What Use Are Wittgenstein’s Language-Games?

Introduction

Wittgensteinians have a tendency to say that it is time to stop merely analyzing Wittgenstein’s works and start applying his method ourselves.[1]  There is little sign of their doing so, however, and one reason why Wittgenstein has fallen out of favor with many analytic philosophers is that he seems to offer little that can be used.  I want to offer an explanation as to why this might be.  I will also try to show just how difficult it is to do philosophy as Wittgenstein thought we should.  In particular, I will try to bring out the problems inherent in the concept of language-games from the point of view of philosophers other than Wittgenstein himself who might want to use his concept.

Wittgenstein does not, I will argue, offer a theory of language-games.  Nor does he say enough about what they are or what we can do with them to allow us to be at all sure that we are using the concept as he intended.  This is not to say that they play no role in Wittgenstein’s work, nor that we cannot make use of something that we might call language-games.  My point is simply that it seems to be virtually impossible for others to do what Wittgenstein did with them.  In this paper I will explain what language-games are in the context in which Wittgenstein introduces them, then look at both the theoretical and the methodological/therapeutic interpretations of the concept before reaching my largely negative conclusion that what language-games are and how they might be used outside the context of Wittgenstein’s own work is shrouded in obscurity.

Wittgenstein’s Philosophical Investigations starts with a discussion of language and quickly introduces the notion of “language-games.”  The idea has become one of Wittgenstein’s most famous, and is clearly important to anyone interested in the subject(s) of logic, games and philosophy.  But what is the point of the idea?  I think there are two obvious possible answers: that Wittgenstein has a theory of language that somehow involves what he calls “language-games” and, alternatively, that Wittgenstein offers no such theory but uses the concept of a language-game as a kind of tool to achieve some therapeutic goal.  I will argue that neither answer is as plausible as its defenders might wish.

            Wittgenstein is widely taken to present a kind of antirealist philosophy (or theory) of language in which contingent human practices, i.e. language-games, play a central part.  Theories of this kind have various undesirable features, as realists and anti-relativists are fond of pointing out.  This in itself does not mean that Wittgenstein was not committed to antirealism, of course, but it does give us a motive to try, for the sake of charity, to find some other interpretation of his work.  And evidence that he did not have a theory of language-games can be found in such passages as §7 and §109 in the Investigations.  The latter is one of the main places where Wittgenstein insists that advancing any kind of theory is not what he means to do.  The former is where he defines language-games as 1) “games by means of which children learn their native language,” 2) primitive languages, 3) processes such as those of naming different stones “blocks”, “slabs”, and so on, or repeating words after someone else, 4) much of the use of words in games such as ring-a-ring-a-roses, and 5) “the whole, consisting of language and the actions into which it is woven.”  There is both variety and vagueness here, neither of which is really desirable in, or perhaps even compatible with, a theory.  So we should consider the possibility that Wittgenstein offers not a theory but, as he implies in §133, instead demonstrates a method.[2]

            There are two problems with this idea.  The most obvious is that Wittgenstein seems so often to be putting forward arguments in defense of theses, despite his insistence to the contrary.  That idea is well known and I will not rehearse here the arguments in its favor.  Instead I will focus on the problematic nature of Wittgenstein’s alleged method.  Bob Plant, in a recent defense of a therapeutic reading of Wittgenstein’s later philosophy, refers to his way of doing philosophy as “inimitable.”[3]  This seems to me to be important and correct.  Wittgenstein’s ‘method’ is one that time and again has been imitated unsuccessfully, which is not surprising since Wittgenstein really does not tell us what his method is.  The goal of course is clarity (of a somewhat obscure kind) but the method itself is demonstrated rather than described.  To apply it oneself one must try to copy Wittgenstein.  But he was a tormented genius who, despite his best efforts, failed to complete his mature project to his own satisfaction before his death.  What hope is there for the rest of us to be able to go on in the same way?  And what would that even mean?  Wittgenstein himself has done more than anyone else to bring out the difficulties of what it means to follow another’s lead.  Are we to use concepts such as language-games, or invent other concepts of our own, as Wittgenstein did?  Perhaps both.  I do not mean to exaggerate and say that following Wittgenstein is impossible, but I do want to bring out the difficulty of understanding what it would mean to be a Wittgensteinian.[4]  The concept of language-games can be thought of as a tool for Wittgenstein, but not necessarily for us.  He invented language-games as objects of comparison, to show how they differed from and how they were the same as actual uses of our language (see §130), but whether we can do the same thing, or should try to do so, is another question.  Without a clear and precise idea of what a language-game is, and Wittgenstein (quite deliberately, I think) does not give us this, it is hard to see how we could answer this question affirmatively.

