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

Philosophical Investigations 206:The Common Behaviour of Mankind

Imagine you came into a strange country. You did not understand the language of the people - if their verbal utterances were a language at all. Now:

(Q) "In what circumstances would you say that the people gave orders, understood them, obeyed them, rebelled against them, and so on?" [PI 206]

This Wittgensteinian question (Q) is commented by Savigny with the thesis:

(i) "The answer to the question in [PI 206]b demands a theory."¹

(i) can be understood at least in two different ways:

(i') (Q) demands as an answer a theory, that is: only a theory could be accepted as an answer.

(i'') The answer to (Q) - whatever it will consist of - demands a theory.

(With regard to (i'') one could, and perhaps would, tacitly add : "... as a justification of the answer.")

What Savigny writes a bit later is not thesis - for the reason that it is, in my eyes, undebated:

(ii) "Wittgenstein does not tell us, what his theory is." In short: Wittgenstein does not give us his theory.²

From (i') and (ii) follows in my view:

(iii) Wittgenstein does not answer (Q).

(And with regard to (iii) one could, and perhaps would, tacitly add: "... as so often in the PI." (i'') and (ii) do not exclude that Wittgenstein did in fact answer (Q), but - again only tacitly - without a sufficient justification.)

Undebated is, as I said above, (ii): Wittgenstein really does not give us a theory. Not only is it lacking in the immediate context, it is not to be found in the entire PI. At least he does not give us a theory in that sense of the word, in which it is taken by him. With regard to (i) we can choose: we can take one of the two alternatives, both or neither of them. (i'') is in my view relatively uninteresting for an interpretation, whereas it could be very interesting for a critique, or a systematically oriented elaboration, of the sparse Wittgensteinian remarks - assuming, that is, that Wittgenstein gave an answer. But since the problem here is interpretation and neither critique nor further elaboration, (i'') shall not further be mentioned, without justifying why (i'') is probably false.

(i') becomes more plausible if we smuggle a word or two into (Q): (Q') "In what circumstances would you be justified in saying ...?"

For the sake of greater clarity and precision (Q') could be reformulated into:

(Q')' "What could you give as a justification for saying that the people ...?"; or complementary:

(Q')'' "How would the people's behaviour have to be, in order for you to say with justification that they gave orders ...?".

The more we reformulate (Q), the clearer, but also the more problematic, the results will be, when we compare them with Wittgenstein's text.

I think (i) is simply inappropriate. And this impression becomes stronger if we pass from (Q) via (Q') and (Q')' to (Q')''.³

This impression is highly welcome, for especially (iii) seems to me to be too unplausible.⁴ Let us therefore forget everything with the exceptions of (Q) and (ii), and register as our first result that all the more or less extended commentaries on PI 206c - which without exception take PI 206c and other parts of the PI to be in some way Wittgenstein's answer to (Q) - are basically right. Now it is time to have a look at this answer:

(A) "Die gemeinsame menschliche Handlungsweise ist das Bezugssystem mittels welches wir uns eine fremde Sprache deuten."⁴

The following questions are immediately connected with (A), if one is interested in interpretation:

- What does "interpret" mean?

- What does "system of reference" mean?

- What does "common behaviour of mankind" mean?

The last question has been central to the discussion in the literature. Therefore let us take it as our starting point. The interpretations which have been proposed in the literature can be ordered by grouping them according to the domain they take to be "common":

(1) the common behaviour is a (regular) behaviour that is common to the members of the observed group - this is Savigny's interpretation;

(2) the common behaviour is not only common to the members of the observed group, but furthermore common to the group the observer stems from - this more or less sums up Schulte [1990];

(3) the behaviour is common to all human beings, to mankind -in this is the standpoint adopted by Baker/Hacker and Haller [1979;1984;1988]⁶.

