'Oxford Philosophy', Philosophical Review, 1953: and
P.A.S., Supp. Vol. XVIII, for Austin's first characteristic publication.

3 PQ t 1952, p. 5 (top).



PHILOSOPHY AND LANGUAGE 11

us only what most of us in a way know, making no distinction
not already provided for in familiar words, to be an exercise
in precisification of thought and in improvement of usage
for all those who work through it ; and not merely for those,
like the students mentioned above, whose word training has
been conspicuously deficient. Consider the effects of de-
scribing the differences and analogies between threats,
promises^ and predictions ; to draw example from a recent
Oxford examination paper. Though often such examina-
tions of present correct usage will show that we need not
only to bring our actual usage more into line with correct
usage, but also to go further by suggesting improvements.
* Essential though it is as a preliminary to track down the
detail of our ordinary uses of words, it seems that we shall
in the end always "be compelled to straighten them out to
some extent' (Austin). 1

(ii) It gives ground for hope that "philosophers, including
always and especially ourselves, who misuse or tolerate the
misuse of certain words and expressions, 2 or who give or
accept incorrect accounts of their rationes applicandi> may
be led by suitable attention to their correct usage and actual
use to realize and remedy their mistakes. This phrase
ratio applicandi is modelled deliberately upon the ratio
decidendi of the lawyers : the principle under which all
previous decisions can be subsumed and upon which, as the
fiction has it, they were in fact made. For just as it is per-
fectly possible to make decisions consistent with such a
principle without having actually formulated it : so it is
possible, and even usual, to be able to apply a word cor-
rectly in unselfconscious moments, without being able to
discern its ratio applicandi^ or even to do so when positively
in error about it ; though of course anyone making such a
mistake will have some inclination to misuse the word.

(iii) But it also makes it possible to misrepresent present

* 'How to Talk' in P,A.S., 1952-3, p. 227.

2 For some subtle and very important examples of a sensitivity to ordinary
correct usage see G. J. Warnock's Berkeley ', especially in chaps. 7-10.



12 ESSAYS IN CONCEPTUAL ANALYSIS

correct usage as nicer, more uniform, and more stable than
it in fact is : ' the assumption being that the necessary
rules and regulations are already embodied in ordinary par-
lance, requiring only inspection, or the production of a few
trivial examples, to make clear what is allowable and what
is not 3 . 1 To do this is especially tempting perhaps for
philosophers in strong reaction against the contempt shown
by many of their mathematically minded colleagues for the
rich and subtle instruments provided by all but the most
beggarly of the natural languages to those willing and able
to use them with care and skill (see (v) below). The extent
to which the 'logical geographers' have in fact succumbed
has perhaps been exaggerated ; but it is well to be on
guard.

(iv) It is this alone which makes it possible to speak at
all of misuses When philosophers are attacked for misusing
an ordinary, or even an extraordinary, word this is rarely
an 'attempt to convict perfectly respectable philosophers of
illiteracy, or of the perpetration of ungrammatical gibber-
ish*, but rather 'what is complained of is not lack of
grammar , even \$ic\ in the text-book sense, but incoherence
or absence of meaning* (my italics); even though some
(like Wittgenstein who perhaps discovered it) given to
pressing 'the familiar and overworked analogy between
logical and grammatical rules' 2 occasionally omit the pre-
fix 'logical' where the context makes clear that it is logical
grammar that is at issue. The point is, usually, that the
philosopher under attack has somehow been misled into
misusing a word in a way which generates paradox, con-
fusion, and perplexity. Hume was scandalized that a con-
troversy 'canvassed and disputed with great eagerness since
the first origin of science and philosophy' had 'turned
merely upon words'. 3 But the skeleton solution he sug-
gested depended, fairly explicitly, upon recalling to mind
with the help of simple concrete examples, just what the