Wittgenstein’s Point

Before anything else I should explain how Wittgenstein introduces the concept of a language-game.  I have set out already the five meanings of ‘language-game’ that he describes in §7 of the Investigations.  What I have not done is to explain the context of this remark.  It comes as part of a commentary on what might be called “a primitive idea of the way language functions” or “the idea of a language more primitive than ours.”[5]  And this begins with a quotation from Augustine’s Confessions, in which Augustine, near the beginning of his spiritual autobiography, describes how he learned his native language in a way that Wittgenstein treats as problematic.  At least part of the problem is that: “Augustine describes the learning of human language as if the child came into a strange country and did not understand the language of the country; that is, as if it already had a language, only not this one.”[6]  That is, one might say, Augustine gives a Platonic account of language-learning.  Some such account is not necessarily something that Wittgenstein would reject.  After all, Augustine might have believed that God created him with the ability to think and speak before embodying his soul on earth.  This would be a religious belief, and Wittgenstein repeatedly emphasized that he did not intend to criticize religious beliefs.[7]  The problem with Platonism seems to be, in Wittgenstein’s mind, that it is philosophical rather than religious.  That is to say, it is confused, misrepresenting not so much the truth as what the speaker really means to say.  If it has a meaning at all, it is not one that the speaker really wants.[8]  A possible example of this is Gottlob Frege’s tendency to slip into Platonism in his philosophy of mathematics, and it is interesting that one of the examples Wittgenstein gives immediately in response to the quotation from Augustine involves the number five and how we can use this number without there being any question of what the number “five” stands for or represents.[9] 

            Another problem with Platonism is the danger of idolatry.  In Wittgenstein’s favorite novel[10], Fyodor Dostoevsky’s The Brothers Karamazov, the spirit of empiricism is shown to lead to absurdity[11] and to murder as the servant Smerdyakov sees no reason not to destroy his buffoonish and hateful father.  This murder is linked to Immanuel Kant’s belief in the moral necessity of belief in the immortality of the soul, but Smerdyakov’s apparent conviction that there is no immortal soul can be traced back through Ivan Karamazov to their father[12] and his skeptical empiricism.  He will believe in nothing that he cannot sense, which leads him to think of the soul, heaven, hell, and the like as things that cannot be sensed, and hence as fictions.  Credulous believers in demons and miracles, conceived in similarly naturalistic ways, are also mocked in the novel.  Presumably, Wittgenstein was sympathetic with the view of such people that Dostoevsky presents.  His own religious beliefs were certainly not of this superstitious kind.  In a diary entry from 20^th March 1937 he writes of his praying: “There is no one here: & yet I speak & thank & petition.”[13]  Wittgenstein could be said to oppose both naturalism (scientism[14] and empiricism[15]) and supernaturalism.[16]  What he sees in Augustine, I think, is the root of such thinking hidden or disguised in a seemingly innocuous passage from one of the great works in the western religious tradition.  If it can lurk there, it can lurk anywhere, which could be why Wittgenstein considered his abstruse and seemingly abstract work to have real, practical importance.[17]

            In short, I am suggesting, the point of Wittgenstein’s philosophical work is to liberate the mind from a tendency toward both naturalism and supernaturalism, a tendency that he thought of as essentially philosophical, a tendency we might call Platonism, and which shows itself quite clearly in the quotation from Augustine with which Wittgenstein begins his Investigations.[18]  Now, how might the concept of language-games help with this project?

