This article was downloaded by: [Ohio State University Libraries]

On: 03 June 2012, At: 15:19

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T
3JH, UK

Inquiry: An Interdisciplinary
Journal of Philosophy
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/sinq20

Perfect Pitch and Austinian

Examples: Cavell, McDowell,
Wittgenstein, and the
Philosophical Significance of
Ordinary Language
a
Martin Gustafsson
a
Uppsala University, Sweden

Available online: 19 Aug 2006

To cite this article: Martin Gustafsson (2005): Perfect Pitch and Austinian Examples:
Cavell, McDowell, Wittgenstein, and the Philosophical Significance of Ordinary
Language, Inquiry: An Interdisciplinary Journal of Philosophy, 48:4, 356-389

To link to this article: http://dx.doi.org/10.1080/00201750510022853

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-

and-conditions
This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,
reselling, loan, sub-licensing, systematic supply, or distribution in any form
to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any
representation that the contents will be complete or accurate or up to
date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not
 be liable for any loss, actions, claims, proceedings, demand, or costs or
damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012
 Inquiry,
Vol. 48, No. 4, 356–389, August 2005

Perfect Pitch and Austinian

Examples: Cavell, McDowell,
Wittgenstein, and the Philosophical
Significance of Ordinary Language
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

MARTIN GUSTAFSSON
Uppsala University, Sweden

(Received 20 December 2004)

ABSTRACT In Cavell (1994), the ability to follow and produce Austinian examples of
ordinary language use is compared with the faculty of perfect pitch. Exploring this
comparison, I clarify a number of central and interrelated aspects of Cavell’s
philosophy: (1) his way of understanding Wittgenstein’s vision of language, and in
particular his claim that this vision is ‘‘terrifying,’’ (2) the import of Wittgenstein’s
vision for Cavell’s conception of the method of ordinary language philosophy, (3)
Cavell’s dissatisfaction with Austin, and in particular his claim that Austin is not clear
about the nature and possible achievements of his own philosophical procedures, and
(4) Cavell’s notion that the temptation of skepticism is perennial and incurable.
Cavell’s reading of Wittgenstein is related to that of John McDowell. Like McDowell,
Cavell takes Wittgenstein to be saying that the traditional attempt to justify our
practices from an external standpoint is misguided, since such detachment involves
losing sight of those conceptual and perceptual capacities in terms of which a practice is
understood by its engaged participants. Unlike McDowell, however, Cavell consistently
rejects the idea that philosophical clearsightedness can or should free us from that fear
of groundlessness which motivates the traditional search for external justification.

It was familiarly said that the point of Austin’s stories, those examples
apart from which ordinary language philosophy has no method, required
what you might call ‘‘ear’’ to comprehend (as in, more or less at random,
setting out the difference between doing something by mistake or by
accident, or between doing something willingly or voluntarily, carelessly

Correspondence Address: Martin Gustafsson, Department of Philosophy, Uppsala University,

Box 627, SE-751 26 Uppsala, Sweden. Email: [email protected]
0020-174X Print/1502-3923 Online/05/040356–34 # 2005 Taylor & Francis Group Ltd
DOI: 10.1080/00201750510022853
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 357

or heedlessly, or between doing something in saying something or by

saying something, or between telling a bird by its call or from its call).
My mother had something called perfect pitch, as did one of her
brothers. That I did not was a source of anguished perplexity to me, one
of the reasons I would eventually give myself for withdrawing from
music, particularly after I found that the only role I conceived for myself
in music was as a composer. Yet I felt there must be something I was
meant to do that required an equivalent of the enigmatic faculty of
perfect pitch. Being good at following and producing Austinian examples
will strike me as some attestation of this prophecy.1
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

I.
When I was fourteen, I went to a class in elementary music theory where one
of the students had perfect pitch. The teacher would often begin his lessons
by playing a tune on the piano, a tune which we were supposed to write
down in musical notation. On one occasion, someone asked what one
should do if one heard a tune and wanted to write it down, but did not have
access to a tuning-fork or something else by means of which one could
decide the key. The teacher suggested using the telephone: at least in
Sweden, the tone one hears before one starts dialing is a 440 Hz A.
However, after the class, the student who had perfect pitch told me, in all
seriousness, that the telephone should not be trusted: ‘‘Often the tone is not
an A, but almost an A flat. That’s really annoying.’’
I remember finding this remark highly intriguing, but for reasons I could
not make entirely clear to myself at the time. Now, I think I can say that at
least part of my bewilderment had to do with the following insight: having
perfect (or absolute) pitch involves not just the ability to identify the pitch of
a given tone without making use of a tuning-fork or some similar device.
Moreover, if he who has perfect pitch is confident enough in his talent, he
might even pass judgment on particular tuning-forks. After having listened
to the tone it produces, he might claim that a tuning-fork is misconstructed
since the A which it is said to generate is, in fact, too low or too high. For
me, who does not have perfect pitch, delivering this sort of criticism is
impossible. My way of arguing that a particular tuning-fork is wrongly
constructed is, rather, to compare its alleged A to the A produced by other
tuning-forks (and similar such devices). Only if, on such comparison, I
found that the tone generated by the tuning-fork deviated from the
other tones, would I dare to claim that the tuning-fork does not produce a
genuine A.
The nature of my bewilderment over the student’s remark can be further
explored by means of the following thought-experiment. Imagine a world in
which everyone has perfect pitch. In this world, tuning-forks would not have
 358 M. Gustafsson

the function they have among us, since people would decide what pitches
tones have simply by listening to them. (That is to say: in this world, there
would be no tuning-forks.) These people always, or very nearly always,
agree in their judgments. When they teach their children to identify pitches,
one thing they do is to produce sample tones, by singing or whistling, and
say things like ‘‘This is a D flat’’, ‘‘This is an F’’, and so on. This sort of
training does, in fact, work; after a while, the children are able to identify
pitches as directly and as competently as their parents. (This is not to say
that such ostensive exercises will be the only training required for the child’s
learning to use names of pitches.)
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

Now there is a temptation, at least for someone who lacks perfect pitch,
to argue that such a community’s way of identifying pitches is, as it were,
strangely free-floating. One might be tempted to say that there seems to exist
no solid foundation on which this practice rests. On hearing a tone, these
people spontaneously agree that it is, say, a B. But if we asked them what
reason they have for thinking so, it might happen that all they would say
would be something like: ‘‘We simply hear, directly, that the tone is a B. We
don’t need any further justification. Indeed, any attempt at such further
justification would be superfluous; nothing could lend additional support to
our identification of the pitch. To us, that the tone is a B is already fully
transparent.’’
This sort of answer may appear unsatisfactory. Borrowing a phrase from
John McDowell, one might want to say that all that seems to be going on
here is ‘‘a congruence of subjectivities.’’2 Hearing tones causes beliefs in
these people, beliefs that, luckily enough, concur intersubjectively. To be
sure (one might want to continue), this spontaneous consensus must have a
natural explanation: nature has endowed these people with similar
perceptual apparatuses. But nonetheless, the consensus seems in a certain
sense like mere luck, for what they agree on is not supported by any rational
procedure of justification.
Consider what would happen if these people suddenly started disagreeing
with one another—if the peaceful consensus which has so far characterized
their practice were replaced by widespread discord. Someone says a given
tone is a B; someone else calls it a C; a third listener identifies it as a C sharp;
a fourth listener thinks it is a D; and so on. In this case, there might be
nothing available which could restore the consensus, no agreed-upon
procedures or criteria by means of which the true answer could be identified.
If such disagreement became the rule, the practice would break down.
Indeed, it would then perhaps be too generous to characterize the prevailing
chaos in terms of disagreement, for the breakdown would be so deep-going
that the sort of determinacy of content required by the notion of
disagreement would be threatened. The old words would cease to have a
clear function within that community. It would no longer be clear what it
would mean for these people to ‘‘identify the pitch’’ of a given tone.
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 359

Such considerations might strengthen the feeling that a community where

everyone has perfect pitch would be a community in which the identification
of pitches is a strangely free-floating affair, based on nothing, and
dependent on a consensus which just happens to be there. Indeed, somewhat
ironically, one might be tempted to argue that the practice of our own
community, where most people do not have perfect pitch, is on firmer
footing. For, among us, there are ways of justifying or falsifying the claim
that a certain tone that one is listening to is, say, a B. Suppose you and I
hear a tone, and are asked to guess what pitch it has. Imagine that we come
up with different hypotheses. Perhaps your guess is that the tone is an A,
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

whereas my guess is that it is a B. So, we disagree. Our disagreement,

however, is easily resolved: consensus can be restored, for example, by
comparing the tone with the reference tone generated by a tuning-fork.
Hence, it might appear that our practice has the sort of rational foundation
that the practice of the community imagined in the thought-experiment
seemed to lack.
But this is an illusion. The difference between the two practices is not that
one is irrationally free-floating whereas the other rests securely on rational
procedures of justification. The difference is better characterized by saying
that in these practices, justification comes to an end at different places. For
us, the end-point is, roughly: measuring the tone against the tone produced
by a tuning-fork. Consider the simplest sort of case. You guess that a given
tone is an A, whereas I guess that it is a B. In order to decide who is right, we
use a tuning-fork. When we measure the pitch of our tone against the pitch
of the reference tone (say, a 440 Hz A), we hear that the pitches are the
same. So, I admit you were right: the tone is, indeed, an A. Certainly, this
procedure constitutes a justification of your hypothesis (and a falsification
of mine). And one might call this justification procedure ‘‘rational’’, if one
likes. But still, this procedure presupposes a kind of recognition that does
not rest on any further justification, namely, our hearing that the pitch of
our tone is the same as the pitch of the tone produced by the tuning fork.
This recognition of sameness of pitch is no less ‘‘direct’’ than the
direct recognition, by someone who has perfect pitch, that a certain tone
is an A.
To clarify this point, let us imagine a tone-deaf person—by which I mean:
a person who cannot decide the pitch of a given tone even if he makes use of
all the tuning forks in the world; a person who is unable to perceive the
difference or similarity in pitch between two tones which are played
simultaneously or right after one another. Such a person will be just as
puzzled by our ability to identify the pitch of a given tone by means of a
tuning-fork as we are puzzled by the ability of someone who has perfect
pitch to identify the pitch of a tone without making use of a tuning-fork.
Suppose the tone-deaf person asked us what reason we have for saying that
two tones that are played right after one another have the same pitch. Our
 360 M. Gustafsson

best answer would be something like: ‘‘We simply hear, directly, that the
tones have the same pitch. We don’t need any further justification. Indeed,
any attempt at justification would be superfluous; nothing could lend
additional support to our recognition that the pitch is the same. To us, the
identity of pitches is already fully transparent.’’ At this point, the tone-deaf
person might be tempted to respond that all that is present here is a
congruence of subjectivities: hearing tones causes beliefs in us, beliefs that,
luckily enough, agree intersubjectively. And he might find support for this
analysis in the observation that if we suddenly started disagreeing with one
another about the results we get when we measure pitches by means of
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

