SIMON BLACKBURN
THE INDIVIDUAL STRIKES BACK
In this paper I address some of the points Saul Kripke makes in his
treatment of the 'rule-following considerations' in the later Wittgenstein.
I There are two different quarries to track down. There is the
question of whether Kripke's exegesis of Wittgenstein is correct whether
KW is LW. And there is the distinct question of the real
significance of the considerations, as they are put forward by KW.
Kripke himself is carefully agnostic about this second issue.2 KW is not
Kripke in propria persona. And Kripke is also careful to distinguish the
exegetical issue from the question of significance. The two issues only
connect like this. If KW's arguments have some property which we are
convinced cannot belong to any argument which LW would have used,
we shall suppose that KW is not the real LW. And admirers of
Wittgenstein will suppose that significance is such a property: if KW's
considerations are faulty, then for that reason alone KW cannot be LW.
LW would not have used faulty considerations. But without commenting
on this optimism, I want to discuss two other properties which might
distinguish KW from LW: KW's use of scepticism, and his attitude to
facts. Each of these aspects may legitimately raise worries about his
identity with LW. And I shall also offer some thoughts about the
significance of KW's arguments for our conceptions of meaning.
Our topic is the fact that terms of a language are governed by rules
that determine what constitutes correct and incorrect application of
them. I intend no particular theoretical implications by talking of rules
here. The topic is that there is such a thing as the correct and incorrect
application of a term, and to say that there is such a thing is no more
than to say that there is truth and falsity. I shall talk indifferently of
there being correctness and incorrectness, of words being rule-governed,
and of their obeying principles of application. Whatever this is, it
is the fact that distinguishes the production of terms from mere noise,
and turns utterance into assertion - into the making of judgement. It is
Synthese 58 (1984) 281-301. 0039-7857/84/0583-0281 $02.10
 1984 by D. Reidel PublishingCompany
282 SIMON BLACKBURN
not seriously open to a philosopher to deny that, in this minimal sense,
there is such a thing as correctness and incorrectness. 3 The entire
question is the conception of it that we have.
KW pursues this issue by advancing a certain kind of scepticism. As
Kripke is well aware, this might provoke immediate protest. Surely LW
is consistently scornful of scepticism? But this reaction misunderstands
the function of scepticism in a context such as this. The function is not
to promote a conclusion about knowledge or certainty, but to force a
reconsideration of the metaphysics of the issue. That is, we begin with a
commonsense or unrefined conception of some kind of fact. We think
we have an understanding of what that kind of fact consists in. The
sceptic tries to show that on that conception we would have no way of
distinguishing occasions when the fact obtains, from those when it does
not. Now we might conclude from this that our conception was correct
and that there is therefore a definite kind of proposition whose truth
value we can never reliably judge. This would be a traditional sceptical
conclusion. But we might alternatively conclude that since we do know
the truth about the kind of thing in question, the conception was at
fault. The things we know do not have the kind of truth condition we
took them to have; the facts are not quite the kind of thing we took
them to be. This is a metaphysical conclusion, and the sceptical
dialogue is an instrument for reaching it. LW may have had no time for
sceptical conclusions. But he may well have had thought processes
which can be revealed by using a sceptical instrument to reach a
metaphysical conclusion. This is what KW does. So, as far as the issue
of using a sceptical weapon goes, KW may well be LW.
It is clear that on this account there are two parts to the business;
attacking the old conception, and producing a replacement. The
negative part might be successful, although the positive part is not. If
LW was successful, then in the positive part he produced a conception
on which public rule-following (in some sense of public) is possible,
whereas private rule-following is not. I shall be arguing that KW is not
successful in doing this. He does not succeed in describing what it is for
there to be a rule in force, with the property that this can obtain in a
public case, but not in the private case.
1. SCEPTICAL SOLUTIONS
Kripke describes KW as adopting a 'sceptical solution' to the sceptical
THE INDIVIDUAL STRIKES BACK 283
considerations, modelled upon Hume's sceptical solution to his own
doubts. To assess this idea it is important to separate various strands in
Hume's extremely complex position. If KW is understood to be taking
over the wrong parts of Hume, he may too easily be rejected as a
pretender. I suspect that parts of Kripke's presentation will encourage
this - particularly those where he talks of the sceptical solution.
Hume calls Section V of the First Enquiry the 'Sceptical Solution of
these Doubts'. The doubts in question were introduced in Section IV:
they concern operations of the understanding. In particular they
concern our ability to reason a priori about what must cause what, and
the impossibility of justifying inductive reasoning. The sceptical solution
of Section V consists in denying that processes of reasoning have
the power and the place hitherto assigned to them. They are replaced
by processes of custom. This is why we have a sceptical solution: Hume
offers a view of ourselves which in part he shares with traditional
sceptics. The shared part is the denial that we can justifiably reason to
our beliefs. (Hume differs from tradition in his estimate of the consequences
of that.) When Kripke introduces the analogy with Hume,
this is what he first mentions. But this is not the aspect of Hume (nor the
section of the Enquiry) that actually matters. What matters is his
reinterpretation of the concept of causation - the topic of Section VII
of the Enquiry.4 It is here that Hume has a (fairly) pure example of the
process I described: a sceptical argument forcing us to revise our
conception of 'a kind of fact. It is here that he parallels KW. But the
reinterpretation does not deserve to be called a 'sceptical solution' to
anything, nor did Hume so call it. It is at most a proposal prompted by
sceptical problems. But in principle it might have been prompted by other
considerations altogether. And in fact Hume's reinterpretation of causation
is only partly motivated by scepticism. It is at least as firmly seated
in the theory of understanding: problems with our Idea of the causal
nexus. In this Dart of Hume, scepticism is subsidiary, even as a tool.