1 PQ, 1952, p. 6. * PQ, ibid,

3 Enquiry concerning Human Understanding , VIII, pt. i.



PHILOSOPHY AND LANGUAGE 13

ordinary use of the word 'free' actually is: and that it is
not its ordinary job (not what it ordinarily means), nor yet
any part of its ordinary job (nor yet part of what it implies)*
to attribute to actions unpredictability in principle, 1 If this
is so then it is not contradictory to say that some action was
both predictable and performed of the agent's own free will :
always assuming, of course, that the key words are being
used in their ordinary senses. And in any case complaints
about 'pseudo-problems', 2 'a petty word-jugglery J , 3 or the
tendency 'for philosophers to encroach upon the province
of grammarians, and to engage in disputes of words, while
they imagine they are handling controversies of the deepest
importance and concern' 4 all miss the point. For Hume is
broaching a conceptual solution to a philosophical problem,
which cannot thereby lose whatever importance it may have
had before.

Such a brief outline example may suggest facile crudity :
'a very simple way of disposing of immense quantities of
metaphysical and other argument, without the smallest
trouble or exertion'.' 5 This is inevitable perhaps in the terse
cartoon simplicity needed in incidental illustrations but the
suggestion could scarcely survive : either an examination of
such contributions to the free-will problems as R. M. Hare's
'The Freedom of the Will 7 , 6 W. D. Falk's 'Goading and
GuildingV and H. L. A. Hart's ' Ascription of Responsibility
and Rights ' ; 8 or an awakening to the fact that no one has
asked to be excused from dealing with whatever arguments
may be deployed in support of such philosophers' misuses.
(We give these examples because they are concerned with
the problem we have chosen as an illustration : but of course

1 His views had been substantially anticipated : by Hobbes, Leviathan,
chap, xxi, and in his *Of Liberty and Necessity' ; and no doubt by many others
long before.

. 2 See JLL, II, pp. 5-6 for an attack on this once popular misdescription of
philosophical problems; Ryle, Concept of Mind, p. 71 (top), lapses by writing
'largely spurious problems'.

3 Kant, Critique of Practical Reason, trans. T. K. Abbott, p. 188.

4 Hume, E.P.M., App. iv. s pQ } 1952, p. i

6 P.A.S., Supp. Vol. XXV. 7 Mind, 1953. * LL, I.



14 ESSAYS IN CONCEPTUAL ANALYSIS

the rest of this book provides many others.) Perhaps Kant
was discouraged from recognizing the merit in Hume here
by Hume's own misleading talk about mere words as well
as by the aggressive way in which he misrepresented a good
start as the end of the affair. Certainly we find him two
pages later very grudgingly conceding part of Hume's point,
but insisting that at any rate transcendental freedom cannot
thus be reconciled with scientific determinism. 1

(v) After so much has been said about misuses and mis-
constructions, it must be mentioned that interest originally
directed at the uses of words only inasmuch as this brought
out what were misuses and misconstructions, is sometimes,
by a familiar psychological process, partly diverted to the
study of use for its own sake. Before suggesting that, how-
ever psychologically understandable, such interests do not
become a philosopher in his working hours we should cast
our minds back to Aristotle and reflect whether all his
studies of the concepts of moral psychology were in fact
wholly directed to some ulterior end even within philosophy ;
or, more generally, ask ourselves whether an interest in
concepts is not one of the things which makes a philo-
sopher.

But whatever are the rights and wrongs about ulterior
and ultimate ends and whatever the jurisdictional pro-
prieties, disputes about these here turn out to be largely
unnecessary. For in elucidating the ordinary uses (as
opposed to philosophers' suspected misuses) of some of the
rather limited range of words around which our contro-
versies tend to cluster, 2 it has been noticed that the concep-
tual equipment provided by ordinary (here opposed par-
ticularly to technical) language is amazingly rich and subtle ;
and that even the classical puzzles cannot be fully resolved
without elucidating not merely the formerly fashionable elite
of notions but also all their neglected logical hangers-on.