            In Wittgenstein’s own case the answer is fairly obvious.  What he does is to use the concept to subvert the Platonist tendency of philosophers.  Thus for instance Investigations §24 tells us that: “If you do not keep the multiplicity of language-games in view you will perhaps be inclined to ask questions like: “What is a question?””  And this is the general form of the Platonic/philosophical question: What is a _____?  Such questions invite not only essentialism, which is not always misplaced, but also either hard-headed naturalistic dogmatism (e.g. “A soul is not something I can sense, therefore it is nothing at all”) or else unrealistic supernatural fantasy (e.g. “A soul is a Cartesian object of some mysterious or comforting kind”).  Early and late, Wittgenstein held that philosophical problems arise from misunderstanding the logic of our language.[19]  A remark on p. 224 of the Investigations suggests that the concept of language-games is essential to the solution of this problem: “We remain unconscious of the prodigious diversity of all the everyday language-games because the clothing of our language makes everything alike.”  So we need to see the diversity of language-games that exists in our actual language, and we can be helped to see it by considering fictional language-games.  Thus Investigations §130: “Our clear and simple language-games are not preparatory studies for a future regularization of language….  The language-games are rather set up as objects of comparison which are meant to throw light on the facts of our language by way not only of similarities, but also of dissimilarities.”  Investigation of the builders’ language in §2 or the model of language presented in the Tractatus, perhaps, can bring out how different, and more varied, our actual uses of language are than these models.  Then we might not expect there to be one answer to the question “What is a mind?” or “What is time?”  Nor might we expect an answer to such a question to be anything like the answer to questions such as “What is a hammer?” or “What is a meter?”  Then we will be free of what Wittgenstein seems to have regarded as the major source of spiritual corruption today (i.e. what we might call Platonism, scientism, naturalism, supernaturalism, or simply philosophy).

            Indeed I think a similar method can be found in the Tractatus itself.  Wittgenstein stresses there the different dimensions of logical space,[20] the differences between Newtonian and other kinds of mechanics,[21] and the difference between logic, science, and, for instance, ethics.[22]  He is primarily concerned here with the difference between logic and metaphysics or science, i.e. statements about how the world is.  The concern, as in his later work, is to draw attention to differences in linguistic use in order to overcome metaphysical tendencies.  Obviously, though, there could not be another Tractatus.  There could not be another set of sentences that constitute a ladder through which, on which, over which, one climbs out of the very nonsense those sentences seem to propound.[23]  At least such a book would have to consist of quite different sentences.  My question now is could there ever be another Philosophical Investigations?  Could someone else do work that was essentially the same as Wittgenstein’s, employing his concept of language-games in just the way he intended it to be used?  By this I mean: (1) can we identify language-games in our actual language, (2) can we invent fictional language-games for philosophical purposes, and (3) would there be any point in our doing so?  It is question 3 that I think presents the biggest problem.

What We Can and Cannot Do

It is quite obvious that we can identify language-games, in at least some of Wittgenstein’s five senses, in the language that we use.  For instance, “games by means of which children learn their native language,” are easy to identify, as is much of the use of words in games such as ring-a-ring-a-roses, if this is anything different.  Perhaps it is different, since children who play these games have already learned quite a bit of language.  Primitive languages are harder to identify, partly because it is hard to know what this expression means.  Perhaps chimpanzees have a primitive language.[24]  Perhaps very young children do.  Or perhaps people engaging in new practices, that need new uses of language, have such a language.  Fourthly, processes such as those of naming different stones “blocks”, “slabs”, and so on, or repeating words after someone else seem to be common enough.  And finally, “the whole, consisting of language and the actions into which it is woven” is all around us.  So we can find language-games easily, but those identified so far seem to have little relevance to philosophy, even as Wittgenstein practiced it.

