Essay 3: While Reading Wittgenstein PI §§ 1-19 (1): A Cautionary Tale


This is an attempt at a fairly close reading of some central concepts that crop up in the first paragraphs of PI. I read that beginning in ways that resist common ways of understanding what Wittgenstein was up to. In the Postscript I try to find flaws in a modern classic, deeply admired.

Essay 3: While Reading Wittgenstein PI §§ 1-19 (1): A Cautionary Tale

Table of contents

    1 Introduction1

    A strange thing happened to me a while back: idly looking into the latest edition of a Norwegian translation of PU, for once having the text as a handsome book, it struck me how odd the first paragraphs are.

    That I was not prepared for, having had my reactions to the work formed while a very young student. My teacher, Jakob Meløe, used, repeatedly, to spend long hours with us over the beginning of the book: setting out in great detail the various early language-games, and more than anything else, filling out details in the shop example; always earnestly, and always in an even, insistent, tone, as I recall. It was as if Wittgenstein had revealed a wondrous explanation of our ordinary, day to day, linguistic behaviour, an explanation now passed on to us.

    Thirteen years ago, I gave a talk to a French-Norwegian conference much like this one.2 Two main claims of the talk were, first, that unlike the Tractatarian view of language, Wittgenstein’s PU view was incomplete, in the sense that linguistic meaning, i. e. the existence of meaningful languages, and language users, wasn’t accounted for, even though that seems to be what the later view can be read as providing. Specifically, I critiziced the opinion that the concept of language-game provided such an account. My point here was that a verbal string could be introduced and drilled in as part of learning a game, without thereby becoming a meaningful string. Just think of cases of nonsensensical rhyme. Secondly, I suggested that what was needed is a biological element, along evolutionary lines, so that the possibility of specific meaningful utterances were tied to our natural aptitudes and needs as a species. I proposed the work of Ruth Millikan as directed towards that purpose. Without arguing the point, I treated PI’s statements about language as statements about human languages. None of my listeners challenged that assumption.

    Needless to say, the talk was not well received by some listeners we may call, echoing a famous dichotomy, the right-wing Wittgensteinians. Times have changed however, so perhaps it is now more acceptable to take a critical look at an issue connected to my previous talk, namely the status of some concepts introduced in the opening paragraphs of PI, §§ 1-19(1).3 Specifically, I believe the time has come to bring to the forefront the assumption in my previous talk. For the truth is that early remarks in PI detach the concept of language from the concept of human being. I am not saying this is wrong in general, or that we should read the text as saying those two concepts belong together. Clearly we cannot do the latter, since Wittgenstein says that animals may have language. But I am saying that when constructing, or imagining, a language, it is natural and necessary to ask who, or what, can have that as a language. I am also of the opinion that readers, by and large, assume (fictional) humans to be partaking in those early language-games, and in doing so, may be reading too much into the text.

    Even though commentators have objected to some parts of the text, most have treated it with great respect, almost reverence. The main conclusion I draw is this: PI, in early paragraphs, is quite an obscure, or ambiguous, piece of philosophical writing. Surprisingly so, given that Wittgenstein endlessly fussed over it, leaving behind page after page of slightly different formulations. (A sign of unsettled thinking?) In order to substantiate this verdict, let me start at the beginning.

    2 The discussion of St. Augustine

    Wittgenstein seems to think of the passage quoted in § 1 as expressing what Gilbert Ryle once called a Fido-Fido theory of language. Words are proper names, standing for objects. (I shall leave aside what one may do while speaking such a language, as I shall not discuss the final sentence in the Augustine quotation, where St. Augustine says that such learning of the language enables one to express one’s desires.) Such an image of the nature of words fits well an idea of objects as, more or less, middle-sized physical objects, cluttering space around us. Since Wittgenstein further claims that this picture sees sentences as combinations of such names, couldn’t he find such a simple view in real life, – apart from the awkward version we find in the Tractatus? 4

    1) For Augustine himself does not hold the view, as is clear from the quotation:

    “Thus, as I heard words repeatedly used in their proper places in various sentences I gradually learnt to understand what objects they signified;”

    “In their proper places”: this means sentential structure. Minimally, a word can play different roles, maximally, there are different types of words. As we can see, especially in his On Christian Doctrine, Augustine considered grammar very important in shaping and grasping the meaning of what was uttered.

