For me the problem has been where to allow in a model the possibility of
the meanings that haven't yet been meant, which must be somehow in the
potential available in the present, as a result of the meanings immanent
from the past. So in configuring a model of context, I set field, mode
and tenor choices as simultaneous, and potentially freely combining - so
at the level of register the model wildly overgenerates - allowing all
sorts of texts the culture hasn't got round to taking up, or perhaps ought
not to because they'd be too disturbing. So when I look up on the bus one
day and see alongside the ads at eye level a story about a sexual
encounter during which someone catches AIDS, and realise that something
has changed, there's a new register configuration there - that level of
detail in the field of sexuality acting as that kind of interpersonal
warning (command not to) in that public advertising mode as part of a
public campaign... then the level of register has in a sense made way for
this possibility in its wild overgenerations... but its up til now absence
has also been configured at another level through genre theory, which
wouldn't have predicted a narrative combining that field with that tenor
with that mode. Of course as soon as it manifests, then the genre level
is reconfigured, and this kind of narrative in this space becomes
something I expect, not something novel, surprising...
So the issue is, how to represent the new in relation to the old...
Re the relation between activity theory and linguistics, I've sometimes
wondered whether the best theory we have of activity is the one we've
evolved through language, as opposed to ones we might design as academics.
So that common sense everyday language (without grammatical metaphor in
technical terms) is a good theory of activity. Would that be a meeting
ground between activity theory and functional linguistics, or something
that would get in the way.
Jim Martin