tuning-forks, there might be no rational way of restoring the earlier

consensus. Our practice would then collapse.
We were unsatisfied with the way pitches were identified by people in a
community where everyone has perfect pitch. That practice seemed to us
‘‘free-floating’’, whereas we wanted a practice which rested on a solid
foundation. In other words, we wanted a kind of practice where justification
comes to an end, not just at a different place, but, as it were, at an absolutely
stable place. But now, it is becoming increasingly clear that the idea of such
an ‘‘absolutely stable’’ foundation is an illusion. Justification within any
practice has to end at some point, and this point will always be such that
questions for ‘‘further reasons’’ will be met by responses which will seem
completely uninformative to the person who wants additional grounds:
‘‘Don’t you hear—this is an A’’, or ‘‘Don’t you hear—this tone has the same
pitch as that’’ etc. This is true even of a world in which everyone is
completely tone-deaf, but where, for some reason, they still want to talk
about ‘‘pitches’’. Let us assume that for such people, the final arbiter is the
result shown on the displays of portable frequency meters. Roughly: when a
tone is played, they look at the display of such a frequency meter, and if it
says ‘‘A’’ then it has been decided that the tone is an A. This kind of
procedure presupposes the ability to identify letters simply by looking at
them: to distinguish the expression ‘‘A’’ from the expressions ‘‘B’’, ‘‘C’’, and
so on. And this ability is no less ‘‘free-floating’’ than the ability to identify
the pitch of a tone simply by listening to it. If these tone-deaf people
were asked to provide reasons for saying that the letter shown on a certain
display is an ‘‘A’’, their answer would be analogous to the one given by
someone who has perfect pitch in response to the question how he knows
that a certain tone is an A: ‘‘I just see, directly, that the letter is an ‘A’. I
don’t need any further reason for making that identification. Indeed, any
attempt at justification would be superfluous; nothing could lend additional
support to my saying that the letter is an ‘A’. To me, this is already fully
transparent.’’
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 361

II.
I have been discussing what, to my mind, is a rather suggestive instance of a
sort of philosophical puzzlement that plays a central role in the writings of
the later Wittgenstein. I have not been talking about rules, but
Wittgenstein’s so-called ‘‘rule-following considerations’’ are very much
related to the sort of worries I have pondered above. Worries of this sort
also play an important role in Stanley Cavell’s writings, and this is, indeed, a
point at which Cavell is directly influenced by Wittgenstein.
In an often-quoted passage from his early essay, ‘‘The availability of
Wittgenstein’s later philosophy’’, Cavell describes a ‘‘vision’’ he thinks can
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

be found in the Philosophical Investigations:

We learn and teach words in certain contexts, and then we are

expected, and expect others, to be able to project them into further
contexts. Nothing insures that this projection will take place […], just
as nothing insures that we will make, and understand, the same
projections. That on the whole we do is a matter of our sharing routes
of interest and feeling, modes of response, senses of humor and of
significance and of fulfillment, of what is outrageous, of what is similar
to what else, what a rebuke, what forgiveness, of when an utterance is
an assertion, when an appeal, when an explanation—all the whirl of
organism Wittgenstein calls ‘forms of life’. Human speech and activity,
sanity and community, rest upon nothing more, but nothing less, than
this. It is a vision as simple as it is difficult, and as difficult as it is (and
because it is) terrifying.3

One who likes to quote this passage is John McDowell. If I read McDowell
correctly, however, he would want to qualify its final sentence. According to
McDowell, the vision Cavell sketches seems terrifying only in so far as it has
not been properly understood. McDowell says, ‘‘[t]he terror of which Cavell
speaks at the end of this marvellous passage is a sort of vertigo, induced by
the thought that there is nothing but shared forms of life to keep us, as it
were, on the rails. We are inclined to think that is an insufficient foundation
for a conviction that when we, say, extend a number series, we really are, at
each stage, doing the same thing as before’’ (MVR, p. 61). But this vertigo,
he continues, is based on a misunderstanding. If we get clear about the sense
in which we are dependent on our shared ‘‘whirl of organism’’, we will
realize that:

Contemplating the dependence should not induce vertigo at all.

We cannot be whole-heartedly engaged in the relevant parts of
the ‘‘whirl of organism’’, and at the same time achieve the detach-
ment necessary in order to query whether our unreflective view of
 362 M. Gustafsson

what we are doing is illusory. The cure for the vertigo, then, is to
give up the idea that philosophical thought, about the sorts of
practice in question, should be undertaken at some external
standpoint, outside our immersion in our familiar forms of life.
(MVR, p. 63)

Or, as he puts it elsewhere, ‘‘we would be protected against the vertigo if we

could stop supposing that the relation to reality of some area of our thought
and language needs to be contemplated from a standpoint independent of
that anchoring in human life that makes the thoughts what they are for us’’
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

(MVR, p. 211).
I cannot here explain all the details of what McDowell takes to be the
‘‘cure’’ for, or the ‘‘protection’’ against, this vertigo. One of McDowell’s
central ideas, however, is that occupying the sort of external, detached
viewpoint from which a given practice might seem to lack a ‘‘solid
foundation’’, involves losing sight of (or being blind to) those thoughts and
perceptions which are made possible by the practice, and in terms of which
the practice is characterized from the inside, by its engaged participants.
According to McDowell, being unsatisfied with the place at which
justification comes to an end within the practice—feeling that justification
should end ‘‘further down’’, at a more ‘‘solid’’ place—means: not having
access to, or perhaps having denied oneself access to, those conceptual and
perceptual capacities in terms of which the practice is understood by those
who are ‘‘normally immersed’’ in it (MVR, p. 211). This means that a
question can sensibly be raised about the significance of this sort of
disappointment with a practice. For if this disappointment is based on a
description of the practice which is alien to the practice itself—in particular,
if the description is not faithful to those conceptual and perceptual
capacities in the light of which the procedures of justification which are
immanent to the practice will seem perfectly sufficient—then there is reason
to ask why this sort of disappointment is anything to worry about at all.
My earlier discussion of perfect pitch, normal pitch, and tone-deafness
provides an instructive illustration of McDowell’s point. Our dissatisfaction
with the way pitches were identified by the inhabitants of the world where
everyone has perfect pitch had to do with our construing the ability to hear
directly that this is an A in terms of the ability to say, directly, ‘‘This is an
A’’, on occasions where everyone else in the community would say the same
thing. Precisely because we lack perfect pitch, we have problems recognizing
any substantial difference between these two abilities. This might be said to
mark the limited extent to which someone who lacks perfect pitch can
understand what it means to have perfect pitch. One might say that not
having perfect pitch means lacking the perceptual and conceptual capacities
required in order to understand clearly how hearing directly that this is an A
can be a matter of perceiving how things are, rather than just a matter of
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 363

being caused, by some natural event, to believe something that one’s peers
also believe. In this sense, perfect pitch is indeed an enigma to he who does
not have it. This is not to say, à la Nagel, that someone who lacks perfect
pitch does not know ‘‘what it is like’’ to have perfect pitch. The relevant
difference is not one between subjective qualities, in Nagel’s sense. Rather,
the point is epistemological: he who lacks perfect pitch is unable to situate
perceptions of the relevant sort within the logical space of reasons (rather
than within the space of causes). He has difficulties conceiving perceptions
that do not involve the measuring of a given tone against a reference tone as
genuine justifications of the claim that this is an A. And similarly for the
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

tone-deaf: his inability to hear differences and similarities in pitch is an

inability to conceive perceptions that do not involve the use of a frequency
meter (or some similar device) as genuine justifications of the claim that this
tone has the same pitch as that tone.
Cavell definitely agrees with much in McDowell’s general analysis of
our inclination to look for a ‘‘solid’’ external foundation for our practices.
But it is also worth mentioning, even at this early stage of the present
investigation, that there seem to be differences between Cavell and
McDowell at this point, differences that are hard to identify precisely but
which might nonetheless be quite significant. In brief, what I have in mind is
this: reading McDowell, one often gets a sense that he thinks the temptation
to search for an external foundation, and the terror one might feel when one
realizes that this search is bound to be unsuccessful, can be kept at a
distance once the falsity or meaninglessness of the ideas on which this
temptation and this feeling of terror are based has been made fully evident.
Again, he suggests there is a ‘‘cure’’ from, and ‘‘protection’’ against, such
terror. According to McDowell, Wittgenstein’s vision of language is
frightening only to someone who is still held captive by the sort of pictures
Wittgenstein meant to combat. McDowell seems to think that if you have a
clear grasp of Wittgenstein’s vision, you realize that such terror is misplaced,
and you will know how to get rid of it were it to begin to haunt you again.
Cavell, by contrast, never so much as suggests that this terror is
something one can be cured from or protected against. Perhaps he would
even hesitate to call it misplaced. He seems to think that no matter how
philosophically clear-sighted we are, our normal immersion in practices
would not be what it is unless the possibility of not being thus immersed
could scare the hell out of us. Of central importance for an adequate
understanding of this aspect of Cavell’s thinking is his idea that there is a
sense in which our very immersion in familiar practices tempts us to
disengage from those practices. He talks of ‘‘the possibility of the
repudiation of ordinary concepts by, as it were, themselves’’, a possibility
he describes as ‘‘our possibility of repudiating our agreement in terms of
which words have criteria of relation (to the world, of the world) given them
in the human life form’’ (PoP, p. 97). He also speaks of the ‘‘attempt, or
 364 M. Gustafsson

wish, to escape […] those shared forms of life, to give up the responsibility of
their maintenance’’, adding, famously, that ‘‘[n]othing is more human than
the wish to deny one’s humanity.’’4
Admittedly, there are passages in which McDowell suggests that
philosophical peace of mind might be no more than a practically
unattainable ideal. ‘‘Interesting philosophical afflictions’’, he says, ‘‘are
deep-seated. Even after temporarily successful therapy, they re-emerge,
perhaps perennially, in new forms. If the risk of re-emergence is perennial,
peace is always beyond the horizon.’’5 Cavell, however, questions the idea
that such peace of mind makes sense, even as an ideal. While McDowell
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

might regard the perennial re-emergence of philosophical terror as a

plausible conjecture, Cavell suggests it is somehow essential to what it
means to lead a human life.
I am gesturing here at one of the central and most controversial parts of
Cavell’s philosophy. Providing more than such vague gestures would be
premature at this point, since further precision presupposes a deeper
understanding of the significance that Wittgenstein’s vision of language has
for Cavell. In particular, it will be necessary to get clear about one important
aspect of what this vision means to Cavell, namely, its import for his
conception of so-called ordinary language philosophy. Achieving such
clarity is the overall aim of sections 3 and 4. Further discussion of how
Cavell might differ from McDowell is postponed until section 5. This issue
will then be related to Cavell’s dissatisfaction with Austin, and, in
particular, to Cavell’s claim that Austin is not clear about the nature and
possible achievements of his own philosophical procedures.