    “[W]hat objects they signify”: this can be fairly innocuous. Augustine talks of signs, as does Wittgenstein himself, a traditional way of conceiving words, in a tradition running at least from Aristotle to Condillac. Wittgenstein, in § 13, talks of what we may call the innocence of using the word “signify”, how little is said thereby, and throughout he uses “sign”. As Aristotle would say, terms that are in the categories signify, the exceptions to the categorical terms are the syncategorematic expressions. Frege, who held that all sentence parts reached through analysis “bedeuten”, explicitly says that not all “bedeuten” objects. The difference between him and the tradition may be different opinions as to what it takes to be an object.”

    It is of interest that Wittgenstein in the first paragraph writes of names for actions and properties! And later, in § 15, he is curiously insensitive when he states that names typically work as labels attached to things. If we think of sticking labels on objects, labels such as these I am about to mention are common: Black & Decker, to be fetched tomorrow, needs seasoning, five pounds to pay, half price. Hardly a name among them! And in most of their uses names are not stuck on what they name.

    In § 325 Wittgenstein expresses a far better insight into what Augustine actually thought. Holding the view that language is given us by God, and that men cannot teach men to speak, to be creatures using language, Augustine had a subtle interest in difficulties of expressing the same thought in different languages. A fact we can detect from his developed interest in problems of translation.

    So perhaps Wittgenstein did pick on Augustine because Augustine held a view on the origin of language that Wittgenstein thought false. An issue we see as different from that of language acquisition in children.

    2) The quotation from St. Augustine in itself expresses something perfectly acceptable, as a description of how children master words. Wittgenstein says that the Augustinian description fits the way children learn to speak, which is exactly what the quotation claims, given that Augustine would mean: learning to speak that particular language (see § 5).

    3) It is important to have one thing clear: Wittgenstein is not denying that the Augustinian description is a description of a language, a possible language. – Whatever that means; Wittgenstein does not specify necessary requirements. (§ 65 is very illuminating on why Wittgenstein does not give such a specification.) What Wittgenstein says is that the picture does not capture the essence of human language. What that should mean, on a reasonable reading of his remark, is that the picture quoted tries, but fails, to describe (the essence of) any human language anyone could call their only one. His further complaint is that the description does not fit our language. It is not clear what “our” means here: it cannot mean the same as “one among the ones we have”. Nor can we say that “our” means our only. Perhaps he is thinking of first language, “Muttersprache”, which word does occur in § 7. The concepts of first language and natural language, English, German, etc., seem absent from the beginning of PI.6

    Anyway, as he says: it is the description of a language more primitive than ours. It fits a language, but one more primitive than ours. So, Wittgenstein sees languages more or less as ranked according to primitiveness. His concept of primitive language goes together with some other terms characterizing languages. I shall now turn to that.

    3 Language versus language use – speech acts

    1) Throughout our text Wittgenstein varies between talking of language, and talking of use of language, now speech acts are in focus, now words, with no apparent attention given to the contrast. Think of his line that the languages in §§ 2 and 8 consist of orders. This statement, in § 18, carries over into a discussion of the notion of a complete language.

    2) Wittgenstein develops the idea of complete language, giving as an example what we find in § 2: the language there is a complete primitive language. Complete and primitive are two important concepts. Another is whole (“ganze”) language, used in § 6: I believe Wittgenstein means that the whole language of someone, a person or a tribe, is always complete, no matter how primitive it is. “Whole” here means only, or total. There is no limit to how primitive, or simple, a total language, the only one had, may be.

    Wittgenstein compares in § 18 a language to a city, as cities may grow, so may languages. They may be added to. This idea of adding to, or extending, a language is very different from logicians’ idea of extension to a logical system. Complete logical systems cannot be added to, except conservatively, without creating inconsistencies. Wittgenstein’s plausible view is that any language may be added to, in a non-conservative way, without thereby being seen as logically incomplete. An ordinary language is never a finished product, speakers may add to the language in ways which make possible sayings not previously expressible. At the same time we cannot use this possibility of enlargement as a proof that a language is incomplete until the enlargement has taken place. That would lead to the conclusion that no language is ever complete. Whatever Wittgenstein meant by “complete”, it is not what the logician means by a complete system. A language, for Wittgenstein, is not a logical system. A language does not get more complete by extending its range of possible sayings. That opinion seems reasonable: for an ordinary, common language to be complete entails that added pieces do not fill in perceived gaps or replace missing pieces. His position can be put in a couple of catching phrases: completeness is compatible with possibilities for enlargement!7 An only language, of a person or of a tribe, is never incomplete!