Kripke can rejoin that Hume's reinterpretation of causation is, in
itself, deservedly called sceptical for a further reason. It denies that
there is a 'facť whether there is a causal connexion between two events.
At least, that is how Kripke takes Hume. 5 But here too there are
subtleties in the offing, and they matter to the parallel with KW. This is
because LW's attitude about 'facts' is going to be crucial, and crucial to
many philosophers' belief that KW differs from him. The philosopher
we have described ends up reinterpreting some kind of fact. This leaves
284 SIMON BLACKBURN
various options. The first might be called 'lowering the truth-condition'.
This asserts that sentences in the area can only, legitimately, be given
such-and-such a truth-condition. This can be combined with the view
that in our ordinary thought we confusedly attempt to do more, or
misunderstand what we are actually doing enough to make mistakes
(although there is always a problem about how we can attempt to do
more, if we are supposed to have no conception of anything more to
do). Lowering the truth condition is then a reforming view, and entails
an 'error' theory of ordinary thought. But it can be combined with the
view that the lowering really reveals what we meant all along, and we
have reductionism. Often it does not matter very much which combination
is offered, and indeed, since a decision depends on a fine
detection of ordinary meaning, we would expect some degree of
indeterminacy.
Quite distinct from lowering the truth-condition, there is the option
of denying one altogether. The sentences in question are given some
other role than that of asserting that some fact obtains. This option is
familiar from expressive theories of moral commitment, or from views
that try to see arithmetical theses as rules rather than descriptions or
propositions, and so on. Now I say that this option is distinct from
lowering the truth-condition, and indeed in its initial stages it certainly
is. It is a confusion, for instance, to muddle together expressive theories
of ethics with naive subjectiv~theories that give moral commitments a
truth-condition, but make it into one about the speaker. Arguments
against this latter view often have no force against the former. But
Hume is responsible for the very complication that makes it so hard to
keep these options properly separated. This is the view that the mind
spreads itself on the world: the view I call projectivism. According to
projectivism we speak and think 'as if' the world contained a certain
kind of fact, whereas the true explanation of what we are doing is that
we have certain reactions, habits or sentiments, which we voice and
discuss by such talk. Hume was quite clearly a projectivist in moral
philosophy, and it is plausible to see his metaphysics of causation as in
essence identical.6
Like the option of lowering the truth-condition, the option of denying
that there is one can be combined with either of two attitudes to our
ordinary practice. One would be that it embodies error. Ordinary talk is
conducted as if there were facts, when there are no such facts. The talk
is 'fraudulenť or 'diseaseď. The other option is less familiar, but much
THE INDIVIDUAL STRIKES BACK 285
more attractive. It holds that there is nothing illegitimate in our
ordinary practice and thought. The respects in which we talk as if there
are, for instance, moral facts, are legitimate. (I have called this view
'quasi-realism'.) 7 If LW were a projectivist, he would have to be this
kind, for it preserves the doctrine that ordinary talk and thought, before
we start to philosophize about the nature of our concepts, are in perfect
order as they are.
Kripke acknowledges the strand in LW which matters here. He also
realizes that it goes far - far enough to stop LW from endorsing any
such judgment as this: 'there is no fact of the matter whether a term is
rule-governed or noť. But KW expresses himself differently: KW
believes that there is 'no such fact, no such condition in either the "internal"
or the "external" worlď (p. 69). This talk will outrage friends of
LW. KW denies the existence of a certain kind of fact altogether; LW
would never so express himself, ergo one is not the other. But now recall
that Hume himself says that 'those who deny the reality of moral
distinctions may be reckoned amongst the most disingenuous of disputants'.
8 Why? Because insisting on this reality is part of normal
thinking. It is part of the way of life, or way of thought or talk, which
quasi-realism can protect for us. So there must be room for a different
version of LW. This one would abandon his hostility to facts. He should
accept that talking of facts is part of our legitimate way of expressing
ourselves on the difference between terms which are rule-governed,
and terms which are not. He has to do this if he doubles for LW. For
recall that the very passage (Investigations 137) that begins these
considerations is a version of the redundancy theory of truth. And on
that theory (whatever else it holds) there is no difference between
saying that it is true that p and that p or between saying that it is a fact
that p and that p. So anybody prepared to assert that terms are
rule-governed - and as I explained initially, that must mean all of us can
equally be heard saying that there are facts, truths, states of affairs
of just that kind.
Kripke says something strange about this. On p. 86 he imagines
someone saying that it cannot be tolerable to concede to the sceptic
that there are no truth-conditions or ~corresponding facts to make a
statement about someone's meaning true: 'Can we not with propriety
precede such assertions with "It is a fact that" or "It is not a fact that"?'
But Kripke puts this complaint in the mouth of an objector to LW. He
then reminds us that LW accepts the redundancy theory of truth, and
286 SIMON BLACKBURN
says that this gives him a short way with such objections. The dialectic
seems to be the wrong way round. Because he accepts the redundancy
theory, LW can assent to the locutions of fact and truth. They add
nothing to the fact that we judge other people to be following rules, and
applying terms in accordance with standards of correctness.
