Thursday, 20 October 2011

Wittgenstein's Toolkit

One rough but helpful way of considering the structure of the Investigations might be as follows: §§1-133 specify a new role for philosophy and introduce the necessary tools for the job. The rest of the book then gets on with actually putting those tools to work. Our understanding of them certainly grows richer as we see their application, but substantially everything needed for the task is there in the book’s first quarter.

With this in mind, it’s perhaps useful to take stock of the “tools” provided. The comments below are not intended to supplant my earlier posts. Hopefully, though, they’ll highlight the positive aspect of the Investigations and show that it’s not simply about taking a hammer to traditional philosophy.

If I had to sum up the central philosophical insight of Wittgenstein’s later philosophy, I’d say it was this: language is an activity. More specifically, it is a rule-governed activity performed by agents in particular circumstances. As he says at §25, “Giving orders, asking questions, telling stories, having a chat, are as much a part of our natural history as walking, eating, drinking, playing.” This consideration stands behind all the strands of his later philosophy and it’s the reason he seized upon the analogy between language and games.

Viewing language in this way highlights its active and varied role. Language is something we do, and we do it in many different, yet subtly interconnected situations. To understand it properly, therefore, it must be considered in context. Attempts to abstract language in the hope of discovering its essence merely falsify the phenomenon. Indeed, the whole urge to produce explanations sub specie aeternitatis is misguided because explanation is itself a language-game that only makes sense when viewed in situ: “If we say, ‘Every word in the language signifies something’, we have so far said nothing whatever; unless we explain exactly what distinction we wish to make” (§13). In such cases language is no longer working; it has gone on holiday. And the only result is nonsense.

Another benefit of likening language to a series of games is that it brings out the importance of two related features: training and rules. Language, like a game, is something we learn. We learn it in a variety of ways: sometimes we’re required to memorise a formal set of rules (at school, for example), but often we work out correct usage from observing others, and sometimes we just copy what everyone else is doing. This is all part of what Wittgenstein calls “mastering a technique” (§150).

Learning and application are intrinsically linked. We can’t do things with language until we’ve grasped the rules (mastered the technique), but the whole process of learning is shaped by the application. This is another way of saying that language can’t be understood when detached from its context. The rules and definitions which are supposed to explain meaning seem mysterious or puzzling when they’re extracted from the situations that give them purpose. Returning them to their rightful home dispels the mystery, and this is the intention behind Wittgenstein’s assertion at §43 that “the meaning of a word is its use in the language”.

Clearly, approaching meaning in terms of use is in keeping with the idea of language as an activity. Its philosophical advantage is that it brings the question of meaning down to earth. By deriving meaning from practice we are less prone to abstracting the concept and viewing it as something occult. In particular, we avoid the temptation to think meaning must be a thing – either the physical object correlated with a word or some kind of mental picture.

More generally, we’re able to trace the complex grammatical rules governing concept-words (including “meaning”) as they are applied in our lives. We see distinctions and similarities that are usually overlooked, and this prevents us from reaching for limited or partial definitions (eg, “knowledge is justified true belief”) and then being puzzled when they conflict with what actually happens.

What is less clear, however, is the precise status of the claim that meaning is use. Wittgenstein cannot be proposing it as a theory (unless he’s made a shocking blunder), and yet as a description of the use of “meaning” it is hardly self-evidently true. This, presumably, is the reason for the caveat in §43 (“For a large class of cases […] though not for all”). But the caveat in itself is troubling because it seems like he’s saying “in some cases the meaning of a word is an object rather than use” and yet (a) he elsewhere brings forward powerful arguments to show that meaning cannot be an object and (b) use does appear to play some part in all explanations of meaning (eg, the difference between “This→ is red” as a definition and as a complaint (because I’d asked for something orange) is a matter of context, although this raises a further question as to how far “use” and “context” are synonymous).

Further thought is needed. For now, though, I think we can say this: Wittgenstein invites us to consider meaning in terms of use because (a) we often do explain meaning by describing the use of a word and (b) from a philosophical viewpoint it helps avoid certain temptations that are more strongly exerted in other forms of explanation – and it does this because it keeps language fixed in the human activities that give it its life.

It is easy to suppose that all items falling under a particular concept must have something in common in virtue of which they qualify as examples of that concept. Wittgenstein maintains, however, that if we look at the way concepts function in our lives (ie, derive their meaning from use) we will see that this is not always the case. Sometimes concepts exhibit a family of affinities that have developed over time, “as in spinning a thread we twist fibre on fibre” (§67). The requirement for commonality is a hasty assumption that often leads to confusion.