III.
I now turn to the analogy Cavell makes in the passage quoted at the
beginning of this paper: between being able to practice with understanding
the method of ordinary language philosophy and having perfect pitch. Let
me start by saying that when I first read this passage I found the analogy
striking but also very puzzling. I could not make clear to myself what the
alleged similarities between these two abilities were supposed to be, and
nonetheless I felt sure that such similarities existed. A few pages after the
quoted passage Cavell again mentions perfect pitch, in a discussion of the
relation between philosophy and autobiography. ‘‘If’’, he says, ‘‘philosophy
is for me finding a language in which I understand philosophy to be
inherited, which means telling my autobiography in such a way as to find
the conditions of that language, then I ought even by now to be able to
begin formulating some of those conditions.’’ He then goes on to give a list
of linked conditions, among which he includes ‘‘a version of perfect pitch’’
(PoP, pp. 38–9). Eight pages later, he says of these conditions that ‘‘the
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 365

feature of perfect pitch is apt to be the hardest to recognize, and the most
variously or privately ratified.’’ And then:

I mean it as the title of experiences ranging from ones amounting to

conversions down to small but lucid attestations that the world holds a
blessing in store, that one is, in Emerson’s and Nietzsche’s image,
taking steps, walking on, on one’s own. (PoP, p. 47)

This is not easy to comprehend, and I do not claim to have anything like a
complete understanding of what Cavell is trying to say here. But I do find
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

his talk about perfect pitch as the title of the experience that one is taking
steps, walking on, on one’s own, illuminating. The best way for me to
explain what kind of elucidation I find here, is by relating the passage I just
quoted to the significance of the fact that the remarks of ordinary language
philosophers are, typically, in the first person plural: ‘‘We say …’’, ‘‘We
don’t say …’’, ‘‘When we say … we imply …’’, and so on. According to
Cavell, this feature is a central clue to the nature of such remarks. (See, for
example, MWM, p. 14.) He argues that the use of the first person plural
shows that such remarks combine two aspirations that, at first, might seem
difficult to reconcile. On the one hand, such remarks are not statements of
empirical science: they are not based on polls, statistical studies of linguistic
behavior, or other such evidence. One might say that this constitutes their
first-person aspect. On the other hand, in making remarks of this form, the
ordinary language philosopher does claim to speak for others. This is
crucial, since ‘‘the primary fact of natural language is that it is something
spoken, spoken together’’ (MWM, p. 33). According to Cavell, if the
philosopher claimed to speak only for himself, as someone completely
detached from the linguistic community of which he is a member, he would
be speaking for no one—not even himself. Language, as we know it, is
something we share.
Some philosophers have felt that the first-person aspect and the plural
aspect of such statements are straightforwardly incompatible. They have
argued that if you claim to say something about how language is used by a
community, the only way to justify those claims is to make empirical, socio-
linguistic observations. According to Cavell, this argument ignores that the
relevant statements about ordinary language are made by native speakers of
that language, speakers who,

do not, in general, need evidence for what is said in the language; they
are the source of such evidence. It is from them that the descriptive
linguist takes the corpus of utterances on the basis of which he will
construct a grammar of that language. To answer some kinds of
specific questions, we will have to […] count noses; but in general, to
tell what is and isn’t English, and to tell whether what is said is
 366 M. Gustafsson

properly used, the native speaker can rely on his own nose; if not, there
would be nothing to count. (MWM, p. 4)

It is easy to misunderstand the idea expressed here. Cavell’s point is not that
the native speaker has some sort of intuitive and infallible insight into the
linguistic dispositions of his fellow speakers. On the contrary, Cavell freely
admits that the speaker’s claims about what ‘‘we’’ say might be mistaken,
and that the speaker’s peers might disagree with them. The point is, rather,
that if such disagreement arises, what happens is quite different from, and in
an important sense much more troubling than, when a socio-linguistic
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

hypothesis is falsified by statistical polls or other empirical data. When my

claims about what we say are questioned, what is questioned are both
my own self-understanding and my sense of what I share with my
peers, these two things being inseparable. Even occasional and corrigible
mistakes may be worrying in that sense, not to mention persistent
disagreement. Such disagreement might undercut that sense of community
inherent in the notion of a common language, and the sense of authority
inherent in the notion of linguistic competence (the authority for which we
take on responsibility when, e.g., we teach our children to speak). Cavell
writes:

The claim that in general we do not require evidence for statements in

the first person plural does not rest upon a claim that we cannot be
wrong about what we are doing or about what we say, but only that it
would be extraordinary if we were (often). […] If I am wrong about
what he does (they do), that may be no great surprise; but if I am
wrong about what I (we) do, that is liable, where it is not comic, to be
tragic. (MWM, p. 14)

This means that practicing the ‘‘method’’ of ordinary language

philosophy involves both a rather particular sort of self-reliance and a
rather particular sort of humility. Indeed, such self-reliance and such
humility are internally related: they are exhibited simultaneously, and
cannot be present apart from each other. This, says Cavell, is why, and the
sense in which, the ordinary language philosopher cannot shun the
autobiographical, but must use it as the ground from which he dares to
speak for others:

Philosophers who proceed as Austin suggests will not be much

interested to poll others for their opinion […]. Then why do such
philosophers say ‘‘we’’ instead of ‘‘I’’? With what justification? They
are saying what the everyday use is […]. And by whose authority?
Their basis is autobiographical, but they evidently take what they do
and say to be representative or exemplary of the human condition as
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 367

such. In this way they interpret philosophy’s arrogance as the

arrogation of the right to speak for us, to say whatever there is to
say in the human resistance to the drag of metaphysics and of
skepticism; and authorize that arrogation in the claim to representa-
tiveness, expressed autobiographically. There is a humility or poverty
essential to this arrogation, since appealing to the ordinariness of
language is obeying it—suffering its intelligibility, alms of common-
ness—recognizing the mastery of it. (PoP, p. 8)

Now where does perfect pitch, as the title of the experience that one is
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

walking on, on one’s own, come in here? Well, in something like the
following way: having perfect pitch involves a peculiar mixture of self-
reliance and vulnerability. He who has perfect pitch trusts his own
perceptual capacity. His having perfect pitch means that he cares little
about consulting tuning forks. Nor does he ask others for their opinion—
after all, that this is an A is not settled by polls. However, if people started
disagreeing with him about the pitches of tones, that would be troubling. It
might be a sign that he is no longer part of that special community of people
who have perfect pitch (his talent may be gone.) Or, it might mean an even
more profound disintegration of his, or his fellow-speakers’, sanity. This
situation is analogous to the case described in the following remark by
Wittgenstein:

How do I know that the colour that I am now seeing is called

‘‘green’’? Well, to confirm it I might ask other people; but if they did
not agree with me, I should become totally confused and should
perhaps take them or myself for crazy. That is to say: I should either
no longer trust myself to judge, or no longer react to what they say as
to a judgement.6

In a rather (but, as we shall see, not altogether) similar sort of way, the
ordinary language philosopher must dare to rely on his own sense of how we
speak. His remarks can be neither confirmed nor falsified by polls. Nor is a
dictionary, or some other officially sanctioned collection of linguistic rules,
of much help. For his investigation takes place at a level where such
officially sanctioned rules do not function as normative standards in the
relevant sense of the word (this point will be further clarified below). It
might turn out that his peers do not want to agree with what he says they
should say. But this lack of agreement does not, by itself, constitute a
falsification of his remarks. Rather, it might make unclear the extent to
which these people are his peers; or, it might be a sign that it is unclear to
him what is involved in his being part of the community in question. If
disagreement becomes prevalent and persistent enough, he might even doubt
his own sanity.
 368 M. Gustafsson

Consider Austin’s famous donkey example:

You have a donkey, so have I, and they graze in the same field. The
day comes when I conceive a dislike for mine. I go to shoot it, draw a
bead on it, fire: the brute falls in its tracks. I inspect the victim, and
find to my horror that it is your donkey. I appear on your doorstep
with the remains and say—what? ‘‘I say, old sport, I’m awfully sorry,
&c., I’ve shot your donkey by accident’’? Or ‘‘by mistake’’? Then again,
I go to shoot my donkey as before, draw a bead on it, fire—but as I do
so, the beasts move, and to my horror yours falls. Again the scene on
the doorstep—what do I say? ‘‘By mistake’’? Or ‘‘by accident’’?7
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

Even if Austin here speaks in the first person singular, his intent is in
the relevant sense plural: his example is designed to remind us of how we
speak. And he succeeds, brilliantly. We all agree that in the first scenario
the donkey was shot by mistake, whereas in the second scenario it was
shot by accident. But what if people did not agree with this? What if
they said the opposite: no no, the first donkey was shot by accident and
the second by mistake? Well, supposing that they are serious, their
disagreement would be utterly weird—much more so than the falsification
of a socio-linguistic hypothesis. For you (who, I suppose, agree with Austin)
would not understand to what their disagreement amounted. It could not
just be a local rupture, for you would then also want to know what places
these people assign to a host of other concepts, such as ‘‘intention’’,
‘‘responsibility’’, ‘‘blame’’, ‘‘excuse’’, and so on and so forth. Of course it
might happen that such an inquiry makes them change their minds about
Austin’s story, and admit that the first donkey was shot by mistake and
the second by accident. If so, there is perhaps not much to worry about. But
it might also turn out that those people want to place the relevant concepts
in a pattern that is foreign to you, or that you do not succeed in
understanding these people at all. In either case, the result is more or less
radical estrangement: these people may become enigmas to you (Cf. MWM,
p. 67).