I don't know that the redundancy theory of truth should have this
soothing power. It may be that there are explanations of why we make
some judgements, and why we apply the calculus of truth-functions to
the judgements we make, which still leave us queasy when we consider
if there are facts in the case. The quasi-realist construction of ordinary
moral discourse is a good example. Even if it protects everything we do,
it can still leave us uneasy when we contemplate the question: 'Yes, but
are there any moral facts on this account?' I shall not consider that
question further in this paper. I just want to note that on a redundancy
theory, it is a bad place to become puzzled, and that LW held a
redundancy theory. Hence it seems that there is space for a persona who
profits from KW's arguments, but draws a rather different conclusion
from them. With due humility, I shall call this character BW. He is
going to share a great deal of the argumentative strategy of KW. In part
his difference is relatively cosmetic: the belief that scepticism is only in
play, if at all, as an instrument, and that the eventual conception of rulefollowing
that must emerg6 does not deserve to be called sceptical. But
in part he differs more substantially. Because he hopes to preserve the
implications of the redundancy theory in LW. And he hopes to cement
a more particular relationship between LW and the real Hume, who
matters to metaphysics. If he can preserve our right to talk of the fact
that words obey principles of application, this may be as much because
he forces us to revise our conception of a fact as anything else. Alas,
however, BW has one radical flaw. On his philosophy, there is no
particular reason to discriminate against the would-be private linguist.
But before advertising his intended end-point further, it is necessary to
review some of the arguments by which he gets there.
2. THE CRITIQUE OF RULE-FOLLOWING
It may seem outrageous to the touchy friends of LW that anything in
Hume could be a model for LW's attitude to rule-following. But
perhaps we can stem some of the outrage by reminding ourselves that it
is an essentially normative judgement that we are chasing. It is the
THE INDIVIDUAL STRIKES BACK 287
judgement that something is correct or incorrect. When this fact proves
fugitive, as KW shows that it is, its normative nature is largely the
problem. So it cannot be that outrageous to apply our best explanations
of value judgements to it, and this is what BW hopes to do. But why is
the fact fugitive in the first place?
In Kripke's development, we start by considering the understanding I
had of some term at some time in the past, which we can arbitrarily call
'yesterday'. Suppose the term is some arithmetical functor, 'plus' or
'+'. I understood it by grasping a principle of application. We then
consider my position today. When I come to do a calculation, which we
suppose I have never done before, I certainly believe myself to follow a
principle. I believe myself to be faithful to yesterday's rule, by adopting'
the same procedure or principle in determining answers to problems
expressed using the functor. Thus I believe that if I am faithful to
yesterday's principle, I should say '57 + 68 is 125'. Notice that this is not
quite the same as claiming that I should say this unconditionally. I may
wish to change my allegiances. There is no impropriety in deciding that
yesterday's principle of application is not the best one for this term, and
to consciously start to use it according to different rules. It is just that if
I am faithful, then I ought to give that answer. I most certainly should
not say that 57 + 68 is 5. Nor of course should I say that there is more
than one answer to that problem, or that the problem is indeterminate,
so that there is no answer at all. The sceptical dialogue then commences.
The sceptic asks me to point to the fact that I am being faithful
to yesterday's rule only by saying one thing, and not these others. And
this proves hard to do. For any fact which I tell him about myself
seems compatible with the 'bent-rule' hypothesis, as I shall call it: the
hypothesis that the rule that was really in force yesterday was a
Goodman competitor. In other words, the kinds of fact I am apt to
allege are compatible with a story in which I really understood by 'plus'
a function with a particular singularity at x = 57 and y = 68. For
example, I might yesterday have meant that we should now express by
saying:
x + y = the sum of x, y, except when x = 57 and y = 68 and
5 otherwise.
This hypothesis is not refuted by my present staunch denial that I had
any such thought in mind yesterday, or that if asked I would have used
words like these to explain myself. For the sceptic will urge that as well
288 SIMON BLACKBURN
as having had a bent interpretation of' +' yesterday, I could have had a
bent interpretation of these other terms as well. So pointing out that I
would have presented no such explanation to myself does not refute the
sceptic. It would really be a question of my using another rule to
interpret the first (e.g., the rule for interpreting various synonyms for
' +' or for interpreting terms that would have occurred in any explanation
of the functor that I would have proffered). So the fact that
yesterday I would have said, for instance 'x + y is always a number
greater than either x or y, with x = 57 and y = 68 no exception' is
consistent with my having been a secret bent rule user. The singularity,
by current lights, would have been there in the way ! took the words
involved in such an affirmation. Perhaps 'exception' meant...
This argument (which is much more forcefully and thoroughly
presented in Kripke) undoubtedly corresponds to a central negative
point of LW's. The point is that taking a term in a certain way is
something different from presenting anything as an aid to understanding
it, or from accepting anything as aids to understanding it. He says in
"201 that he has shown 'that there is a way of grasping a rule which is
not an interpretation but which is exhibited in what we call "obeying the
rule" and "going against it"'! The negative point, that we gain no
approach to the required fact by embarking on a potential regress of
interpretations, is quite clear. The more positive claim that the fact is
exhibited in what we call obeying a rule, must wait.
