Language games

Suppose we simplify our picture of language so that we can fit 'saying things' into a game analysis grid - with 'statements' or 'utterances' as strategies and a dimension for each player.  We could count each utterance opportunity as a sub-game.

We would need some way of ranking outcomes for players, in order to analyse the game.

If we wanted to retain our conception of the game as a language game, the 'moves' would have to translatable into the language we use to describe the game (the one we're using now) as intelligible utterances - as moves in 'the language game' we are now playing.  There could be more than one way of doing this, and perhaps no clear way to select between translation schemas.

Describing the rules of the game would not be a possible move in the game.  (Although describing the rules of some sub-games would be).

Given a translation schema, we could attach a value to 'truth-telling' - i.e. maintaining general intelligibility and individual credibility.  Perhaps.

But our selection of a schema would depend on assuming that players already attached a value to intelligibility?

(Davidson and decision theory?)