The ‘paradox’ and ‘resolution’ (201-242)

							Wittgenstein: notes 2/23/05
Philosophical Investigations 201-300

The „paradox‟ and „resolution‟ (201-242)

At 201 Wittgenstein formulates the paradox. Each course of action seemed to be able to
accord with the rule, for every course of action seemed to b e interpretable in conformity
with the rule. The language-game of providing rules itself does not determine – in the
way that we wanted it to – the correctness or incorrectness of our action. Here we are
tempted to an infinite regress, to explain the rule we‟ve given – which does not yet
elimiante all misunderstanding – with another rule, which itself does not eliminate
misunderstanding, etc., etc. But we can see that this was a misunderstanding from this
tendency itself. What we call „obeying a rule,‟ following it correctly, going on as we do,
etc., is not explained or described through this language- game of successively
introducing rules and their interpretations. Nevertheless we can say that someone
followed a rule, and we can say that they followed it correctly. This is possible, though,
only in connection with the vast and interconnected situations of practice and context that
give shape to our following a rule in each case.

There is a way of grasping a rule which is not an interpretation, but which is shown
in what we call „obeying the rule‟ and „going against it‟ in particular cases. (Here we
should pay attention to the use of „shown‟, reminiscent of the Tractatus). What it is to
grasp a rule, and follow one, takes shape only against this variety of particular cases, and
is shown only in it, only insofar as there exists a regular practice or a technique that
determines what counts as following a rule in each case. “And hence it is not possible to
follow a rule privately. Otherwise it would be the same thing to think one is following a
rule as actually to follow it.”

In 202, Wittgenstein suggests what has appeared to some to be a “solution” to the
skeptical rule-following paradox. “And hence also „obeying a rule‟ is a practice.” It
can seem as if he is here suggesting that what it is to obey the rule correctly is essentially
determined by something that we can call a practice, for instance by the totality of
socially regulated and normed behaviors that we follow in determining correctness or
incorrectness. And it is true that Wittgenstein is suggesting that what counts as following
a rule, or not, is shown in what others in our society will say about what we do, and
cannot be determined in advance of this by a standard independent of social practice. But
this does not require us to interpret Wittgenstein as holding a “social theory” or a
“pragmatist theory” of rule- following. Though he is saying that whether we have
followed correctly or not will be shown in social reactions, he is not s aying that
correctness simply consists in society‟s judgment that we have gone on correctly.
Indeed, his therapeutic aim here, at least in part, is to remove the temptation to think there
must be anything, factual or normative, in which following a rule correctly consists. If
we take him to be holding a social theory of rule- following here, we miss one of the
largest points of his application of the philosophical dialectic: against the thought that
there is anything that “ultimately grounds” the foundatio ns of rationality, even if it be
something social rather than something individual. This also has important consequences
for our interpretation of the „private language‟ argument, to which we‟ll turn.

What it is to follow a rule is shown in, and only makes sense in the background of, a
system or set of practices and ways of understanding one another. “The common
behaviorur of mankind is the system of reference by means of which we interpret an
unknown language.” (206) We might think that we could detect the presence of a
language by finding a certain degree of order, or regularity, in their responses, even if we
do not understand the language. But really we are only prepared to describe the language
as “ordered” at all if we already understand a great deal about what is going on. (Here
there are points of comparison with Davidson‟s argument that we must, in order to make
sense of another language at all, construe its speakers as already accepting a great deal of
what we believe.) What it is for a pattern of behavior to be “ordered,” regular, only
shows up against the backdrop of the complex variety of natural and usual ways in which
we instruct children in a language (208). To say this is not to define “order” in terms of
regularity but to gesture towards the ways in which we “learn” how to go on “regularly”
in ordnary practice. Here, Wittgenstein‟s remarks are constantly in dialogue with the
conception according to which there is a fact of the matter for us to grasp, independently
of our own socialization into a community, about whether somebody has followed a rule
or not. It is possible to see the alien speaker as following a rule, as meaning what we
mean – but only if we can portray the alien speaker as doing what we do. And the basis
for our ability to do this is not some thematizable grasp of what we do, but rather its
concrete recognition. The success-conditions for this recognition depend on our being
able to see the other as like us, in some relevant respect. Without this, there is no hope of
our being able to see him as meaning what we mean, doing what he do.

We are still tempted to think that the rule must reach beyond its finite expression to
determine all of the cases. We want to pose the justificatory question (NB the question
that Kripke wants to pose?) and we don‟t take the response “I was instructed in the usual
way, etc.” to be the right sort of answer. But how do I explain the meaning of the word
“rule,” of “regularity”? By means of examples, and practice, and without avoiding the
essential moment where the student must go on for himself. (208). But my reasons will
run out, and then I will just act, without reasons (211), (217), (219). A doubt was
possible; but this doesn‟t mean that I did doubt. The justificatory question assumes that
where a doubt is possible, it must be problematic; we must come up with a philosophical
account that eliminates its possibility. But there is nothing to demand this.

