From the WAB archives: A selection from the Bergen Wittgenstein archives' Working Papers and audio-visual materials

I

Wittgenstein's remarks which are in question here belong to the first sections of the PI. The word "complete" occurs first in PI 2. I take it to be appropriate first to consider the context of this occurence.

In PI 1 Wittgenstein quotes a passage from Augustine on his learning (and use of ³) language. In these remarks we get, in Wittgenstein's view, "a particular picture of the essence of human language", according to which "the individual words in language name objects - sentences are combinations of such names." In this particular picture we find, according to Wittgenstein, "the roots of the following idea: Every word has a meaning. This meaning is correlated with the word. It is the object for which the word stands." For the sake of exegetical clarity we have to register that Wittgenstein is concerned with three different things:

(i) Augustine's description of (his) learning etc.,

(ii) a picture of the essence of human language, and

(iii) an idea of the meaning of a word.

The phrase "these remarks, ..., give us" indicates that (i) presupposes (ii)⁴ and the phrase "in this picture ... we find the roots..." indicates that (iii) also presupposes (ii). (ii) is then, one could say, at the heart of the whole section. Nevertheless, the presuppostion (ii) does not exhaust its respective antecedents (i) and (iii).

The next three steps taken by Wittgenstein are:

- to register that Augustine "does not speak of there being any difference between kinds of word",

- to give a diagnosis according to which somebody describing the learning of language in Augustine's manner is "thinking primarily of nouns like 'table', 'chair', 'bread', and of people's names, and only secondarily of the names of certain actions and properties; and of the remaining kinds of word as something that will take care of itself" (), and

- to describe a special use of language - the well-known shopkeeper-example.

The first two comments are given from the point of view of our language. In our language there are different kinds of words like nouns, proper names, names of actions, of properties, etc. The first two steps are furthermore, as I would put it, taken from an internal point of view with respect to Augustine's remarks, that is, in these points Wittgenstein restricts his own comments to Augustine's description. These comments are also restricted to the topics which are circumscribed by (ii) and (iii). The functions of these comments are:

(a) to make explicit what - with regard to (ii) and (iii) - gives Augustine's description (i.e. (i)) a prima facie plausibility, and

(b) to remind the reader that there are more kinds of words in our language than Augustine in writing and the reader in reading may have been thinking of.

With (a) Wittgenstein implicitly accepts that a description of the learning of language must be brought into accordance with, or must correspond to, a picture of the essense of human language - that is: with a picture of what it is that has been learned - and with an idea of the meaning of a word. To see why Wittgenstein says that somebody describing the learning of language in Augustine's manner is thinking primarily of nouns like "table" and of proper names one must register a characteristic of Augustine's description. He speaks not only about objects but furthermore about the elders moving towards something, the play of their eyes, their "seeking, having, rejecting, or avoiding something", about their meaning to point out, the expression of their faces, their state of mind, his "seeing what they tried to name by the sound they uttered". Augustine's description is full of mental predicates⁵, especially those which are connected with visual experience of things and with attitudes towards material objects. Tables, chairs, the Chinese Wall and Marilyn Monroe are - contrary to red and five - paradigmatic examples for entities one can move towards, see, reject, avoid, seek, etc.

The third step is taken from an external point of view, that is, Augustine's description of his learning (and use of) language is contrasted with a description of a simple example of language in use - the well-known shopkeeper-example. The connection between the first two steps on the one hand and the third step on the other is at least twofold:

- whatever the learning of language may consist of, it should enable the child to use the language, that is to "operate with words" [PI 1]⁶,

- our operation with words includes operating with words like "apple", "red", and "five" as described in the shopkeeper-example in PI 1. Whereas Augustine's description of the learning of language seems to fit to the use of, or operation with, the word "apple" the following questions arise with regard to the other two words: "'But how does he (i.e. the shopkeeper -R.R.) know where and how he is to look up the word 'red' and what he is to do with the word 'five'?'" [PI 1] These questions arise because it is not prima facie plausible that the uses of the words "red" and especially "five" could be learned in the way indicated by Augustine's description.⁷

Instead of giving an answer, Wittgenstein repudiates the question: "Well, I assume that he acts as I have described." [PI 1] What gives Wittgenstein the right to do this? He says: "Explanations come to an end somewhere." [PI 1] At first sight this doesn't seem to be a good reason for the repudiation, for this phrase, one could object, is true despite our understandable need for explanations in various situations. For this remark to be a reason for repudiating the question one should imagine what an answer - in the context of the discussion of PI 1 - could consist of. In the style of Augustine the answer could be a refined description of the learning of language. But this would nevertheless have been provoked by the description of the use of, or the operation with, the words given by Wittgenstein, for the question quoted above arises just because it is not prima facie plausible how Augustine's report could fit the use of the words "red" and especially "five". We see, the refined description of the learning of language would be dependent on the description of the use of language.⁸ But nevertheless, the repudiation of the question is something like a break in the argumentation. The effect and the point of this move is that it turns one's thoughts to the description of the use of, or the operation with, words as the most important subject of a search for the essence of human language in a Wittgensteinian manner. An effect of this redirecting of interest consists in the concentration on what people are doing when speeking, not on what they may think, feel or wish.⁹