(3) is the strongest interpretation in the sense that, with (3), it is presupposed that there is (at least) one (kind of) behaviour that is common to all human beings, whereas this domain can not be so wide in the case of (2) and must be even smaller in the case of (1). From an extensional point of view (3) entails (2) and (2) entails (1). Hence (1) is the extensionally weakest position. With regard to (1), (2) and (3) stand in need of justification. Anyone who tries to defend (only 1)⁷ shows that he is interested in exactly that behaviour in regard of which (1) is distinct from (2) and/or (3). Here the point is an intensional one. Anyone who tries to defend (only 1) would thereby defend the position that, whatever may be subsumed under (2) and/or (3), is not sufficient, perhaps even not necessary. In this sense (only 1) is the intensionally strongest position. In his interpretations of PI 206,207 Savigny tries to justify this position. In his essay [1989] this position is further elaborated. I will refer to both, but primarily to the commentary.

With regard to (Q) Baker and Hacker give an interpretation which is similar to Savigny's: "What justifies the judgement that alien people are giving orders, understand them, obey of flout them?" The difference with regard to Savigny is that Baker and Hacker (not only at this place) do not talk about (the need of) theories. But the more important difference for our topic is their interpretation of (A): "The common behaviour of mankind provides an essential leverage for interpreting an unknown language." This common behaviour of mankind is characterized as "not merely that behaviour which manifests our animal nature, our natural needs for food, drink, warmth, our sexual drives, our physical vulnerability. etc. It also includes the diverse species-specific forms which such behaviour may naturally take for human beings.It is part of the natural history of mankind that we are impressed by fundamental features of our lives (birth, death and procreation), by elemental features of our natural world (the sun and moon, the cycle of seasons, the fecundity of nature, its fury and tranquility), by the basic patterns of human relationship arising out of sexual differentiation, parenthood, the overlapping of generations."

Savigny is right in mentioning that Baker and Hacker do not found their interpretation on the text of the PI, especially PI 206 and its context. Instead of this they refer to two passages from the manuscripts. And also Haller does not refer to the context of the passage in question. Therefore (3) must be taken not to be justified by its proponents according to the principles of "textimmanent interpretation" as put forward by Savigny.⁸ But (3) is taken by Savigny not only to be not justified enough by the context of (A), it furthermore obstructs Baker and Hacker from giving an appropriate interpretation of PI 207a, 1,2.⁹ This means: even if (3) cannot be justified in terms of PI 206c - although it might be in terms of other passages of the PI - it is an obstruction for the interpretation of at least one passage from the PI. If one does not want to attribute inconsistency to Wittgenstein's text, one could be forced to say that (2), (1) or (only 1) are the only appropriate interpretations.

PI 207a, 1,2 go as follows:

"Let us imagine that the people in that country carried on the usual human activities and in the course of them employed, apparently, an articulate language. If we watch their behaviour we find it intelligible, it seems 'logical'."

Baker and Hacker comment the passage "carried on the usual human activities": "... puzzling, since if it turns out that the noises they emit are not speech, then can they carry on the usual human activities at all? Without a language they would just be hairless apes." In this comment Savigny views "an interpretation interesting for the feuilleton", namely the supposition Wittgenstein would have argued for an "anthropocentric language-imperialism".¹⁰

Wittgenstein continues:

"... But when we try to learn their language we find it impossible to do so. For there is no regular connexion between what they say, the sounds they make, and their actions; but still these sounds are not superfluous, for if we gag one of the people, it has the same consequences as with us; without the sounds their actions fall into confusion-as I feel like putting it.
     Are we to say that these people have a language: orders, reports, and the rest?
     There is not enough regularity for us to call it 'language'."