We see that the way that we talk about rules is tied up with the way that we ta lk about
“regularity”; with “the same.” If we need an intuition to continue the series 1, 2, 3, 4, we
also need an intuition to continue the series 2 2 2 2 . (214) One skeptical response is to
think that rules “bottom out” with the recognition of the ide ntity of the same. (This is
also, in a sense, the response at the bottom of metaphysics. For metaphysics thinks of the
rule as self- identical in all of its instances of application; the most pure and simple
instance of this would be the endless repetition of the same in accordance with the law of
identity). At this point, we are tempted to think that the bottom- level explanation of the
rule is the infallible identity of the thing with itself. But although it is true that the use of



                                                2
the word “rule” and the use of the word “same” are interconnected, we can think that the
self- identity of the rule helps us here only if we think that it gives us something. We feel
like here, finally, we have the guarantee we wanted: “Here at any rate there can „t be a
variety of interpretations. If you are seeing a thing you are seeing identity too.” But
really there is nothing happening here but a seductive play of the imagination. We want
to use the law of identity to get us the reassurance we desire, the reassurance that the rule
really does underlie all of its concrete instances, that it stays the same despite the
heterogeneity of its applications. But we can only picture the law of identity as doing this
work if we can picture it as something substantial, a kind of guarantee of self- identity.
We will think of this guarantee as present whenever we see something (that is identical
with itself). But if it were a guarantee at all, it could fail to hold as well. When we say
that “a thing is identical with itself,” we are just engaging a certain play of the
imagination; we are not saying something that has any substantial truth. The last
recourse of the metaphysician, the self- identity of the rule in the heterogeneity of its
applications, finally fails to provide the assurance that was desired.

The metaphysical idea of the rule as determining all of its infinite number of instances is
itself just such a play of the imagination. Where does this play of the imagination come
from? We want to ask the question about what justifies us; we don‟t feel content with the
answer “this is simply what I do.” So we indulge a certain picture of the answer, a
picture, for instance, of rails laid to infinity (218). Our symbolic picture gives expression
to the way things strike me – in particular, to the fact that when I obey the rule, I do not
have the feeling of making a free choice. I obey blindly, and the picture of rules as rails
to infinity tries to figure this phenomenology. But this is all that my symbolical
expression, my play of the imagination, can do. It can only bring into focus this
distinction between being causally determined and being logically determined; it can do
nothing to provide the justification that I seek. The symbolical expression of the rule
representing it as a rail laid to infinity was a mythological description of my practice, of
my use of the rule in concrete cases (221). The task of the grammatical investigation is
to demystify the picture, to show us the concrete reality of the facts that we are trying to
capture with it anyway. With the symbolic expression we are trying to bring something
into focus, the fact that “we look to the rule for instruction and do something, without
appealing to anything else for guidance.” But under the pressure of philosophy this
comes to seem to be the justification for a picture of the rule as determining all its
instances all by itself, of the rule as the same thing, one thing, despite the heterogeneity
of its instances. Our description of our own phenomenology comes back to us, mystified,
as an explanation of the reason for that phenomenology. But here the symbolic
description really says nothing more than the phenomenological fact we had set out to
accommodate.

Whence arises the sources of the temptation to think that the rule intimates or directs me
in going on the way that I do? This experience is not quite like that of hearing an inner
voice (232); I don‟t need to await it, and I can teach it as a technique. If it were simply a
matter of hearing an inner voice, then calculating would be like composing: each would
just hear his own (233). We might still (234) imagine calculating as we do, agreeing, but
having the feeling of being guided as if by a spell, feeling astonishment at the fact that we



                                              3
agreed. The picture of “intimation” goes along with the idea – that is not appropriate
here – that we are surprised by the fact that we can go along in the way that we can. In
fact, though, “the rule can only seem to me to produce all its consequences in advance if I
draw them as a matter of course.” (238). We go along, we draw inferences; for the most
part we agree, and if we disagree there are in any case regular procedures to resolve the
disagreement. We are typically not surprised. We can give voice to this phenomenology
in various ways, but there is really nothing that explains our not being surprised. This is
just how things are. Disputes don‟t break out (among mathematicians, etc.) over the
question whether a rule has been obeyed or not (240). This is part of the framework of
language, of what we always already presuppose, without justification, in the concrete act
of speaking and understanding one another.

Here Wittgenstein alludes again to the vast complexity of agreements and practices, of
ordinary and natural dispositions and proclivities, that provide in all of their complexity
the “foundation” for human behavior and agreement. Wittgenstein‟s word for this is
“language”; and the kind of agreement that is basic for our practices is not agreement in
beliefs or opinions, but agreement or attunment in language itself, in “forms of life.” If
language is to be used for communication there must be agreement not only in
definitions, but also substantive agreement, on judgments. The consistency that we find
in our practices is not only explicable in terms of the consistency of stipultations or
agreements we have entered into beforehand; it is, rather, the case that what we call
stipulation and agreement is partially dependent on what we call consistency in practice.
As with the rule, the use of the word “judgment” is tied up with the use of the word
“consistency.” They describe the same phenomenon, and one cannot be used to explain
the other.

Some commentators have read Wittgenstein, in these passages, as gesturing at something
that is foundational in the justificatory sense for our practices, perhaps something like the
Husserlian “lifeworld” that provides the origin and concrete basis for all of our varied
epistemological and intersubjective activities. But it is important to realize that
Wittgenstein‟s intention is not to gesture to any thematizable totality of attunements or
any explanation of their basis, but simply to remind us of the fact of this attunement, a
fact that has no justification or basis in the general sense that the philosopher expects.
When we point to the vast and complex attunement that organizes our daily lives and
characteristic practices, we are not gesturing toward any determinate fact or totality of
facts. Even if there were such a totality, it would be possible in each case to misinterpret
it and go on differently, as the rule- following investigations have shown. We are,
instead, pointing to the reality of attunement as a kind of article of faith, as a completely
unjustified assertion that we will go on agreeing, that we will keep doing what we have
done. Where particular justifications arise, they function in the same element and
according to the same presuppositions: they always only make sense within a concrete
practice or language-game. But there is no general justification for language or
communication in the sense that the philosopher expects. The upshot of the rule-
following investigations makes us feel the vertigo that we will initially feel when we
discover that the deepest ground of our practices is groundless. But the detailed teaching
of Wittgenstein‟s consideration of rules also tended to show that the kind of justification



                                              4
that we thought we needed here was actually useless, that what we responded to was not
a need for foundations at all, but rather a fundamental mystification of our own existence.