We saw that Augustine's description of his learning of language runs into trouble if contrasted with even a simple example of an operation with words, at least it loses its prima facie plausibility and therefore part of its attractiveness. It seems that the general idea of meaning brings about similar problems. At least Wittgenstein's response to the question "But what is the meaning of the word 'five'?" [PI 1] seems to be similar to his response to the question above: "No such thing was in question here, only how the word 'five' is used." [PI 1] Notice that the question is only concerned with the meaning of the word "five" and not with the meaning of the word "red". Wittgenstein chooses the most obvious case to show that there is something unclear with the general idea of meaning due to Augustine - even in such a simple case like that of the shopkeeper-example. If the general idea of the meaning of a word surrounds the working of language with a haze, then we should expect that the picture of the essence of language, in which the roots of this general notion lie, could also be criticized. "That philosophical concept of meaning has its place in a primitive idea of the way language functions." [PI 2] If, as we said above, (iii) presupposes (ii), and if (iii) runs into trouble when applicated on an operation with words like "five", then one could expect first to have to change (ii) in order to solve the puzzles. This is surely true, and it is exactly what Wittgenstein tries to do with his invention of primitive language-games like that of the shopkeeper. But: "If we look at the example in §1, we may perhaps get an inkling how much this general notion of the meaning of a word surrounds the working of language with a haze which makes clear vision impossible. It disperses the fog to study the phenomena of language in primitive kinds of application in which one can command a clear view of the aim and functioning of the words." [PI 5]¹⁰

Therefore, even if (iii), to be true or false, depends on the truth of (ii), in philosophizing (iii) plays a crucial role. It is a prejudice which gives rise to the assumption that the language must function in accordance with it. Where this seems not to be obvious, a search for explanations and hidden entities is produced. What produced the haze and the fog (that is: the false picture) in the case of Augustine's description and its respective idea was the fact that it is not clear how they could be brought into accord with the variety of different kinds of words and the variety of different kinds of use of words in our language. As long as we try to follow the idea we are not able to get a clear view and a correct picture.

The questions posed in PI 1 were, as we saw, concerned with the explanation of the use and meaning of some of the three words, not all of them. Augustine's description of the learning of language and the respective picture of the essence of language and the general idea about the meaning of words are not useless in all respects. Therefore Wittgenstein says that the philosophical notion of meaning due to Augustine stems from a primitive idea of the way in which language functions, from a primitive picture of the essence of human language. But, one could also say that this idea is an "idea of a language more primitive than ours." [PI 2] These two things are not the same. The primitive idea of the way in which language functions is primitive with regard to the way our language functions. The philosophical notion of meaning brings about problems - for instance it produces a need and a search for refined explanations - when it becomes applicated to fragments of our language which had not been part of what its proponents have been thinking of. In the case of a language more primitive than ours there is no need for explanations, there are no puzzles. Insofar as one can characterize such a language as a language consisting only of the examples the proponents have been thinking of. Therefore Wittgenstein's remarks on a language more primitive than ours are not the expression of a new observation or something similar, but a methodological reformulation of the first remarks. Its methodological point is the following: if a general notion of the meaning of a word and its respective picture of the essence of human language produce - when they are compared with the whole or with particular fragments of our language - the need for explanations and further refinements like for instance the postulation of hidden entities etc., and if they have therefore to be taken to be inadequate, and if finally this need is not brought about if they are compared with (fictitious) primitive languages, then is seems possible to get an adequate conception, an appropriate picture by looking at our language as a system, a whole of such (real) primitive languages, which is itself - as a whole or a system - not primitive. If this is true, then this sheds light on the repudiation of the question in PI 1 and the remark that explanations come to an end somewhere: everything that is necessary in philosophy is available without explanations, without theories, without postulates, etc.

Such a "language more primitive than ours" is the language described in PI 2 - the language of the builders. Wittgenstein writes:

"Let us imagine a language for which the description given by Augustine is right. The language is meant to serve for communication between a builder A and an assistant B. A is building with building-stones: there are blocks, pillars, slabs and beams. B has to pass the stones, and that in the order in which A needs them. For this purpose they use a language consisting of the words 'block', 'pillar', 'slab', 'beam'. A calls them out; - B brings the stone which he has learnt to bring at such-and-such a call. - Conceive this as a complete primitive language."

For this language the picture of the essence of human language which is implicit in Augustine's description is right. But it is not right for the whole of our language. Insofar as a correct response to Augustine's description is: "'Yes, it is appropriate, but only for this narrowly circumscribed region, not for the whole of what you were claiming to describe.'" [PI 3]¹²

In PI 8 and PI 15 Wittgenstein describes expansions of the language of PI 2. In PI 8 the expansion consists of "a series of words used as the shopkeeper in (1) used the numerals (it can be the series of letters of the alphabet); further, let there be two words, which may as well be 'there' and 'this' (because this roughly indicates their purpose), that are used in connexion with a pointing gesture; and finally a number of colour samples." In PI 15 Wittgenstein assumes that A uses tools, and he introduces marks which are borne by the tools which A uses in building. "When A shews his assistant such a mark, he brings the tool that has the mark on it."