Savigny points out that "their language" in the first sentence is used attributively and not referentially. The people do not have a language. "Their language" is that, which seems to us to be their language. But then they are - if Baker and Hacker are right - hairless apes. Of course it is clear that they are people. Here, like at many places in the PI, Wittgenstein introduces a fictitious community of people to contrast particular sides of our psychological concepts and thereby to make them visible. But the question of Baker and Hacker can be given sense by considering that it is not a priori clear where we would like to give, or to accept, limits for the property "being a human being". If to carry on the usual human activities and to use in the course of this a language belongs to that property, then Baker's and Hackers's question is correct. Savigny does not see any problem here, for he writes: " ... contrary: the behaviour is described as 'usual human activities', their behaviour as 'seems 'logical''... So we find a human way of acting ("Handlungsweise" in German, see above - R.R.)...". Here, I think, one has to look closer at the text.

Firstly, it is surely clear that with PI 207a not only orders are meant - which are the ostensible topic of (Q) - but language as discussed in PI 206c and 207b. It is also clear that it is a fictitious situation ("Let us imagine"). Now, Wittgenstein uses the phrases "wie es scheint" (in the English translation: "apparently") and "it seems" (in the German original: "erscheint uns"). What we are to imagine is, that it seems the people use an articulate language and that their behaviour seems 'logical'. So far there is no difference between this and Savigny. But now the question arises, whether "usual human activities" is also used attributively and not referentially, as it is, according to Savigny, in the case of "language". Exactly this problem is touched by Baker and Hacker with their question. Savigny seems to suppose that this is not the case.¹¹ At least his whole argumentation against (3) bases on this assumption. But if this assumption should be true, then the situation would be strange. For what we are to imagine then are usual human activities, which - as we are to imagine at the same time - only seem to be connected with an articulated language. This double demand only makes sense if the things demanded are different. Usual human activities do not only seem to be connected with an articulated language. Normally they are. At least the games of ordering, of reporting, etc. - as we play them - are connected with an articulated language. And they are connected with it in such a way that, to say they seem connected with an articulated language has no clear sense. On the contrary, they are paradigmatic examples of what we call "activities connected with an articulated language", paradigmatic of - to say it with Austin - how to do things with words. Something can seem to be a usual human activity which is connected with an articulated language (can seem to be ordering, to be reporting, etc.) - which we are to imagine according to Wittgenstein - and here we can err. That is: it may be that what seemed to be such an activity connected with language is not really one. But it cannot happen that we imagined a usual human activity which seemed to be connected with an articulated language, but this only seemed so. What could happen - and what therefore could be demanded by Wittgenstein - is that something seemed to be a usual human activity which is connected with an articulated language, but that we register while trying to learn it, that it is not connected with an articulated language, but only something that appears to us as such. But then it is not the imagined usual human activity, if an articulated language essentially belongs to it, as it is the case with ordering, reporting, etc. In our imagination the articulated language need not be regularly connected with the activity. This is one of the important point of Wittgenstein's philosophy of language.

Here one could refer to PI 207a, 4, where Wittgenstein says that the sounds the people utter are not superfluous. Savigny writes with regard to this remark: "If there is no regular connection between sounds and actions, then the gagging doesn't have exactly the same consequences as it does for us. (For us there is a regular connection.)" But, if there is no regular connection between sounds and actions, how can it be then a usual human activity?

If we interpret PI 207a, 1,2 in the way proposed above then also the "aber"¹² in the next sentence becomes clear. For there are two things we are to imagine:

- an explorer who only watches, and

- an explorer who furthermore tries to learn the language. Now it should be clear why at the beginning I made such a trouble about (Q). This question is not concerned with the problem in what circumstance we would be (theoretically) justified in saying that the people give orders, obey them, etc., but with thecircumstances. The circumstances are also different with regard to the explorer. What he wants to say while watching may be different from what he wants to say while learning.