IV.
In the previous section I said something brief about the level at which the
investigations of the ordinary language philosopher take place. I said this
had to do with what I take Cavell to mean when he talks about perfect pitch
in terms of ‘‘walking on on one’s own.’’ Let me try to clarify what I meant
by reflecting on a passage from a paper by Stephen Mulhall. In his reply to
Steven Affeldt’s criticism8 of his book Stanley Cavell: Philosophy’s
Recounting of the Ordinary, Mulhall discusses Cavell’s notions of grammar
and criteria, and he argues that there is a perfectly good sense in which
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 369

‘‘there is a given grammar which represents the normative standard against

which a particular judgment is to be assessed.’’ Mulhall proceeds by giving
the following example of what such givenness and determination might
involve:

If someone were to dispute my claim that the colour of my living room

wall is blue rather than green, we might try to settle the dispute by
asking the views of others, but we might also refer to the colour chart
in an interior decoration book—checking the match between wall
colour and sample. This is what I would mean by talk of assessing a
judgment against a grammatical standard.9
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

What Mulhall says here is just not right. Typically, if someone were to
dispute my claim that the color of my living room wall is blue rather than
green, we would not try to settle the dispute by consulting the color chart in
an interior decoration book. A color chart might be of help if we were
quarrelling about, say, whether a certain shade of green is called (by a
certain paint maker) ‘‘Apple Green’’ or ‘‘Pear Green.’’ But a chart will not
help us decide whether a given color is blue or green. In our world, color
charts just do not have that kind of authority in this sort of case. If two
people really disagree about whether the color of a wall is blue or green, and
if they are at all confident in their ability to distinguish these colors, they will
continue to disagree about whether the relevant color sample is blue or
green. It matters little whether such a color chart counts that sample as a
shade of green or a shade of blue. If I am really convinced that the color is
blue, the color chart’s calling it, say, ‘‘Sea Green’’, will not make me change
my mind; rather, I will argue that what this color chart calls ‘‘Sea Green’’ is,
in fact, a shade of blue.
Nor will asking other people be of much help. Suppose I think the color of
the wall is a borderline case, but if I had to make a decision I would call it
‘‘blue’’ rather than ‘‘green.’’ Then, if almost everyone else told me that they
would rather call it ‘‘green’’, I might change my habits of talking. But if I do
not, I will still be counted as a member of that linguistic community. My
continuing to classify this shade as ‘‘blue’’ does not mean that people will
regard me as incompetent in my use of color words. At most, they will think
of me as just a little bit idiosyncratic. By contrast, if I am absolutely
convinced that the color is not a borderline case, but definitely blue, the fact
that other people disagree with me would be very troubling. But it would
not be troubling in the way a straightforward falsification of an empirical
hypothesis is troubling. It would be worse. As Wittgenstein says in the
passage quoted above, such disagreement would make me totally confused
and I might then perhaps take my peers or myself for crazy. It might, of
course, turn out that I have become color-blind; but this will then show itself
in my being unable, in regular and predictable ways, to distinguish between
 370 M. Gustafsson

green and blue in a series of other cases. This scenario is not, I suppose,
what Mulhall has in mind.
So ironically, the example Mulhall offers illustrates precisely the opposite
of what he takes it to show. If, by ‘‘a normative standard against which a
particular judgment is to be assessed’’, we mean something like a color
chart, then there are many uses of language which are not governed by any
such standard. Indeed, many of the most basic, everyday uses of language
involve no such authoritative standard. Now what I am arguing is that,
according to Cavell, the ordinary language philosopher is interested in
precisely these very basic and elementary strata of linguistic practice. That is
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

the level at which his investigation takes place. Indeed, how could it be
otherwise? If he were interested in uses of language where authoritative
standards analogous to color charts did play the kind of role Mulhall
describes, his task would be relatively simple but also relatively uninterest-
ing. All he would have to do then would be to inform us about the
authoritative charts, rule-books, and so on. But this is precisely what he
cannot do; it is precisely the sense in which his task is not simple. Instead, he
has to remind people of what they already know (but cannot be informed
about), and the only basis on which he can claim to do so is his own position
as competent language user, his own sense of the language he has inherited
and claims to share with others.10
This is related to what Cavell regards as another central feature of
philosophical investigations of ordinary language, namely, that such
investigations ‘‘can as appropriately or truly be said to be looking at the
world as looking at language’’ (MWM, p. 99). Considering the cliché that
Austin’s central philosophical concern lies in drawing distinctions, Cavell
discusses how examples and (‘‘most characteristically’’) stories set the stage
for Austin’s distinctions. He contrasts Austin’s use of examples with that
which can be found in the writings of more traditional philosophers, such as
Russell, Broad and Moore,

whose distinctions do not serve to compare and (as it were) to elicit

differences but rather, one could say, to provide labels for differences
previously, somehow, noticed. One sometimes has the feeling that
Austin’s differences penetrate the phenomena they record—a feeling
from within which the traditional philosopher will be the one who
seems to be talking about mere words. The differering role of examples
in these philosophies is a topic whose importance cannot be
exaggerated, and no amount of words about ‘‘ordinary language’’ or
‘‘make all the distinctions’’ will convey to anyone who does not have
the hang of it how to produce and test such examples. (MWM, p. 103)

On reading Austin, says Cavell, ‘‘what we learn will not be new empirical
facts about the world, and yet illuminating facts about the world. It is true
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 371

that he asks for the difference between doing something by mistake and
doing it by accident, but what transpires is a characterization of what a
mistake is and (as contrasted, or so far as contrasted with this) what an
accident is’’ (MWM, p.104).
Consider a case when you need to use a color chart. You and your friend
are having a dispute about the color of his living room wall: you say it is
Apple Green, he says it is Pear Green. Your problem might then be
described as follows: you want to know about a set of rules telling you which
one of the labels ‘‘Apple Green’’ and ‘‘Pear Green’’ is rightly applied to an
already identified bit of reality. Now this is very different from the kind of
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

problem with which Austin’s donkey example is designed to deal. Austin

does not give us rules for how to apply the terms ‘‘doing something by
mistake’’ and ‘‘doing something by accident’’ to already identified bits of
reality. Rather, he tells us stories which remind us of what reality might be
like, and once these stories have been told it is clear to us that there is a
difference between doing something by mistake and doing it by accident.
The question, ‘‘Do these words apply to this piece of reality?’’ never comes
up. The problem is, rather, that we have no clear conception of what ‘‘this
piece of reality’’ is or can be like. But once Austin has told us stories that
describe such pieces of reality, we immediately recognize the scenarios he
describes as instances of doing something by mistake and doing something
by accident, respectively. Hence, what we needed was not rules telling us
that those scenarios are instances of those concepts; what we needed was to
be reminded of the very possibility that reality might involve such scenarios.
This makes it somewhat clearer why it is so difficult to come up with
stories like Austin’s donkey example. Finding rules of application is not so
hard, and presupposes that we are clear about what the rules are going to be
applied to. But if what the rules are going to be applied to is precisely what
we are not clear about, then we have a much more challenging task, the
accomplishment of which requires a form of sensibility that cannot be
conveyed by simple instruction.11 Cavell suggests it is no less troublesome to
teach someone who lacks the required sensibility the materials and methods
of ordinary language philosophers than to teach someone who lacks musical
sensibility to master an instrument. ‘‘Perhaps what is wanted’’, Cavell says,
‘‘is a matter of conveying ‘the hang’ of something, and that is a very
particular dimension of a subject to teach—familiar, for example, in
conservatories of music, but also, I should guess, in learning a new game or
entering any new territory or technique or apprenticing in a trade’’ (MWM,
pp. 103–4).

V.
I now return to the theme of section II: the similiarities and differences
between Cavell’s and McDowell’s ways of understanding Wittgenstein’s
 372 M. Gustafsson

vision of language. According to McDowell’s Wittgenstein, the worry that

established practice amounts to nothing but a congruence of subjectivities
depends on the idea that philosophical reflection should be undertaken ‘‘at
some external standpoint, outside our immersion in our familiar forms of
life.’’ What can be learnt from Wittgenstein, McDowell says, is that such
external reflection loses sight of, and thus fails to do justice to, those
conceptual and perceptual capacities in the light of which customary
procedures of justification seem perfectly sufficient. This is the reason why
philosophical detachment makes it hard to understand how familiar
concepts and perceptions can tell us how things really are. Seen from
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

within their anchoring in human life, however, the objective purport of such
concepts and perceptions is unproblematic. Indeed, McDowell’s
Wittgenstein thinks such anchoring is a precondition for objectivity.
Weighing the anchor means closing oneself off from the world.12
So far, there is perhaps not much disagreement between McDowell and
Cavell. In any case, the disagreement I am interested in here has to do with
the question of whether Wittgenstein’s vision of language is, as Cavell puts
it, ‘‘terrifying.’’ I suggested before that McDowell would not, without
qualification, agree with this characterization. According to McDowell, if
Wittgenstein’s vision appears terrifying, it is only because we have not fully
understood it. By contrast, Cavell’s Wittgenstein does not think that
philosophical clear-sightedness offers protection against such terror. On the
contrary, his view appears to be that the dream of such protection is
fundamentally misguided.
In order to get clearer about this issue, it is useful to consider some of the
ways in which the relation between having perfect pitch and lacking perfect
pitch is different from the relation between being ‘‘normally immersed’’ in
familiar practices and being philosophically dissatisfied with such practices.
Here is one such difference. In relation to a community where everyone has
perfect pitch, I would be an outsider. For I cannot identify pitches without
making use of something like a tuning-fork. Similarly, someone who is tone-
deaf is an outsider in relation to people who have normal pitch. And the
colour-blind are unable to use colour words in that direct sort of way that
characterizes the practice of people endowed with normal colour vision. In
all these cases, the outsider’s position is, as it were, totally involuntary: his
inability is a natural deficiency he can do little or nothing about. Roughly
speaking, he has no choice but to view the practice from without. By
contrast, in the sort of case on which both McDowell and Cavell focus,
where someone is philosophically dissatisfied with familiar practices, the
situation is different and more complex. For in this sort of case, the person is
perfectly able to participate in the relevant practices. Indeed, most of his
time he is as ‘‘normally immersed’’ in those practices as anyone else.
Consequently, his philosophical alienation must somehow be self-imposed,
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 373

and thus quite different from the non-participation of the tone-deaf or the
color-blind.
Moreover, this philosophically dissatisfied person is not clearly aware of
his detachment: he does not take himself to have lost sight of the concepts
whose objective purport he wants to legitimize. On the contrary, he thinks
his reflections lead to a firmer grasp of those concepts. His sense is that the
procedures of abstraction and generalization that typically characterize his
approach serve to clarify or purify the concepts, rather than to hide them
from view. So here we have a second disanalogy between this sort of
philosophical detachment and the way in which someone who lacks perfect
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

pitch is detached from practices that involve the exercise of perfect pitch. He
who lacks perfect pitch knows he is an outsider. Whereas, if Cavell’s and
McDowell’s diagnosis is right, the philosophically dissatisfied person is not
aware of his having dissociated himself from the conditions under which the
concepts he claims to be clarifying are meaningfully applied.
Note how extremely provocative this diagnosis is, or should be. The
provocation is largely due to the fact that the supposedly detached
philosopher is a master of the practice from which he is said to have
alienated himself. He is a competent language user. And yet, Cavell and
McDowell dare to claim, when such a person thinks he is gaining a reflective
understanding of certain concepts of that language, what happens is the very
opposite: those concepts slip through his fingers.
How is such self-inflicted yet unconscious alienation possible? If the sort
of diagnosis that Cavell and McDowell are offering is going to seem at all
plausible, they must give a credible answer to this question. At this point,
one can begin to sense a distinctive difference between the two philosophers.
McDowell can be read as providing an answer according to which the
temptation to alienate oneself from familiar practices is contingent upon
certain events in relatively recent intellectual history. In Mind and World, he
argues that the philosophical anxieties he discusses require the background
of a conception of nature as ‘‘disenchanted’’, a conception which ‘‘was made
available only by a hard-won achievement of human thought at a specific
time, the time of the rise of modern science.’’13 Admittedly, it is not
altogether clear if the root of the anxieties McDowell refers to here is
identical with the sort of self-inflicted alienation I have been describing
above, or if those anxieties should rather be seen as resulting from one
specifically modern form of such alienation, other forms being imaginable
under different historical circumstances. The latter alternative would leave it
open for McDowell to argue that the temptation to alienate oneself from
familiar practices is not, in general, as historically conditioned as he
sometimes seems to be suggesting. On the other hand, the fact remains that
the only account he gives of the origin of this temptation is in terms of the
emergence and metaphysical exploitation of our modern conception of
 374 M. Gustafsson