In PI 18 Wittgenstein responds to a possible objection to his comment in PI 2: that this language cannot be complete because it consists only of orders. Wittgenstein's argumentation goes as follows: if one wants to object that the language of PI 2 (and also, we could add, the expanded languages) is incomplete because it consists only of orders, then one should ask further whether our language is complete, whether it was complete before the symbolism of chemistry and the notation of the infinitesimal calculus were incorporated. The point of this response is, that, in making the objection above, one presupposes at least that a language could only be complete if it consists of more than just orders. But what could function here as a criterion of a language's completeness? Is the inclusion of the symbolism of chemistry such a criterion? And was our language then incomplete before this symbolism was incorporated? It is clear, I think, that the question serves as a reductio ad absurdum. The point of Wittgenstein's response is that the concept of a language's completeness is - at least - not related to the abstract degree of structural elements included in the language. Therefore, the use of "complete" in the objection is not a correct use.

But the remarks in PI 18 bring about a problem. What is meant by "language" in PI 2? If the objection in PI 18 should be an objection at all, then one must accept that only orders belong to the language. But in PI 2, as well as in PI 8 and PI 15, Wittgenstein does not speak explicitly about the language's consisting of orders, but only about its consisting of words and how they are used. Is there not a category-mistake in Wittgenstein's remarks? Someone who says that a book consists of a cover, pages, cardboard and paper makes a category-mistake, for the proposition that a book consists of a cover and pages and the proposition that a book consists of cardboard and paper are not propositions on the same category-level. From one point of view - say a structural one - a book consists of a cover and of pages. From another point of view - say a material one - a book consists of cardboard and paper, namely the cover consists of cardboard and the pages consist of paper. Whereas it is questionable whether something is a book which does not consist at least of a cover and pages, something can surely be a book even if, rather than cardboard and paper, it consists of, say, plastic or leather. To consist of a cover and pages seems then to be a necessary condition for being a book¹², but to consist of cardboard and paper is not. If what is true for books, is also true for languages, then the question must be answered: What takes the place of the cover and the pages on the one hand, and the cardboard, paper, plastic, or leather on the other? What seems to support the analogy, and what seems to give the answer, is that we can imagine that instead of the words "block", "pillar", etc., A and B use other signs, for instance different pieces of paper, or - as Wittgenstein himself introduces - marks which are used by A and B and which are borne by the tools A uses in building [PI 15], or colour samples [PI 8, see above]. And instead of ordering by uttering "Block!", A could also order by showing a piece of paper. So it seems as if the words, pieces of paper, marks, and colour samples play for language the role which cardboard, paper, plastic, and leather play for books.

But there is a difference between words, colour samples, etc. on the one hand and cardboard, paper, etc. on the other hand. Whether something is cardboard, or paper, etc. can be decided independently of the identification of its being a cover, a page or part of a book. The criteria of identity of cardboard, paper, etc. are independent from the criteria of identity of covers, pages, and books. But whether something is a word, or a colour sample, etc. cannot be decided independently of the identification of its being used in language.¹³ The criteria for the identity of a word are not independent of the criteria for the identity of a language¹⁴. The proposition that this or that is paper entails nothing about pages. But the proposition that that or that is a word entails that it is in some way used in language, for instance in the way of giving, obeying, etc. orders.¹⁵ But one problem remains then: does the process of building belong to the way the words are used, or not? And if it belongs to this, then we could ask whether Wittgenstein's "Conceive this as a complete primitive language." in fact means "Conceive this as a complete primitive activity."

This question seems to be answered in PI 7. Here Wittgenstein writes that "the whole process of using words in (2) (can be thought of - R.R.) as one of those games by means of which children learn their native language. I will call these games 'language-games' and will sometimes speak of a primitive language as a language-game." But in the last sentence of PI 7 Wittgenstein writes: "I shall also call the whole, consisting of language and the actions into which it is woven, the 'language-game'." Here it is clear that "language" cannot mean "language-game", for if it did so, the whole sentence would become obscure. What one has to register here is that Wittgenstein does not use "language" and "language-game" as termini technici with a clearly defined meaning, that is according to explicitly given rules, but in the way the word "language" is used in ordinary discourse. But this does not mean that it is used without any rules. Normally the context makes clear how a word has to be understood. When Wittgenstein speaks in PI 2 about a language for which the description given by Augustine is right, he uses "language" in the sense of "the whole process of using words". When he speaks about a language consisting of the words ... (or consisting only of orders), he uses "language" like he does in "language and the actions into which it is woven". But how are "this" and "language" used in "Conceive this as a complete primitive language."? I think both senses just mentioned are possible, but both give rise to special problems.¹⁶

If we take "language" and "this" in "Conceive this as a complete primitive language." in the sense of "the whole process of using words" (including the activities), then the "Conceive this ..."-demand has to be understood in the sense of "Conceive this as a complete primitive activity." What could "complete" mean in this case? To answer this it is fruitful to ask what it could mean to say a given language-game (in the sense discussed here) is incomplete. Remember Wittgenstein's answer on the incompleteness-objection in PI 18. The point there was that to call our language (in)complete presupposes a criterion according to which it could be (in)complete. The objection was rejected because there is no such thing that could function as a criterion.¹⁷ The only way I see to give this proposition a sense despite this rejection would be to understand it in the sense that the language does not function, work, or the like. That is: it has some aspects which make it resemble a language-game, but it is not really one. It is incomplete in itself. The demand "Conceive this as a complete primitive language." would mean then: "Do not look at this as if it could not function, work, or the like. Look at this as if it does."¹⁸ For the whole second part of PI 2 stands under a let-us-imagine clause, the demand would say: this is an imaginable, functioning language-game.