That for an observer it may only seem that the people are engaged in usual human activities which are connected with an articulated language, bases on the fact that knowledge of the rules which characterize our use of language is neither a necessary nor a sufficient condition for following them.¹³ That the explorer registers that there is something wrong with the "language" of the people, must not be connected with his insight into the rules we follow. Here it is sufficient if he registers that he cannot - while trying to learn the language - take part in the games in the usual way. Then he can perhaps come to the result that the people do not have a language, that it only seemed so. That he cannot learn the language is neither the cause nor the reason for the unknown people's not having a language. On the contrary, for they have no language, there is nothing he can learn. But that he cannot learn it is the cause or the reason for his judging that they do not have a language. This point is similarly accentuated by Savigny: "Wittgenstein describes realistically in which way we find out that they do not have a language. (Before a philologist can get to grips with describing a living language he must first learn it.) But it has to be emphasized that the language fails as such because there is a lack of regularity, not because we cannot learn it."¹⁴

Before I try to give an answer to the other two questions from the beginning, those concerning "interpret" and "system of reference", I want to draw attention to a section close to PI 206/207 that is in more than one respect similar to them. If the following emphasis of the resemblance is correct, then at least it is not true of (2) that "nowhere in the preceding text can even the vaguest hint" (Savigny) of (2) be found.

In PI 200 Wittgenstein writes:

"It is, of course, imaginable that two people belonging to a tribe unacquainted with games should sit at a chess-board and go through the moves of a game of chess; and even with all the appropriate mental accompaniments. And if we were to see it we should say they were playing chess."

The "tribe" here corresponds to "the people" in the "unknown country" of PI 206. The games of the former correspond to the language of the latter; the game of chess corresponds to the giving, obeying, of orders etc. That games are unknown in the one corresponds to the fact that the people do not have a language in the other. "The common human way of acting"¹⁵ is then that the two people - as we do - sit - as we do - at a chessboard - as we do - go through the moves of a game of chess - as we do - and show the appropriate mental accompaniments - as we do. (That there is so much correspondence is, of course, not a surprising result, so that one could say: Look, here Wittgenstein created an example which resembles our life enormously. The correspondence is constructed.) In these circumstances, where there is for us as the observers ("if we were to see it" - my underlining) everything is as we are used to, we would of course say that the two people were playing chess.¹⁶ But what if we tried to play a game of chess with one of the two people?

Wittgenstein continues with PI 200:

"But now imagine a game of chess translated according to certain rules into a series of actions which we do not associate with a game-say into yells and stamping of feet. And now suppose those two people to yell and stamp instead of playing the form of chess that we are used to; and this in such a way that their procedure is translatable by suitable rules into a game of chess. Should we still be inclined to say they were playing a game? What right would one have to say so?"

There are at least two possibilities to interpret this passage. The first bases on the supposition that "those two people" are the "two people belonging to a tribe unacquainted with games" (my italics). Then, of course, they do not play in the second case.¹⁷ The second interpretation which seems possible, takes the "instead" literally: "those two people" play really, only not a form of chess we are used to. "Those two people" can then, of course, no longer belong to a "tribe unacquainted with games". The point of this interpretation is that we should not exclude that something really is a game even if it is far from resembling one. On the other hand something might not be a game even if it looks exactly like one.¹⁸

Whichever interpretation one chooses influences the interpretation of the question at the end of the whole section. But independent of both is an interpretation of the answer Wittgenstein suggests to the preceding question: we would not be inclined to say they were playing a game. The visible differences, the superficial differences, are too strong. In this sense there is also no common human way of acting, no system of reference, no interpreting.¹⁹

But there are not only resemblances between PI 200 and PI 206/207. One important difference is that the examples given in PI 200 concern one token of behaviour, whereas it seems clear that in PI 206/207 types are meant. If Wittgenstein doesn't explicitly say at the beginning of PI 200 that in the tribe games are unknown, then the question whether what the two people are doing is playing or not would hardly be answerable at all. The cases would be "temporally under-determined".²⁰ In this view PI 206/207 are further elaborations of PI 200; here the answer is not impossible because there is not enough time for the "explorer". If observation is sufficient, then the observer has enough time to observe. The new element is: learning is emphasized, instead of observing.