When presenting the sceptical challenge this way, we should not lose
sight of the fact that the case can be made without instancing a rival
principle of application, a bent rule, at all. A sceptic might just doubt
whether there was, yesterday or today, any principle at all behind my
application of '+'. Perhaps all that happened was that I would look at
things, such as triples of numbers, and after a process that was
phenomenologically just like one of being guided by a rule, declare
'z = x + y' or the reverse. I would be in the same case as a lunatic who
thinks he is doing sums, when all that is happening is that he is covering
pages with symbols. Or, I would be like the man whom Wittgenstein
considers at *237 who with great deliberation follows a line with one leg
of a pair of dividers, and lets the other leg trace a path, but one whose
distance from the original line he varies in an intent but apparently
random way. He might think he is tracing a path determined by a rule
relating its course to the first line. But his thought that this is what he is
doing does not make it true. In some ways this is the primary weapon
THE INDIVIDUAL STRIKES BACK 289
that LW uses against the private linguist. He forces him back to saying
that he has only his own conviction that he is following a rule at all, and
this private, phenomenological conviction that one is following a rule is
not enough to make it true.
I have followed Kripke in concentrating upon the normative aspect
of the fact we are looking for. So I agree with him that the answer to the
problem is not going to be given just by talking of dispositions we
actually have. However, and crucially for what follows subsequently, I
do not think the dispositional account falls to all of Kripke's objections.
The analysis he considers (p. 26) says that I mean some function 4~by
my functor if and only if I am disposed, whenever queried about the
application of the functor to a pair of numbers, to give the answer that
actually is that function, ~b,of them. Kripke attacks this on the grounds
that my dispositions are finite. 'It is not true, for instance, that if queried
about the sum of any two numbers, no matter how large, I will reply
with their actual sum. For some numbers are simply too large for my
mind - or my brain - to grasp' (p. 26-27). So, according to Kripke, my
dispositions fail to make it true that I mean addition by '+' and not
quaddition -a function that gives different results just when x and y are
so big that I cannot do sums involving them.
There are difficulties here. It is not obvious that dispositions in
themselves are either finite or infinite. The brittleness of a glass is a
respectable dispositional property. But there is an infinite number of
places and times and strikings and surfaces on which it could be
displayed. Does this glass have a disposition that covers, for example,
the fact that it would break if banged on a rock on Alpha Centauri?
What if scientists tell us that this glass couldn't get there, because it
would have decayed within the time it takes to be transported there?
Perhaps I am not disposed to give the answer faced with huge sums. But
perhaps also I have dispositions that fix a sense for the expression 'the
answer I would accepť. The answer I would accept is the one that
would be given by reiterating procedures I am disposed to use, a
number of times. (The notion is doubly dispositional.) The fact that I am
not disposed to follow those procedures that number of times seems like
the fact that the glass cannot get to Alpha Centauri. Now, a sceptic
might maintain that we do not know of a dispositional fact about me
that is described in this way. Perhaps ! am only disposed to say that
'3 + 5 = 8' when the calculation is not embedded in really huge calculations.
But this is just scepticism about dispositions. It is like
290 SIMON BLACKBURN
supposing that the glass may be not brittle but 'shmittle', where x is
shmittle if it breaks when struck except... In effect this is inductive
scepticism about the concept of a disposition, querying whether we can
legitimately take dispositions to cover what would have happened on
unobserved occasions. So it cannot be used to argue that even if we
accept the concept, it permits no answer to the problem of huge
calculations. A similar complication might answer Kripke's second
objection (p. 28-29). This is that what I mean cannot be read off from
what I am disposed to do, since I may be disposed to make mistakes.
The dispositionalist would have to read off what I meant from a table of
answers I actually give, and this might involve saying that I was
computing (correctly) a bent function - 'skaddition' - and not making a
mistake in attempting to add. But this seems to ignore surrounding
dispositions. Kripke rightly dismisses any view that simply takes for
granted a notion of the function it was intended to compute, or which
defines user's competence, since it presupposes the ideas we are looking
for. But at least it is true that a calculator can have, in addition to
dispositions to give answers, dispositions to withdraw them and
substitute others. And it is possible that putting the errant disposition
into a context of general dispositions of this sort supplies the criterion
for which function is meant. The equation would be: By '+' I mean that
function 4~ that accords with my extended dispositions. An answer
z = 4~(x, y) accords with my extended dispositions if and only if (i) it is
the answer I am disposed to give and retain after investigation, or (ii) it
is the answer I would accept if I repeated a number of times procedures
I am disposed to use, this being independent of whether I am disposed to
repeat those procedures that number of times.
Kripke's point about mistakes can be illustrated if we consider a
calculating machine. There is no physical or dispositionai difference
between a machine that is 'supposeď to compute addition, but because
of a mistake in the hardware computes a bent function, and one that is
designed to compute the bent function in the first place. The two may
be perfectly identical. This strongly suggests that the notion of correctness,
the notion of the rule to which we are to be faithful, has to
come from outside the thing itself. In the case of the calculator, it
certainly does. But this need not entirely destroy a dispositional
account, provided it can look for dispositions outside the simple
disposition to give answers.
So is it a real truth that the right rule for ' +' is in force amongst us?