    But his way of thinking about complete primitive languages goes the other way. Take an ordinary, natural, language, remove parts of it, keep going until only a small fragment is left: this fragment is itself a complete primitive language. Fragment here means: just a few words and a few occasions on which to use them.

    So we have two very different ideas: the first is that we may enlarge any given language; the second, the problematic one, is that we may take powers of expression away from a language, and whichever way we do this, the left-overs are still a language.

    In the text we are looking at, Wittgenstein doesn’t state either of these ideas in those words, but he expresses closely enough the first with the metaphor of a city, and his notion of complete, primitive language is not licensed by this first idea. It is the second idea, in one form or other, which is required here. That idea drives the construction of very simple linguistic exchanges, claiming that the verbal exchanges, with the resources used, make up a complete language (I leave other actions out of it). And that second idea is unreasonable, which we see when asking who has the fragment as their (only) language. Remember that when we enlarge a given language, the natural habitat is doing this with human languages; we see that it is natural to think it possible, generally speaking and thinking of past developments, for us to increase expressiveness. And a reduction of expressiveness also typically seems to make sense when we start out from a human language, and imagine ourselves shedding parts of it, however well we understand how far that may go. But, there comes a point when we have to ask: could this be an only language for us? Or, can we picture creatures having this as their only language? We must tell ourselves that the important test as to whether we are dealing with a possible (complete) language is if we can see any kind of creature having that as their only language.

    3) We must further tell ourselves: our interest is in human languages; and Wittgenstein’s characterizing judgement on the examples in §§ 2 and 8 is not convincing when they are judged to be specimens of human language. The natural conception of them is to see them as isolated cases of language use, in odd circumstances, – a very different idea. If we came upon such examples, as isolated cases or in other texts, the natural conception might be a good interpretation of what the author intended, I do not see that to be so in the present case.

    When Wittgenstein says that the builders’ language in § 2 may be a complete language, and their whole, total or only, language, it is tempting to press that claim. Do they only speak those words, and only while building; or is it possible that when the lads knock off from work and come home, “slab” means: what is for dinner? And later, in bed, “slab” means: not tonight dear, I have a headache?

    Wittgenstein sets up the example in such a way that we are supposed to know the sense of “slab”; it is our word after all that is used by the builders, isn’t it? But, unless we know the range of use for “slab” in the builders’ language, we do not know what word, or words, the inscription “slab” is in their language. Their slabs might not all be our slabs. Used the way Wittgenstein describes, the words are not our words.

    4 Imaginary exercises

    Wittgenstein is quick to say things like “we can (could) imagine” or “let us imagine”: for example, we can imagine a language consisting entirely of orders. (§ 19) Well, can we? Wittgenstein, of all people, ought to be, and in other places is, aware of the pitfalls here.8 We do not imagine something just by saying “Yes, I am imagining, or picturing it”. “Imagine pigs flying, – yes, yes, I picture a pig now, up there, flapping its short legs, propelled across the sky by powerful gases exploding from its behind”. As Hegel says: we have to distinguish between real possibilities and superficial ones, the latter, we may say, are expressed by any non-contradictory sentence. There is a difference in credibility between saying “I can learn some Italian” and saying “I can jump over the Moon”.

    When Wittgenstein says that we could imagine the language in § 2 being the whole (“ganze”) language of A and B (see § 6), this leads to one conclusion. If such a primitive language as that in § 2 could be a complete language, and the whole (i.e. only) language of A and B, the total and only linguistic means at their disposal, then they are not humans. Here is the argument.

    The argument

    Pr 1(1)The imputed Augustinian language description, call it IALD, doesn’t fit
    a human language
    Pr 2(2)IALD fits BLIT (the builders’ language in § 2)
    1,2(3)BLIT is not a human language (i.e. not a complete, or only, language spoken by humans)
    Pr 3(4)BLIT is (or can be conceived as) a complete language
    Pr 4(5)BLIT can be (conceived to be) the only language of the builders in § 2
    3,4,5(6)The builders in § 2 have a language, but do not speak a human language
    Pr 5(7)Human beings who speak have a human language, whatever
    else they speak
    (8)The builders in § 2 are not human beings

    The first three premisses come more or less directly out of Wittgenstein’s text. The fourth is a reasonable consequence of his claim that the language in § 2 may be the whole language of A and B. The fifth I consider a necessary truth, one Wittgenstein cannot very well deny. We have to handle “human language” with care: humans may speak BLIT in the sense that we can get humans to say what is said in § 2, but that does not entail that BLIT is a human language.