Private Language:

AT PI 243 Wittgenstein turns to the consideration of a “private language”: “a language
in which a person could write down or give vocal expression to his inner expe riences
– his feelings, moods, and the rest – for his private use? -- Well, can‟t we do that in
our ordinary language? -- But that is not what I mean. The individual words of this
language are to refer to what can only be known to the person s peaking: to his
immediate private sensations. So anothe r pe rson cannot understand the language .”

It is important to be clear, at the outset, about just what Wittgenstein is critiquing here.
This is one point on which he has been very commonly misunderstood, so it is important
to understand just what he is saying. Obviously, he is after some idea of privacy, of „the
mental‟ as a special, private realm separate from the physical world. On this idea,
„mental items‟ like pains, sensations, thoughts, and the like, have a special kind of
existence in that they are private or proprietary to their owners, unlike ordinary, „public‟
items. But he‟s not just after just any idea of privacy. It is important to see at the outset
that his goal is not to deny the “inner” or the “mental” outright, or to say, for instance,
that mental events and happenings are really just physical events and happenings, either
physical states of the brain or states of behavior and dispositions to behave.

He‟s after a specific, philosophically prejudicial and problematic picture of the „inner‟,
one that he thinks leads to characteristic philosophical confusions. On the philosophical
idea that Wittgenstein is after, the denizens of the „inner‟ – pains, sensations, thoughts,
memories, etc. – are special in that they – or their character – can essentially only be
known to their owners. There is in this sense a special privelege in the relationship that
the owner has to the pains or sensations. Because they or their character can only be
known to the owner, the language he uses to refer to them can be meaningful only to him.

There are various components to traditional pictures of the existence of the mental or the
subjective, some but not all of which come under the scope of Wittgenstein‟s attack:

privacy: A mental state – or its qualitative character – can be known only to the person
who has it. It is impossible for someone else to know it; in this sense, it is private.

priveleged access: A person has better or more direct access to his own mental states than
anyone else can have. This is connected with the idea of introspection: the idea of a
special faculty that gives us access to our mental states, and is in some way more direct or
immediate than our perception of the external world.

incorrigibility: Our relationship with mental states is such as to give us indefeasible
certainty about their character. We can be mistaken about the nature of something we




                                              5
perceive in the external world, but not about the character of our own pains and
sensations

immediacy: Our knowledge of (relation to) mental states is in some sense immediate, so
that there is no conceptual work (inference or deduction) that has to go on, even
implicitly, in order for us to know them or their character.

ineffability: The „raw feel‟ or experiential quality of a mental state is ineffable, incapable
of being expressed in language, but is somehow prior to language or outside it.

Non-conceptuality: Our mental states such as sensations present themselves to us prior
to, and independently of, their being interpreted through concepts.

We‟ll see that Wittgenstein wants to critique some, but definitely not all, of these ideas.
He‟s not going to deny, for instance, that our reports of pains and sensations have a
certain kind of indefeasibility: we are automatically entitled to them in a way we are not
entitled to reports of external states of affairs. But he will target a philosophical
conception of privacy that underlies most kinds of attempts to make a foundational use o f
the idea of private experience – for instance in sense-data theories, or in traditional
Husserlian phenomenology; and that still underlies the picture of consciousness that is
often presupposed in contemporary debates.

What is the target of the private language argument? Who actually believes in a private
language, and why? It might not seem obvious who Wittgenstein is attacking, because
even if one can identify certain ideas of privacy as clustering around our idea of the
mental, it‟s not obvious why this should have to be connected with the idea of a private
language. But actually theorists of the twentieth century have made quite a bit of use of
this idea. It is at the root of „sense-data‟ theorists‟ attempts to define sense-data, known
immediately and incorrigibly, as the immedate basis of empirical knowledge.

               Qualia are subjective; they have no names in ordinary discourse but are
               indicated by some circumlocution such as „looks like‟; they are ineffable,
               since they might be different in two minds with no possibility of
               discovering that fact and no necessary inconvenience to our knowledge
               of objects or their properties. All that can be done to designate a quale
               is, so to speak, to locate it in experience, that is, to designate the
               conditions of its recurrence or other relations of it. Such location does
               not touch the quale itself; if one such could be lifted out of the network
               of its relations, in the total experience of the individual, and replaced by
               another, no social interest or interest of action would be affected by such
               substitution. What is essential for understanding and communication is
               not the quale as such but that pattern of its stable relations in experience
               which is what is implicitly predicated when it is taken as the sign of an
               objective property. (C. I. Lewis, 1929)


               My companion and I are convinced that we both see the same field; but
               each of us has a particular sense impression of green. I glimpse a


                                                6
                strawberry among the green strawberry leaves. My companion cannot
                find it, he is colour-blind. The colour impression he gets from the
                strawberry is not noticeably different from the one he gets from the leaf.
                Now does my companion see the green leaf as red, or does he see the red
                berry as green, or does he see both with one colour which I am not
                acquainted with at all? These are unanswerable, indeed really
                nonsensical, questions. For when the word „red‟ is meant not to state a
                property of things but to characterize sense impressions belonging to my
                consciousness, it is only applicable within the realm of my
                consciousness. For it is impossible to compare my sense impression with
                someone else‟s. For that, it would be necessary to bring together in one
                consciousness a sense impression belonging to one consciousness and a
                sense impression belonging to another consciousness. Now even if it
                were possible to make an idea disappear from one consciousness and at
                the same time make an idea appear in another consciousness, the
                question whether it is the same idea would still remain unanswerable. It
                is so much of the essence of any one of my ideas to be a content of my
                consciousness, that any idea someone else has is, just as such, different
                from mine. (Frege, 1918)