1 Loc, cit. p. 190.

2 See Waismann, 'Language Strata' in LL, II, and Ryle, 'Ordinary Lan-
guage ', for suggestions about this clustering.



PHILOSOPHY AND LANGUAGE 15

In formulating and attacking free will puzzles, philosopher
with the outstanding exception of Aristotle, have been
inclined to concentrate on a few ideas : free will, compulsion,
choice, necessity, responsible, and one or two others. Whereas
we have available in our ordinary vocabulary of extenuation
and responsibility a great range of notions, which it would
be wise to master and exhaust before thinking of adaptation
or invention : * automatically, by mistake, unintentionally,
by force of habit, involuntarily, unwillingly, on principle,
under provocation, to mention a few. Philosophers have
tended to ignore all this richness and variety, assuming that
it could all be satisfactorily assimilated to a few most favoured
notions. But to do this is clumsy and slovenly. While pro-
posals to jettison ordinary language in favour of new-minted
terms overlook the crucial primacy of the vernacular :
ordinary, as opposed to technical, language is fundamental
in the sense that the meaning of terms of art can only be
explained with its aid ; and it is a perennial complaint
against such lovers of jargon as Kant and the Scholastics
that this essential work is so often botched, skimped, or
altogether neglected. The upshot of all this is that it is im-
probable that the elucidation of the logic of any term at all
likely to engage any philosopher's attention will fail some
day to find application to some generally recognized philo-
sophical problem, however 'pure' his own interests may have
been : the implied comparison with the pure scientific re-
search which so frequently finds unexpected and unintended
application is suggestive and, up to a point, apposite. It is
to such often seemingly indiscriminate interest in the uses
of words that we owe such fruitful logical explorations of
neglected territory as R. M. Hare's 'Imperative Sentences', 2
and J. L. Austin on performatory language in ' Other Minds '.
Contrast the old 'fetich of the indicative sentence' (Ryle)

1 Not that such adaptation or invention, may not be called for : see P. D.
Nowell-Smith in The Rationalist Annual, 1954, for suggestions designed to
accommodate the discoveries of psychoanalysis.

2 Mind, 1949: and incorporated with additions and improvements into
his The Language of Morals.



16 ESSAYS IN CONCEPTUAL ANALYSIS

formulated by Hobbes : ' In philosophy there is but one kind
of speech useful . . . most men call it proposition, and [it]
is the speech of those that affirm or deny, and expresseth
truth and falsity* (Works, Vol. I, p. 30).
If one quoted

Others apart sat on a hill retired
In thoughts more elevate, and reasoned high
Of providence, foreknowledge, will and fate ;
Fixt fate, free will, foreknowledge absolute
And found no end, in wandering mazes lost

(Milton, Paradise Lost, Bk. II)

to those who have learnt most from Austin, the reply would
be that the Devils in Pandemonium found no end precisely
because they insisted on ' reasoning high'; that they should
have begun with a meticulous and laborious study of the
use of 'free will* and all the terms with which it is logically
associated. Such an examination, which is certainly no
quick and easy matter, is, as Austin has said, if not the be
all and end all, at least the begin all of philosophy.

(E) A derisive brouhaha has been raised about the
notion of 'Standard English 5 'Why it should have been
thought to deserve consideration as a philosophical prin-
ciple it is by no means easy to imagine 7 . 1 Those who have
emphasized the frequent philosophical importance of ' unex-
plained and unnoticed distortions of standard English' and
'deviations from standard English to which no sense has
been attached ' 2 have not, of course, been claiming that
there is or ought to be an absolute, unchanging, universal,
inflexible standard of correctness applicable to all users of
the English language, past, present, and to come. The
strange idea that they have seems to derive: partly from
failing to appreciate the force of the emphasis on uses, etc.
(see (A) and (C) above) ; partly from a significant though
perhaps seemingly trivial misrepresentation, whereby a con-
cern for 'Standard English' is attributed to those who have

* PQ, 1952, pp. 2 and 3. * LL, I, p. 9.