            What Wittgenstein does with the concept of language-games is primarily two things.  First of all, he invents a series of variations on the builders’ language in order to make clear just how different our language is from this primitive one, even though the primitive language matches the kind of account of language that we find in Augustine and the Tractatus.  Thus this use of language-games is a way to criticize Wittgenstein’s earlier(-presented) theory.  If we criticize the picture theory in this way we will merely be repeating Wittgenstein’s point, not doing anything new.  And the criticism will not work if applied to other types of theory about language.  Secondly, Wittgenstein goes beyond his (almost) formal definition of “language-game” in §7 to refer to all manner of uses of language as language-games.[25]  His famous discussion of games and introduction of the notion of family resemblance seems to be expressly designed to emphasize the variety of uses of language.  §65 explicitly denies that there is an essence of language-games or of language.  While the Tractatus had tried to give philosophy peace by putting an end to all its problems,[26] the Investigations demonstrates a method, namely how to employ the concept of language-games.  Thus in §71 he explains how “the language-game with the word “game”” is taught: “One gives examples and intends them to be taken in a particular way.”  But this itself is part of an explanation of language-games.  So how to use this concept, we must think, is taught to us by a series of examples intended to be taken in a particular way.  And indeed in §133 Wittgenstein tells us that he is demonstrating a method by examples, and that this demonstration is offered instead of the “real discovery,” i.e. “the one that makes me capable of stopping doing philosophy when I want to.”  It is such a discovery, I would argue, that he had tried to provide with the Tractatus.  In other words, the point of the idea of language-games is either the same as or an alternative to the point of the Tractatus.[27]  And the concept of language-games, understanding what they are and what we are meant to do with the idea of them, is to be grasped through a discussion of ideas from the Tractatus.  Indeed, it is one of the ideas that Wittgenstein says “could be seen in the right light only by contrast with and against the background of [his] old way of thinking.”[28]  It does not follow immediately from the fact that ideas are developed and learned in connection with the Tractatus that they cannot therefore be used outside this connection, but the question is worth raising, and I will return to it later.         

What about our ability to make up new language-games as objects of comparison?  There are two problems here.  One is that the imaginary language-games that come to mind are so similar to Wittgenstein’s own builders’ language that they scarcely seem to count as inventions of our own.  The other is that it will obviously depend on what we want to do with the comparison.  Invented language-games should be modeled on confused philosophical uses of language.  One would seem to have to identify the confusion first before one could construct the appropriate language-game to help one see it as an instance of confusion.  But in that case the language-game would be superfluous.  It might be used to help others to see the error of their ways, but Wittgenstein’s method seems to involve the presence of an interlocutor.  In the Investigations he acts as his own interlocutor, but then he has a public self, the author of the Tractatus, with whom he can have a public dialogue, of interest to a public audience.  As he says in Culture and Value: “Nearly all my writings are private conversations with myself.”[29]   Few other people’s private conversations are of interest to a general audience.  Elsewhere, he says that “The philosopher strives to find the liberating word” and that only if the reader, or one’s interlocutor, acknowledges it as such has the philosopher found the right expression.[30]  This makes the successful practice of Wittgensteinian philosophy hard unless an interlocutor is actually present, the reader is someone very familiar to one, or else one’s philosophical work is really aimed at oneself.[31]  So if one has developed a simplified model or theory of language, one can liberate oneself from its grip by inventing a primitive language that fits the theory and then seeing how different this is from actual language.  This would seem to require a somewhat split personality, involving one part that adheres, up to a point, to the theory, and another that sees through it well enough to develop a plausible language-game based on it, designed specifically to show its inadequacy.  Perhaps such philosophy could be practiced only on a few friends (as Wittgenstein sometimes described his work[32]) or else as a series of guesses as to what the reader might accept.  In short, to use this technique in philosophy one would have to be facing a rather Tractarian (or, admittedly, Augustinian) interlocutor whom one knows well, and one must free this person’s mind from the grip of their simplistic picture of language by means of invented languages like the one used by Wittgenstein’s builders.  How likely this is to happen unless one is Wittgenstein I cannot say, but the chances might well be small.

