Wittgenstein and Idealism
Abstract: I argue, on both textual and philosophical grounds, that, contra GEM Anscombe and David
Bloor, it is misleading to characterize the Wittgenstein of the Philosophical Investigations as a linguistic
idealist. First, such an interpretation conflicts with Wittgenstein’s therapeutic method; second, the textual
evidence adduced in favor of this reading is ambiguous at best, and finally, it misses the fact that
Wittgenstein rejects the underlying assumption, common to both realism and idealism, that language and
reality can be clearly distinguished.
Was Wittgenstein an idealist? Some commentators, most notably GEM Anscombe and,
following her, David Bloor, argue the he was. They attribute to him a view they call ‘linguistic
idealism’, which says, briefly, that some things are created by, and dependent for their existence
on, human linguistic practices.
One problem with claiming that Wittgenstein was an idealist is that there is ample
evidence from the remarks Wittgenstein makes in the Philosophical Investigations and elsewhere
that he saw philosophy as essentially a matter of showing that philosophical problems, and
therefore the theories advanced to solve them, are based on ‘a false picture’ of how things are, or
how they must be. Once this false picture is exposed, both the problem, and the proposed
solutions, are dismissed. Thus, he tells us that philosophy consists in ‘showing the fly the way
1
out of the fly-bottle’, and that philosophy should be understood as something akin to
psychological therapy. By showing the philosopher that certain basic, unexamined, a priori,
questionable assumptions generate philosophical problems, and showing that these assumptions
are not forced upon us, philosophy, perhaps paradoxically, allows us to stop philosophizing when
we want to.
It would be surprising if, having claimed a therapeutic method for philosophy in the way
just described, Wittgenstein were to go on to take sides in the debate over realism and idealism.
What is more likely is that both the idealist and the realist stand in need of philosophical therapy,
and that neither view is ultimately satisfactory because the debate itself is the product of a
misleading picture of the relationship between language and the world.
Bloor defines linguistic idealism as follows:
…the claim that some truths or realities are created by our linguistic practices. The contrast is
with cases where language transmits or reflects an independent reality. 1
So the distinction between realism and idealism is the distinction between the view that language
reflects or mirrors or maps an independent existing reality, whereas idealism is the view that
reality is at least in part a creation of language. According to Bloor, Wittgenstein was a full
linguistic idealist with regard to some things, and a partial linguistic idealist with regard to
others. To be a partial linguistic idealist with regard to some concept is to think that the objects
that fall under the concept exist independently of human thought or belief, even though the
linguistic practice or concept is a human construct that cannot be ‘read off’ the reality. For
instance, horses (to use Anscombe’s example) would still exist even if human beings failed to
recognize or interact with them at all. But the concept horse is all the ways in which horses are
David Bloor, “The question of linguistic idealism revisited”, in Sluga,Stern (eds) The Cambridge Companion to
Wittgenstein 354-382.
1
2
incorporated into human social practices, for instance, in the institutions of racing and sport, of
farming, and as pets. So while the concept of horse is humanly created and constituted by all the
ways in which humans use horses, horses are not. With regard to natural objects, Wittgenstein’s
supposed idealism is only partial, because although the concept in some sense creates the kinds
of horses there can be (for instance work horses or race horses or even horses-as-artworks1),
horses exist even if the concept of horse does not.
To be a full linguistic idealist with regard to some thing is to think that the objects that
fall under the concept would fail to exist independently of human convention. Laws, languages,
economies, political systems, games, and all human institutions and their objects exist only
within and as a product of human agreement. For example, without the convention of making,
keeping, and enforcing promises and contracts, there would be no such thing as promises or
contracts. Promises and contracts are not, as horses are, the sort of things that are simply waiting
for humans to come along and incorporate them into their activities and institutions. Thus, with
regard to at least some things such as promises, Wittgenstein’s supposed idealism is full.
Bloor characterizes Wittgenstein’s insight as follows:
Of the various shortcomings in what he [Wittgenstein] called the “object and designation
model”, the most dramatic is its inability to deal with the case where the “object” has no
existence at all outside the talk about it and the reference to it. It breaks down entirely where our
talk is about other talk, and the linguistic practice has a wholly self-referring character. This is
where the label “idealism” has its most plausible and revealing application.
Thus, for Bloor, Wittgenstein’s realization that there are cases in which language creates a reality
that did not previously exist undermines the object and designation model of language that he
previously held. It does so because that model treats language, to put it crudely, as a kind of
picture or map of the already existing world. If some bits of the world are created by our
language, as in the case of promises or runs batted in, this shows that the object and designation
3
model of language is to be rejected. On this view, Wittgenstein’s idealism is part of his rejection
of his early picture theory of language.