    This is what the text permits us to conclude, unless we take Wittgenstein to make a modal claim: there are no human languages as primitive as that in § 2, but there could be. That is, unless we read the first premiss as “IALD does not fit any existing human language, but it is a description of a possible complete human language.” And this reading expresses what would be an odd claim to make by Wittgenstein, a claim I also believe is false and aim to undermine.

    In order to reach the conclusion, I referred, amongst other pieces of text, to the first sentence of § 6. But that sentence has an ending that directly contradicts my conclusion:

    “die ganze Sprache eines Volksstamms”

    Since the word “Volksstamm” is used to speak only of human beings, Wittgenstein here says that we can imagine the language in § 2 being the whole language of human beings, that is, A and B can be imagined to be human beings. I cannot incorporate what Wittgenstein says here otherwise than that his text contains incompatible claims.

    The conclusion of the argument, I wager, is deeply counter-intuitive to most followers of Wittgenstein. (I shall later turn to a surprising exception, one that can be said to be based on philological support for the above conclusion.) Most readers probably conceive A and B as esoteric types that could come out of a habitat such as a Polynesian island. But, if A and B are not human, what are they? Well, in § 25 Wittgenstein admits that animals may use very simple, primitive languages. So, we may ask Wittgenstein: are you really of the opinion that animals can use language, that is, play language-games? And further, is the builders’ example such a case?

    5 Language-game § 7

    If ‘the Augustinian theory’, of the uniformity of words, is mistaken, what should ‘replace’ it, in Wittgenstein’s view? Quite obviously, an account showing the various ways in which words do mean, in other words: how they function in language-games, and how this functioning shows up the differences among words. Wittgenstein needs to bring forth the actual uses of words, and then we see how they contribute, in different ways, to meaningful sayings. This puts a heavy explicatory load on the concept of language-game. So, let us look at it.

    Wittgenstein explains what he means by the expression in § 7. He there gives four different cases of language-games, plus two simpler processes resembling language:

    • 1) The whole process of using words as in § 2 to learn one’s mother tongue, or native (first) language [Muttersprache]
    • 2) primitive languages, as the one in § 2 – whatever that is
    • 3) processes of calling out names, as in ring-a-ring-a-roses
    • 4) a whole consisting of language and the actions into which it is woven.

    It is strange that such a central notion (word, concept) is given such a disjointed account. The first three we may understand more or less, but the fourth is different, what that could mean is difficult to fathom. At the same time, 4) gives the most complex sense, with the widest ramifications. 4) comes from the final part of § 7. According to Schulte’s edition, this part is an insertion found only in the Spätfassung (late draft), handwritten by a strange hand. (Our text is to a remarkable extent identical with earlier drafts, going back to the one found in the Urfassung, MS 142, from 1936.) My guess is that most readers use a fifth sense of the word, or a sixth, or....., for one sense go to § 1309 . Later, at § 290 and in Part II, when Wittgenstein talks of the beginning of a language-game, and what is said then, I think we are dealing with something much more complicated than his characterizations in § 7 indicate, something they do not prepare us for.

    Wittgenstein uses the word seemingly everywhere in later texts; no matter how variously it may be read. It is curious that one known as so anti-theoretical chooses to employ centrally a word one may rightly consider proto-theoretical, or quasi-explanatory.

    I would like to stop here for a moment and raise a question natural to consider at this point: how does Wittgenstein see the relationship between language, linguistic meanings and language-games? It would be nice to say that we have the inkling of a clear answer in the text, – but: A plausible answer looks impossible within the paragraphs occupying us. (So I shall look a bit beyond.) The question also opens up serious disagreements among some important writers deeply involved in Wittgensteinian exegesis.

    Rush Rhees has a very decided opinion: “[Wittgenstein] says that any language is a family of language games” (Discussions of Wittgenstein, p. 73).

    Rhees considers this a weak point in Wittgenstein’s philosophy, since it does not account for the unity a language has. According to some informed scholars, Rhees is (almost) alone in holding this opinion. Cora Diamond has stated explicitly that Wittgenstein is “not working with a language-game theory of language.”10 Whatever else Diamond intended, I read the remark as opposed to a view of language as a collection of meaning-independent language-games. But other close readers and interpreters of Wittgenstein, e.g. Peter Winch according to Diamond, hold the view that language-games are separate, with no crossover effect with respect to word meaning.11 What the view of Winch leaves obscure is what role language plays, in giving language users the unity they need to have as thinkers, through the employment of stable word meanings that transcend particular language-games.