Finally we come to what I suggested might be the biggest problem of all: what would be the point of such an exercise?  That is, either of mirroring Wittgenstein’s critique of the Augustinian or picture theory of language by means of fictional linguistic practices, or else of drawing people’s attention to the great variety of uses of language that actually exist.  It will help to recall what Wittgenstein’s goal was, namely clarity.  By clarity he means the complete absence of idols, and by this he means also not making an idol of the absence of idols.[33]  I assume he has in mind dogmatic anti-religionists (and perhaps dogmatic or complacent anti-metaphysicians) such as Friedrich Nietzsche and Bertrand Russell.  Dogmatism and metaphysics (conceived as a confused sort of pseudo-religion) are the enemies.  We can use his method without sharing this goal, but we are unlikely to want to do so.  There are significant problems with the idea that one might use this method on oneself, and it is anyway more likely to be used on others.  However these others must accept not only the diagnoses offered but also their subjection to the treatment itself.  And this is not likely.  Wittgenstein’s later philosophical method, to which the concept of language-games belongs, seems fruitless to most contemporary philosophers.  Wittgenstein foresaw this problem, predicting that his work was unlikely to do any good since it was so contrary to the grain of current thought.[34]  He said that:

The person with a ‘healthy human understanding’ who reads a former philosopher thinks (and not without right): ‘Mere nonsense!’  If that person hears me, he thinks – rightly, again – ‘Nothing but boring truisms!’[35]

 Certainly linguistic philosophy has often been dismissed as trivial, and this remark of Wittgenstein’s is reminiscent of Tractatus 6.53, which says that the only strictly correct method in philosophy would not be satisfying to the other person, because it would not seem to be philosophy.  This method would also have the disadvantage of seeming to require the physical presence of the interlocutor, as James Conant has pointed out.[36]  But the later method faces very similar problems.  How can boring truisms be satisfying?  Perhaps by being contrasted with certain philosophical theses.  But which ones?  Ones held by the audience would be of most interest to them.  But then how can one know what these are if the audience is not present?  It seems to me almost inevitable that the application of the method must be done at a great level of generality (as has already been done, in the Investigations) or else aimed at a public, well-known set of the author’s own thoughts (as the Tractatus also does).  That is to say, the most promising kind of application of Wittgenstein’s later method would be to problems more specific than those dealt with in the Investigations, something smaller scale than rule-following or thinking, say, and hence problems specific to a fairly small audience, which would almost have to be known personally by the philosopher.  This is not an impossible situation, but the results are unlikely to be relevant to many people, and so are not likely to be publishable.

Rush Rhees tells us that Wittgenstein wrote in 1948:

These difficulties are interesting for me, who am caught up in them, but not necessarily for other people.  They are difficulties of my thinking, brought about by my development.  They belong, so to speak, in a diary, not in a book.  And even if this diary might be interesting for someone some day, I cannot publish it.  My stomach-aches are not what is interesting but the remedies---if any---that I’ve found for them.[37]

Now what can Wittgensteinians do?  Their stomach-aches are not (likely to be) interesting to others.  And the remedy has already been found.  Whatever we do with Wittgenstein’s concept of language-games, then, is unlikely to help us (because if we can see our way clear to use the concept then we do not need it) or to interest others (since they might reject the therapy and we might fail to find the liberating word).

Conclusion

I am not saying that it is impossible to be a Wittgensteinian philosopher.  I am saying that to use Wittgenstein’s concept of language-games, rather than some other concept derived from it, one has to be a Wittgensteinian philosopher.  And since that means not doing science but something much more like therapy, which is quite an individual matter, it is at least highly unlikely that it will be publishable in even sympathetic philosophical journals.  Almost all that can be of interest to them is non-Wittgensteinian philosophy or scholarly work on Wittgenstein himself.  Which is why, I think, we see so little application of Wittgenstein’s method in print, despite repeated claims by Wittgensteinians that that is what they require.                        

Notes