It is fairly clear that Wittgenstein does indeed reject the object-designation model of
language. What is not clear is that this commits him to anything like linguistic idealism, for that
leaves him still too much in the grip of the idea that language and reality can be clearly
distinguished such that we can meaningfully ask what the relationship between them is. It is my
interpretation of Wittgenstein that he aims to reject this distinction, and with it the philosophical
problem to which idealism and realism are supposed to be solutions.
What textual support do Anscombe and Bloor give for their interpretation of
Wittgenstein? First, there is this passage, which occurs at the end of the Investigations:
If the formation of concepts can be explained by facts of nature, should we not be interested, not
in grammar, but rather in that in nature which is the basis of grammar?- Our interest certainly
includes the correspondence between concepts and very general facts of nature. (Such facts as
mostly do not strike us because of their generality). But our interest does not fall back upon
these possible causes of the formation of concepts; we are not doing natural science; nor yet
natural history – since we can also invent fictitious natural history for our purposes.
I am not saying: if such and such facts of nature were different people would have different
concepts (in the sense of a hypothesis). But: if anyone believes that certain concepts are
absolutely the correct ones, and that having different ones would mean not realizing something
that we realize –then let him imagine certain very general facts of nature to be different from
what we are used to, and the formation of concepts different from the usual ones will become
intelligible to him…
(PI xiii)
Bloor rightly wonders where in the passage Anscombe sees support for linguistic idealism, since,
as he rightly suggests, the natural interpretation of the passage would seem to be directly at odds
with the claim that some realities are humanly constructed. Indeed, the phrase ‘general facts of
nature’ would seem to suggest that there are facts independent of language. The point seems to
4
be while the idealist is right to remind us that that our concepts are not simply ‘read off’ the
general facts of nature, they are nevertheless dependent upon and constrained by them, such that,
in a different world (one in which different general principles hold), different concepts would
become imaginable. No doubt the converse is also true: were the general facts of nature
different, certain concepts we now have would not be imaginable. That sounds realist as much
as it sounds idealist, insofar as it affirms the underlying realist impulse to remind us that there is
a world that is, and that it is within this world, with these constraints, that human life and
language has evolved.
Bloor appeals to PI 142 in favor of the idealist interpretation:
It is only in normal cases that the use of a word is clearly prescribed; we know, are in no
doubt, what to say in this or that case. The more abnormal the case, the more doubtful it
becomes what we are to say. And if things were quite different from what they actually are – if
there were for instance no characteristic expression of pain, of fear, of joy; if rule became
exception and exception rule; or if both became phenomena of roughly equal frequency –this
would make our normal language games lose their point. The procedure of putting a lump of
cheese on a balance and fixing the price by the turn of the scale would lose its point if it
frequently happened that such lumps suddenly grew or shrank for no obvious reason…
Here again, though, Wittgenstein seems to want to insist that our concepts are not arbitrary, that
they are at least constrained by the general facts of nature, so that, for instance, if objects such as
hunks of cheese were apt to randomly change their size and weight, the practice of selling them
by weight would not be possible. If anything, this captures the realist’s main insight that the
world exists independently of our concepts.
Thus the textual evidence for Wittgenstein’s supposed idealism is ambiguous at best.
Furthermore, the distinction between full and partial idealism is questionable. Even in the case
of promising, the general facts of nature constrain what is possible. Consider, for instance, the
following passages from Nietzsche’s Genealogy of Morals:
5
To breed an animal that is entitled to make promises—is that not precisely the paradoxical task nature
has set itself where human beings are concerned? Isn’t that the real problem of human beings? . . .
... In order to organize the future in this manner, human beings must have first learned to separate
necessary events from chance events, to think in terms of cause and effect, to see distant events as if
they were present, to anticipate them, to set goals and the means to reach them with certainty, to
develop a capability for figures and calculations in general—and for that to occur, a human being
must necessarily have first himself become something one could predict, something bound by regular rules, even
in the way he imagined himself to himself, so that finally he is able to act like someone who makes
promises—he can make himself into a pledge for the future!
As Nietzsche points out, promising assumes that human beings can distinguish causes from
coincidences, that they can predict and plan, that they can calculate, that their behavior can be
made predictable, that they fear punishment. Were these things not so, promising would not be
possible. They are themselves dependent on the very general fact that humans are pliable, and
can be bred in certain ways. These are examples of general facts of nature, and they bear the
same relationship to our ability to make conventions as horses bear to horse racing or the
cavalry. In both cases, the fact that the world is such-and-such a way makes it possible for us to
have the practices and concepts we have. And were these facts to be radically different, if there
were no large animals that could bear human weight, or if human behavior were flighty and
unpredictable or incapable of being manipulated the corresponding practices would cease to
exist. Calvary horses and promises both depend on human convention, though horses and human
nature exist independent of such conventions. So while there are many different kinds of
institutions and practices that can be ‘built up’ around horses and human beings, there are limits,
imposed by nature, in what is possible. There are no ‘fully idealistic’ human concepts.
