From the ALWS archives: A selection of papers from the International Wittgenstein Symposia in Kirchberg am Wechsel

1. How do we institute (and bind ourselves to) norms?

The enlightenment has equipped us with a strong intuition that there would be no right and wrong in the world, were it not for the attitudes of intentional beings like us. Norms, we now insist, are “our creatures” (Brandom 1994, 626). Unfortunately, however, Ludwig Wittgenstein's “rule-following considerations” (PI 138-242) have since destabilised our confidence in our capacity to beget such creatures – to frame them, to follow them, and to base our assessments of one another's behaviour on them.

The problem is that anything that can be said or done in an attempt to fix the demands of a norm can be interpreted in such a way that any (future) course of action not (yet) explicitly considered turns out to accord with it. The trouble, of course, is that an interpretation of what is said or done in order to fix a norm is itself subject to norms, producing the same problem over again and threatening an infinite regress. At the same time, it won’t do to equate a norm with a piece of, say, neurological machinery, as this approach either begs the question how we tell whether the machinery is working properly, or fails to capture the fact that a norm specifies what should happen, not what will happen. We are left wondering whether it is at all possible to fix a norm and bind oneself (and others) to it.

Today, the debate on rule-following is dominated by community verdict theories (Kripke 1982, Wright 2001), by dispositionalist accounts (e.g. Pettit 1990), and by quietist or therapeutic takes (McDowell 1998). However, the first seem overly revisionary, the second seem unable to capture the normativity of norms, and the third seem to reject a quite legitimate call for elucidation. In this essay, I want to discuss a recent approach which seeks to avoid the problems of its competitors, namely the strategy adopted by Robert Brandom in his Making It Explicit (“MIE”, Brandom 1994).

2. From practical attitudes to normative statuses

Brandom's strategy to demystify normativity rests on the idea that an agent's normative status – i.e. the property of being bound to a norm, so that some of her performances count as correct, and others count as incorrect – can be explained in terms of her and her peers' practical attitudes. If true, this would be good news for naturalists, because practical attitudes are easier to place in a naturalistically conceived world than normative statuses are.

Practical attitudes are dispositions of differential response to or interaction with specified patterns of stimuli. They can involve complex algorithms (and thus require the availability of memory) but they can also be very simple. At any rate, we can specify them without relying on intentional notions. Brandom's account starts with the idea of multiple beings displaying practical attitudes towards one another (MIE, 30ff.) and revolves around the claim that if the attitudes have the right structure, we can see them as giving rise to actual normative statuses, i.e. to proprieties and improprieties in the beings' conduct. Since Brandom takes treating as correct/incorrect to be a legitimate description of an attitude, he writes that it is possible for beings who mutually treat each other as correct/incorrect to make it the case that their performances are or become correct or incorrect.

Initially, the strategy's prospects seem dim. In particular, it is hard not to be impressed by the dilemma which is standardly diagnosed as its main problem (see Rosen 1997): If practical attitudes are really just dispositions of the sketched kind, then all we can hope to get from them is regularity, not normativity. This is true even if they involve sanctioning manoeuvres (Hattiangadi 2003). In order to get proprieties and improprieties of conduct, it seems, we have to interpret practical attitudes as already endowed with normative significance, i.e. see them as properly or improperly adopted. If we do this, however, the account becomes viciously circular. In the end, it is hardly more illuminating than the statement “[T]hat is an authentic Vermeer just in case it is correctly attributed to Vermeer” (Rosen 1997, 167).

3. The theorist as bearer of attitudes

Brandom, however, is not only aware of this charge, but actually has a two-part response to it. The first part is to take sides for one of the supposedly dilemmatic theses:

The work done by talk of ... statuses cannot be done by talk of ... attitudes actually adopted … nor by regularities exhibited by such adoption… Talk of ... statuses can in general be traded in only for talk of proprieties governing adoption and alteration of ... attitudes – proprieties implicit in social score keeping practices. (MIE, 626, see also xiii and 58ff.)