PHILOSOPHY AND LANGUAGE 17

in fact written of ' standard English ' ; l and partly from, the
sheer errors that standards must necessarily be universal,
inflexible, unchanging, and absolute. These last may be
dispelled by the reflection that makers of cars may offer
fresh standard models yearly, different ones for different
markets, and with a standard choice of fittings and colours
for each. About standards such as these there is presum-
ably nothing normative, whereas with standard linguistic
usage there certainly is. For, for the reasons already given
(see (C) ad init.\ everyone ought in general to conform with
the usage accepted as correct by the language group or
sub-group of which his linguistic and non-linguistic be-
haviour makes it reasonable to presume he is, tacitly or
explicitly, claiming membership. This is not to say that
usage ought to be absolutely rigid, uniform, and static
among all users of any language ; this would be to impose
an embargo on improvisation and innovation, growth and
decay.

It is common enormously to exaggerate the amount of
variation in usage which there in fact is. People often write
as if usage were so fluid, irregular, and varied that it must
be impossible to say anything about the meaning of any
word, except perhaps as employed by one particular person
on one particular occasion. As we have argued above (in
(C)) and elsewhere, 2 if this were in fact the case, verbal com-
munication would be impossible. These exaggerations, like
the linguists' analogue that different languages are all so
very different that there is no equivalent of any word at all
in any other language, arise from the understandable and
inevitable preoccupation of philologists with differences and
changes, and of translators with their more intractable diffi-
culties. They are perhaps encouraged by vested interests, in
obfuscation generally, and in the pretence that knowledge of
foreign languages is even more important than it actually is.

1 Compare JPQ, 1952, pp. 2 and 3 with LL } I, p. 9, of which it purports to
be a criticism.

2 LL> II, pp. 8-9 : on ( the unwitting allies of a revolution of destruction 1 .



18 ESSAYS IN CONCEPTUAL ANALYSIS

(F) Ryle has so recently distinguished again between,
and redeployed some of the arguments in favour of, the
various policies which have sometimes been confused to-
gether as 'the appeal to ordinary language' that there is
no need here to develop at length the relations and differ-
ences between these. 1 First, appealing to the ordinary use(s)
of terms to elucidate philosopher's possible misuses. Second,
appealing for plain English as opposed to jargon and high
abstraction in philosophical prose. Not that anyone is sug-
gesting an embargo on technical terms and abstraction :
only a bias against, except where they prove essential. Third,
concentrating upon everyday as opposed to technical con-
cepts and their problems. Yet the fact that a large proportion
of classical problems centre round such notions as cause,
mistake, evidence, knowledge, ought, can, and imagine is
no reason at all for neglecting those which arise from
p si-phenomena, collective unconscious, transubstantiation,
economic welfare, and infinitesimal. But whereas (most of)
the former group can be tackled with no knowledge other
than that minimum common to all educated men, even to
understand the latter one must acquire some smattering of
the disciplines to which the notions belong. Fourth, a protest
'that the logic of everyday statements and even ... of
scientists, lawyers, historians, and bridge-players cannot in
principle be adequately represented by the formulae of formal
logic'. 2 Oxford philosophers who incline to all four policies
together may be thought of as trying to preserve a balance :
between this 'formalizer's dream* that non-formalized lan-
guage really is, or ought to be replaced by, a calculus; 3
and the Humpty Durnpty nightmare that there is, at least
in those parts of it which most concern philosophers, no logic
or order at all.

1 'Ordinary Language', loc. cit. 2 Ryle, loc. cit,, p. 184.

3 Though there has been a long tradition of this sort of thing, e.g. Leibniz's
characteristic a universalis and similar ideas in his century. Nowadays it is
usually a matter of an over-estimation of the philosophic value of the techniques
of symbolic logic ; found in the Vienna Circle, in Russell, and to-day particularly
prevalent in the U.S.A.



