You are here : Home Page / Theories / Wittgenstein / Language Games

Available Languages

Follow us on

SignoSemio's Facebook SignoSemio's Twitter

Wittgenstein's Language Games

By Nicolas Xanthos

Université du Québec à Chicoutimi



Ludwig Wittgenstein

In their later acceptation (beginning with the Philosophical Investigations), Wittgenstein's language games established some notions that have extremely important implications for the theory of signs, in that they cover the entire range of semiotic practices. The language games can be understood as the shared conceptual parameters that make it possible to identify and produce signs, and to establish relations of signification and representation.

In this section, we introduce three interdependent notions: 1) the language games (semiotic practices which, despite the term "language", are not restricted to verbal language); 2) the moves of the language games (concrete actions performed in a given language game, the raw data of semiotic theory); 3) the grammar of the language games (the conceptual architecture that determines how the signs are used).

To give an example (which should not be taken as an exclusive paradigm), we could roughly say that the interpretation of legislative acts, at its most basic level, is a language game, a rule-guided way of attributing meaning. A particular interpretation of a particular law, composed of a defined set of arguments, would be a series of moves in the language game of interpreting legislative acts. Their interpretation as laws presupposes the concepts of rights, duties, obligation, possibility, responsibility, action, etc., which compose the grammar of this particular language game.

This text may be reproduced for non-commercial purposes, provided the complete reference is given:
Nicolas Xanthos (2006), « Wittgenstein's Language Games », in Louis Hébert (dir.), Signo [online], Rimouski (Quebec),


2.1 The Language Game


Although the concept of the language game is central to Wittgenstein's later philosophy, the notion is still an elusive one to define and to grasp, for at least two reasons. Firstly, from The Blue Book to On Certainty, one can pick out several distinct usages of the term "language games". Sometimes it refers to the fictive examples that Wittgenstein invents to explain how language ordinarily functions, sometimes it refers to children's language-learning games, and sometimes to semiotic practices, that is, the socially shared ways of using signs, of signifying and of representing. Secondly, the notion is never explicitly defined. Wittgenstein preferred to proceed by example, using fragments of short, dense analyses to convey what the language games are.

In this chapter we will present the language games in their later acceptation, the one that began to emerge in the Philosophical Investigations, then moved to the foreground in On Certainty: language games as semiotic practices. Although we cannot directly solve the problem of not having a definition, we will nonetheless try to present an analytical trajectory that will help convey the key concepts – language game, move and grammar –, which are all related. In order to have a complete picture, we would also have to introduce Wittgenstein's "form of life" concept,  which is the cultural environment in which the language game occurs, the "community which is bound together by science and education" (Wittgenstein, 1969, p. 38e). For reasons of brevity, we will leave it aside.


In paragraph 23 of the Philosophical Investigations, Wittgenstein presents a list of examples to give us a sense of "the multiplicity of language-games":

"Giving orders, and obeying them –
Describing the appearance of an object, or giving its measurements –
Constructing an object from a description (a drawing) –
Reporting an event –
Speculating about an event –
Forming and testing a hypothesis –
Presenting the results of an experiment in tables and diagrams –
Making up a story; and reading it –
Play-acting –
Singing catches –
Guessing riddles –
Making a joke; telling it –
Solving a problem in practical arithmetic –
Translating from one language into another –
Asking, thinking, cursing, greeting, praying."
(Wittgenstein, 1958, pp. 11-12)

These are all semiotic practices that exhibit regularities, often with language playing an essential role. As we will see, language-game analysis is a specific way of looking at these practices: they are viewed as operations governed by a set of discrete concepts that the analysis must seek to express.

If we were a bit simplistic, we could say that making statements is equivalent to lining up a series of words. However, this activity is not randomly performed; it follows a certain number of rules described in the English grammars that we sometimes consult. In conjunction with the lexicon (the units being combined, e.g., words), these rules constitute the possibility conditions for our innumerable statements. On an infinitely larger scale, the language games gravitate toward a common idea: our ways of interacting with signs are rule-guided, and we need to bring these rules to light. Although it can be compared to the concept of the "convention", Wittgenstein's notion of the "rule" is still distinct, in that the rules are essentially conceptual in nature, and they are not (and cannot be) agreed upon or discussed in advance. They are not agreed upon in the sense that one cannot disagree with the rules of a language game: that would simply mean not playing that particular language game; there is no choice in the matter. And they are not discussed in advance, with a very partial exception for what we usually mean by "game", where rules are set out in advance (to be more exact, a small portion of them are). But we must emphasize that this is the exception and not the norm.