From the perspective of his critics, of course, this stance pushes Brandom's story right into the pit of vicious circularity. Brandom's reaction to this charge (his response's second part) consists in a trick pulled off in the last few pages of MIE , namely to change the theorist's position from one of detached observer to one within the social phenomenon theorised about:

We are always already inside the game of giving and asking for reasons. We inhabit a normative space, and it is from within those implicitly normative practices that we frame our questions, interpret each other, and assess proprieties of the application of concepts. … With the collapse of the external interpretation [which pictures the theorist as a detached observer of essentially alien agents practically relating to each other] ... those proprieties are assimilated to the score-keeping proprieties in our own discursive practices. The norms turn out to be … here. (MIE, 648)

Now, this passage certainly provokes the suspicion that Brandom falls back on a form of quietism or even platonism. However, I do not think that this is the case, and I want to propose a reading which actually strengthens Brandom's official strategy. It rests on two ideas. Firstly, practical attitudes can be jointly self-validating. Secondly, normative statements, including the theorist's statements about the appropriateness of practical attitudes, themselves exhibit practical attitudes. It is in this sense that we are “always already inside the game”. There is still circularity in this proposal, but it is not vicious. Let me elucidate the proposal with a simple example, paper money, and then apply it to the phenomenon of language.

4. Paper money and value

It is quite obvious that in order to carry value, a Euro note (say) need not be edible (or consumable in any other way). Also, it hardly relies on people with guns. All that a Euro note needs in order to be valuable is to be widely valued, where “valuing a note” signifies the disposition to exchange it for goods or (other) notes which are designated as of equal value (or more for buyers and less for sellers). The qualification “widely” is meant to block the idea that my privately valuing a piece of paper is sufficient to confer value on it all by itself. What is needed, in addition to me, are other agents with the same disposition. After all, I can only buy something for a Euro if there are sellers, and sell something for a Euro if there are buyers.

As soon as these others are present and equipped with the right dispositions, we can speak of a qualitative change. It now becomes warranted to say that the Euro note carries actual value. What this means, among other things, is that if I stopped valuing it accordingly, I would be making a mistake, whether or not the mistake was noticed – let alone sanctioned – by anyone. In fact, once the Euro note has attained value, this fact underwrites (in an unassuming way) an infinite number of proprieties and improprieties: a buyer can always try to demand more goods for her money, and a seller try to offer less.

By now, it is probably clear that the idea I am after is that a paper note's value exemplifies a normative status, while valuing the note exemplifies a practical attitude, and that in an unproblematic way, the latter explains the former. More precisely, it looks as though multiple agents’ attitudes towards paper money are jointly sufficient for the latter to attain the status of being valuable. Its status of being valuable, in turn, is sufficient to make it appropriate for each agent to take the attitude of valuing it. What we see, here, is a circle, but not a vicious circle, for it is not our explanation which manifests it, but the system of jointly self-validating attitudes. Of course, the circular pragmatic system rests on various layers of non-normative matter. These layers are economic subjects with property rights and market strategies, the neurological bases of the relevant algorithms, the neurons' chemical make-up, and so on. In an obvious way, the emergence of value relies on the stability of these layers of structured matter. However, it is clear that value cannot be reduced to any of it.

One important feature of the example is that it elucidates the theorist's position within the system. This was the second part of my defence of Brandom's strategy. The point is that the theorist's claim that a Euro note is (now) worth, say, one loaf of bread, occupies the same logical space as other people's valuing of it and is thus best interpreted as an explication of her own valuing. Consider a theorist who comments on a particular marketplace and its paper currency. Imagine that she claims that a note is worth one loaf of bread, while other people consider it worth two loaves. In this situation, people would rush to make deals with our theorist – bringing it about, if the market was small enough, that the note becomes worth something like one loaf, or if it is big, that she gets poor. The theorist has no reason to ward them off with the remark that she is “just a theorist”: if her statement about the value of the note is truthful, and if she is rational, then she must be happy to make the deals.