"Completeness of a language", if "language" is understood in the second sense above, means: a language consisting of this and that is complete if this and that are sufficient¹⁹ tools for the activities which this and that are used for. This means for the situation described in PI 2, that A is able to build with his calls of the words, that B is able to pass the stones A called for, and if they are sometimes not successful then not because there are not enough words or because there are no descriptions and only orders.²⁰ That A is able to build with his calls, and that B is able to pass the stones A called for, means then - with regard to completeness - that there are enough words (and perhaps also that they are different enough from one another, not too long, etc.) and/or that orders are sufficient. The situation may change, and that may bring about a need for further words, or language-forms. But now the language is complete. We see that to say a language is complete means to say that it is complete relative to a practice, not in itself, or absolutely. The practice functions as the criterion of completeness.²¹

On the basis of this interpretation we can say that somebody making the incompleteness-objection above, uses the predicate "incomplete relative to our language", without registering the hidden part "relative to our language" in this predicate. "Complete" is as well a predicate with one place more than obvious, as is "appropriate". Augustine's description is neither appropriate, nor inappropriate; it is appropriate relative to x, and not appropriate relative to y, where "x" and "y" stand for different entities. The language of PI 2 is neither complete, nor incomplete; it is complete relative to the language-game of PI 2, and incomplete relative to our variety of language-games. So one cannot object to Wittgenstein's remark that the language is incomplete. It is incomplete relative to our language-games, but this is not the question of PI 2. It is complete relative to the game of PI 2. These are not surprising insights; the game of PI 2 was designed for this purpose. Notice that the language of PI 2 is complete relative to the language-game of PI 2, not relative to Augustine's description.²²

Wittgenstein compares Augustine's description of the essence of language with a description of the essence of games according to which a game consists in "moving objects on a surface according to certain rules".[PI 3] This definition is incorrect with regard to the variety of games we play; it can be made correct "by expressly restricting it to those games (i.e. board-games R.R.)." (] The point of introducing the game-analogy is in our context twofold²³: - it is much easier to see the relatedness of definitions, descriptions, or the like, of games to specific games than it is in the case of language,

- if, for example, one were to describe two players, A and B, playing a game with four kinds of card, one would hardly object that the set of kinds of card is incomplete. It is simply only another game of cards than bridge, poker, or skat. And this is obvious, not least because games of cards are in many respects played according to explicit rules, some of them often prescribing with which and how many different kinds of cards the game has to be played. A set of cards of the game skat would be incomplete if it consisted only of 31 kinds of cards, instead of 32. Similarly, the language of PI 2 would become incomplete if, for example, A and/or B forgot the word "pillar" (or, in the cases of the extended versions of PI 2, if they lost a colour sample or a mark) but still had to build with pillars which were still available. The language would be incomplete with regard to the purpose of the language of PI - to build with four kinds of stones.²⁴ An objection that a given set of cards of skat is incomplete, would be true or false. But an objection that the set of kinds of cards of an unspecified game is incomplete (that is in the sense of the possible objection in PI 18) is senseless. Nevertheless it shows that one, making this objection, is familiar with other cardgames, like Augustine was familiar with - cum grano salis - our variety of language-games.

With the help of the game-analogy Wittgenstein is able to illuminate his use of primitive languages/language-games. It serves to disperse "the haze" which "surrounds the working of language" [PI 5], and which arises when we look at it through the general notion of the meaning of a word as found in Augustine. Instead of primitive, fictitious kinds of word application one can also take the primitive forms of language which "a child uses ... when it learns to talk. Here the teaching of language is not explanation, but training ("Abrichten" in German -R.R.)." [Ibid.]²⁵ So we must register another important distinction Wittgenstein makes, a distinction between:

- "the practice of the use of language", and

- the "instruction in the language" [PI 7].²⁶

That distinction has furthermore the aspect of bringing into account the fact that people have to be instructed in their language. Wittgenstein writes in PI 6 that an "important part of the training will consist in the teacher's pointing to the objects, directing the child's attention to them, and at the same time uttering a word; for instance, the word 'slab' as he points to that shape." Wittgenstein calls this "ostensive teaching of words" and contrasts it with "ostensive definition" (]²⁷. Whatever ostensive teaching of words consists of, it can surely not be identified with giving orders. Wittgenstein himself emphasizes the contrast between the orders of PI 2 and the instruction in the language of PI 2: "In the practice of the use of language (2) one party calls out the words, the other acts on them. In instruction in the language the following process will occur: the learner names the objects, that is, he utters the word when the teacher points to the stone." [PI 7]²⁸

The importance of the distinction between the practice of the use of language and the instruction in the language - in our context - lies in its capacity to solve an implicit tension in the remarks above. For, if people have to be instructed in their language and the instruction entails ostensive teaching, then the language of PI 2 would not consist only of orders, unless we were to make a categorical distinction like the one above, but still - which would be in accordance with Wittgenstein's position and which is, Wittgenstein aside, true - accepted that people have to be instructed in their language. Therefore, the objection of PI 18 would be based on a false presupposition - that the language consists only of orders - and would have to be answered otherwise.