THE INDIVIDUAL STRIKES BACK 291
My own answer would be that we do have dispositions that enforce this
judgement. They make it the only possible judgement about ourselves,
when we describe each other's thoughts. The concealed bent-rule
follower is a theoretician's fiction. Whenever we try to fill out the story,
of a person or a community that really adopts a bent rule, it turns out
that the singularity in the rule (by our lights, of course) must affect the
dispositions to behave that the community or individual shows. I have
argued this elsewhere in connexion with Goodman's paradox, s The
concealed bent-rule follower is often thought of as though nothing
about him is different until the occasion of bent application arises. But
this is wrong. Someone who has genuinely misunderstood a functor is
different, and the difference can be displayed quite apart from
occasions of application. Consider, for instance, the bricklayer told to
add bricks to a stack two at a time. If this means to him 'add 2 up to
1000, and then 4' his reaction to the foreman may be quite different.
Perhaps he cannot carry four bricks at a time.
However, I am not going to pursue this defence of dispositions. I
share Kripke's view that whatever dispositions we succeed in identifying
they could at most give standards for selection of a function which we
mean. They couldn't provide us with an account of what it is to be
faithful to a previous rule. It is just that, unlike Kripke, I do not think
dispositions are inadequate to the task of providing standards. Indeed, I
think they must be. For notice that the problem of finitude applies just
as much to any community as to any individual. If the finiteness of an
individuaFs dispositions leaves it indeterminate whether he means one
function or another by some functor, then so must the finiteness of
community disposition. So although communities induct their members
into '+' using practices and go through all the corrections and
imitations that constitute community use, still all that is consistent with
the 'skaddition' hypothesis too. We don't find communities disposed to
calculate numbers which are just too big. But prosecuting this point
takes us into BW's divergence from KW.
3. THE COMMUNITY AND THE INDIVIDUAL
The individual has a hard time against Kripke's sceptic. How does the
public fare better? The individual couldn't make the sceptic appreciate
the kind of fact it was, that he was being faithful to a principle, or rule or
previous intention, when he gave some answer. The sceptic charges
292 SIMON BLACKBURN
that there is no fact of the matter whether the bent rule or the natural
rule was the one intended, or whether one principle of application or
another was in force yesterday. And if there is no fact of the matter of
this sort, then there is really no fact of the matter that any principle at all
was in force. Any answer to the new sum can be regarded as equally
'righť and that just means that we cannot talk about right. Faced with
this impasse the individual thankfully turns to the community. He can
point to his inculcation in a public practice, his gradual conformity to
patterns of behaviour accepted by others, and his acceptance as a
competent operator with ' + '.
The sceptic might allow all this to make the difference. But he has
suddenly gone very soft if he does. He can easily specify bent principles,
with points of singularity where neither I nor anybody else used the
terms yesterday. If such points worried the individual, then they should
equally worry the community. So how does mention of the community
give us the determinate rule?
KW's answer is that in a sense it doesn't (see for instance p.111). The
community is as much at a loss to identify the fugitive fact as the
individual was. The position is supposed to be no different metaphysically.
The difference is that the community endorses or accepts the
competent operator. They or 'we' allow him to be using '+' to mean
addition. He is 'seen' or dignified as a rule-follower. We, the community,
have justification conditions for doing this. To gain the title, the
individuaľs practice (on the finite number of occasions on which it will
have been shown) must accord in some suitable way with the community
practice (on the finite number of occasions on which it will have
been shown).
Now, merely citing the fact that we 'see each other as' obeying the
same principle of application makes no headway at all. Remember that
the point of the original worries about rule-following was not epistemological.
The aim was not to suggest that we cannot provide a
foundation of some sort for the judgement that a rule is in force, or that
we cannot provide principles for inferring such judgements from others
more basic. If this were the aim, then replying that there is no inference
and no foundation, but just a basic fact that we make such judgements
of each other, 'seeing each other as' following principles, might be
relevant - at any rate, as relevant as this move is in other areas of
epistemology. But the point does nothing at all to suggest how we are
seeing each other when we say such a thing. We know well enough what
THE INDIVIDUAL STRIKES BACK 293
it is to see something as a duck, because we know what ducks are. But
we don't know what it is to see someone as obeying a principle of
application, unless we know what it is to follow one, and this is the fact
of which we still have, so far, no conception. To put it another way, we
do not know what a community would be lacking if its members failed
to see each other this way, or if they continually saw each other in the
light of potential bent-rule followers.
So the sceptic is still liable to feel short-changed. He has pointed out
bent rules which might have been in force yesterday, compatibly with
all that the individual could point to. He has pointed out that for all that
the individual could show, there might be nothing but his onward
illusion that rules are in force, and that his dispositions to respond are
correct or incorrect. The individual now links arms with others. The
sceptic attempts to point out the same two things to them: for all they
can describe about themselves, bent rules (and perhaps differently bent
rules) might have been in force, underlying their fortuitous coincidence
of behavior over the finite samples they have come across. Since this,
means that any answer to a new problem might be as right as any other,
the sceptic suggests that again there is just the onward illusion that
there is correctness or incorrectness. The community replies that it has
a practice of dignifying its members as saying things correctly or
incorrectly, and in the light of this practice it says that all its members
do mean the same, and that what they mean provides a principle of
application of a term.