Frege says that the word „red‟ is applicable only „within one‟s consciousness‟ – at least if we use
it in one way. The idea of a private language itself played an essential role in the foundationalist
projects of theorists like Russell, Schlick and Carnap who thought of our empirical knowledge of
the world as based on experiences that are immediate and known incorrigibly. For Russell in Our
Knowledge of the External World, for instance, all knowledge by description is based on
immediate knowledge by acquaintance of sense-data or immediate qualia. There must be a
language to refer to these qualia, because our knowledge by description is said to be logically
constructed or derived from it. The original project of logical positivism also made use of the
idea of a language that could refer only to immediate experience. Protocol sentences were to be
sentences to describe the “immediately given”: the basis of scientific knowledge, which is present
to consciousness prior to our conceptualizing it. The Vienna Circle later underwent a vigorous
and decisive debate about the form of these protocol sentences, and the sense in which they could
be said to be true or justified.

But even outside the foundationalist projects that made particular use of the idea of derivation of
public statements from a language referring, in the first instance, only to private phenomena or
events, the idea of privacy that underlies the private language picture still probably figures in
most contemporary appeals to the idea of “qualia” or phenomenal raw feels. What is not
captured in the totality of public language is – so the idea seems to go – what could only be
referred to by a private language, the qualia or „what it‟s like‟ to be something. This arises
particularly in connection with the physicalist attempt to explain the mind: what can‟t be
explained by the totality of physical sentences, so it seems, is the immediate, subjective raw feels.
Philosophy of mind over the last twenty years has witnessed a vigorous debate about qualia,
about their existence and properties, and about their relation to scientific or physicalist
explanation. But by considering Wittgenstein‟s considerations of private language, we can begin
to see our way to a way of understanding the “inner” that does not opt for the Cartesian picture of
a realm of experience incorrigible or inaccessible to public knowledge, but at the same time does
not eliminate the inner by holding to an unadulterated physicalism, behaviorism, or materialism
either.




                                                  7
It will be important, as well, to distinguish Wittgenstein‟s position from behaviorism.
Behaviorism also attacks most components of the traditional picture of the inner. But it does so
with a spirit that is quite alien to Wittgenstein‟s concerns. Behaviorism comes in various
versions, but according to the kind of behaviorism that Wittgenstein is most often accused of –
so-called “logical behaviorism” – a sensation or a mental image just is a host of behaviors or
dispositions to behave. Behaviorism is often also discussed as verificationist, because it holds
that what we cannot directly verify – the existence of another‟s sensation – is not real. But
behaviorism is a bad option, because it seems to deny the reality of something that is very real
and important to us, our inner mental life. We‟ll see that, although Wittgenstein is denying a
traditional picture of what this life consists in, he‟s not at all denying its existence. His position is
nothing like the behaviorism he has often been accused of.



At 244, Wittgenstein launches the attack. The private language theorist believes that the words of
the private langauge connect directly with the private sensations they name. Now, how do the
words refer to the sensations? Here we should keep in mind the whole consideration of reference
that has preceded this discussion, and remember the moral that the unique connection between a
word and the thing it stands for is set up only in the context of a large amount of stage-setting and
of systematic connections with the rest of our practices and behavior. Well, how do we learn the
names of sensations? Wittgenstein considers one possibility: there are first primitive, basic
utterances (like a cry of pain). Later, we learn to replace these utterances, in various
circustances, with characteristic words. The replacement is only possible because of the
“primitive, the natural …” expressions that they replace. If this is the case, then the private
linguist‟s fantasy of a special act of naming the sensation is just that: a fantasy. Is this a
behaviorist theory of pain? Such a theory would claim that what it is to be in pain just is to give
the characteristic signs of it, to cry out and grimace, etc. But it is not W‟s intent to give such a
theory of pain. Rather, he is telling us a possible story about how we might learn the technique
we call “sensation-language”. This story seems to make sense; it might be the only story we can
tell. What role do these “primitive, natural expressions” have? It might seem as if Wittgenstein
is here inviting a “naturalist” picture of the “basis” of our sensation-reports. But actually it is not
obvious that he intends anything like a foundationalist naturalism. He seems to intend, instead, to
remind us of how things in fact are with us: what actually happens, what it is to be socialized into
a human form of life, in all of its complexity and inherent lack of guarantees.

246: “Only I can know whether I am in pain; another person can only surmise it.” This is a
characteristic attempt to put into words the thought that underlies much of the picture of privacy
Wittgenstein will oppose. This thought is perhaps the core idea of the private language
conception, underlying not only the idea of privacy but also immediacy and incorrigibility, etc.
Wittgenstein will criticize it in much the same way he criticized the propositions of philosophy in
the Tractatus: as attempts to say what is actually only tautologous, as attempts to give empirical
expression to truths that are not characteristic of anything empirically, but are really just
“illustrated turns of phrase” of our language.

“Only I can know I am in pain.” In one sense this is false. Very often another person can know
that I am in pain; for instance if I tell them, or cry out. But now we will feel as if, whatever the
other person knows, he cannot know it with the immediacy and certainty that I know it. But what
justifies me in saying that I know that I am in pain at all? This just means that I am in pain.
Talking about “knowing” doesn‟t add anything, even on the incorrigibilist‟s picture. Here, the
crossing of pictures, or the turns of phrase our language allows, confuses us. We can talk about
another person‟s knowing or not knowing I‟m in pain; but this makes us think it must make sense


                                                   8
to talk about our own knowing or not knowing we are in pain. But really it makes no sense to say
that I am in pain but I don‟t know it. (That is, unless we take „pain‟ to mean something like
„objective bodily damage‟, and then there is again no problem). Here, the language has forced a
picture on us; we sublate this picture, and apply it as the picture of subjective certainty. Here, it
doesn‟t make sense to say that I doubt that I am in pain; but this doesn‟t mean that I am certain
that I am in pain. Doubt just doesn‟t get a foothold here. It doesn‟t have a role in the language-
game in this way; so far we haven‟t given it one.