But what gives Wittgenstein the right to make such a categorical distinction? The first thing one should notice is that Wittgenstein speaks about "instruction in the language", not about "instruction in the use of language".²⁹ The point of this difference is that Wittgenstein agrees with his interlocutor that the effect of the ostensive teaching of the words can consist in establishing "an association between the word and the thing."[PI 6]³⁰ This can mean various things, for instance evoking images; but one thing is clear: "... in the language of §2 it is not the purpose of the words to evoke images. (It may, of course, be discovered that that helps to attain the actual purpose.)" [Ibid.] To be a learning of the use of that language ostensive teaching is not sufficient. What a child learns in this way, one could say, is naming an object. But, in parenthesis of PI 49, we can say: "... naming and calling in the game of PI 2 do not stand on the same level: naming is a preparation for calling the words in the game of PI 2. Naming is so far not a move in the language-game any more than putting a piece in its place on the board is a move in chess. We may say: nothing has so far been done, when a thing has been named. It has not even got a name except in the language-game." [PI 49]³¹ That naming and calling (in the game of PI 2) do not stand on the same level means that instruction in the language is not something specified enough to be an instruction in the use of that language. And if it is not connected in this or that way with this or that language-game, then it is perhaps no instruction at all. But given the context of, for instance, the game of PI 2 its components are "processes resembling language." [PI 7]

To become an instruction in the use of a language, the training must be specified. Associations, mental images, and the like are not sufficient. Necessary and sufficient for the training in the use of a language(-game) is that the child can "act upon it in such-and-such a way ... Doubtless the ostensive teaching helped to bring this about; but only together with a particular training. With different training the same ostensive teaching of the words would have effected a quite different understanding." [PI 6] For something, we see, to be an instruction, or a training, in the use of a language, the relatedness to the language(-game) in question is essential. So we can say that such a training is not self-sufficient, nor self-defining; it is defined by the language-game. Whatever the training in the use of a language may consist of, it does not change the essence of the language(-game) in question³²; moreover, "it may be all one to us whether someone else has learned the language, or was perhaps from birth constituted to react ... like a normal person who has learned ..." [PI 495, my italics.] And even if the person has learned, it could be that he has learned only by watching [cf.: PI 31].

That the language of PI 2 has to be regarded as complete (in both of the senses above), does not mean that is has to be regarded as the whole language of A and B.³³ Otherwise the following passage from PI 6 would make no sense: "We could imagine that the language of § 2 was the whole language of A and B; even the whole language of a tribe."³⁴ If we could imagine this, then we should also be able to imagine that the language of PI 2 is not the whole language of A and B, or of a tribe.³⁵ This point can also be illuminated with the help of the game-analogy. We can imagine both:

-two players, or even a tribe, playing only one game, and

-two players, or even a tribe, playing the same game, but also other games.

And in both cases we could as well say that this and that are the complete elements of the game, or of one of the games respectively, as we could say that the game/each game is complete in itself.³⁶

But there is still a problem here. The problem is that Wittgenstein does not simply assert that the language of PI 2 is complete, but demands of the reader to conceive it as a complete language. So one should expect that Wittgenstein connects a certain purpose with this demand. What purpose? The question is the same as: What role does the demand of PI 2 play in the course of the argumentation in the PI? If we accept that (2), that is the essence of human language, lies at the bottom of the whole discussion at the beginning of the PI - which doesn't imply that other topics are not to be found among the PI, for instance the nature of mind - then it is not hard to see that the essential role of the demand is to serve as a preparation for the discussion connected with the concept of family resemblance. For, if we conceive the language of PI 2 as complete, then we immediately see that this language is very different from our language, that it is much more primitive. But we also see that the language of PI 2 is different from the language of PI 8, and from many other languages too, which Wittgenstein describes in the PI. What becomes a problem then is the assumption that there is (or must be) something common to all language-games. For, even if we call the words of both PI 2 and PI 8 "names", we see that the uses of the names are "absolutely unlike" [PI 10]. The same is true for language-forms such as orders, questions, and the like. [Cf.: PI 19ff.]³⁷ Therefore, if we accept that languages such as those of PI 2, 8, 15, etc. may be taken to be complete languages, then we are also prepared to see that there is nothing common to them all in virtue of which we call them all language.

Also this can be illuminated with the help of the game-analogy. "Consider for example the proceedings that we call 'games'. I mean board-games, card-games, ball-games, Olympic games, and so on. What is common to them all? Don't say: 'There must be something common, or they would not be called 'games'" --but look and see whether there is anything common to all. - For if you look at them you will not see something that is common to all, but similarities, relationships, and a whole series of them at that. "[PI 66] Wittgenstein calls these similarities "family-resemblances" [PI 67]. Therefore, the demand of PI 2 to conceive the language of A and B as complete, is the demand to put aside the prejudice that there must be something common to all (language-)games, and it is the demand not to theorize, but to look and see. To conceive the language of PI 2, and others too, as complete is a condition for seeing the family-resemblances, or - if seen - to give them their right weight.