My conclusion so far is that Anscombe and Bloor have not made the case that
Wittgenstein can profitably be considered an idealist, even in cases like promising. This is not to
say that Wittgenstein is therefore a realist. Things are not so simple: while it is true that human
concepts create new realities (here is the idealist’s fundamental intuition), their ability to do this,
and what realities are possible, is itself constrained by the very general facts of nature which hold
6
independently of our concepts (here is the realist’s fundamental intuition). And here there are
both elements of realism and idealism, but no commitment to either. I want to now suggest that
Wittgenstein rejects the underlying picture of language upon which both realism and idealism
rely.
Recall Wittgenstein’s famous example of the builders, who appear near the beginning of
the Philosophical Investigations. Wittgenstein makes it clear that the example of the builders is
supposed to help cure us of the idea, (explicit in the allusion to St. Augustine that opens the
Investigations), that words are names for things. Wittgenstein asks us to imagine two builders
(A and his assistant B); A is building with building stones. The stones consist of blocks, pillars,
slabs, and beams. When A wants a pillar, for instance, A calls out “pillar!”; it is B’s job to pass a
stone to A when he hears A call for one, etc. The example of A and B is an example of what
Wittgenstein calls a “language game”.
…I shall also call the whole, consisting of language and the actions into which it is woven, a
“language game”. (PI7)
The phrase “language and the actions into which it is woven” gives us an important clue in
uncovering the picture of language that Wittgenstein holds in the Philosophical Investigations. The brick,
the men, the sounds the men utter, their facial and physical gestures, all of these things are woven into a
seamless whole that constitutes the language game of building a structure. But now, can we say with
certainty which parts of this story constitute the linguistic parts and which constitute the world? Later in
the Investigations Wittgenstein asks us to imagine a more sophisticated language game, in which
(among other things) a set of colored slabs are incorporated into the game, so that B can distinguish slabs
according to their color. Wittgenstein says at P16
What about the colour samples that A shews to B: are they part of the language? Well, it is as
you please. They do not belong among the words; yet when I say to someone: “Pronounce the
word ‘the’”, you will not count the second “the” as part of the sentence. Yet it has a role just like
7
that of a colour-sample in language game (8); that is, it is a sample of what the other is meant to
say.
It is most natural, and causes the least confusion, to reckon the samples among the
instruments of the language. 2
The word “the” in the sentence to which Wittgenstein refers is the object of the sentence, a
sound-token. If we had to choose whether to characterize it as part of ‘language’ or part of the
‘world’, it is not clear what we should do. On the one hand, insofar as it is a word, it generally
counts as part of the language; on the other, insofar as it is the subject of the statement, and not a
word in the sentence, it counts as part of the world. Similarly, the color samples are, on the one
hand, simply bits of colored cardboard, and, as such, would seem to most comfortably fit in the
‘world’ category. And yet, they function as symbols for their respective kinds, which seems to
put them in the ‘language’ category. There seems to be no clear criterion for deciding, in either
case, what the correct classification would be.
Artistic examples help illustrate the point. Ernest Gombrich’s “Truth and Sterotype”
provides us with a good example, this time in the context of a discussion of Plato’s distinction
between the carpenter, on the one hand, who makes a bed, and the artist who, on the other,
merely imitates it:
Must it always be true that the sculptor’s couch is a representation? If we mean by this term that
it must refer to something else, that it is a sign, then this will surely depend on the context. Put a
real couch into a shop window and you thereby turn it into a sign. It is true that once this is its
only function, you may choose a couch which is not good for anything else. You may also make
a cardboard dummy. In other words, there is a smooth and even transition, dependent of
function, between what Plato called “reality” and what he called “appearance”.
The point of these examples is that the border between language and reality is a porous one.
Bits of what we would normally characterize as part of the extra-linguistic world easily become
linguistic symbol. Linguistic symbols can be the subject of discourse, and thereby part of the
2
Ludwig Wittgenstein, PI 16 7e
8
world the discourse is presumably about. Is the bed in the window part of the world, or part of
our language? Again, it is as you please; there are grounds for claiming both.
These examples remind us that, at any given time, we cannot neatly and clearly draw the
line between what is, and what is not, part of language as opposed to part of reality since, as
Gombrich points out, that depends on the context. We may think that this line can or must be
neatly drawn if we think of the relationship between language and reality as either a relationship
of the one mirroring the other or the one creating the other. But once we see that this distinction
cannot be neatly maintained, the idea that the world on the one hand and language on the other
are two things such that they can have a relationship, be it mirroring, or creating, is called into
question.
9