Here language misleads us in a way that is typical for it. (See 251). We say that we can‟t
imagine the opposite, but here we‟re just mistaking a grammatical proposition (a proposition
about what our language allows; a tautology) for an empirical proposition.

253. The interlocutor tries a different tack. “Another person can‟t have my pains .” Here, the
emphasis isn‟t on knowledge, but rather on the proprietary character of pains. This thought – that
another can‟t have what I have – might even be taken as a definition of the subjectivity of the
subject. But again, here we have a grammatical proposition masquerading as an empirical one.
“In so far as it makes sense to say that my pain is the same as his, it is also possible for us both to
have the same pain.” The proposition seems to evoke sameness and difference of pains, but when
we get to asking what the criteria for sameness and difference are, we find that we had no
particular criteria in mind at all. Either we knew what would count as „same‟ and „different‟ – in
which case we could after all talk about the two of us having the same pain – or we didn‟t; in
which case what we said wasn‟t meaningful.

256. Again, what about the language that is supposed to stand for my inner experiences and
which only I can understand? The question is still: how do I set up the connection between words
and sensations. If I use these words as I ordinarily do, they are bound up with our natural
expressions of pain and they are as understandable as any words. But now the interlocutor
suggests the possibility that there is no natural expression, and I simply come up with a new name
to stand for a particular sensation.

258. Now we get the private language scenario in its clearest case. I keep a diary and write down
“S” every time I have a certain sensation. Now: what is the definition of “S”? It can‟t be defined
in a way that could be explained to someone else; otherwise it wouldn‟t be private. But it can still
seem that I can give it a kind of “private definition” by way of focusing my attention on it
privately. This kind of “private ostension” has often seemed to be the essence of our relationship
to experience, of the way in which experience could be the basis of language. But
characteristically, Wittgenstein immediately asks what this ceremony of private attention-
directing accomplishes. In particular, we have not yet given ourselves a criterion of correctness.
Whatever seems to me to be right is right. So here we can‟t talk about getting the sensation
wrong – but we can‟t talk about getting it right either. We don‟t have a criterion that would give
us the difference that this kind of talk would demand. So far, this mark “S” has no function.
(260).

We are seduced into thinking that the ritual of writing “S” can suffice to name a sensation,
something that is really private to me. But if we can speak of naming and re-identifying
something, we can speak of getting it right or wrong; but here we can‟t do so. Again, why should
we think that the writing of “S” should be described as the naming of a sensation at all?
“Sensation” is, after all, a word of the public language, in which we have criteria that are statable
and describable for the identification and description of sensations. (261). What we want to say,
in calling the ritual the naming of the sensation, is really something that we can‟t say at all, even
by our own lights.


                                                  9
We think that – as it were – we can make the difference between correctness and incorrectness
by appealing to the sensation itself. But this is just what is at issue. For an appeal that tests
correctness to make sense, we have to appeal to something that could itself be tested for
correctness (265). But here the case is conceived in such a way as to rule just this out. If the
language really is a private language, and the sensation it names genuinely private, there is
nothing else to which we can appeal to tell whether we got our reference to it right. Our only
reference is our decision, and the correctness of our decision drops out of the picture.

In order for naming to take place – in order for the minimal difference it presupposes between
correctness and incorrectenss to obtain – we have to connect the ritual to further practical
consequences. (The naming of the sensation is, so far, like the „left hand giving the right hand
money‟ (268) – we can‟t suppose, even hypothetically, that this takes place, unless we say what
difference it would make). So let‟s imagine a real use for the entry “S” (270): I find that on every
day on which I enter „S‟, my blood pressure is up. Now I can use my entering of “S” to predict
my blood pressure; now the sign has a meaning. But now – precisely where I wanted to give my
naming of the sensation meaning by attaching it to a further significance, the sensation itself
drops out. It now doesn‟t matter whether I have identified the sensation right or wrong; what
predicts the blood pressure phenomenon is not whether the sensation is there, but just whether I
think it‟s there. And this alone shows that there was really no sense to the hypothesis that I might
“get the sensation wrong.” In this sense, the sensation doesn‟t even have the minimal
independence of our practice of describing it that would allow us to take ourselves to be
describing it at all. Our only reason for thinking that the same sensation was there all along was,
in fact, that we write the same letter. But this just shows that we wrote the same letter.

At 272, W. sketches the possibility that is often called the „inverted spectrum.‟ We can‟t know
whether everybody – or anybody – else sees the same thing (has the same „experience‟, the same
quale) when I see green as I do. It could be that, where I see green, everyone else sees (what I
would call) red. But they learned all the characteristic behaviors that go with the public objects,
so the difference doesn‟t show up behaviorally. So it can seem as if there must be qualia –
private exemplars – even if we could never tell that there are. But there is a confusion here,
insofar as the inverted spectrum case makes us think that “red” must have (as it were) two
meanings – one for the public objects (fire engines, apples) and one for the private exemplar. But
the second meaning is idle (274). Saying this doesn‟t help us to understand what „red‟ means.
Even when we use expressions like “look at the blue of the sky” (275) we don‟t mean to point to
something inner – unless we‟re doing philosophy. Here, as usual, we take ordinary propositions
like “I know how the color green looks to me” (PI 278); “I know what it‟s like to be an
unemployed bachelor” ; and then we give them the strangest kind of significance.