We can also draw a parallel between language games and speech acts. In Anglo-Saxon pragmatics (and we are primarily alluding to John L. Austin (1962) and John R. Searle (1969)), when we speak, we are performing illocutionary acts, such as affirming, promising, asking, suggesting, refusing, etc. To take one example, we do not promise just anything in just any way: we only promise something prospective, something that will not necessarily come about, but which we intend to do and which has a positive value to the person to whom the promise is made. Thus we cannot promise to have been good in the past, or that the sun is going to rise tomorrow (except in obviously apocalyptic circumstances), cannot promise that we are going to take the next space shuttle flight within the hour, or promise our interlocutor that we are going to torture him horribly and at length (unless our sexual habits are somewhat "alternative"). In other words, speech acts are ruled-guided language practices.

That is the basic idea behind the concept of the language game: Whether language plays the central role or some more or less apparent role, our semiotic practices can be thought of as rule-guided practices. They are not chance actions nor randomly proffered words, but actions that owe their legitimacy, relevance and even existence to a set of rules determining their use.

It is also enlightening that Wittgenstein's concepts are rooted in a comparison with games. In order to examine our practices, we must approach them in the same way we would an unknown game whose rules we want to learn. Watching a chess game, if we know nothing about it, we would conclude that the actions performed by the participants are not random, that not all moves are equally possible in all circumstances, that not all moves are equivalent, and so on. We would gradually come to understand the value of the pieces, how to move them, the purpose of the game, and other elements. In short, we would apprehend bit by bit the rules that give meaning to this particular restricted space, these particular objects, these particular movements – in a word, to this practice, or (language) game. The same concept applies to all of the language games cited above by Wittgenstein.

But these are certainly not the only possible language games. In On Certainty, Wittgenstein shows that history (as a field of study), for example, can be viewed as a language game: it is a rule-guided way of attributing meaning to events.


Given the current complexity of the field of history as such, it would be more accurate to say that the discipline consists of a set of related language games set apart by their objects (social, economic, cultural and political history) and the perspectives they use in constructing these objects and making them signify (the Marxist, Foucauldian, Annales or other approach).

We could also pursue our investigation into the field of literature, where we would find a vast mosaic of language games for both production and reception. Writing a novel, a poem or a scientific article is a language game – and so is reading a novel, a poem or a scientific article. Doing a Greimassian analysis and a psychoanalytical analysis on a short story by Maupassant also makes use of two different practices.

In athletics, to give one last series of examples, one can also discern a panoply of language games. Firstly, each sport is itself a language game to start with, since it is a practice governed by a set of concepts (players, the field, goals, points, etc.). And if we look at sports writers and sports announcers, we also find several language games: describing a sporting event and analyzing it are two completely distinct language games.

This is not saying much, obviously, and we need to pursue the  discussion further. In particular, we will need to define the nature of the rules that generate and give meaning to the various language games. But before we get into the notion of rules, or the grammar, we must talk about moves in language games.

2.2  Moves

This notion is very simple, but is an important one in Wittgenstein's theory. Ordinarily, we are not in contact with language games as such, but with actions performed as part of a language game: we do not see "chess", but a game of chess; instead of "promise", we see a specific promise; instead of "novel", a particular novel; rather than "textual analysis", a particular textual analysis. In one sense, the language game is a hypothesis that we are making about the basis of individuals' semiotic behaviour, assuming that this behaviour is not random, but a function of specific rules.

Although for certain language games there are explicit, precisely formulated rules that we can learn prior to the game, this is not the case for most language games. Taking Wittgenstein's list of examples, one can observe that there is no "rulebook" for the games involved in reporting an event, speculating about an event, or making a joke. All we have in these cases (the majority) are events reported, and speculations and jokes made in different circumstances from which we must infer both the language game and its rules. Most of the time, then, we are in contact with actions performed in language games yet to be identified: these actions are what Wittgenstein calls "moves" in the language games.