    Rhees doesn’t quote Wittgenstein saying what he attributes to him, but is it a reasonable attribution? One difficulty in deciding is that the issue itself is not very clear. Let us say that language-games as particular incidents are occasions for use of language. Such particular incidents, or tokens, fall under various descriptions. They are, in multiple ways, describable by general terms, themselves typically not used on those occasions. Some of those general terms describe the activities of the language user, and derivatively a type of occasion. A language will contain the words used on these occasions, but also the general terms that describe the language-games, both tokens and types. To ask about the relationship between a language and the occasions when its words are used can now be centered on one specific question: do words bring with them into occasions of use what we call dictionary meanings? Put thus, the question seems simple minded, with an obvious yes-answer. And many remarks by Wittgenstein fall easily into line with such a natural position, including the remark from § 116 about the use of words in the language which is their home.

    This remark was mistranslated prior to Anscombe’s final, third, edition. In the first two, a word was said to have its home in a language-game. (Strangely, all three editions use the qualifier ‘original’: what does it mean to say, in this context, that a language is the original home of a word?). Anyone working solely from the English text would thus have a reason for taking Rhees’ line. And there are other remarks, including one I believe is very important for understanding Wittgenstein’s discussion of private language, namely PI § 261. I find Wittgenstein’s thinking about private language very illuminating, and informative as to his thinking about language, and it connects in an interesting way Wittgenstein with Herder and Hegel.12 I do not consider it misplaced to use § 261 for the purpose I have here. At the end we find this:

    “[an inarticulate sound] is an expression only as it occurs in a particular language-game”.

    Unfortunately, PI’s translation of § 261 is possibly misleading, PU has:

    “[einen unartikulierten Laut] ist ein Ausdruck nur in einem bestimmten Sprachspiel”.

    A better translation would be:

    “[an inarticulate sound] is an expression only in a specific language-game”.

    (The German text does not authorize “occurs”.)

    The German word “Laut” I see as a word for the auditory version of inscriptions, the word for which is “Zeichen”. Considered in isolation, the distinction, if that is what it is, between inarticulate sound and expression, is reminiscent of the Tractarian one between “sign” and “symbol”. The present distinction is between, say, a mere physical mark (or inscription) and a vehicle of meaning. Wittgenstein is generally careful about not putting anything between the sign and the fact, see § 94:

    “The tendency to assume a pure intermediary between the propositional signs and the facts. Or even to try to purify, to sublime, signs themselves”;

    and this from § 503:

    “If I give anyone an order, I feel it to be quite enough to give him signs. And I should never say: this is only words, and I have got to get behind the words.”;

    and this, § 504:

    “But if you say: “How am I to know what he means, when I see nothing but the signs he gives?” then I say: “How is he to know what he means, when he has nothing but the signs either?””

    We should be careful and not identify his distinction with the one held in hardcore analytical philosophy (David Lewis): a word is a mark plus a meaning. Lewis seems to consider the distinction natural, but it is artificial, since the mere sound, or inscription, is reached by abstraction. Words in any language are not mere sounds or marks to speakers of that language. And speakers will not see words of an unknown language that way, once they see them as words. The question is: does Wittgenstein reach the signs by abstraction?

    The whole of § 261, as I read it, stresses the inevitability of using a common (shared?) language when communicating, so private signs, were they possible, could not be used to tell others. (Hegel would have liked this, given his treatment in the Phenomenology of the early shapes of Geist.) The question now is whether the quote, to return to our concern, does express the view that an unusable sound is not an expression, or whether is says that an unused sound is not an expression, i.e. a sound is a part of language only when used. And this second one I read as denying that words have dictionary meanings.

    I cannot tell what the quote expresses. But it seems to give Rhees, and Winch, some support for their contentions. However, Wittgenstein also often talks in a general way of using an expression (sign, word), without thereby seeming to use it in a particular language-game, and still he treats the expression as meaningful, – which is the natural conception. I cannot believe that Wittgenstein considers ‘something’ a word only when used in a language-game. If he were to do that, his opinion would be that words are, as it were, held in storage until activated on occasion, that is, whenever they are used, discussed or thought of. This would shift the problem to the one of explaining what ‘words’ are when in storage, are they words?

    Perhaps any distinction between the translations comes across as artificial. But the new translation opens up a different focus for the remark: it can be read as laying stress on specific, and so indicate conditions on the adequacy of being a language-game. If we read the quote as saying that a word must be in use somewhere, if it is to be meaningful, then § 261 offers no support for Rhees’ contention. See also § 96:

    “(But what are these words to be used for now? The language-game in which they are to be applied is missing.)”