PHILOSOPHY AND LANGUAGE ig

One pattern of argument, a particular application of the
first policy, demands special attention. Talk, mainly de-
riving from Moore, of the Plain Man and his Common
Sense, has now been largely replaced by emphasis upon the
ordinary uses of words. But many philosophers have been
as reluctant to abandon their reasoned paradoxes because
they offend the Plain Man in his capacity as arbiter of
ordinary language as they were to abdicate in face of Moore's
protests on behalf of his common sense. And not without
reason. The clue to the whole business now seems to lie In
mastering what has recently been usefully named, The Argu-
ment of the Paradigm Case. 1 Crudely : if there is any
word the meaning of which can be taught by reference to
paradigm cases, then no argument whatever could ever prove
that there are no cases whatever of whatever it is. Thus,
since the meaning of 'of his own freewill' can be taught by
reference to such paradigm cases as that in which a man,
under no social pressure, marries the girl he wants to marry
(how else could it be taught ?) : it cannot be right, on any
grounds whatsoever to say that no one ever acts of his own
freewill. For cases such as the paradigm, which must occur
if the word is ever to be thus explained (and which certainly
do in fact occur), are not in that case specimens which might
have been wrongly identified : to the extent that the mean-
ing of the expression is given in terms of them they are, by
definition, what ' acting of one's own freewill ' is. As Runyon
might have said : If these are not free actions they will at
least do till some come along. A moment's reflexion will
show that analogous arguments can be deployed against
many philosophical paradoxes.

What such arguments by themselves will certainly not
do is to establish any matter of value, moral or otherwise :
and almost everyone who has used them, certainly the pres-
ent writer, must plead guilty to having from time to time
failed to see this. For one cannot derive any sort of value
proposition : from either a factual proposition about what
1 By J.O, Urmson in * Some Questions concerning Validity*, Chapter VI, below,



20 ESSAYS IN CONCEPTUAL ANALYSIS

people value; or from definitions, however disguised, of
value terms. This applies to any sort of value : indeed we
might distinguish a Special (in ethics) from a General (any-
where) Naturalistic Fallacy. 1 There is a world of difference
between saying that it is reasonable in certain circumstances
to act inductively (which is a value matter, one of commend-
ing a certain sort of behaviour) ; and saying that most people
regard it as reasonable so to act (which is a factual matter,
one of neutrally giving information about the prevalence of
that kind of ideal). Thus that too short way with the
problem of induction, which tries to deduce that induction
is reasonable from the premise that people regard it as so
or even that they make inductive behaviour part of their
paradigm of reasonableness, will not do. It is necessary for
each of us tacitly or explicitly actually to make our personal
value commitments here. Most of us are in fact willing
to make that one which is involved in making inductive
behaviour part of our paradigm of reasonableness. But
as philosophers we must insist on making it explicitly and
after examining the issues. Mutatis mutandis the same
applies to attempts to derive ethical conclusions simply from
what we (as a matter of fact) call reasonable behaviour or
good reasons to act ; without the introduction of an explicit
commitment to accepted moral standards. These must
involve versions of the (Special) Naturalistic Fallacy. 2

To see the power, and the limitations, of the Argument
of the Paradigm Case is to realize how much of common
sense can, and how much cannot, be defended against
philosophical paradoxes by simple appeal to the ordinary
use of words ; and why.
UNIVERSITY COLLEGE OF NORTH STAFFORDSHIRE

1 This is the fallacy of trying to deduce valuational conclusions from purely
factual premises which in its Special form was first named by G. E. Moore.

* Compare R. M. Hare, The Language of Morals, with S. E. Toulrnin, The
Place of Reason in Ethics, and see Hare and J. Mackie on the latter in PQ 1951
and AJP 1951, respectively. T. D. Weldon in The Vocabulary of Politics gives a
crisp example of the false move involved, pp. 42-3. I have tried to say something
myself about it in 'The Justification of Punishment', Philosophy, 1954,