This is why in most sign production and interpretation practices, the raw material for a Wittgenstein-style analysis is the action, the move (or the set of moves), which we can trace back to the language game and its grammar. The text you are currently reading is a set of moves in a language game that we could provisionally call "introduction to (or simplification of) a theory". The way in which semiotic relations operate between this text and Wittgenstein's work are directly related to the rules of this language game. A cartoon in the print media is a move in the language game of cartoons. When a neighbour, colleague or friend acts in an uncharacteristically aggressive way, if we attribute this to stress he is experiencing at work, we are making a move in the language game of interpreting human behaviour (or some more or less recommendable version of it, such as biological, psychological, sociological, political, religious, or racial interpretation).

There are two reasons why the move is the preferred way to access the actual language game. Firstly, as Wittgenstein clearly implies through a geometric metaphor he gives in On Certainty, most of the rules of a game are not learned explicitly, but are discovered a posteriori by examining the moves: "152. I do not explicitly learn the propositions that stand fast for me. I can discover them subsequently like the axis around which a body rotates. This axis is not fixed in the sense that anything holds it fast, but the movement around it determines its immobility" (Wittgenstein, 1969, p. 22e).

Secondly, and this is more basic yet, the link between moves and the grammar is a close one: the moves only acquire meaning by existing within the area of discourse and action defined  and delimited by the grammar.

2.3 Grammars

The grammar of a language game – what we have also called the "rules" here – is truly the keystone of Wittgenstein's theory, and uncovering it is the purpose of the analysis. To begin with, we should make it clear that the term "grammar" is not to be understood in its usual acceptation. We must emphasize that grammar has a basically conceptual character for Wittgenstein, although the concepts themselves can sometimes be expressed as propositions. These concepts, or grammatical propositions, are the possibility condition for the moves made in the language games (sometimes called "empirical propositions").

The analogy with games and sports is once again enlightening, as well as the role played by the rules. The rules interdefine the elements that make up the game; they assign a role and a meaning to each element, they define the game's space and time, the participants' functions and goals, and so on. In short, they create and give structure to an area of potential discourse and actions that owe their meaning to the rules. Any specific action in soccer, bridge or checkers owes its meaning and its very existence to the entire set of rules for the game. The rules impose their order on that portion of reality in which the game unfolds. Even an apparently stable empirical object such as the human body can end up segmented into areas that have distinct meanings that vary from one sport to another: in boxing, the hips have a signification that they do not have elsewhere; the hands and the feet do not have the same meaning in hockey and soccer; in fencing, the torso has a signification that it does not have in judo. As Wittgenstein says in the Philosophical Investigations: "373. Grammar tells what kind of object anything is" (Wittgenstein, 1958, p. 116e).


It would be helpful at this point to mention the distinction made by Searle between constitutive rules and normative rules (1969 and 1995). Constitutive rules create the game and define it; without them it would not exist. Normative rules indicate which actions are legitimate and which are not within the area created by the constitutive rules. For example, playing poker with a few aces up one's sleeve is against the normative rules; however, trying to win at poker by building up the largest possible number of spades in one's hand is against the constitutive rules. When the normative rules are broken, we conclude that the player made a mistake or cheated; when the constitutive rules are broken, we feel slightly perplexed, and may think that the player is playing some other game that we will never quite grasp. When Wittgenstein talks about "grammar", he is referring to the constitutive rules.

The game's constitutive rules are the possibility condition for the actions performed in these games and sports, just as grammatical propositions are the (conceptual) possibility condition for the moves in the language games. And as Wittgenstein says in the Philosophical Investigations: "our investigation, however, is directed not towards phenomena, but, as one might say, towards the 'possibilities' of phenomena. We remind ourselves, that is to say, of the kind of statement that we make about phenomena. […] Our investigation is therefore a grammatical one" (Wittgenstein, 1958, p. 42-43e)

Therefore, our language behaviours (among others) are moves in language games (of which we are often unaware), and they draw their meaning from the grammar of these games. And an analysis of them examines the moves in order to arrive at their grammatical possibility conditions.

One of the difficulties encountered in this analysis is our familiarity with the language games, which obscures the existence of the moves, even in the clearest cases of games. For example, when  we "take the opponent's rook with our bishop" in a chess game, we think we are dealing in raw facts; it does not occur to us that these movements of objects through space can be seen as we see them only if we have integrated the interdefined concepts of pieces, movement, chessboard, square, player and capture, to name a few. The grammar of our other language games, some of which are not even named, has a similar sort of familiar invisibility.