    He may point to other remarks though. When Wittgenstein in § 19 opens with this:

    “It is easy to imagine a language consisting only of orders and reports in battle.”,

    then it is tempting to see here a blurring of any distinction between sentences of a language and utterances of such sentences. It is a telling feature of § 19 that when Wittgenstein discusses the nature of words versus sentences, he puts exclamation marks into his examples. Thus the discussion is conducted with sentences-in-use. The meaning that, say, “slab” has is a question of what is meant by saying it. So Wittgenstein marks what it is said as: “Slab!”

    As a general observation one may say that Wittgenstein does not impart to the reader any feel for why a particular language-game exists. (Actually, he makes us feel there is no “why” to be answered.) Why would language users behave in those ways, and not in others? As a consequence of this lack of interest, readers also miss a sense of how the author sees language-games as different, or not, and how they may be connected. It is no defence here to say that he need not answer any why-question, that he is only describing. It is precisely his description in the first paragraphs that is in question. What he does is talk of Lebensform, and so must we.

    6 Lebensform

    1) If the notion of language-game does not seem to provide much in the way of making us understand how we come to be language users, and how to understand what a language is, what about Lebensform? After all, § 19 links Lebensform directly to language.

    The word “Lebensform” occurs five times in PU, according to the index in the bilingual editions, The first comes in § 19, the second in § 23. The others are § 241, and two places in Part II. Schulte lists six occurrences, but the sixth, attributed to § 325, is in error.

    Here are the five, in the English PI, but leaving the German word in;

    § 19 “And to imagine a language means to imagine a Lebensform

    § 23 “...the speaking of a language is part of an activity, or of a Lebensform

    § 241 “That is not agreement in opinion but in Lebensform” [that is, to agree in language we use is to agree in Lebensform]13

    p. 174 “[T]he phenomena of hope are modes of this complicated Lebensform

    p. 226 “What has to be accepted, the given, is – so one could say - Lebensformen.”

    How are we to read these occurrences: four in the singular, and only the last a plural case? Given what we have been led to understand previously, Wittgenstein is asking us in § 19 to look upon the existence of the totality of particular language-games played by users of the same language as the existence of a Lebensform. If so, how are we to understand the word?

    My two modern German dictionaries, Duden and Wahrig, both give “Lebensform” two sharply separate meanings, the same as the word “livsform” has in Norwegian. One is what I shall call the biological sense, using the word as a general term for kinds of living organisms, roughly the same as the word “species”. The second sense, or rather set of senses, is roughly way of life, or way of living. For this second sense both German and Norwegian have alternative words, e.g. “Lebensweise”, “Lebensart” and “livsførsel”, “levemŒte”, respectively. Both languages have a variety of words in this area, with subtle differences. The second sense, in both languages, is open to a wide variety of interpretational slack, at the minimal end we have what the English would call “life-style”: e.g. healthy, vegetarian, sporty, alternative14 , etc. One may try, on this basis, to picture, or build up, a maximal sense, providing what some philosophers call a “thick” concept, in this case an anthropological one.15 What this would be is less than clear. My guess is that most readers of PU have a more or less developed idea of such a maximal sense in mind when they come across “Lebensform”.

    Wildhagen, in his German-English dictionary from 1953, has only one entry for “Lebensform”, which he translates as meaning life-pattern. The sense is clearly the biological one. Wildhagen did the work from the twenties onwards, and died in 1945.

    It seems that a ‘thick’ life-style sense is (was) neither easy to define, nor attach to the word

    “Lebensform”.16 Such a sense cannot without debate be given when translating the German word. It is therefore worth reflecting on whether Wittgenstein, or others, succeeded in expressing such a sense. At a minimum, we ought to find an acknowledgment that an explanation is owed us of authorial meaning. One doesn’t identify a phenomenon by using a simple phrase where a long, theoretical explanation is needed, and none has been given.

    2) We should be careful about judging how far people are moulded by their time and their culture. Until one knows the minds and capabilities of language users, one cannot say which transgressions are within their ken. The dangers of attempting to use a word as “Lebensform” with a ‘thick’ life-style sense, are twofold: it is wide open to more or less easy readings; with an added danger when coupled with sentiments such as Lebensform being the given, what has to be accepted, constituting bedrock. His opinions, thus interpreted, make Wittgenstein sound like a glib philosopher of culture.