Now we can see the importance of the point of PI 18 more clearly. To ask whether our whole language is complete is either itself an incomplete question, or is the question whether it is complete relative to the language-games we play. In the first case it is an incomplete question because it is not clear with regard to what our whole language should be (in)complete. That means that we have to decide according to which criteria we want to know whether our whole language is complete or not. If we want on the one hand to know whether our whole language is complete with regard to the variety of language-games we play - that is: whether our terminology, syntax, etc. are complete - or with regard to their respective "purposes" - then we want at best to know something that can only be answered by an empirical investigation. Such an empirical investigation could for example provide us with the insight that there are some language-fragments in our whole language which are incomplete.³⁸ The practical result could be then a proposal of a reform. "Such a reform for practical purposes, an improvement in our terminology designed to prevent misunderstandings in practice, is perfectly possible. But these are not the cases we have to do with (in philosophy - R.R.). The confusions which occupy us arise when language is like an engine idling, not when it is doing work." [PI 132]

II

But, there is a point in Wittgenstein's remarks in PI 2 that seems to bring about problems and to give reason for Rhees' and Øfsti's criticisms. To see this point better, we must go back to PI 2 and look at it more closely.

Who is speaking here? It is only A, who is speaking. What is B doing? He only has to pass the stones A called for. There is no difference in principle here between what A and B are doing and what a hunter and his dog are doing. Indeed, many dogs, which are trained for hunting, are able to understand - "understand" just in the sense of PI 6 - much more words than only four, and are able to do much more than only to pass things the master calls for. But nevertheless, dogs do not - for instance - philosophize, even if some dogs look very skeptic. Dogs, and animals in general, just do not talk. "... animals do not talk because they lack the mental capacity. And this means: 'they do not think, and that is why they do not talk.'" [PI 25]³⁹ To think, one could say, is to be reasonable (or rational), or - weaker - only thinking beings are reasonable (rational) beings. Therefore, whatever PI 2 may show us, it is not sufficient with regard to thinking/reasonable (rational) beings, that is with respect to (most of) us. The game of PI 2 lacks an essential aspect: that human beings think. Therefore, the language of PI 2 can be neither complete nor the whole language⁴⁰ - in a not trivial sense of these concepts - of (a tribe of) human beings. We couldn't even imagine this, if our imagination were in accord with our fundamental intuition about what it is to be a human being.

In this point Augustine was much nearer to the truth than Wittgenstein, for he uses mental predicates in his description of the learning of language. Whatever may be wrong with Augustine's conception in detail, it simply bases - contrary to Wittgenstein's - on a fundamental intuition we cannot miss. This is the point which Øfsti also wants to stress: "On a special level one can say, that the language-subjects are reasonable beings and that 'the extent' of their language (notwithstanding which 'suburbs' this language now entails or not) is unimportant." [1985, p. 589]⁴¹ A language can only be complete - in a not trivial sense - if it allows the users to be reasonable (rational). Similarly Apel writes: " Wittgenstein once said: 'If a lion could talk, we could not understand him.' This seems to me not very plausible since it is precisely linguistic competence and not ... the conditions of life (birth, death, sexuality) conceived of as independent of linguistic competence - that separate us from lions."⁴²

But not only are the "conditions of life (birth, death, sexuality)" not enough, neither are the particular activities which Wittgenstein considers with his language-games. As Aristotle wrote: "Just as for a flute-player, a sculptor, or any artist, and, in general, for all things that have a function or activity, the good and the 'well' is thought to reside in the function, so would it seem to be for man, if he has a function. Have the carpenter, then, and the tanner certain functions or activities, and man has none? is he naturally functionless? or as eye, hand, foot, and in general each of the parts evidently has a function, may one lay it down that man similarly has a function apart from all these? What can this be?" His answer is: "the activity of soul in accordance with, or not without, rational principle".⁴³ Therefore, it is not enough to presuppose a certain "transcendental language-game" or the like, but one has to make a clear distinction between language-games (in the sense of PI 2 and other sections) on the one hand, and the language on the other.⁴⁴ Even if Wittgenstein's concept of completeness is correct, it is valid only for language-games like that of PI 2, not for the language. But the language-game of PI 2, as we saw, is a language-game which neglects the differentia specifica of human beings. It is not only comparable with a hunter's and his dog's interaction, but also with interaction among animals alone, for instance with a cocks calling his hens by crowing. Therefore it is by no means clear, "when ... the different language-game competences of a person (are) sufficient to constitute linguistic competence ..." [1990, p. 138]. This is the question of "formal completeness", which is "rather avoided by Wittgenstein." [1990, p. 139]

This question is not one of complexity, like Wittgenstein seems to suggest (e.g. in the Blue Book), but rather a question of different levels. However complex the plurality of language-games which a child has to learn may be, "sooner or later we will have to reach a level where the subject becomes a subject in the full sense, a level which must have to do with its recognition of itself and others as autonomous, 'equivalent' subjects." [1990, p. 144] Even if we conceded that the builders in PI 2 think, it is "too meagre to capture the full sense of 'thinking'. And this would still be so however much we extended the repertoire of common language-games in the person's home country ..." [190, p. 145] For the language this means that it must be constructed in a way which allows us