Consider the following (empirical) proposition: "Having been a great coffee drinker since his teens, John went to get cream at the grocery store at 9:30 p.m. so that he would have some on hand the next morning." This statement, easily recognized as a fairly rudimentary narrative utterance, is a move in a language game: an action game. An analysis of it must try to reveal the concepts that constitute its grammar. This particular grammar is often described, and is made up of the concepts of intention, goal, agent, motive, cause, etc. In this statement, then, John is the agent, having cream on hand for his coffee the next morning is his goal, going to the grocery store for cream is his action, and his long-standing love of coffee is his motive.

To give another type of example, a sentence like the following one is perplexing for strictly grammatical reasons: "Imagine a slightly bluish-red green color, lighter than greyish yellow." This empirical proposition is a move in the language game of color. The grammar of this game involves certain relations between the colors, and excludes certain other relations. What we call yellow cannot have the property of being darker than what we call green – in much the same way that tenderness cannot be irritable, pity cannot be likeable, and courtesy cannot be desirous. The grammar of our language games excludes these moves that are not within the realm of possible moves for discourse and action dictated by the grammar.


We should add that fiction, whether in literature, film, theatre or even philosophy, can become a place in which to explore the boundaries of our language games, and even to challenge them to some degree. Thus fiction can improve our understanding of the area defined by our language games, and provide a place to experiment with developing atypical grammars. For example, Éric Chevillard's work has played this sort of role in more than one instance (see The Crab Nebula (1997 [1993]) or Les absences du capitaine Cook (2001)).

The grammar of a language game is not something we learn, and its propositions are unquestioningly accepted when we are playing the game. They are not explicitly learned, but rather absorbed over time by practicing a language game. They are implied logically by the examples through which we learn the game, and are never specifically brought up, except in philosophical discussion. And while they are unquestionable, this is because they create the very possibility of the game being played: questioning them is equivalent to being out of the game. For instance, a psychoanalyst cannot question the existence of the unconscious without thereby ceasing to be a psychoanalyst.


Wittgenstein's famous distinction between saying and showing takes on its full meaning with the grammar issue. Any move in a language game, or any empirical proposition consists both in saying something and in showing the grammar of the language game. Wittgenstein's position concerning what is being shown seems to have changed. In the Tractatus Logico-Philosophicus, he is unequivocal: "4.1212. What can be shown cannot be said" (Wittgenstein, 1972, p. 51). Impossible, then, to talk about the grammar or name its contents; we can only try to give an idea of it by giving examples of not playing the game, as we did above, in order to convey a sense of the grammatical area of the game. In On Certainty, his position is less radical: "88. It may be for example that all enquiry on our part is set so as to exempt certain propositions from doubt, if they are ever formulated. They lie apart from the route travelled by enquiry" (Wittgenstein, 1969, p. 13e).

2.4 Language games and semiotics

Wittgenstein's theory of language games can be instructive on several accounts when applied to semiotic discussion. Indeed, any interaction with signs, production of signs, or attribution of meaning owes its existence to its status as a move in a language game – that is, a conceptual architecture, a grammar, that we must uncover.

Consider the Augustinian definition of the sign: something put in the place of something else (to which it is imperative to add: in a relation of meaning or representation). Wittgenstein tells us that of the elements that make up the semiotic relation (sign, modes of representation or signifying, the sign's referent, etc.), none exists outside a language game. In an interpretive act, nothing is "intrinsically" a sign: the grammar of the language game is what makes it possible to identify the sign, its way of being a sign and what it is a sign of.

The grammar of psychoanalysis is what turns a failure of memory into a parapraxis, and the parapraxis into a sign of some unconscious desire; this justifies inferring desire from a failure of memory. The grammar of the language game of reading fiction is what makes it possible to see a particular printed object as a fictional discourse, which then allows us to imagine the fictional world that this discourse represents. The grammar of psychological interpretation of facial expressions is what makes it possible to see a frown as a sign, and to read the frown as a dysphoric expression of incomprehension, disagreement or scepticism. Even identifying a sign, regardless of its degree of complexity, is a move in the language game that will lead to its interpretation; and a mere description of the sign cannot help revealing the grammar of the game being played.