280. Painting a picture for others; using it as a “picture” of my private image. In what sense do I
do the latter? it‟s not a representation or a piece of information.

281. Behaviorism. Interlocutor asks if W. is a behaviorist; but the point is different. It‟s that
only of a living being and what resembles one can we say: it has sensations, it sees, etc.

281-283. First, we project onto the other, imaginatively. Then, we ascribe pains.

288. Can we doubt that we are in pain? It makes no sense to doubt that we are in pain. It can
only seem to make sense – for a minute – when we think of ourselves, e.g., turned to stone and do
not know whether we would still be in pain or not. But that does not mean that we ordinarily
know, with absolute certainty, that we are in pain. “The expression of doubt has no place in the


                                                 10
language-game; but if we cut out human behavior, which is the expression of sensation, it looks
as if I might legitimately begin to doubt afresh … My temptation … arises from this: If I assume
the abrogation of the normal language-game with the expression of a sensation, I need a criterion
of identity for the sensation…”

289. When we say we are in pain, we say so without justification. But this does not mean that it
is unjustified. “To use a word without a justification does not mean to use it without right.”
Ordinarily, we are justified in saying that we are in pain, etc. But does this mean that we have to
tell a story about the source of our justification: either that we are justified by the presence of a
private object, or that we are justified by the tolerance of our peers, their agreement to allow us
these expressions. Here, we don‟t describe something – at least not in the way a I describe a
machine – because I don‟t use a diagram for a particular purpose (291) (cf. TLP picture theory).

293. The beetle in the box. This is one of the more famous remarks of the Investigations, despite
the fact that it has been unclear what it means. As far as the public language is concerned, the
private item is like the beetle in the box, which nobody else can see. Now, what about: “I know
what the beetle is only from my own case.” But as far as the public language is concerned, the
beetle could be anything – or nothing. “One can „divide through‟ by the thing in the box”. That
doesn‟t mean that there‟s nothing there – but just that nothing at all would do as well as what is
there. “If we construe the grammar of the expression of sensation on the model of „object and
designation‟, the object drops out of consideration as irrelevant.” Now, note that this does not
mean that it is wrong to talk about sensations. The claim is just that when we do so, we are doing
something other than designating an object with a word or name. Our language game with
sensations doesn‟t consist in doing this; it‟s something else. Wittgenstein seems to want to get us
to see the real complexity of our language game, over against our tendency to think of it as the
relatively simple designation of objects by words.

If the inner sensation is like the beetle in the box, we really have no idea of what is before us
when we think of ourselves as using the private language to describe our sensation (294). What
reason do we have anymore to say that we have “something” at all? But it is not as if we do not
have this picture: we have the sense – we think of ourselves – as having inner objects. [These
remarks, which are often ignored, are actually important in showing that Wittgensteins is not (as
it were) denying the reality of the „inner‟ or the „mental.‟] When we say, for instance, that “I
know … only from my own case” … we don‟t give any information. The idea that we thereby
introduce an informative sentence into the language is wrong. But we might still want to say this;
it might be a kind of exclamation; or a “grammatical” summary of “what we do.” Its source is the
ordinary phrases of language and our confused interpretation of their significance. “When we
look into ourselves as we do philosophy, we often get to see just such a picture. A full-blown
pictorial representation of our grammar. Not facts; but as it were illustrated turns of speech.”
(295)

What happens when we think of the private language as informative is like what Wittgenstein
imagines in 297 (the „parable of the pictured pot‟). We have a picture – the picture of the „inner
object‟ – and it has its uses for us. But we are tempted to think that there is more to it than this,
that there must “really be” something inside – and not just in the “ordinary sense” of „inside‟ –but
in the special sense that our sublimation of our own picture makes us believe in. “The image of
pain certainly enters into the language game in a sense; only not as a picture” (300). It is
important – essential to our language game – that we can talk about our own pains and those that
others have, that we can know when someone else is in pain, and also that we can conceal or hide
our pains – at least some of the time – if we want to. But what supports this is not the picture of a
pain (like the picture of boiling).


                                                 11
What we have in the language-game with sensation, so it seems we have to say, “is not a
somehting, but not a nothing either! (304). The conclusion was just that if we conceive this
object on the model of object and designation, it drops out. But Wittgenstein does not want to
deny that there are sensations; only there grammar is not as the picture of private language would
make it appear. To see the real grammar, we need to break radically with the idea that there is
only one form of language – description – and that we use this form in the most various contexts.
Our language game with pain is not a descriptive one. And pointing this out allows us to see its
complexity.

Wittgenstein can seem to be denying something substantial – for instance the existence of
sensations. But all that he is denying is really a certain picture of them, that can seem to be
forced on us by our grammar (305). Are we denying, then, that there is an „inner process‟? But
to say that I go through an „inner process‟ of remembering is just to say that I remember. And of
course we don‟t deny this. All that we deny is that the notion of the „inner process‟ helps us to
see the use of the word „remember‟ as it is. The picture forces itself on us – it is an artifact of
philosophical inquiry, aided by the misinterpretation of certain turns of phrase. Once we see the
real use of these terms, however, we can demystify the picture, and return to a clear
understanding of the way we actually use these terms.