For example, if we consider all the different instances of what’s called “understanding” we may become puzzled as to how these varied items all fall under the same concept. Here it’s tempting to posit a phenomenon (perhaps some kind of mental act) that will provide the unity we think we need. But far from solving the problem, the phenomenon creates a host of further puzzles and our confusion only deepens.

The notion of family resemblance concepts frees us from this impasse. It shows how concepts can retain integrity without commonality: they do it by remaining useful. Thus our preconceived need for unity is trumped by our actual behaviour and we no longer have to invent what we do not see.

Forms of Life

[NB: consider this section a “first go” at describing the importance of forms of life. I’m making some pretty contentious claims here and absolutely reserve the right to change my mind later on.]

Approaching language from the point of view of use inevitably brings the agent into the picture. If language is something that gets done then who is doing the doing? Wittgenstein’s favoured term for this “doer” is the strikingly non-specific “form of life”. Why choose this expression?

Well, for one thing, using “form of life” rather than “person” or “human being” discourages us from dragging in any distorting preconceptions we might have about ourselves. Nothing is assumed. Rather, the form of life is whatever happens to emerge from descriptions of linguistic activity. This allows for distinctions and connections both within and across any particular species, as well as change over time in the practices of individuals and linguistic communities. We might link the term to language-games here and say that insofar as two individuals share a language-game they share a form of life.

This non-presumptive generality of form arises out of the focus on use, and is linked to the very nature of philosophy as Wittgenstein saw it. The philosopher’s task is to outline the concepts we have. To go further and ask why we have them is to cross the boundary between a conceptual investigation and an empirical one. For the philosopher, therefore, the form of life represents a terminus (PPF §345: “What has to be accepted, the given, is – one might say – forms of life”). It is here we find the foundation of our concepts, not in theories about the a priori structure of the world, nor in theories about the a priori structure of experience. We simply have to accept the form of life as manifested through activity. “This is how we think. This is how we act. This is how we talk about it.“ (Zettel, §309). We can go no further; the buck stops here.

A major source of philosophical error involves faulty presuppositions (or theories) about the way language is actually used. We assume commonality, essence, similarity, and so on. The only way round this is to get a clear view of language. In other words, we must abstain from theory and simply describe its use (§66: “don’t think, but look!”). Perhaps surprisingly, this description is not preparation for developing better answers to our problems; it is the answer itself.

Why does description resolve philosophical problems? Because it brings our attention back from a situation where a concept is puzzling (ie, philosophy) to one where it is not (ie, our lives). The concept of knowledge, for example, puzzles philosophers. Is it possible? And if it is, then how? And yet every day we use the term without any difficulty at all. Wittgenstein suggests the problem arises when we lose sight of the position that such concepts occupy in our day-to-day affairs. “The civic status of a contradiction, or its status in civic life – that is the philosophical problem” (§125). Description restores clarity, and “the clarity that we are aiming at is indeed complete clarity. But this simply means that the philosophical problems should completely disappear” (§133).

Apologies for going over (mostly) old ground, but I’ve been keen to do two things: first, to highlight the fundamental importance of §§1-133 with regard to the rest of the Investigations; and secondly, to show the unity of approach behind the positive aspects of Wittgenstein’s later philosophy.

Speaking from personal experience, it’s tempting to gloss over §§1-133. Arguments about logical atomism and propositional form can seem more relevant to historians than anyone engaged with philosophy today, and it’s easy think, “When do we get to the sexy stuff? When do we get to rule-following or the private language argument? That’s where the action is.” But the discussion of the “sexy stuff” flows directly out of the approach and methodology developed in the book’s opening quarter. If you don’t have a clear grasp of that then it’s going to be next to impossible to see the later arguments in their right light. (For now I shall merely flag up the additional point that you probably can’t get a clear grasp of §§1-133 until you’ve read the rest of the book.) If my experience here is atypical then: good. If, however, it reflects a more general trend then: be warned.

Regarding the unity of Wittgenstein’s approach, I think it’s a corollary of its radical nature; the whole thing points away from more traditional philosophical methods. For me, this puts a deep question mark over attempts to cherry-pick his ideas or blend them with other approaches. How far, for example, can you pray-in-aid family resemblance while maintaining a broadly naturalist approach? Family resemblance ramifies into notions such as “meaning as use”, grammar and the descriptive method – and these notions work against the fundamental thrust of naturalistic analysis. The reason why such cherry-picking happens is interesting and (I believe) often has as much to do with disarming the threat of Wittgenstein as anything else. But that’s for another post.