We can see one way in which this could silence the sceptic. If the
mutual support itself provides the standard of correctness, then a
community can answer him. To understand this, consider the analogy
with an orchestra. Suppose that there is no such thing as an individual
playing in accord with instructions coming down from the past, by way
of scores or memory. Suppose too that on an individual instrument
there is no standard for the way a piece ought to go (all melodies are
equally acceptable). Then an individual cannot play well or badly in
isolation. Nevertheless, the orchestra may have standards of harmony
across instruments. And if at a given time most instruments are playing,
say, notes from the chord of C major, then the individual who hits a
dissonant C ~ is incorrect by the standards of the orchestra. They can
turn on him. Unfortunately, this provides only a poor analogy for
communities and their relations to their own rule-following. For in the
orchestra, harmony with others provides a direct standard of correct-
294 SIMON BLACKBURN
ness. This is just not so with judgement. If my community all suddenly
started saying that 57 + 68 = 5, this fact does not make me wrong when
I continue to assert that it is 125. I am correct today in saying that the
sun is shining and daffodils are yellow, regardless of what the rest of the
world says. Obviously any solution to these problems must avoid the
disastrous conclusion that it is part of the truth-condition of any
judgement that a community would make it (unless of course the
judgement is itself not about the sun and daffodils and so on, but about
the community).
If the community cannot turn to the orchestral metaphor then how
have they answered the sceptic? And why cannot their answer be taken
over by the isolated individual? Remember that there is a distinction
between the overall practice of a community, thought of as something
defined by principles and rules, and the exposed practice, thought of as
only a partial, finite segment of applications. If a community practices
addition, meaning one thing and not another by some functor ' + ', then
the exposed practice will cover only a small proportion of the applications
whose correctness follows from the overall practice. As the
possibility of bent functions is supposed to show, the exposed practice
does not logically determine the question of which overall practice is in
force. Then we can imagine what we might call a 'thoroughly Goodmanned
community' in which people take explanations and exposure to
small samples - yesterday's applications - in different ways. The sceptic
who won against Kripke's individual will now win against a community,
showing that they have no conception of the fact which makes it true
that they do not form a thoroughly Goodmanned community.
It may be helpful to think of it like this. The members of a community
stand to each other as the momentary time-slices of an individual do. So
just as the original sceptic queries what it is for one person-time to be
faithful to a rule adopted by a previous person-time, so the public
sceptic queries what it is for one person to be faithful to the same rule as
that adopted by another. Now if the public sceptic can be by-passed by,
in effect, saying that this is what we do - we see each other as mutually
understanding the same rule, or dignify or compliment each other as so
doing, provided the exposed practice agrees well enough, then the
private sceptic can be by-passed in the same way. His doubts admit of
the same projective solution. When LW denies that 'we have a model of
this superlative facť ('192) we can, as far as the metaphysics goes,
shrug and say that this is how we see ourselves. Then, as I have already
THE INDIVIDUAL STRIKES BACK 295
explained, the entire problem is to explain how we are seeing ourselves
when we go into this mode. In particular we need to cite some standard
for saying this - and, if LW is to win through, a standard which
separates the public from the individual. For when I write the sensation
term in my diary I can and will see myself as being faithful to a previous
intention to apply it only in a determinate range of circumstances. And,
paying me the compliment as it were, you can do so too. It might be
different if it somehow came naturally to us to dignify fellow members
of the same public with the title of rule followers, and natural to hesitate
over any purely private attempt at self-description. But generally
speaking there is little difference in naturalness. Until LW supposedly
argues that we shouldn't do so, most people would find it quite natural
to believe that a putative private linguist might be following or failing
to follow previously formed intentions to apply principles of ap-
plication.
The dilemma so far is this. If the presence or potential presence of a
community of persons practising the same way enters as part of the
truth-conditions, part of the analysis, of what it is to follow a rule, the
sceptic who won against the private individual looks equally set to win
against a community which has the benefit of mutual support. But if
mention of the community comes as part of a Humean or projective
solution, allowing us to by-pass troubles over our conception of the
superlative fac't, then a similar side-step is in principle available to the
individual. To split the individual from the public we need enough
understanding of how we are regarding ourselves to be able to specify
standards which need to be met by a candidate for rule-following, and
we have not yet got this.
The simple move, endorsed by KW and, for instance, by Peacocke, is
to say that in the case of the individual there is no distinction between it
seeming to him that a rule is in force, and there being a rule in force.
The trouble with this is that in any sense in which it is true (see next
section), it is equally true that there is no distinction between a
community being thoroughly Goodmanned, but seeming to itself to
have a unified practice, and its actually having a unified practice. And
when the community says 'well, we just see ourselves as agreeing
(dignify, compliment ourselves as comprehending the same rule)' the
individual just borrows the trick, and compliments himself on his
rapport with his previous times. The community needs to say that the
compliment is empty in the case of the individual but meaningful in the
296 SIMON BLACKBURN
case of a set of them. How is this to be argued?
4. PUBLIC AND PRIVATE PRACTICE
Peacocke writes:
In the end, Wittgenstein holds, the only thing that must be true of someone who is trying
to follow a rule, so long as we consider just the individual and not facts about some
community, is that he is disposed to think that certain cases fall under the rule and others
do not. But this is something which is also true of a person who falsely believes that he is
conforming to a rule. His general argument is that only by appealing to the fact that the
genuine rule-follower agrees in his reactions to examples with the members of some
community can we say what distinguishes him from someone who falsely thinks he is
following a rule. 1°
The lynch-pin of this interpretation is the absence of a 'distinction' in
the case of an individual, between the case where he believes that there
is a rule, when there is none, and where he believes it and there is one.