    3) Now to the piece of interpretational support mentioned earlier. At the end of her life Anscombe did some revisions in her translation of PU. Since a number of serious flaws were not corrected, she may have left the revision work unfinished, but she made one change which ought to cause alarm bells to go off for many readers of Wittgenstein. The old translation by Anscombe translates “Lebensform” as “form of life”. In the first two occurrences of “Lebensform” she changed her translation from “form of life” to “life-form”, leaving the last three alone.17 (The index doesn’t register the changes.) What does “life-form” mean in English? According to the latest Shorter Oxford English Dictionary “life-form” in English carries only the biological sense, meaning roughly species18 . The expression has a fairly technical biological sense: it is in use when talking about plants or living organisms.

    Anscombe is not to be ignored, but it is curious: did she go through all five occurrences? I am not aware of any one making a reference to this change in translation, but, unless her mind had gone, Anscombe must have done a major rethinking of what Wittgenstein could have meant. Maybe I ought to stress that understanding “Lebensform” as the English understand “life-form” does not, in itself, exclude human affairs from consideration, it opens up for other species.

    7 Conclusion

    Anscombe has left readers with a choice. When reading the builder examples we see them either as cases about human beings, or as cases about other life-forms. What we do may depend on what we ignore in the text.

    Whichever way we read them, we have problems. This can be seen if we return to the remark in § 19 (1): to imagine a language is to imagine a life-form. Equipped with that, let us return to the builders. What are they? Two options: they are human, or they are, say, a kind of very large insect. What are they building? A wall, a house, a hive? Can we see an end to their putting slab on slab? Do they make other moves in their building project? We cannot ask them, for they cannot answer. But we can ‘study’ them. Do we see them as having blueprints? Do they really have no other words, and if human, do they build, i.e. are they builders, like the carpenter next door? When they stop working for the day, what happens then? Do they freeze up? Or do they just stop talking and walk away, carrying on dumbly, in both senses? Can both utter the same words? When I try to recall what I previously thought when reading the first paragraphs, I draw a blank. I cannot even recall that in my early years as student we ever made guesses as to what they were doing, –beyond putting slab on slab, that is. Or, did they even do that? How could we have gone into the examples, in such great detail, and not even have asked?

    The first example in the book, the shop story in § 1, is later referred to twice by Wittgenstein, but not referred to as a language-game. I believe we, the readers, do look upon it as such. If any one sees the fourth case in § 7 as including the shop story on the strength of descriptions in § 519 , that is fine with me.

    Not realizing that the shop story works as a sleight of hand, it dazzles us. I think the conjuring trick is there: accepting it as a use of language, we have taken the bait. And when we much later come across the hints in § 130 of how to conceive the language-games, how well served by those hints are we in regard to the notions criticized here? Well, – § 130 talks of our language, not human language nor just plain language. Not much help there. Generally speaking: anything can be compared to any other thing, there are always similarities and dissimilarities. The question to ask is: what telling similarities are there between the example in § 2 and our language? If we cannot see the game in § 2 as a language, then the question is not asked of two objects in the same category. Whatever Wittgenstein later says, the early ways of describing language-games lead to mistaken perspectives on human languages. Which is my point here.