nature. Insofar as he thinks the temptation is capable of taking other, less

scientistic forms, he tells us nothing about its general allure.
Cavell, by contrast, tends to focus on features that make our urge to
alienate ourselves from familiar practices seem less contingent, or better,
more ineradicably human. He would not deny that the view of nature
as disenchanted is significant to certain ways in which this urge manifests
itself in contemporary thinking. But such disenchantment does not play
the sort of pivotal role in Cavell’s writings that it seems to play in
McDowell’s. Typically, Cavell’s descriptions of what self-inflicted yet
unconscious alienation may involve are not restricted to philosophical
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

anxieties whose ultimate source is a modern conception of nature. Rather,

what he tries to make us recognize is what he takes to be the more
fundamental allure of such alienation—something about which McDowell
has little to say.
The best way of getting clear about the details of Cavell’s analysis is to
look at a concrete and representative example. Consider the following a
story.14 Yesterday night, Peter witnessed a magical performance. ‘‘In his
right hand’’, he tells his friend Paula, ‘‘the magician held two ping-pong
balls. They were in full view, right under my eyes, no concealment or
anything. Suddenly, he moved his hand a little, and one ball just
disappeared.’’ Paula, who knows something about conjuring tricks, replies:
‘‘Did you really see the whole balls? You know, magicians often use shells.
It’s a simple idea, you can make two such shells yourself by cutting an
ordinary ping-pong ball into halves. By displaying the convex side, you
make people think they see a whole ball. If the magician holds one real ball
and one shell in his hand, and then rapidly slides the ball into the shell, he
creates exactly the kind of illusion you describe.’’ Peter pauses and then
says, ‘‘Yes, you’re right. Come to think of it, I didn’t see the whole balls, but
only the front surfaces of what I took to be two balls. Clearly, one of them
might have been a shell. That’s the explanation of course.’’
This is an example of an everyday use of the expression, ‘‘I didn’t see the
whole thing, but only its front surface’’. Now, imagine an epistemologist
who wants to make what appears to be a rather similar (even if much more
general) application of the same expression. This epistemologist argues that
whenever we look at an object, all we really see is its front surface. Of course
the epistemologist knows that in everyday life we say things like ‘‘I see the
whole building from here’’ or ‘‘If you remove your hand I will be able to see
the whole teapot’’. But these ways of talking, he argues, are strictly speaking
false: it never happens, it couldn’t happen, that we see a whole building or a
whole teapot.
This epistemologist might be said to invoke a model or a scheme that is as
simple as it is persuasive. According to this scheme, the part of an object
seen by an observer at a given time is identified only in terms of the
geometrical and physical circumstances of the situation. The basic idea can
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 375

be rendered as follows: the part of the object that is seen consists of all the
points of the object which are such that you can draw straight lines from the
observer’s eye to those points without interruption. If no non-transparent
matter blocks the way from the eye to a given point, then that point belongs
to the visible part of the object. The eye is like a searchlight directed toward
the object: the part that gets illuminated is the part which can be seen, while
the part which remains dark is out of sight. Since any instance of seeing
something is possible to picture in this sort of way, it seems to follow that
whenever we see an object, we see only that part of it which faces us in the
above mentioned sense—the object’s front surface. Hence (the argument
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

continues), to the extent that we want to say that a certain object consists of
something more than such a front surface, we will have to admit that we
never see the whole object, but only a part of it.15
This argument may seem indisputable. But Cavell thinks there is a
fundamental problem with it. This problem becomes discernible once we
dare to raise the following question: does the model the epistemologist
invokes really manage to identify and distinguish between parts of an object?
At first, this question may sound ridiculous. Clearly, one wants to say, the
suggested scheme distinguishes between parts. One part is the part made up
of all the points to which straight lines can be drawn without interruption,
and the other part is the part to which such uninterrupted lines cannot be
drawn. That seems obvious enough.
But are these so-called ‘‘parts’’ really parts in any substantial sense of the
word? Think again about Peter’s and Paula’s discussion. To understand
what it means for Peter to be struck by the fact that he did not see the whole
ball, it is crucial to realize that the identity of the part he did see is not
established by reference to the fact that this part happens to be facing him.
Suppose Peter grabs the object and turns it around. That would mean that
the part that he just saw gradually disappears from view, and yet it remains
the same part as before. Some other part of the thing, a part that was not
seen earlier—the back side—will then come into view. If the thing is a ball,
that other part will have the same convex shape as the part that was visible
earlier. If the thing is a shell, Peter will instead be looking into a concave
hollow.
So when Peter and Paula talk about ‘‘the back side’’ of what Peter took to
be a ball, the identity of that part ‘‘is established independently of the
(merely geometrical-physical) fact that it is then and there not visible from
[Peter’s] position’’ (CR, p. 200). And analogously for the front surface. This
makes their usage significantly different from the one suggested by the
epistemologist’s scheme. In fact, it is crucial to the general aspirations of the
epistemologist—his will to conclude that we can never see any part of an
object other than its front surface—that no such independent identification
of parts occurs. If the identity of a part is established independently of the
geometrical-physical facts of a particular moment of seeing, then it is
 376 M. Gustafsson

perfectly possible to see the back side of an object: just walk around the
object, or turn it around.16
Considered merely as a formal move within his model, the epistemolo-
gist’s conclusion, ‘‘We never see any part of an object other than its
front surface’’, is indeed indisputable. But it is also both trivial and
empty: it is a tautologous consequence of the conventions that define the
model, and has no bearing on anything else. The problem is that the
epistemologist wants his generalization to amount to something more
than such an empty move. His conclusion is supposed to result from
an application of the model to reality. He intends to make a substantial
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

claim.
This means that the epistemologist has run into a dilemma. His general
aspirations require that no independent identification of parts be made. His
wish to make a genuine claim requires the opposite: that such an
independent identification of parts is made. The validity of the epistemol-
ogist’s conclusion depends on the model’s being a model of nothing;
whereas the substance of the conclusion requires that the model is actually
applied.
This might be difficult to realize. One might think that applying the model
cannot be a problem, for it seems to fit (automatically, as it were) all
instances of someone’s seeing an object. Again, it is trivial that each such act
of seeing takes place under certain geometrical and physical circumstances,
and the scheme seems to tell us how those circumstances determine what is
actually seen. But, again, the problem is with the expression, ‘‘what is
actually seen.’’ There is of course a sense in which it is true that geometrical
and physical circumstances determine what part of an object is visible from
a given point of view. It is essential, however, that we can specify the seen
part, the ‘‘what’’, without reference to those geometrical-physical circum-
stances. Otherwise, it is totally unclear what is being ‘‘determined’’ by the
geometrical and physical circumstances; all we get is a redescription of those
circumstances in terms of ‘‘visible points.’’
And if we try (by force, as it were) to make reality fit this empty
redescription in a more substantial sense, the result is arbitrary distortion.
For example, how are we going to account for the fact that real-life
observers can and must move around in the world? If the identity of the
parts of an object is established solely in terms of their geometrical relation
to the observer, such moving around becomes incomprehensible. After all,
for a creature with visual organs, moving around among objects means
coming to see new parts of the objects, parts that were previously out of
sight. But seeing parts that were previously out of sight is precisely what the
redescription seems to leave no room for, since it entails that we always see
the same part of an object, ‘‘the front surface.’’ So, the upshot is that we do
not move around. Allegedly, what actually happens when we, as we usually
put it, ‘‘walk around objects in order to take a look at their back sides’’, is
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 377

that the front surfaces of objects change their visual appearance. The picture
is that it is the passing show around us which alters, somewhat like when
watching a film, whereas the position of our eyes remains fixed.17
I hope my discussion of this example has given the reader a sense of the
complexity Cavell sees in the relation between our familiar use of concepts
and the philosophical dissatisfaction with that use. Moreover, the example
should suggest why it is there, in that complex relation, that Cavell finds the
explanation of how a master of the language can take himself to be
clarifying familiar concepts when, in reality, he loses sight of those very
concepts. It is, for example, highly significant that our imagined
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

epistemologist starts from familiar language. He uses ordinary words, even

if his use of those words is not entirely natural. Cavell says the
epistemologist makes a projection: he imports familiar words into a not so
familiar context, somewhat like when the word ‘‘feed’’ gets projected from
familiar constructions such as ‘‘feed the baby’’ and ‘‘feed the dog’’ into new
ones, like ‘‘feed the meter’’ and ‘‘feed his pride.’’
Now, Cavell does not think there is anything wrong with projections from
familiar contexts into less familiar ones. In particular, he does not think that
the unfamilarity of a new context means that the projection involves a
distortion of established concepts. ‘‘Feed the meter’’ is a perfectly fine and
meaningful expression. And, at least at first sight, the epistemologist’s way
of talking seems equally all right. As Cavell notes, ‘‘it doesn’t seem obvious
that an object can’t (and even oughtn’t to) be taken to be something whose
front ineluctably conceals its back’’ (MWM, p. 251).
In fact, Cavell thinks meaningful language use is not just compatible with,
but dependent on our projective abilities. According to Cavell, the pervasive
significance of our unregularizable projective imagination manifests the
extent to which keeping language alive and the world in view is not a matter
of passive conformity, but a continuous undertaking which requires the
employment of those interests, feelings, modes of response, senses of humor
and significance and fulfillment that he refers to in his earlier quoted
description of Wittgenstein’s vision of language. This undertaking is our
task, as language using creatures; we are burdened with this responsibility. It
cannot be transferred to, say, a machinery of rules the application of which
is fixed independently of human modes of response.
What is special about the epistemologist’s projection is that it serves to
hide the fact that we are responsible in this sort of way. This makes it
different from a projection like that from ‘‘feed the dog’’ to ‘‘feed his pride’’,
where it is more or less clear that our ability to go on in the relevant sort of
way involves an unregularizable employment of human sensitivity. The
epistemologist’s projection, by contrast, appears to make such human
sensitivity redundant. By abstracting from the particular reasons people
have for asking what parts of an object are visible, and instead providing
systematic principles for distinguishing between the seen and the unseen
 378 M. Gustafsson

parts in any particular case, the epistemologist’s model seems to cleanse our
ordinary ways of making that distinction from the shifting concerns which
make us raise the issue in real-life situations. In real cases, we have specific
worries: we suspect that what looks like a ball is, in fact, a shell; we want to
know whether John sees that the vase before him has a crack in it; we use an
ultra-sound scanner to look at a fetus in order to check its development; and
so on and so forth. In each such case, saying ‘‘I don’t see the whole thing’’
has a specific point, the appreciation of which requires, precisely, a sharing
of interests, feelings, and perhaps senses of humor, significance, and
fulfillment. The epistemologist’s conclusion, by contrast, is in that sense
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

pointless. Indeed, it is this very pointlessness that constitutes its attraction.