And my question has been: do we think we have yet been given a
'distinction' between a thoroughly Goodmanned community, accidentally
agreeing in exposed practice, and a real community of
understanding? And then, if the distinction is given using materials
from Hume - the projection upon one another of a dignity - I ask what
the standards are whereby it takes several people to do this, when one
cannot do it to himself. I shall now try to exhibit the force of this
question, by considering the individual further. This is to illustrate the
gap between anything that has happened in LW, at least as far as "202,
and the application to exclude the private linguist.
Let us recall the basic point, that giving an explanation, either in
words, or using other aids, such as pictures or models, does not logically
determine the rule that governs one's understanding of a term: 'Any
interpretation still hangs in the air along with what it interprets, and
cannot give it any supporť ('198). To avoid the threatening paradox,
that nothing can accord nor conflict with a rule, because anything can
be made out so to do, LW introduces the positive suggestion that our
rules are anchored in practice. In the remainder of this paper I shall
assume that this is along the right lines. That is, dignifying each other as
rule-following is essentially connected with seeing each other as
successfully using techniques or practices. This at least begins to isolate
the nature of the judgement. It suggests a direction from which to find
standards for making the judgement, and in my view it connects
THE INDIVIDUAL STRIKES BACK 297
interestingly with pragmatic and coherence aspects of judging. Now,
having introduced the notion of a practice or technique, LW immediately
goes on to draw two famous conclusions - that to think one is
obeying a rule is not to do so, and that it is not possible to obey a rule
privately. These are not the same, and the second is not warranted. The
question is, what kind of thing does a practice have to be, if it is to block
the sceptical paradox? Perhaps, for instance, the concept of a practice
can include that of someone setting out to describe his mental life, even
on a highly private conception of the mental. The basic negative point
certainly does not rule that out. To do so we would need further
thoughts about what a practice must be, and the connexion it is
supposed to need with actual or possible publics. It is therefore a great
pity that, with one eye cocked on the later applications, commentators
simply assume that by *202 the publicity of practice has been satisfactorily
argued (Peacocke for instance just announces that 'by "practice"
he here means the practice of a community.' (p. 72)). Of course,
many believe that the later discussion of sensation S and so on justifies
the restriction. But that is another issue; to invoke those is to abandon
the hope that the rule following considerations provide wider, more
general thoughts from which the anti-private language conclusion
independently emerges. To illustrate the gap we can consider the
half-way house (to full privacy) provided by Robinson Crusoe cases.
Consider the'example (due to Michael Dummett) of a born Crusoe
who finds a Rubik's cube washed onto his island, and learns to solve it.
The fact is that he does it. He certainly doesn't solve it randomly, for he
can do it on demand. It is natural to say that he follows principles (when
there is a last corner left to do...). Perhaps he has some rudimentary
diagrams or other mnemonics which he consults. With these he can do
it, and without them he cannot. Kripke considers what LW's attitude
should be to such a case. The attitude he offers, on behalf of KW, is that
we can think of Crusoe, in such a case as this, as following rules. But,
'If we think of Crusoe as following rules, we are taking him into our
community and applying our criteria for rule following to him.' The
sceptical considerations are supposed to show 'not that a physically
isolated individual cannot be said to follow rules; rather that an
individual considered in isolation (whether or not he is physically
isolated) cannot be said to do so'. 11 Again there is a nice parallel
(although not quite the one Kripke makes) with Hume. One conclusion
that might be drawn from Hume is not that a pair of events in a universe
298 SIMON BLACKBURN
with no others cannot be related causally (which is how Kripke takes
Hume). Rather, if we suppose that they are, we are thinking of them as
members of a (potential) family of other regularly related events. We
are not considering them in isolation, but, for instance, are thinking of
what would happen if there were others like them.
Still, it is not clear what the compromise is. Certainly I, or we, are
doing the thought-experiment. I have to consider whether Crusoe is a
rule follower by using the normal, community-wide way I make the
judgement. But that would be true of any situation I seek to describe.
And then, just as (contra Berkeley) I might conclude that an island
considered in isolation has a tree on it, might I not conclude that
Crusoe, considered in isolation, was following a rule? How does the
phrase 'considered in isolation' bring the community further into the
picture in this case than in the case of the tree? We are apt to retort that
Crusoe would have been a rule-follower in this situation whatever I or
we or any other community in the world had thought about it - just like
the tree. And the reason is that all by himself he had a technique or
practice.
Now reconsider the private linguist, meaning someone who believes
that he has given an inner state a semantically essential role. Suppose he
believes that yesterday, in the presence of the state, he defined for
himself a qualitative similarity that other states might bear to it; that he
is on the look-out for recurrence, and involved in the practice of
judging if such a recurrence takes place. We seek to show that this is
not a real practice. Let our private linguist accept the basic point,
admitting that the mere offering of words, images and so on, does not
determine a rule of application, or principle that is really in force. But
he does not accept that his candidate practice is unsatisfactory. What
has LW got to show him?
'Whatever is going to seem right is righť - there is no distinction
between his seeming to himself to follow a rule, and his genuinely doing
so. It has often been suggested that this charge is unargued, or, if
argued, only supported by overtly verificationist considerations. My
endeavour has been to show how difficult it is to release the public from
its attack. In the light of this, let us reconsider the projective and
naturalistic elements that assisted the public: the point that we naturally
and perhaps usefully regard ourselves as mutual possessors of the same
understanding. We see ourselves so, and this attitude is, in Humean
vein, immune from sceptical destruction. Furthermore, there is no
THE INDIVIDUAL STRIKES BACK 299
lowering of the truth-condition of this judgement. It sits with its own
vocabulary and scorns any 'accounť of it in other terms. It is just that if
a public failed to see itself in this light, it would mean that it could only
see the ongoing patterns of noise and reaction, in which no principle, no
genuine judgement, no truth and falsity, is visible.
