In my previous entry, I discussed Hugh Knott’s presentation of Rhees’s ideas and his critique of Wittgenstein. According to Knott, Rhees thought that Wittgenstein had “failed to follow up adequately the connections between the reality of discourse and our reality as persons”, later he adds, in explicating Rhees, “the situation in which people are making remarks and grasping the connections between them has as its background a life in which words have obtained a currency in multifarious contexts and connections…” (p. 242).
The perspective in the latter quotation is one that I can wholeheartedly embrace. Rhees and Knott are emphasizing the importance of looking farther, of raising our eyes from the immediate situation, in the effort to get a clear view of what goes on when someone is speaking. This is important advice. What makes it important is the fact that it offers an antidote to certain well-entrenched presuppositions of Western philosophy.
One such notion is that all that is relevant for an account of what a person is saying is what is the case with him at the moment of speaking – the mental contents accompanying his words, or the stimuli to which he is responding. To overcome the problems of this (“psychologistic”) view of meaning (which ignores the fact that speaking shapes our lives, that it is something we do together), other philosophers have proposed that we should focus instead on the system of symbols being employed. According to them, to speak is to make moves that are defined by the rules of something like a calculus. The problem of this latter perspective is that it abstracts from the question what a speaker is doing in addressing someone – it ignores the way in which what is to be understood as having been said is dependent on the particular context of speaking. (Rhees, by the way, was quite right in pointing out that the language game metaphor may not do enough to discourage this way of thinking about speech.) In the first case, as we might put it, we get life without logic, in the second case logic without life; the problem is to see the logic in the life.
The whole idea of a structure or system, like the idea of a logical connexion, depends on what speaking is. But this is hardly even a statement, because speaking is not one thing, and “having meaning” is not one thing either. This is perhaps the hardest thing to understand about language, and it is for this that [Wittgenstein] refers to the language games. (“Wittgenstein’s Builders”, Discussions of Wittgenstein, p. 75.)
Here I am at one with him.
Knott speaks about “the emphasis [Rhees] gives to the mutually constitutive aspects of the reality of language and our reality as persons” (p. 238, my italics), and Rhees talks about the “reality” or the “possibility” of discourse, or of the “unity of language”. Rhees also writes about our speaking the same language in the different language games. Here I am getting unsure about how to read him. My difficulties have to do with the role of the word “language”. Rhees says that “Wittgenstein did not always distinguish between ‘language’ and ‘speaking’, and sometimes this brings trouble” (“Wittgenstein’s Builders”, p. 71.) I am not sure why Rhees finds this distinction important, or why he needs the concept of language to make his points. Why could he not have made them by talking about speaking? (I think I have an analogous problem with what Knott says about the role of abstraction in learning to speak, p. 240 f. “Entities should not be multiplied beyond necessity.”)
The idea of “a language”, it seems, only comes to have a role when we are confronted with different languages, or different dialects – with cases in which the same thing will be said in different ways (where something that sounds “wrong” to us can still be “right” to others). We could imagine someone who never came across other languages, we could even imagine a world in which nothing like “different languages” existed. We may ask what use there would be in these cases for the word “language” (or what need for the concept), yet speakers would understand as well as we do how what was said in one context might bear on, or depend on, what was said in other contexts.
(It seems to me that the word “language” has a more prominent role at the beginning of Rhees’s essay, whereas “speaking” becomes more central from Section 4 onward. I find what he says in the later part much more striking.)
It was because I could not see why the unity of language as such is important that I tried to suggest, in my paper on Rhees, that the important unity was that of the speakers’ lives. I’m not sure whether that’s a plausible suggestion. (By the way, Knott is probably right in arguing that what I say about the “moral dimension” of the unity is beside the point. I did express some uncertainty about this myself in the paper.)
In sum: where Knott speaks about the mutually constitutive aspects of the reality of language and our reality as persons, what I would talk of, more modestly, is the connection between what speaking is and what it is to be a speaker.