Behaviorism and the “myth of the given.” Analytic philosophy began with the phenomenalist
hope of reducing our empirical knowledge to a foundation of what is immediately given, the
immediate character of primitive, basic experiences or sensations. Russell and Ayer‟s sense-data
theory and the logical positivists‟ epistemological foundationalism alike saw our knowledge of
external objects as reducible to solipsistic, immediate knowledge of items known only to one
person. The logical positivists actually engaged in a decisive debate about the nature of the
“protocol sentences” that were supposed to express the immediately given: what was it that
verified these sentences, and how could we conceive of the contact between a sensation or a
“given” fact and a linguistic item? The debate proved decisive in changing the minds of several
of the logical positivists, almost to the opposite position, a physicalistic one according to which
there are no private or solipsistic facts, but all facts are from the beginning part of the public,
intersubjective language of science. [W. accused Carnap of plagairism]

Starting in the 1930s, Carnap and Hempel articulated a “logical behaviorism” according to which
what it means to say that someone is in pain, is having a sensation, etc. just is the same as what it
means to say that they are exhibiting characteristic behaviors. There are analytic identities
between sensation-ascribing sentences and behavior-ascribing sentences. From the beginning,
though, logical behaviorism was problematic. For can‟t we hide our pains and sensations? And
why should we think that our way of knowing our own pains is by observing our behavior?

Following on Wittgenstein, in the 1950s several philosophers articulated critques of the
traditional sense-data theory and of parts of the picture of experience that it represents. None of
these critiques are actually examples of logical behaviorism, though they have often been read as
such: Sellars “Empiricism and the Philosophy of Mind” Ryle “The Concept of Mind”; Austin
“Sense & Sensibilia”. Sellars criticized the so-called “myth of the given,” the myth of immediate
experiential episodes that have their own kind of content, and are simply “given” to
consciousness without further interpretation. For Sellars, our discussion of the supposed “given”
faces a insuperable dilemma: either the “given” is already in the logical space of concepts – in
which we must think of it as responsive to our conceptual faculties and the product of
interpretation, in a relevant sense – or it is not, in which case it is a mystery how it could
rationally contrain or interact with other, higher-level propositions. To treat givenness as in the


                                                 12
logical space of concepts – to treat the content of the given as articulated into phrases like “there
is a red dot here now” is essentially to admit the dependence of givenness on our conceptual and
interpretive abilities, on the same conceptual abilities that underlie our use of the public words
“red,” “dot,” etc. But to deny this is to make the very content of givenness deeply mysterious, a
kind of bare presence incapable of further understanding.

Ryle further articulates the critique of the sense-data picture as part of his own linguistic critque
of the Cartesian dualistic picture of mind. For Ryle, the central error of the sense-data picture is a
“category mistake” or grammatical error: it treats perception (of external objects) as if it were
dependent on a kind of observation (of inner objects). Really, there is just perception of external
objects. [consider, e.g., why it makes sense to say that I see a table better or worse, more or less
clearly; but it doesn‟t make sense to say the same about my sense-datum.] But what, then, is the
nature of sensations? Ryle opposes the Cartesian picture of sensations as elements of a „private
mental life,‟ a kind of second biography that goes alongside the public one. For Ryle, to say of
someone that they have a sensation (e.g. a toothache) is to say something, rather, of a
dispositional nature: that is, that they are likely to cry out, to go to the doctor, etc. But it isn‟t to
say that the toothache just is these behaviors. The behaviors are the criteria for the toothache.
To say that someone has a toothache is not to gain access to some puzzling, inner strata of their
mental life, but rather to say something perfectly public and knowable, under ordinary conditions.

All of these responses have been read as a kind of logical behaviorism that denies the reality of
the mental. But they really aren‟t. The reason they aren‟t is that they are rooted, not so much in
denying the reality of some object or phenomenon, as in noting the way that our ordinary
language of sensation actually works. As far as this language goes, talking about pains,
tootheaches, and even sensations (in particular cases) is perfectly all right, and indeed essential to
our form of life; though talking about “sensations” in general is unordinary and questionable.
What all of these philosophers have in common is not the thought that the publicity of language
somehow shows that there is no such thing as subjectivity, but that it is to the public language that
we must look if we want to understand what subjectivity is like. It has, on this story, only been
the egregious failure of philosophers to look at this ordinary language that has caused them to
indulge characteristic fantasies of privacy and solipsism; (althouth the roots of these fantasies are
themselves in our grammar). Looking to language as it actually functions, however, we can begin
to look toward a conception of subjectivity that does not identify it with interiority and privacy,
but rather with the concrete kinds of communication and mutual understanding that are possible
for us. On this story, subjectivity is not so much a matter of my being able to know my own
pains, as my being able to acknowledge the pains of the other, something that is just as important
for our form of life. But this is not to say that all information, all language is simply the objective
description of objective facts. There is, rather, a dialectic of recognition and acknowledgment
that runs as deep in our language, and behavior, as any picture of subjectivity plausibly could.
The turn toward language allows us to see this dialectic, and to resist the characteristically
totalizing phiosohpical pictures of subjectivity that deny it, themselves products of a
mischaracterization of language.

Today, the debate goes on in the form of the problem of explaining consciousness on the two
most prominent theories of mind, physicalism and functionalism. Physicalism holds that mental
states are really physical brain states: for instance particular configurations of neuropsychology.
Functionalism holds that mental states are functionally described states of the brain, for instance
having a causal role, a place in a typical pattern of cause and effect. Functionalism is amenable to
physicalism, for it doesn‟t have to introduce any ontology beyond physical facts and events and
their properties. But then it can seem that neither theory can explain consciousness: can explain,
e.g., what it‟s like to be seeing red; or the raw feel or quale of blue. Nothing like a physical fact –


                                                   13
so we might want to say – can so much as begin to explain that we experience subjective feels.
So the question becomes one of explaining these extra, further facts. (Chalmers).
This problem defines contemporary philosophy of consciousness; but over the years it has come
to seem (to me) increasingly illusory. One reason for insisting on consciousness as recalcitrant to
physicalist or functionalist description is that to talk about consciousness is not – it‟s right to say
– to talk about physicalistic facts like brain states, or indeed any other „objective‟ facts. Our
language game isn‟t like that. But it‟s also not the kind of thing Descartes assumes, the
description of some inner realm or strata. Reflecting on the role of the language of subjectivity in
our lives, by contrast, begins to give us a way into a picture of subjectivity that is not physicalist,
but is not Cartesian dualism either. Indeed, there is a way of seeing physicalism itself as a kind of
sleight of hand: a characteristic assumption of totality that results when we see all of our language
as functioning in one way, and then become puzzled as to how it can do what it seems it must do
in the case of consciousness. But if we remember the complexity of our language and its rich and
complex interweaving with our form of life, we can begin to see past this picture, and the
problem that it seems to require.