My criticism of the fiat reply to the sceptical problem ('this is how we
see ourselves') was, in effect, that it gave no account of what a
community would be missing if it failed to see itself in this light.
Following LW we have accepted that the clue to what is missing is to
come from the notion of a practice. BW therefore uses this thought, and
the fact that dispositions provide satisfactory standards for the making
of the judgement. Like other judgements with normative elements,
there is no attempt to make a lowering of the truth-condition. The
judgement can perfectly well be seen in terms of the projection of an
attitude (this cannot be uncongenial to LW). But that left us unclear
about the standards for applying it. BW therefore accepts most of what
emerges from LW, doing less violence to the redundancy theory and to
the preparedness to talk of facts, even when the underlying metaphysic
is not what that might suggest. But however well BW matches LW,
Blackburn insists on asking: is there any reason why the private linguist
should not so regard himself? And in that case whose is the attitude
'whatever is going to seem right is righť? Not the subjecťs own, for he
dignifies himself as a genuine believer, as having a principle of
application and making a judgement with it. In doing so he allows the
possibility of mistake (it is not something there in the things going on in
his head or in his behaviour: it is something arising as a projection from
an attitude he takes up to his own projects). It is a component of his
attitude that a particular judgement might turn out better regarded as
mistaken.
How can this attitude be appropriate? A technique is something that
can be followed well or badly; a practice is something in which success
matters. Now in the usual scenario, the correctness or incorrectness of
the private linguisťs classification is given no consequence at all. It has
no use. He writes in his diary, and so far as we are told, forgets it. So
when LW imagines a use made of the report (e.g., to indicate the rise of
the manometer) he immediately hypothesizes a public use. He thereby
skips the intermediate case where the classification is given a putative
private use. It fits into a project - a practice or technique - of ordering
the expectation of recurrence of sensation, with an aim at prediction,
300 SIMON BLACKBURN
explanation, systematization, or simple maximizing of desirable sensation.
To someone engaged on this project, the attitude that whatever
seems right is right is ludicrous. System soon enforces recognition of
fallibility.
I conclude then that it is no mistake to see the later sections, from
*240 onwards, as integral to the anti-private language polemic. BW
simply cannot separate the private from the public with any considerations
that are in force earlier on. But I have tried to suggest other
things as well. Following through the problem of answering the paradox
leads to sympathy with a basically 'anti-metaphysicaľ conception of
rule-following. We simply cannot deliver, in other terms, accounts of
what constitutes shared following of a rule, or what the fact of a rule
being in force 'consists in'. In my view this invites a projectivist
explanation of these kinds of judgement, although also in my view we
cannot conclude that it is improper to talk of facts of the case. In any
event, we are left searching for standards whereby to make the
judgement. It is possible that those standards should exactly separate
the public from the private (on some vulnerable conception of the private,
of course). But there is no particular reason to expect them to do
so. The problems with dispositions, either as giving us the missing kind
of fact, or as providing standards for allowing that a rule is in force,
failed to separate the public and the individual. So we cannot now
simply demand from the putative private-linguist an 'account of what
the distinction' (between genuinely following a rule, and only seeming
to himself to do so) amounts to. He can reply. It amounts to all that it
does in the case of the public. Just as a public dignifies itself as
producing more than an interminable flood of words and noise, and sees
itself as making mutually comprehensible judgements, capable of truth
and falsity - so does he. The public doubtless has a purpose in doing
this, and is right to do it. When his putative discriminations are part of a
practice - so does he.
NOTES
1 Saul A. Kripke, Wittgenstein on Rules and Private Language, Basil Blackwell, Oxford,
1982. All page references are tothis work. Paragraph references in the text are to the
Philosophical Investigations
aP. 5.
3 Crispin Wright comes close to denying it in Wittgenstein on the Foundations of
Mathematics, e.g., pp. 21-22, where he seems to attribute to Wittgenstein an error theory
THE INDIVIDUAL STRIKES BACK 301
of determinacy in the correctness of saying anything.
4 Similarly in the Treatise, esp. Bk I, Pt III, §14.
5 pp. 68, 69.
6 For typical statements compare Appendix to Enquiry, also Treatise, p. 167 (Selby
Bigge). Projectivism is widely misunderstood even today, and its true resources are not
easily recognized.
7 See my 'Truth, Realism and the Regulation of Theory', in French, Uehling, and
Wettstein (eds.), Mid- West Studies in Philosophy, Vol V.
8 Reason and Prediction, cambridge University Press, Cambridge, 1973, Ch. 4; and
Spreading the Word, Oxford University Press, Oxford, 1984, Ch. 3.
9 D. Hume, An Enquiry Concerning the Principles of Morals, Section I.
l0 C. Peacocke, 'Rule Following: The Nature of Wittgenstein's Arguments', in Holtzman
and Leich (eds.), Wittgenstein: To Follow a Rule, p. 73.
1~ Kripke says useful things about this, p. 90ff.
Pembroke College
University of Oxford
Oxford
England