(i) to reflect on language,

(ii) to move in and out of certain language-games,

(iii) to imagine and describe language-games other than those we ourselves play,

(iv) to give an "'overview'" on the games we play, and not only to play them,⁴⁵ and

(v) to reach and express a "complete communicative competence" [1985, p. 589.]⁴⁶

This level cannot itself be a "special, more or less separable language-game (as for instance Apel means -R.R.), but rather part, or an aspect, of a structural trait which runs through the whole of a complete language, i.e. through each of its language-games." [1990, p. 146]⁴⁷ The reason for the necessity of a complementary structure of language is that "any intentional act (in the full 'human' sense) must reach two levels":

(i) "the 'game' level"", and

(ii) the level where "the agent masters the concept of what is done." (]

The second level is the one on which descriptions of the game played are possible.⁴⁸ This is the level, which guarantees that responsible actions are possible, the level where "the drama of human interaction takes place." [Ibid.]

These are Øfsti's remarks on what is trivial with Wittgenstein's concept of completeness, and on what a non-trivial concept must include. Now we can look at what Wittgenstein could answer.

III

Section 25 of the PI goes in its entirety as follows:

"It is sometimes said that animals do not talk because they lack the mental capacity. And this means: 'they do not think, and that is why they do not talk.' But - they simply do not talk. Or to put it better: they do not use language - if we except the most primitive forms of language.-Commanding, questioning, recounting, chatting, are as much a part of our natural history as walking, eating, drinking, playing."

Why does Wittgenstein writes "But - ..."? Isn't it accepted in the first two sentences that animals do not talk? The difference between "they do not think, and that is why they do not talk" and "they simply do not talk" consists in the first phrase's being an explanation, and the second phrase's not being an explanation.⁴⁹ "Our mistake is to look for an explanation, where we ought to look at what happens as a 'proto-phenomenon'. That is, where we ought to have said: this language-game is played." [PI 654] To say that animals do not talk means then to say that they do not play language-games. That is why it is better to say that animals do not use language, than to say that they do not talk. And the exception is that they play the most primitive language-games. Whether a creature's uttering of sounds is talking or, say, crowing depends on the circumstances of the utterances. Not even each verbal utterance of a human being is talking, just like a cockatoo's talking is only talking in a very special sense, which is secondary with regard to our normal use of "talking". "We say 'The cock calls the hens by crowing' - but doesn't a comparison with our language lie at the bottom of this?" [PI 493]

To say that animals do not play language-games, if we except the most primitive ones, means that they do not play the variety of language-games we play, if we except the most primitive ones. This is a difference of multiplicity, complexity, structure, richness, content, and purpose in the sense above. It is a difference of social forms of life.⁵⁰ Therefore, even if a lion could talk, we could not understand him because his form of life would be too different from ours. But this is not something special for lions. "We also say of some people that they are transparent to us. It is, however, important as regards this observation that one human being can be a complete enigma to another. We learn this when we come into a strange country with entirely strange traditions; and, what is more, even given a mastery of the country's language. We do not understand the people. (And not because of not knowing what they are saying to themselves.) We cannot find our feet with them." [PI, p. 223]⁵¹

But to repeat: it is a difference, and one could add here, as in the case of the difference between living and not living beings: "this is a case of the transition 'from quantity to quality'." [PI 284] So, the first thing we must register, is that Wittgenstein very well accepts in his philosophy something like a fundamental (perhaps better: qualitative) difference between animals and human beings. However many "levels" there may be, in learning the multiplicity of our language-games a child becomes a being that is qualitatively distinct from animals. We see that the intuition of there being a fundamental difference between human beings and animals (and all other things in the universe we know about) is no argument against Wittgenstein's philosophy. On the contrary, in his PI Wittgenstein tries to give us a clear picture of that difference.

But, one could ask, what about thinking? Do the remarks above mean that Wittgenstein wants to deny that people think, and only accepts a difference in the complexity of verbal behaviour? Here, I think, one should see that Øfsti's argument bases on the primitiveness of the game of PI 2. Otherwise he would have to say that also the games of

"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-
..." [PI 23]

are played by beings which are not reasonable, or have no "linguistic competence". This would be, I think, obviously nonsense. On the contrary, to be able to do things like these above means to be able to think. To think (or to have "linguistic competence") is not something that lies behind, or at the bottom, of playing language-games, but is expressed in playing them. Understood in this sense the phrase "animals do not think, and that is why they do not talk" becomes at best a tautology.⁵² The exception, that animals use the most primitive forms of language - that is, that they play the most primitive language-games - means then that they think only in the most primitive sense of the word, are only primitive reasonable beings, or that they do not think in "the full sense of 'thinking'" (Øfsti). "We say a dog is afraid his master will beat him; but not, he is afraid his master will beat him to-morrow." [PI 650] Here both parts of the sentence are important:

- that we say a dog is afraid, that is: that we attribute mental capacities to animals, but

- that we do not say a dog is afraid he will be beaten tomorrow, that is: that we do not attribute to animals the mental capacities we attribute to people, but only primitive, simpler ones.