Identifying and interpreting signs are actions that take place in language games that can be described by analysis; and the same applies to sign production. From this standpoint, the entire theory of literary genres can be seen as an enormous undertaking in which the grammar of the various genres is revealed. For example, Aristotle's "Poetics", which has been passed down to us, is an attempt to analyze the grammar of the language games of the tragedy and the epic. In Greimas' or Bremond's work, the narrative sequence is a formulation of the grammar of narrative: manipulation, competence, performance, sanction or contingency, and initiating and completing the action would be seen here as elements in the grammar of the language game of representing action. (On this subject, see the chapter on the canonical narrative schema in Signo.)


On the one hand, we will refrain from putting concepts with very different or opposite epistemological foundations into one basket. These examples are aimed at showing the nature of grammatical constructs, not their processes. On the other hand, the example of narrativity guides us to a basic fact: language games are not time-independent; they can change over the course of their history. It would be a difficult endeavour to try and develop a single language game to account for narrative both for popular Russian fairy tales and the contemporary writers published by the Minuit publishing house.  The two language games will have an obvious "family resemblance", just the same.

All sorts of grammars are conceivable, including ones underlying the signs (and thus the language games) that have just recently arisen in our societies, such as the grammar of video game images, of web sites and of hypertext.


Consider the following situation: Jack and John are watching a soccer game in John's apartment. During the game, they hear a loud, heavy thud on the ceiling. Looking slightly annoyed, John comments about his upstairs neighbour:

– Really, he's even clumsier when he's drunk, that guy.

A little embarrassed, Jack says:

– Maybe he's trying to tell us that your TV is too loud.

Their replies are moves in different language games (both of which are intended to give a meaning to the noise), which could be named as follows: for Jack, interpreting a noise produced intentionally with a communicative intent; for John, interpreting a noise made accidentally.

In John's view, the gesture is accidental, allowing him to trace it to the psycho-physiological traits that caused it, the identifying characteristic being a loss of self-control. The grammar of this language game postulates a psychological or physiological interiority over which the subject has limited control, in a causal connection with a particular kind of movement: involuntary. The states of this interiority are also of different durations: the clumsiness is permanent, the drunkenness is chronic. Lastly, they are interdependent, since the drunkenness aggravates the clumsiness.

In Jack's view, the noise (the action, that is) is deliberate, manifesting a communicative intent on the part of the agent. We should note that the intent could have been an entirely different one than the one Jack identified: the neighbour upstairs might have stomped on the floor to chase away a sensation of pins and needles in his leg. The communicative intent is characterized grammatically by its content and its receiver, both of which are in a relation of interdependence (with the receiver helping to identify the content, and vice versa).


AUSTIN, John L., How to Do Things with Words, Cambridge, MA: Harvard University Press, 1962.
CHEVILLARD, Éric, The Crab Nebula, trans. J. Stump, E. Hardin, Lincoln: University of Nebraska Press, 1997 [1993].
CHEVILLARD, Éric, Les absences du capitaine Cook, Paris: Minuit, 2001.
SEARLE, John R., Speech Acts: An Essay in the Philosophy of Language, Cambridge: Cambridge University Press, 1980 [1969].
SEARLE, John R., The Construction of Social Reality, New York: Free Press, 1995.
WITTGENSTEIN, Ludwig, Tractatus Logico-Philosophicus, trans. D.F. Pears and B.F. McGuinness, London: Routledge & Kegan Paul, 1972.
WITTGENSTEIN, Ludwig, Philosophical Investigations, trans. G.E.M. Anscombe, New York: Macmillan, 1958.
WITTGENSTEIN, Ludwig, On Certainty, trans. Denis Paul and G.E.M. Anscombe, New York: Harper Torchbooks, 1969 [1965].

5. Exercises

A. Pick out a chair, a desk or a bookcase. Try to estimate the year it was made, its monetary value, and assign it a well-reasoned aesthetic value. Examine how the moves in three different language games made you structure the object in a different way each time, and how the grammars of these games are revealed in the structures and your comments.

B. We do not read an opinion column the way we read a poem. Identify the grammatical concepts at work in these two language games, and how you segment and structure the mental processing of the opinion column and the poem with them.

C. Which emotion is in between anger and empathy? Which emotion is less intense than fear but more intense than jealousy? When does one feel surprise and bitterness at the same time? What do our difficulties in answering these questions teach us about the grammar of the language game of emotions?

Share this page