    1. Anscombe (1957), G. E. M., “Does Oxford Moral Philosophy Corrupt Youth?” The Listener, February 1957. Reprinted in Mary Geach & Luke Gormally (eds): Human Life, Action and Ethics Essays by G.. E. M. Anscombe, Imprint Academic, Exeter 2005.
    2. Augustine (1991), St., Confessions, translated by Henry Chadwick, Oxford University Press, Oxford.
    3. Azzouni (2005), Jody, Tracking Reason: Proof, Consequence and Truth, Oxford University Press, New York.
    4. Diamond (2003), Cora, “Unfolding truth and reading Wittgenstein”, Nordic Journal of Philosophy, Volume 4, No 1.
    5. Duden (1983), Deutsches Universal Wörterbuch, Duden Verlag, Mannheim
    6. Pichler (2005), Alois and Simo Säätelä (eds), Wittgenstein: The Philosopher and his Works, The Wittgenstein Archives, Bergen.
    7. Rhees (1970), Rush, Discussions of Wittgenstein, (the essay is “Wittgenstein’s builders” from 1959-60), Routledge and Kegan Paul, London.
    8. Shorter Oxford (2002), Shorter Oxford English Dictionary, Fifth edition, Oxford University Press, Oxford.
    9. Whittlesey (1943), Derwent, “The geopoliticians”, in Edward Mead Earle (ed): Makers of modern strategy, Princeton University Press, Princeton.
    10. Wildhagen (1972), Héraucourt, German – English Dictionary, Second edition, Brandstetter Verlag, Wiesbaden.
    11. Winch (1987), Peter, Trying to make sense, Basil Blackwell, Oxford. Reprints both “Im Anfang war die Tat” and “Facts and superfacts”.
    12. Wittgenstein (2001a), Ludwig, Philosophische Untersuchungen, Kritisch- genetische Edition, Herausgegeben von Joachim Schulte, Suhrkamp Verlag, Frankfurt am Main.
    13. Wittgenstein (2001b), Ludwig, Philosophical Investigations, translated by G. E. M. Anscombe, Third Edition, Blackwell Publishing, Malden, Oxford and Carlton.
    I may be here under false pretences, since I do not claim there is much philosophy in my remarks. I like to think of them biographical and philological. “Here”: the main part of the paper was given at a French-Norwegian Conference in Bergen, September 2005. I have made a few changes in the text as it was delivered, and added the final section. Since this is an English text, the work discussed is referred to as PI, but I shall have occasion to distinguish between the German original PU and the English translation, also referred to as PI.
    My contribution was published in a volume from the Wittgenstein Archives, edited by Henry and Utaker. The volume is available from the Wittgenstein Archives’ website at See also the second essay in this book.
    My reason for stopping at the first part of § 19 is simple, I see Wittgenstein as starting a new discussion at that point, having so far ‘set the scene’.
    For very different reactions to Wittgenstein’s early examples of language-games and his use of Augustine, see Postscript below.
    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.
    PI does not index ‘first language’, ‘native language’ or ‘mother tongue’. Schulte’s critical-genetic edition does not index this first occurrence of “Muttersprache”, but only the second in § 156.
    This is arguably the case also for, say, first-order logic. “[T]here are many ways to augment the expressive resources of first-order logic – by certain generalized quantifiers (“There are uncountably many…” etc.), and by other tools, and yet retain abstract completeness.” (Jody Azzouni Tracking Reason, p. 196). The same page makes clear that by “abstract completeness” is meant recursive enumerability.
    “There is a lack of clarity about the role of imaginability in our investigation. Namely about the extent to which it ensures that a proposition makes sense.” (§ 395)
    “The language-games are rather set up as objects of comparison which are meant to throw light on the facts of our language by way not only of similarities, but also of dissimilarities.” (§ 130)
    In a manuscript titled “Unfolding Truth and Reading Wittgenstein”, p. 29.
    Diamond’s reference is to Winch’s “Im Anfang war die Tat” and “Facts and Superfacts”, both in his Trying To Make Sense. “Meaning” is my interpolation, Winch discusses truth. I see the issue of meaning as consequential to that.
    Personally, when discussing Wittgenstein on private language, I like to play down rule-following in favour of his insistence on the need to use common language when communicating. I also do not think commentators have been paying sufficient attention to the occurrence of “interpretation” in § 201. They employ in their discussions a very different concept. See the following essay.
    There must be a distinction between agreeing in opinion and having the same language. “Agree in language we use” seems to blur that distinction. We may share the same language, but disagree on which words to use on particular occasions.
    “Alternative” is actually Duden’s illustration of this second sense.
    Or sociological; the word, used in biology, was given an analogical sense in the new science of political geography, applied on states. They were seen as organisms, needing space and resources: Lebensraum. See Derwent Whittlesey “The geopoliticians” in Earle (ed.) Makers of Modern Strategy, Princeton 1943, pp. 388-411. Whittlesey gives details of how in the 1920s right-wing groups in Bavaria developed the ideology of Lebensraum from the work of the founders of political geography. Spengler lived in Munich at that time.
    Anscombe herself was very sarcastic about the easy use of “way of life”, see her essay “Does Oxford Moral Philosophy Corrupt Youth?” (reprinted in Human Life, Action and Ethics, 2005, p. 166). She obviously saw it as a case of smug self-satisfaction.
    I assume Anscombe is responsible, but other hands were involved, see Kenny in the Pichler & Säätelä volume, p. 341.
    The expression does not occur in the Oxford English Dictionary before the Supplement from 1987, where the first registered use is from 1899. That text explicitly refers to the German word.
    “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. A child uses such primitive forms of language when it learns to talk. Here the teaching of language is not explanation, but training.”
    Harald Johannessen. Date: XML TEI markup by WAB (Alois Pichler) 2011-13. Last change 18.12.2013.
    This page is made available under the Creative Commons General Public License "Attribution, Non-Commercial, Share-Alike", version 3.0 (CCPL BY-NC-SA)


    • There are currently no refbacks.