  1. Hi Philip. You can probably guess which section I want to comment on! You ask "who is the doer?" and suggests that Wittgenstein answers "form of life", but actually what he does is refer vaguely to we! I think it is very understandable that he does so, but it certainly raises a lot of issues. For example, in talking about God are the religious believer and the non-believer playing the same language-game? Should we say that Wittgenstein and Gertie Anscombe (his catholic friend) did not share a form of life, but that Gertie and the fifteenth century Grand Inquisitor Tomás de Torquemada did? Of course, it is not just to do with religion - do me and Lady Gaga share the same form of life? Or me and an astologist? Or someone who sees the unconscious as a very important concept and someone who does not?
    As you suggest, I think the point of the concept is to highlight that explanations come to an end in the way we do things, but then we Wittgensteinians really do owe the world a discussion of "we" and that is certainly going to be quite a complicated discussion. My inclination would be to apply the concept pretty broadly, so that probably everyone I have mentioned by name in this comment shares the same form of life, but I haven't really thought it through. I am certainly much more sympathetic to that sort of conclusion than one that trivialises the notion by suggesting that our civilisation contains hundreds, thousands or even millions of forms of life!

  2. Heh. Even as I wrote that section I thought "I'm digging a hole for myself here", hence the huge caveat at the start. Yes, indeed, my "form of life = language game" equation has some difficult explaining to do (and could well be plain wrong), but I'm not quite ready to renounce it yet.

    A lot seems to turn (as so often with Wittgenstein) on our understanding of the term "same as". And I'm tempted to say that whether I share a form of life with Tomás de Torquemada or with my next-door neighbour might depend on why you're asking.

    But you're absolutely right: a) this requires further thought, and b) it's not an issue that a Wittgensteinian can dodge forever.

    For now, though, I will.

    BTW, the other day I stumbled upon a review of "Wittgenstein and Moral Philosophy" by Simon Blackburn. It was the first page only (in pdf) so I couldn't tell what he'd made of it. But it occurred to me that if I'd written a book on Wittgenstein and moral philosophy then Simon was probably one of the LAST people I'd want reviewing it. Was John McDowell busy or what? ;)

  3. A take on §43. First, I think it's necessary to establish a relationship between "description" and "explanation", the words used by Wittgenstein. I, of course, have no way of knowing what he had in mind. But rightly or wrongly, my guess is that per §10, the description of "what a word signifies" is a catalog of "examples" of its use (as in a dictionary definition). Each "example" could alternatively be called an "explanation", as in §43 - and will be in the following.

    When I first read §43, I made the "all too easy" mistake (per Philip's post "W, Names, Meaning, and Use") of interpreting it as saying something like "The meaning of most words is their use, but there are exceptions." On a more careful reading (motivated by that post), I came up with a revised interpretation (somewhat different from Philip's) that addresses the meaning of "meaning" - ie, the different uses of that specific word - which is the stated target of §43:

    The description of the employment [ie, use] of the word "meaning" will include explanations of its use in linguistics. Eg, "The meaning of a word of a specific language is the word's use in that language.

    In the case of names, sometimes the explanation will involve non-verbal actions. Eg, "The meaning of a name is the use of the name as a substitute for physically pointing at the bearer of the name."

    (The second explanation seems related to the last sentence in the first paragraph of §38 which introduces the Zen-like idea that the verbal pointer "this" is the "real" name. But I'm still fuzzy about how - if at all - it relates.)

    Finally, a reader who has gotten as far as §43 will presumably be pretty thoroughly immersed in the use of "meaning" as it relates to language. If so, in the construction "For a large class of cases ... but not for all" (especially considering the possibly misleading italic emphasis) the "not for all" - ie, non-linguistic - cases [explanations] can easily be assumed to be exceptions. But I think the wording is consistent with comparable numbers of linguistic and non-linguistic explanations (or even with more of the latter). Ie, from the fact that §43 presents only linguistic explanations one should not necessarily infer that in the full description of "meaning" most explanations are linguistic. Understood this way, there seems to be no conflict between that phrase and most (all?) explanations of specifically linguistic uses of "meaning" involving assigning meaning to words via their use.

  4. Thanks Charles. I'm currently preparing another post on Meaning is Use which is not a million miles away from what your saying about explanations of meaning and ties things in with the opening few pages of the Blue Book.

    BTW the remarks on "this" as the only real name (§38) are a reference to Russell's theory of proper names and the role he tried to make them play in his logical atomism. W's account of proper names in the TLP is not exactly the same (he never thought "this" was a name) but it was clearly influenced by Russell's theory in broad terms. Obviously, W rejected both versions in the PI.