But now to the questions: What does "system of reference" mean, and what does "interpret" mean? In which way could "system of reference" be used if (1) is correct? Here is a suggestion: We observe how the people carry on different activities and how, in the course of these activities, they utter different sounds. Or, we observe how they utter different sounds without carrying on any activities. We want to know whether at least some of the sounds they utter are orders, questions, etc. But we only observe tokens of activities and sound utterances. These are not enough for us. What we need are types, or schemata, namely types, or schemata, of sound utterances which are regularly connected with the people's other behaviour, that is with types, or schemata, of their behaviour. What we need is - a system of reference.

This interpretation is, of course, guided by the content of (1) and it does not show more than that one can give associations for "system of reference" which are in accordance with, or at least not contrary to, (1). Therefore the interpretation of "system of reference" is no experimentum crucis which could help us to justify a decision between (1), (2) and (3).²¹ But at least one difficulty connected with the given interpretation of "system of reference" should be seen. In this picture there seems to be no good place for the distinction between observing and learning, that is for the importance of this distinction with regard to the problem in question. This difference seems only to be one of the psychology of the explorer. What sounds more natural therefore is an interpretation of "system of reference" on the basis of (2). Our interaction with the people of an unknown country is that system, that is: that way of acting which is common both to them and to us.

To justify that (1) is correct and that (2) is not, that is to justify that (only 1) is correct, it is not enough to show that (1) is an appropriate interpretation, but furthermore that (2) is not. According to Savigny (2) is compatible with the text, because "interpret" in PI 206c has not to be interpreted as "translate".²² This position presupposes that "to interpret" is not the same as "to translate". But in the commentary on PI 207 one can read: "... translating is, according to (or after ? - R.R.) PI 201 nothing other than interpreting." So, we are saying that, according to PI 201, "to translate" is to be read as "to interpret", and if we ask now how "to interpret" is to be read, we hear that it is to be read as "to translate". That doesn't sound good.

In one sense of "translate" this cannot be taken to be the correct interpretation of "interpret" in PI 206c, even if it seems that this is exactly the point. In PI 23 one example of a language-game is: "translating from one language into another". But "a language quite strange to you" in PI 206b makes sufficiently clear that with "interpret" in PI 206c is not meant "translate" in the sense of PI 23. To play the game of translating from one language into another we have dictionaries, grammars, classical examples, specialists for special fragments, etc. But in this sense we simply cannot - ex hypothesi - (try to) translate the sounds of the people. But in this sense of "translate" "interpret" in PI 201 is also not to be interpreted as "translate". As far as "interpret" in PI 201 is to be interpreted as "translate", "interpret" in PI 206c is also to be interpreted as "translate". There is no conceptual change between these neighboring sections. According to PI 201 "interpret" is not that "which is exhibited in what we call 'obeying a rule' and 'going against it' in actual cases." In the situation described in PI 206b there is nothing we could call "obeying a rule" or "going against it", for the language is "quite strange to us". Here one can only interpret, based on the common human way of acting, as it is described in PI 200. A good translator, on the other hand, only chooses sometimes, but normally he follows the rule blindly. [Cf.: PI 219] That makes him a good translator. Before one becomes a good translator, one has to learn the language.

Let us summarize. (3) is too strong for interpreting PI 206c, for it is not made compatible with the whole section and its context. Whether (3) is fruitful for the interpretation of other passages was not discussed here.

Neither (1) nor (2) is excluded by the text. But the text itself is sufficiently unclear to make a decision between (1) and (2) difficult. But it was shown that (only 1) is too weak, and that the arguments against (2), which are given by Savigny, are not strong enough.

To come to a decision it seems appropriate to widen the context.²³ For instance PI 54, a section not mentioned in any of the commentaries, seems to bring (2) into trouble and to support (1) and (3).

Finally, no independent and systematic argument was given here. Both - the widening of the context and the systematic argumentation - go beyond my aim in this essay.²⁴