That people have to learn their language means then that they have to become reasonable beings. This point is Wittgenstein's main argument against Augustine: "And now, I think, we can say: 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. Or again: as if the child could already think (or were already reasonable - R.R.). And 'think' would here mean something like 'talk to itself'." [PI 32]⁵³

What Wittgenstein attacks is neither the idea that people think, nor the idea that the difference between human beings and animals may be seen to consist in the human being's having mental capacities and the animal's not having mental capacities (with the exception of the most primitive ones), but the idea that we have a clear picture of what it means to think, or to have mental capacities, according to which thinking or having mental capacities is something which is in principle independent from using language; and that this picture serves as an explanation for the human being's ability to talk and the animal's lack of such an ability.

But if Wittgenstein does not deny that there is a (fundamental) difference between human beings and animals, does this not indicate that there is something wrong with his demand in PI 2 and especially with his remark in PI 6? That there is a conceptual contradiction between the demand and the remark on the one hand and our concept of a human being on the other would only be true if to play the variety of language-games we play - and, accordingly, to have the mental capacities we have - were our only criterion for being humans. But since it is already indicated by the fact that people have to learn their language, and accordingly have to become thinking or reasonable beings it is clear that this is not the only criterion. If little children are human beings, then also A and B can be. That animals are also able to play a game like that of PI 2 does not mean that A and B, or children, are not distinct from animals, for we have furthermore lots of criteria for being humans.⁵⁴ Therefore, neither the demand of PI 2 nor the remark in PI 6 stand in contradiction to our basic intuitions about what it means to be human.

But now one could ask: What is the difference on principle between Wittgenstein on the one hand, and Øfsti (Apel, and others) on the other, if Wittgenstein neither denies that there is a fundamental difference between human beings and animals, nor denies that this difference can be seen to consist - understood in the right way - in the human's thinking and the animal's not thinking? The deep difference, I think, is one connected with the conception of philosophy, that is, the philosopher's understanding of himself.

Remember the words of Aristotle. He asked whether there is anything essential to human beings that is not reducible to the various functions they have in social life, anything that is essential to them qua human beings. And whereas all questions about the "good and the 'well'" of this or that certain function or activity belong to the various sciences and arts of these functions and activities, and these sciences and arts get their importance from the importance of their subjects, the question of the essence of a human being qua human being belongs to philosophy. If we know what this essence consists of, it seems, then we can also say which real life human beings should live, and what is wrong with their actual life, how society should be constructed to fit to the real nature of human beings, what individuals should really do - and not only "this language-game is played". Philosophy gets its importance from these important questions and the possible results which answers to them might bring about.⁵⁵

But now it seems as if Wittgenstein denies that there is such an essence, and would only accept forms of life, or language-games. "What has to be accepted ("Das Hinzunehmende" in the German original - R.R.), the given, is - so one could say - forms of life." [PI, p. 226] But, to accept the existing forms of life, it seems, excludes the possibility of (rationally) criticizing them. The exclusion of the possibility of (rationally) criticizing them entails, it seems, the exclusion of the possibility of (rationally) changing them. Wittgenstein's philosophy expresses, one could say⁵⁶, a pure conservatism. In the end philosophy loses its importance. So the question is: "Where does our investigation get its importance from, since it seems only to destroy everything interesting, that is, all that is great and important." [PI 118]

To give an answer to this question, I want to point to a similar problem. Bertolt Brecht writes: "Me-ti taught: Master Ka-meh says that consciousness depends on the actual way in which people produce what is necessary for life. He denies that, in their minds, people can free themselves farther from the economical standpoint than in economy. This sounds at first oppressive. But the simple consideration that in fact all great works came about via this dependence, and that these would by no means be diminished by acknowledging the dependence, puts everything right again."⁵⁷

Two points interest me in these remarks. The first is: that what Ka-meh taught sounds downhearted only for people who view the great works not as important in themselves, but rather as important because they are the expression of something else: the reason or rationality, for instance.

We can register something similar in the case of interpreting Wittgenstein. Singing catches, play-acting, building - all these and many other activities are not taken to be philosophically important. They are perhaps plebeian examples for something higher, but not in themselves worthy of consideration. "Language (or thought) is something unique" - as Wittgenstein lets his interlocutor say in PI 110. But he continues: "-this proves to be a superstition (not a mistake!), itself produced by grammatical illusions.

And now the impressiveness retreats to these illusions, to the problems." Surely, language and thought are something important, but they are not something unique! If the plebeian examples can be called thinking, then nothing is lost, except our delusion. "What we are destroying is nothing but houses of cards and we are clearing up the ground of language on which they stand." [PI 118]

The second point of Brecht's remark is this: to take the great works for what they are in themselves presupposes seeing them as something following their own rules, or with their own underlying criteria. This does not exclude that they are causally, or otherwise, effected by something not based on these criteria, not following these rules, the economy for instance. But to identify the causes presupposes to get a clear picture of what it is that is causally effected.

Something similar, again, we can see in the case of Wittgenstein. What he is interested in are the conceptual confusions which are expressed in philosophy. Therefore he tries to show what it is to think, to feel, to order, and so on. It is the essence of thinking, feeling, ordering, and so on, he is interested in. And the "Essence is expressed by grammar." [PI 371] The grammar is one of a family of word-uses, which is connected by family-resemblances. To see the family-resemblances it is fruitful to conceive the (primitive) language(-games) as complete.