Ruk asked:

How can we apply the concept of 'language game' to a problem of translating texts from one
language to another?

============

The concept of language as being a "game" originated, as far as I know, with the later Wittgenstein,
after he repudiated the Tractatusand wrote the Blue and Brown books.One of the problems I have
with answering the above is that I cannot tell whether you have read these or about these, so I'll have
to assume you haven't. But that entails my giving you a short course on Wittgenstein, which I simply
don't have space for. So I'll give a very brief and superficial summary.

Here goes: language is a "game" in the senses a) that it (i.e., its rules and its members) is arbitrary,
b) that the "goal" of language (perhaps to explicate truth — or describe the world — and you play with
whether those are the same) is arbitrary, in the sense that we only know it through language, and c)
that what supports language, what holds it together, so to speak, is no more than our mutual
agreement to maintain it. But that agreement is tacit, i.e., we aren't usually aware of it, and no one is
aware of all of it, so what it amounts to is that we are and must be constantly interacting with each
other and with various linguistic constructs, like books, dictionaries, etc., to maintain language's
stability, so to speak. So language is a "game" that we all play; we agree to the rules, use them to
describe what we think and see, and all that we can know of the latter is through language... so
there's a kind of inherent circularity to the whole system. Wittgenstein sort of broke that circularity by
claiming that we have knowledge otherthan language, but that latter is basically ineffable so it's just
not part of the system we can grasp hold of, manipulate, communicate, etc. A rather black and white
analysis, I've always thought, but there you are.

So, to your question. When we're translating, we're translating the rules of one game into the rules of
another. Can this be done? Sure, as long as we've got commonalities to use for rule translation. One
of the strange things about this conception, to me, at any rate, is that since there is nothing, really,
except these games, there's not too much problem dealing within one system as between systems.
It's all arbitrary rules, and we just have to get them down through observation and agreement. That's
one very general meaning of your term "apply". If you want to know whether someone has worked out
translation rules on Wittgensteinian principles... an interesting question. I would very much doubt that
anyone has done this explicitly and in detail. However, think about computer translations. Machines
are very good at treating things like "games", using rules, etc. and I think that a comparison of the
principles of machine translation to Wittgensteinian ideas is quite apt. In fact, Daniel Dennett is quite
a proponent of a kind of general approach to Wittgensteinian/ machine synthesis; you might read
Consciousness Explainedfor more ideas along these lines.

Steven Ravett Brown