5. Language and meaning

I want to claim, now, that the phenomenon of meaning can be demystified in quite the same way. My private disposition to use an expression in a particular way – inferentially, substitutionally or non-inferentially – by itself hardly confers meaning on the expression. But if my disposition interacts with the relevant dispositions of other speakers in such a way that we can speak of an ongoing conversation, then it is quite appropriate for me to say that the expression carries a particular meaning settling the correctness or incorrectness of my uses of it. Moreover, my statement about the meaning of the expression itself exhibits a linguistic disposition, thereby affecting the meaning and hence contributing towards its own appropriateness. We thus have another circular pragmatic system of attitudes (this time: linguistic dispositions) and statuses (this time: meanings), and once again, the theorist's place is right within it. On this reading, the social function of meaning talk is the mutual calibration of speakers' linguistic dispositions, just like the social function of money (and value talk) is the mutual calibration of buyers' and sellers' exchanging dispositions.

Consider a group of speakers whose communication runs quite well, but who are also used to utterances which are surprising and slightly off the mark, sometimes indicating real errors. They are all used, that is, to people talking about the “Sahara Dessert” or saying that “dolphins are majestic fish”. They are also used to people telling their doctors that they have “arthritis in their thighs”. The first type of utterance is best interpreted as a malapropism, the second as involving an irrelevant error, and the third as involving a mistake important enough to be pointed out. Of course, the group of speakers is nobody else but us.

In order to find out whether a speaker just speaks weirdly, whether she makes a real but irrelevant mistake, or whether she makes a mistake relevant to the purpose of the conversation – and if so, where the mistake lies –, we have developed a range of techniques. One of them is meaning talk. We ask each other questions of the form “Do you mean A by B?” or short: “Does B mean A?” In our answers, we make statements of the form “I mean A by B” or short: “B means A.”. Imagine that you just ordered water in a restaurant, and the waiter brought tea. On complaining that you ordered water, he replies (with a puzzled face): “Well, this is water.” In this situation, it is natural to ask: “By ‘water’, do you perhaps mean any mixture which has more than 99% of H₂O in it?” – and the reply may well be “Yes, ‘water’ means mostly H₂O”.

What is interesting in this context is the direct analogy to what we have seen in the money example. By replying something like “But ‘water’ means clear H₂O”, we make it the case (albeit defeasibly) that the exhibited use of the relevant terms becomes the appropriate use of the relevant terms – namely by inducing our interlocutors to use them (substitutionally, inferentially, non-inferentially) in the way indicated in our meaning statement, at least when dealing with us. Not only do we induce our waiter to bring us water when we order “water”, we also enable him to translate his own remarks into our ways of speaking. What he formerly meant by “water”, he will, when talking to us, express (perhaps) as “water or tea or juice”.

In order to avoid three obvious complaints, let me enter three equally obvious provisos: Firstly, it is – of course – true that the term “water” is used differently in different settings. Nothing I have said is meant to deny this; it makes sense, however, to speak of different concepts in this context. Secondly, in my scenario, I made common parlance win out against weird parlance. However, I could just as well have had you reply something like: “Oh, if that is what you mean by ‘water’, then I would like to order cold bottled H2O without any added substances.” Within the little language game between you and the waiter, something being “water” would then be compatible with it being “tea”, again showing that the practical attitude exhibited by the meaning statement – this time the waiter's – contributed to its own appropriateness. Thirdly, while in private conversations, people are substantially more free to adjust even to extremely inefficient meaning statements of their interlocutors (perhaps involving terms such as “grue” ), in larger conversations with more than two speakers, there are often good reasons to refrain from giving in too much. Here, it is much more likely that a speaker's individual meaning statement fails in the sense that it is not validated by other speakers. My proposal is not meant to hide this fact; indeed it shows something interesting, namely that the appropriateness (in our judgement) of a meaning statement depends on what is taken to be the relevant language game (that of all English speakers? or just yours and the waiter's?).

6. Conclusion

Of course, much more remains to be said about the emerging picture of normativity. But I do want to claim that the two ideas – the joint self-validation of attitudes, and the attitude-explicating nature of normative talk – give us a way to rescue Brandom's strategy and, in fact, to render our capacity to mean our words in particular ways as unmysterious as our capacity to use money.