Unclouded by real-life concerns, the epistemologist’s scheme seems to allow
us to focus on the issue itself, as it were: what is it that we really see when we
see an object?18
This, then, is the irony of the epistemologist’s endeavor: if a projection is
something we do and for which we are responsible, the epistemologist’s
projection does not appear as a projection at all, but as a necessary
development grounded in and guided by something more fundamental and
firm than that ‘‘whirl of organism’’ which manifests itself in our being able
to understand expressions like ‘‘feed his pride.’’ As Cavell notes, the
epistemologist is likely to think of his way of talking, not just as natural
enough, but as inevitable (CR, pp. 144–5, MWM, pp. 107 and 251). The
invoked scheme seems to be called for by the concepts themselves, as it were.
The feeling is that once you have become clear about what ‘‘seeing’’, ‘‘part’’,
and so on, really mean, you must admit that all you ever see of an object is
its front surface. As a result, the epistemologist’s projection tends to obscure
or even make invisible the responsibility we have as language users.
Language itself seems to take over. It appears as if all we need to do is follow
it in its tracks.
It is here, and not in the mere fact that the epistemologist’s usage deviates
from ordinary language, that Cavell locates the real trouble with the
epistemologist’s way of reasoning. According to Cavell, the epistemologist
does not realize that his projection assigns a purely schematic role to
familiar expressions. The epistemologist’s conclusions are moves within the
invoked model. Considered merely as such internal moves, they are indeed
unobjectionable. The problem, however, is that the conclusions are
conceived as substantial claims: schematic validity gets confused with
general application. As Cavell puts it, the epistemologist fails to see that
‘‘what his conclusions find in the world is something he himself has put
there, an invention, and would not exist but for his efforts’’ (CR, p. 223).
As I have emphasized, Cavell thinks this kind of self-delusion constitutes
a permanent temptation. He thinks we are persistently attracted to the
supposition that if the world can be adequately described at all, the
description has to be made from a position where we have, as it were,
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 379

purified our concepts so that they are no longer tainted by human life. The
struggle with this temptation is what Cavell calls the struggle with
skepticism. He explicitly says this struggle is ‘‘endless.’’19 Our question is:
what are his reasons for saying so? Why, according to Cavell, can’t we put
an end to the struggle, simply by taking Wittgenstein’s vision to heart and,
in McDowell’s already quoted words, ‘‘give up the idea that philosophical
thought […] should be undertaken at some external standpoint, outside our
immersion in our familiar forms of life?’’
To begin with, Cavell thinks it cannot be made fully clear, beforehand and
at a general level, what it is that we are supposed to give up here. What can
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

be learnt from Wittgenstein, he argues, is that the line between what is

‘‘internal’’ to a practice and what is ‘‘external’’ to it cannot be drawn pre-
emptively, as it were. If, in a suitable situation, I say, ‘‘feed his pride’’, I am
still ‘‘inside’’ our practice with the word ‘‘feed.’’ My peers and I sense the
significant continuity between this and earlier uses of the word. But if there
is someone who does not sense this, there is no independent and
authoritative way of demonstrating to him that he is wrong. Conversely,
if someone claims to sense a significant continuity that we fail to see, there
may be no agreed-upon procedure by means of which the issue can be
resolved. The notion of giving up, once and for all, the idea that
philosophical thought should be undertaken at some external stand-
point—not to say the notion of being protected against the terror associated
with this idea—seems to make sense only on the assumption that we can
decide in advance, and with respect to any case of metaphysical perplexity,
what projections are tolerable. According to Cavell, what Wittgenstein
makes clear is precisely that the idea of such a pre-emptive decision is
incomprehensible.
This is related to Cavell’s worry about the tendency, in the writings of
Austin and other ordinary language philosophers, to argue as if certain ways
of talking constitute demonstrably inappropriate deviations from estab-
lished usage. The deep problem with this attitude, Cavell argues, is that the
method of ordinary language philosophy, properly conceived, leaves no
room for such straightforward dismissal. As Cavell says, the vision of
language that should inform the practice of ordinary language philosophers
‘‘must itself prevent flat repudiation’’ (CR, p. 192). Such flat repudiation
would be possible if the disagreement between the ordinary language
philosopher and his interlocutor were analogous to the disagreement over
whether a certain paint-maker calls the color of a certain wall ‘‘Apple
Green’’ or ‘‘Pear Green.’’ But, as I have already emphasized, Cavell’s point
is that the disagreement is of a different sort—akin to, though not identical
with, the more radical sort of breakdown that I talked much about in the
first half of this paper. Restoring consensus, if possible at all, cannot in such
cases be a matter of straightforward proof or demonstration. For the
disagreement goes too deep: demonstration, in the relevant sense of the
 380 M. Gustafsson

word, requires a consensus that, in this sort of case, cannot be taken for
granted without begging the question.
This is not to say that one cannot have a meaningful discussion about the
appropriateness of a projection. In some cases, as in ‘‘feed his pride’’, there
is spontaneous agreement—or ‘‘attunement’’, to use the term Cavell prefers
in this connection. In other cases, however, the agreement or attunement is
less immediate. As Cavell notes, ‘‘a new projection, though not at first
obviously appropriate, may be made appropriate by giving relevant
explanations of how it is to be taken, how the new context is an instance
of the old concept’’ (CR, p. 192). In fact, this should make it even clearer
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

that the idea of deciding beforehand which projections are tolerable and
which are intolerable is absurd. Explanations of the sort Cavell is talking
about are given, for example, when an art form is renewed, or when a
scientific revolution is instigated. It is of the essence of such renewals and
revolutions that their character and appropriateness is unforeseeable. We
need to consider each such case individually, when we are actually
confronted with it. Before such actual confrontation, it is indeterminate
what the adequate diagnosis of and response to the projection will be.20
Consider the following example. A director claims to have made a film
noir. Watching the film makes us puzzled and suspicious, however. For, in
many seemingly important respects, this film seems quite different from
classical film noir. Either it constitutes a radical renewal of the genre, or it is
not a film noir at all—and we suspect the latter. Clearly, deciding the issue
cannot just be a matter of applying pre-established rules or definitions or
criteria. To be sure, many film guides and cinematographic textbooks list
criteria for what constitutes a film noir, but one thing that characterizes the
imagined situation is that such official criteria are useless, since the director
clearly intends to challenge their purported accuracy. The way to resolve our
puzzlement is, rather, to engage in a careful reexamination of what it is that
we want to say when we classify a film as a film noir. Perhaps a sensitive
critic can show us that our spontaneous reaction (to say that the film is not a
film noir at all) is in fact in tension with what we are able to recognize as the
most profound characteristics of classical film noir when we watch them
carefully and perceptively. By reminding us of those characteristics and
experiences, and by making careful comparisons between paradigm
examples of the genre and the new film, such a critic may be able to
identify surprising continuities. Taken together, such comparisons and
reminders may make it clear to us that we should acknowledge that the critic
is right: the film is, indeed, a film noir. It may of course also happen that the
critic’s explanations do not work, but betrays his own superficial grasp of
the genre. In such a case, we might want to conclude that our initial refusal
to count the movie as a film noir was perfectly sound.
There are many other cases similar to this one. In his early essay, ‘‘Music
discomposed’’ (MWM, pp. 180–212), Cavell expresses his misgivings about
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 381

the ways in which the terms ‘‘composition’’, ‘‘improvisation’’ and ‘‘chance’’

are used in the writings of John Cage, Ernst Krenek, and others. Part of
Cavell’s method is to remind his readers of facts they should be able to
recognize as quite familiar. He points out that ‘‘[i]t is, obviously enough,
within contexts fully defined by shared formulas that the possibility of full,
explicit improvisation traditionally exists’’ (MWM, p. 201). He notes that if
we see someone being interested and becoming absorbed in a pin, or a
crumpled handkerchief, ‘‘[t]he situation demands an explanation, the way
watching someone listening intently to Mozart, or working a puzzle […]
does not’’ (MWM, p. 197). And so on and so forth. By putting such
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

reminders together in the right sort of way, Cavell makes us think again, and
think hard, about what it is that we want to say when we talk about
compositions and improvisations and chance. Clearly, he wants to
encourage the suspicion that the theories behind some of Krenek’s and
Cage’s works are problematic. However, he is careful to point out that his
aim is not to prove, for example, that those works are not musical
compositions. Indeed, it is totally unclear what proving such a thing would
amount to (MWM, p. 205). Rather, the aim of his procedure is restricted to
that of urging us to reconsider what is involved, and what we take to be of
value, in that sensitivity and in those modes of response by which we give
music the place it has in our lives.
According to Cavell, the aim of ordinary language philosophy is, or
should be, similar. The disagreement between the ordinary language
philosopher and his more traditional interlocutor is not resolvable by
reference to rules or definitions. On the other hand, this disagreement is
different from a disintegration like that described at the beginning of this
paper, of the practice where all participants have perfect pitch. With respect
to such disintegration, it makes little sense to try to revive the practice by
entering a dialogue with oneself or with one’s peers. If our perfect pitch is
gone, then it is gone; verbal exchange will not make it come back. By
contrast, practicing the method of ordinary language philosophy, as Cavell
conceives it, means keeping the hope for agreement alive. In this sort of case,
entering a dialogue is meaningful. Even if there is no guarantee that a
resolution will be found, it makes sense to strive for it, and it is the ordinary
language philosopher’s job to do so.21
One way of summarizing Cavell’s dissatisfaction with Austin’s methodo-
logical self-understanding is to say that Austin does not take seriously
enough the need for a genuine dialogue in philosophy. He has a wonderfully
sensitive ear for the nuances of everyday language, but his impatience with
those who do not see those nuances or do not find them illuminating
displays a lack of understanding of, or perhaps of interest in, what it is that
pulls those unappreciative interlocutors away from the ordinary. Austin
often argues as if the difference between him and his interlocutor is,
basically, just a matter of discrimination—as if the perceptivity of his
 382 M. Gustafsson