                                                  14
1) I think that Wittgenstein‟s concern throughout the passages we read is not to show that
there is no coherent notion of privacy, but to show the attempt to formulate such a notion
– or use it in philosophical theorizing – is incoherent. He will show this,
characteristically, by imagining various ways of formulating the statements with which
we want to articulate a notion of privacy (for instance: “Only I can know whether I am in
pain” or “sensations are private”) and show that these turn out not to have the substant ial
meaning that the theorist thinks they do. Insofar as they are true, they are grammatical
(logical) truths (PI 248); if we take them otherwise, we‟ll be misled.

2) The “private language argument” is best seen, not as comprising a positive doctrine,
but as a set of interrelated criticisms of a particular view that plays a historically
important role in empiricist epistemology and sense-data theories, and is quite tempting
in its own right. Minimally, the view involves the following three commitments:
i) A „private language‟ is a language, or part of a language, that describes and reports the
status of „private‟ items – those that only the describer, necessarily, can have or know.
(“Sense-data” and qualia, on most theories of them, are such items). Such a language is
one that only the describer can understand. (PI 243)
ii) Some terms in the private language function as names – that is, they sustain direct
referential links to their referents. (PI 256, 258)
iii) Sentences or propositions in the private language can interact rationally with (e.g.,
rationally constrain) ordinary public-lang. statements.
The argument will be against the specific view that comprises all three commitments. It
doesn‟t, of course, aim to exclude anything that anyone would ordinarily say about their
sensations, pains, dreams, etc. Some semi-ordinary locutions are exposed as misleading
– for instance the use of “I know I am having a pain” rather than “I am having a pain” –
but only because they can tempt us to the picture invo lved in i)- iii).

There are already various ways to put pressure on the conjunction of the three
commitments. One is to point out (PI 260) that the private language is supposed to name,
e.g., _sensations_; but “sensation” is a word in our public language. Again, if one takes it
that the private language uses terms like “red”, we will have to say whether they have the
same meaning in the private language as in the public one; if the same meaning, then we
violate i) ; if different, then the private- language sentence gives us no rational inference
to the public- lang. one; but this violates iii).

3) A more subtle line of challenge questions the possibility of ii) by asking how the
private linguist sets up and supports the envisioned referential links. The natural thought
is that the link is set up by a private ostension or focusing of attention on the private item.
(PI 262, 263) The private linguist believes that by doing this he can introduce a name
which directly refers to the item. But now we can put pressure on his claim to have set
up the requisite link. For: a) if the name sustains direct reference to the item, there is a
fact of the matter, in each case, about whether the name is used correctly or not. But
there is no obvious introspectible difference between having a sensation and only
_thinking_ one has the sensation. (PI 270, 271). (Appeal to cases in which we might
want to say that somebody has a pain, but does not know it (or vice-versa) doesn‟t help
here, because the criteria in virtue of which we want to say that they have a pain in these



                                              15
cases are public criteria.) And: b) more generally, the assumption of a direct referential
link between the sign and the item has no use. For anything we can *do* with the sign –
anything we can use it to significantly say – we could still do if the link did not obtain (if,
for instance, the sensation itself were constantly changing, and we kept mistaking its
character) (PI 271).

4) This kind of criticism of the private language theorist is not exclusively
“verificationist.” (Indeed, at P. 272 W. explicitly sketches the inverted spectrum as
“possible, though unverifiable”). It doesn‟t rest (only) on the absence of ways of
knowing or verifying the being/seeming difference for sensations. It rests on the t hought
that the private diarist hasn‟t yet given the sign S a _use_. (And we can‟t say that the
diarist has given the sign a “private” use. For then we don‟t get iii) above). There is
admittedly a background principle about meaning at work here; but it seems quite
plausible. All that we need is that any sentence that says something at all does so by
having an inferential role, by constraining the rational inferences that can be drawn from
it (this is a general point about the relationship between meaning and inference; it‟s more
than just a point about pragmatics). Sentences involving “S” don‟t yet do so, until we
give them an inferential significance.

5) The conclusion (P. 293) is not that there are no sensations, but rather that our reference
to them is not direct reference; it‟s an ordinary possibility in ordinary (public) language.
Our ability to make such reference doesn‟t depend on our ability to introspect the
characteristics of the private items. Even if everyone‟s beetle is different – or if there are
no beetles at all – the public word “beetle” means the same for everybody. (Of course,
someone could still maintain that his use of the word “beetle” is grounded in the
introspectible characteristics of _his_ beetle. But it would now be mysterious why
anybody would want to maintain this, or what they meant by it.)

6) The argument doesn‟t cite any absolute inconsistency for the very idea of a private
language. If the private linguist is sufficiently attached to his picture, he can still
maintain it. (For instance, he can insist that, even if there‟s no way to tell the difference
between being and seeming for sensations, still there _is_ a difference; and that even if
the direct reference fails to set up a public use, still it‟s reference all the same). But the
hypothesis of a private language is now quite useless for any theoretical project in
epistemology; and the private language theorist ought to ask himself why he still wants to
maintain it.

143, 185, 186, 208, 214, 215, 241-242, 246




                                              16