interlocutor were somehow impaired, in contrast to his own perfect pitch, so

to speak. This, however, indicates that Austin fails to understand the
significance of the fact that the projections that his interlocutor is suggesting
are, as Cavell puts it, ‘‘not obviously inappropriate.’’ According to Cavell,
that a projection is ‘‘not obviously inappropriate’’ does not just mean that
its inappropriateness is difficult to perceive. Rather, it means that the
inappropriateness cannot be perceived—or, better, that the charge of
inappropriateness makes no clear sense—unless the interlocutor is given the
opportunity to explain how his suggested projection is to be taken, how it is
supposed to be an instance of the old concept. Not giving him this
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

opportunity amounts to ordinary language dogmatism. It signals a failure to

differentiate between philosophical confusion and mere linguistic deviance.
It is only by asking the interlocutor to clarify what it is that he wants to say
that we can put the adequate sort of pressure on his conception. It is only in
relation to the answers that he then provides that we can be said to have
revealed that his purported projection does not put language to new and
interesting work, but on the contrary, lets it go on holiday.
One thing that is likely to make an investigation of this sort particularly
difficult is the sense of inevitability that characterizes the interlocutor’s
relation to his own projection. As I said before, Cavell thinks such a
person’s dissatisfaction with established practice is internally related to his
feeling that the model or scheme that he invokes is somehow made necessary
by the concepts themselves. This brings us to another aspect of Cavell’s
qualms about Austin, namely what he regards as Austin’s failure to see the
methodological importance of this sense of inevitability. Consider Austin’s
tendency to accuse his philosophical foes of mistakes. Cavell objects that
accusations of being mistaken are usually misplaced in philosophy, since ‘‘it
seems […] that ‘mistaken’ requires the idea of a wrong alternative (either
taking one thing for another, or taking one tack rather than another)’’
(MWM, p. 107). The problem is that this notion of ‘‘wrong alternative’’
suggests that the ‘‘correct alternative’’ is somehow open even to someone
who is held captive by a philosophical scheme. This, in turn, makes it seem
as if such a person is philosophically confused because he has been
overhasty, sloppy, ignorant, stupid, or suchlike. And indeed, Austin often
uses such terms to account for what he takes to be the errors of his
opponents. But, as Cavell points out, accusations of this sort are, at best,
artificial: philosophers worth taking seriously are not hasty, sloppy,
ignorant or stupid in any reasonable sense of those words. Philosophical
confusions go deeper, and should be characterized in ways that do justice to
their irresistable appeal. Once you have come under the spell of a
philosophical scheme or model, there seems to be no alternative. The
scheme tells you how the concepts must work. And it is precisely this
apparent capacity to annul our options (and, thus, our responsibility) that
makes the scheme so attractive.
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 383

Austinian examples may serve to break the spell of such a scheme. But
there is no guarantee that this will happen. Sometimes some people will
experience the examples as immensely powerful and significant. A more
typical reaction, however, is to say that they are irrelevant. The interlocutor
may admit that the examples remind him of how we talk in real life. But
then, he might add that this established way of talking is of little
philosophical significance. He explains away the reminders. This may be
done in various ways. In a Humean fashion, the interlocutor may argue that
the practical demands of everyday life force us to talk with the vulgar, even
if this means saying things that are strictly speaking false. Alternatively, he
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

may decide that what does not fit his model lies outside the realm of
conceptual content altogether: he may classify it as imprecision due to
practical demands or limitations, as a merely ‘‘pragmatic’’ aspect of
everyday usage, or whatever. Since we are here moving at a level at which
talk about language and talk about reality cannot be neatly separated, such
explaining away constitutes a disapproval, not just of everyday language but
of everyday reality: our life-world is reduced to mere ‘‘appearance.’’ It is
part of being held captive by a scheme that one is good at finding such
explanations, explanations that seem able to disarm any attempt to make
what is ordinary significant for philosophical investigation.
The reminders themselves do not determine how the interlocutor will
respond. Nor do they determine how he should respond. In a sense, no
response is wrong, as long as it tells us something about how the
interlocutor wants his words to function. Indeed, this is precisely the
significance of the interlocutor’s reactions: they show us (and him) what he
wishes to say. Austinian examples are not proofs but provocations. They
bring forth responses that clarify the relation between the interlocutor and
the words he wants to use. Clarification of this relation is achieved even in
cases where the reminders are explained away. For such explaining away
also tells us something about the interlocutor’s desires and requirements.
His preferring a certain explanation may make it clearer what the structure
of his philosophical scheme is, what he expects from this scheme, and so
forth. The hope is that if we are careful and patient enough, our discussion
may reach a point at which it is becoming clear to everyone, including our
interlocutor, that the ways in which he wants to use his words do not hold
together. We state reminders; those reminders get explained away; we state
more reminders in response to the ways in which the former ones were
disarmed; and so on and so forth. Eventually, this dialectical process may
make it patent to us and to the interlocutor that his position is unstable, that
he is wavering between different and incompatible requirements. This is
what it means to show that the interlocutor is speaking nonsense, in a
philosophically pertinent sense of that word. Rather than proving that his
way of talking constitutes an inappropriate departure from everyday usage,
what we do is bring the investigation to a point where it is plain, to us and to
 384 M. Gustafsson

him, that his requirements are incompatible and nothing determinate is

being said.22

VI.
According to Cavell, Wittgenstein’s vision of language entails that we are
burdened with the responsibility of keeping language alive and the world in
view. Moreover, Cavell says we fear that responsibility: we would like to get
rid of it. At bottom, it is this fear that shows itself in our tendency to reason
as if being immersed in human life forms is an impediment to (rather than a
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

precondition for) perceiving and describing the world. In a sense, we want a

gap to open up between familiar practices and reality, a gap the bridging of
which cannot be up to us. For the existence of such a gap would seem to
mean that our having access to the world depends on there being some self-
standing mode of justification, the validity of which is independent of what
we do or who we are.
If this is right, it means that our fear of responsibility tends to conceal
itself. It tends to appear under the disguise of what seems like a very
different, and by no means irresponsible, propensity: wanting to find that
rational and self-standing mode of justification which, supposedly, makes
our descriptions and perceptions genuinely objective. One of Wittgenstein’s
most important aims, Cavell argues, is to reveal this charade.
The revelation, however, cannot be definitive. The masquerade is never
over. According to Cavell’s Wittgenstein, ‘‘there is no absolute escape from
(the threat of) illusions and the desires constructed from them, say there is
no therapy for this, in the sense of a cure for it’’; he ‘‘sees illusions of
meaning as something to which the finite creature is subject chronically,
diurnally’’ (MWM, p. xx). McDowell sometimes describes Wittgenstein’s
conception in similar terms. According to McDowell’s Wittgenstein,
permanent peace of mind is implausible since the intellectual roots of our
philosophical anxieties ‘‘are too deep for that.’’ McDowell says we should
not take Wittgenstein ‘‘to be envisaging a post-philosophical culture […]. He
is not even envisaging a future for himself in which he is definitively cured of
the philosophical impulse. The impulse finds peace only occasionally and
temporarily.’’23 But still, there remains a significant difference between
McDowell’s and Cavell’s Wittgenstein. Cavell and McDowell might agree
that the fact that I have taken Wittgenstein’s vision of language to heart and
am prepared to acknowledge my responsibility qua language user does not
mean that the temptation to disclaim that responsibility is dissolved or can
be kept under control. For Cavell, however, this is not just a plausible
conjecture. Rather, he sees it as a constitutive claim. According to Cavell’s
Wittgenstein, being burdened with responsibility for keeping language alive
is inseparable from being tempted to disclaim that responsibility: they are
two sides of the same coin.
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 385

Thus, Cavell’s Wittgenstein thinks the fact that I have become fully aware
of my responsibility in itself means that I am prepared to acknowledge both
that the temptation is still there, and that the possibility of my succumbing
to it is a real and permanent danger for which no secure protection can be
provided. If this sounds speculative or even abstruse, it may be helpful to
note that something similar may be said about various other kinds of human
interaction. For example, consider marriage. As a spouse, I am responsible
for keeping my marriage alive. Passively ignoring the required effort means
killing the relation; the marriage then continues to be a marriage only in the
formal sense of the word. And yet it is tempting to give up this responsibility
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

and behave as if the maintenance of the marriage is somehow guaranteed

even in the absence of my active and continuous devotion. Now, imagine
someone who claims to be fully aware that he is responsible for keeping his
marriage alive, but then adds that this awareness means that he no longer
feels, or that he is protected against, the temptation to passively let the
marriage take care of itself. Arguably, this by itself reveals that he is not fully
aware of his responsibility. For being thus aware involves being prepared to
acknowledge that the fear of responsibility and the associated temptation to
give up that responsibility is still present and is not under full control. What
true awareness of responsibility gives you is not protection against that kind
of fear and that sort of temptation. Rather, the awareness involves an
appreciation of the fact that the dream of such protection is itself a
manifestation of the fear against which one wants to be protected.
Cavell claims that Wittgenstein’s vision of language is terrifying. What is
the nature of the alleged terror? As long as we think of ourselves as engaged
in the search for a rational and self-standing mode of justification for our
familiar practices, what seems worth fearing is certainly not our propensity
to engage in that kind of search, but the possibility that there is no
justification of the sort we want to find. This means that we take our fear to
be directed toward something that is not, as it were, part of ourselves—
something that is independent of our predispositions and respon-
sibilities. We are not to blame if it turns out that the alleged gap between
our practices and the world is unbridgeable. The threat, it seems, comes
from without.
According to Cavell, Wittgenstein’s vision of language entails a sort of
Copernican turn precisely at this point. Cavell reads Wittgenstein as
showing us that the most profound threat does not come from without, but
from within ourselves. At the deepest level, what merits fear is our own
inclination to disclaim responsibility for the maintenance of those human
practices within which language has its life. And, again, Cavell’s
Wittgenstein offers no protection against that inclination. If Cavell is right,
what Wittgenstein’s vision should make us realize is this: we must learn to
live with our fear of responsibility, with the associated inclination to
disclaim that responsibility, and—if we are philosophically clear-sighted
 386 M. Gustafsson

enough—with a fear of that very inclination. Taking Wittgenstein’s vision of

language to heart means understanding that this is the price we must pay for
having a language at all.24

Notes
1. Cavell (1994), p. 21. All future references to Cavell (1994) are made parenthetically in the
text by means of PoP and the appropriate page number.
2. McDowell (1998), pp. 61, 207. All future references to McDowell (1998) are made
parenthetically in the text by means of MVR and the appropriate page number.
3. Cavell (2002), p. 52, note omitted. All future references to Cavell (2002) are made
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

parenthetically in the text by means of MWM and the appropriate page number.
4. Cavell (1979), p. 109. All future references to Cavell (1979) are made parenthetically in
the text by means of CR and the appropriate page number.
5. McDowell (2002), p. 294.
6. Wittgenstein (1978), paragraph VI-35.
7. Austin (1979), p. 185, n. 1.
8. Affeldt’s criticism is in Affeldt (1998).
9. Mulhall (1998), p. 40.
10. I take my point here to be closely related to what Affeldt says about Cavell’s notion of
criteria in the following passage: ‘‘To speak of our sharing an order of criteria is to speak
of an order in our judgments (and conduct). Eliciting criteria reveals the fine-grained
structure of our agreement in judgment. It does not reveal a separate order under-
girding and controlling that agreement. […] It would be a philosophical distortion of
what is involved in our being, as Cavell puts it, initiates of language, to imagine that we
somehow always have or possess (perhaps only tacitly) a catalogue of criteria which we
use in speaking and making judgments. One central reason for Cavell’s insistence that
our agreement in language and our agreement in criteria are the same phenomenon
differently described is precisely to fend off this, philosophically tempting, idea’’ (Affeldt
(1998), p. 15).
11. Perhaps the difference between the two tasks is not very different from Kant’s difference
between formal and transcendental logic: whereas the former takes the application of
rules for granted, the latter constitutes an investigation into the conditions for the
possibility of such application (Cf. MWM, p. 168).
12. One might want to argue that the term ‘‘objectivity’’ is misplaced here, since this term is
too ingrained with philosophical prejudices of the sort which both McDowell and Cavell
are concerned to reject. But McDowell would disagree. He freely admits that the notion
of objectivity involves the idea of inquiry as answerable to something other than
ourselves, but argues that this is a perfectly intelligible idea that does substantial work
within our familiar practices. What needs to be rejected, according to McDowell, is not
the familiar notion of objectivity, but the idea of a gap between the world and us. (See,
for example, McDowell (2000), pp. 110–111.) I take Cavell to ascribe a similar view to
Wittgenstein, for example in the following remark which alludes to Kant’s conception of
things-in-themselves: ‘‘For Wittgenstein it would be an illusion not only that we do
know things-in-themselves, but equally an illusion that we do not (crudely, because the
concept of ‘knowing something as it really is’ is being used without a clear sense, apart
from its ordinary language game)’’ (MWM, p. 65). As the parenthetical remark makes
clear, the point here is not that Wittgenstein takes the concept of ‘‘knowing something as
it really is’’ and its cognates to be somehow illegitimate. Rather, Cavell is arguing that
Wittgenstein thinks those concepts have familiar uses, uses that are perfectly all right.
What Wittgenstein is said to react against is the philosophical attempt to remove those
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 387

concepts from that stream of life within which they do their work. According to Cavell,
it is precisely that sort of attempt which makes it seem as if a gap opens up between
familiar practices and reality.
13. McDowell (1996), p. 70.
14. What follows is a free rendering of the discussions in CR, pp. 197ff., and MWM, pp.
249ff. Thanks to Stina Bäckström for a good conversation about these passages, and to
David Finkelstein for a stimulating discussion about the philosophical significance of
ball shells.
15. I am disregarding the case when the object is made of some transparent material, such as
glass. In this sort of case, the scheme, as specified above, does allow us to say that we
‘‘see the whole object’’ even if the object consists of something more than a front surface.
In fact, reflecting on the phenomenon of transparency may have far-reaching
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

consequences for the epistemologist, and cause considerable modifications of his

original scheme. For example, he might conclude that the very idea of a third dimension
is foreign to our visual sense, arguing, perhaps, that before they are conceptualized by us
visual impressions are inherently two-dimensional. It is not necessary to consider these
complications in further detail here, however.
16. Of course, getting to see the backside of something might in some cases be quite
cumbersome. Consider the moon.
17. The epistemologist might try to leave room for the fact that we can move around in the
world by saying that what establishes the identity of the parts of an object is not just
their position relative to an observer, but also certain other properties that are
independent of the geometry of the situation. That would seem to allow for the
possibility of seeing new parts of an object: even if I always see a front surface, the front
surface I now see might be distinguished from the one I saw, say, two seconds ago. This,
however, has another weird consequence. For let us suppose that the front surface I now
see is not the same as the one I saw two seconds ago. Given this assumption, it seems
difficult to avoid the conclusion that the front surface I saw two seconds ago no longer
exists, and, conversely, that the front surface that I am seeing right now did not exist two
seconds ago. For everything that exists now differs from the front surface I saw two
seconds ago in ways that are relevant to its identity: either with respect to its position
relative to me, or with respect to some of the other criteria of identity. After all, the only
thing that now occupies the same position as the front surface I saw two seconds ago is
the front surface that I am now seeing, and we were assuming that these two surfaces are
not identical. Analogously, the front surface I now see did not exist two seconds ago,
since what occupied its position relative to me was the front surface I then saw, and,
again, the two surfaces were assumed not to be the same. (Cf. CR, p. 202.)
18. For a penetrating discussion of Cavell’s conception of the point of an utterance, see Baz
(2003).
19. Cavell (1989), p. 57.
20. I am presupposing that the aim of serious revolutionaries within art or science is not
limited to the mere overthrowing of old conventions (if their aim were thus limited, their
endeavors should not be characterized as projections at all.) According to Cavell, what a
serious revolutionary wants to do is often to regain the original significance of an art
form or a science. In the eyes of such a revolutionary, the reason for replacing old
conventions is that, under new circumstances, they have come to obfuscate rather than
promote this original purpose. Cavell says, ‘‘deep revolutionary changes can result from
attempts to conserve a project, to take it back to its idea, keep it in touch with its history.
[…] It is because certain human beings crave the conservation of their art that they seek
to discover how, under altered circumstances, paintings and pieces of music can still be
made, and hence revolutionize their art beyond the recognition of many. This is how, in
my illiteracy, I read Thomas Kuhn’s The Structure of Scientific Revolutions: that only a
 388 M. Gustafsson

master of the science can accept a revolutionary change as a natural extension of that
science; and that he accepts it, or proposes it, in order to maintain touch with the idea of
that science, with its internal canons of comprehensibility and comprehensiveness, as if
against the vision that, under altered circumstances, the normal progress of explanation
and exception no longer seem to him to be science’’ (CR, p. 121).
21. What I am saying here and in the rest of section 5 is meant to address the worry that
ordinary language philosophy, as conceived by Cavell, must be impotent as a mode of
criticism. This worry is perhaps reinforced by a misunderstanding of the intimate
connection between, on the one hand, the impossibility of flat repudiation, and on the
other, the first person plural character of the investigations of ordinary language
philosophy. Cavell writes, ‘‘the way you must rely upon yourself as a source of what is
said when, demands that you grant full title to others as sources of that data—not out of
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

politeness, but because the nature of the claim you make for yourself is repudiated
without that acknowledgement: it is a claim that no one knows better than you whether
and when a thing is said, and if this is not to be taken as a claim to expertise (a way of
taking it which repudiates it) then it must be understood to mean that you know no
better than others what you claim to know. With respect to the data of philosophy our
positions are the same’’ (MWM, pp. 239–40). Isn’t the consequence that the conflict
between the ordinary language philosopher and his interlocutor is exactly like the earlier
imagined conflict between people who used to have perfect pitch but who no longer
agree in their identification of pitches? If so, it seems as if all the ordinary language
philosopher can do is to insist that his own ‘‘everyday’’ way of talking is the appropriate
one. And that would not convince his opponent, who we may expect to be an equally
staunch defender of ‘‘philosophical’’ usage. According to the sort of reading I present
here, however, a main purpose of Cavell’s criticism of Austin and other ordinary
language philosophers is precisely to show them a way out of this kind of stalemate.
According to Cavell, to argue that ‘‘philosophical’’ usage constitutes a demonstrably
inappropriate deviation from everyday language is dangerous, precisely because the real
consequence of such belligerence is impotence: the ordinary language philosopher and
his opponent will be talking past one another, and each one will be as convinced as
before that he is (demonstrably) right. (See, for example, CR, p. 146.)
22. For an illuminating discussion of what Cavell thinks it means to lapse into
meaninglessness, see Witherspoon (2002).
23. McDowell (1996), p. 177.
24. Earlier versions of this paper were presented at workshops at Uppsala University and at
The University of Chicago. I thank the audiences for thoughtful discussions. In
particular, I wish to thank Zed Adams, Steven Affeldt, Avner Baz, Stina Bäckström,
Stanley Cavell, James Conant, David Finkelstein, Stephen Mulhall and Lisa Van
Alstyne for helpful comments and criticisms. I am, of course, solely responsible for any
remaining mistakes and misinterpretations. Work on the paper was financed by The
Swedish Research Council.

References
Affeldt, S. (1998) ‘‘The ground of mutuality: Criteria, judgment, and intelligibility in Stephen
Mulhall and Stanley Cavell’’, European Journal of Philosophy, 6, pp. 1–31.
Austin, J. L. (1979) Philosophical Papers, 3rd ed (New York: Oxford University Press).
Baz, A. (2003) ‘‘On when words are called for: Cavell, McDowell, and the wording of the
world’’, Inquiry, 46, pp. 473–500.
Cavell, S. (1979) The Claim of Reason (New York: Oxford University Press).
Cavell, S. (1989) This New Yet Unapproachable America (Albuquerque, NM: Living Batch
Press).
 Perfect Pitch and Austinian Examples: Cavell, McDowell, Wittgenstein 389

Cavell, S. (1994) A Pitch of Philosophy: Autobiographical Exercises (Cambridge, MA: Harvard

University Press).
Cavell, S. (2002) Must We Mean What We Say? Updated edition (New York: Cambridge
University Press).
McDowell, J. (1996) Mind and World. With a New Introduction by the Author (Cambridge, MA:
Harvard University Press).
McDowell, J. (1998) Mind, Value, and Reality (Cambridge, MA: Harvard University Press).
McDowell, J. (2000) ‘‘Towards rehabilitating objectivity’’, In: R. B. Brandom (Ed.), Rorty and
His Critics (Malden, MA: Blackwell). pp. 109–123.
McDowell, J. (2002) ‘‘Responses’’, in: N. H. Smith (Ed.), Reading McDowell. On Mind and
World (New York: Routledge). pp. 269–305.
Mulhall, S. (1994) Stanley Cavell: Philosophy’s Recounting of the Ordinary (New York: Oxford
Downloaded by [Ohio State University Libraries] at 15:19 03 June 2012

University Press).
Mulhall, S. (1998) ‘‘The givenness of grammar: A reply to Steven Affeldt’’, European Journal of
Philosophy, 6, pp. 32–44.
Witherspoon, E. (2002) ‘‘Houses, flowers, and frameworks: Cavell and Mulhall on the moral of
scepticism’’, European Journal of Philosophy, 10, pp. 195–208.
Wittgenstein, L. (1978) Remarks on the Foundations of Mathematics, 3rd ed (Oxford: Blackwell).