[Xmca-l] Chat Is Not Instruction

[David Kellogg]

Happy new year, everybody: may it be far happier and healthier and
humaner than the old one. We completed the latter here in Sydney, by
blowing up half a million dollars in fireworks, with all the artistic eclat
of the invasion of Iraq. After which I came home and read Peter Jones'
chapter in the Cambridge volume "The Transformation of Learning" (van Oers,
Wardekker, Elbers and van der Veer eds.), entitled "Language in
Cultural-Historical Perspective".

Peter starts out noting that for linguists CHAT isn't a philosophy of
science of which linguistics is one application, but rather a vital and
shared set of concerns based on a larger common project. Humanity too is a
common, and very much unfinished, project; watching the fireworks over
Sydney and remembering how the Chinese invented gunpowder as
courtly entertainment but soon found other uses for it, I cannot help
but agree with Peter that that humanity would be a good idea but is still,
in so very many ways, an untested one.

Peter then says that mainstream linguistic theories "speak to this
unfreedom by speaking this unfreedom", by making speech into a genetic
endowment, a modularized component of the mind/brain, and  and he calls
it), and  "fixed, isolated, and mutually unintelligble shards of human
being and potential". This isn't really how I read Chomsky: it ignores a)
Chomsky's insistence that most language is never spoken becasue it is self
directed, an insight that seems to me completely compatible with Vygotsky,
and b) Chomsky's whole rationale for treating language as a formalism,
which is to demonstrate the infinite potential of a very finite number of
elements.

In contrast, I find Peter's idea that the "instruction" is the unit of
analysis for a cultural historical approach to linguistics to be an
instance of speaking to this unfreedom by speaking this unfreedom. It's not
a new idea: Janet also located the origins of language in people telling
other people what to do. But it's a profoundly depressing one. Fortunately,
it is also one that is quite at odds with very simple facts of language.

First of all, linguistics (and any other science, including psychology)
doesn't require or even allow any "unit of analysis" that will fit any
problem at all: specific analytical problems require specific units of
analysis. Secondly, there is, of course, the problem of defining the object
of study of linguistics (a problem that stumped Sassure and Firth and many
other linguists), but instruction would be a very poor unit of analysis for
this problem, since animals can and do instruct each other, but they have
no human language; and thirdly, direct instructions are a very small part
of language and become increasingly smaller as we become more and more
human).

May we all require fewer and fewer direct instances of direct instruction
in the new year!

David Kellogg
Macquaire University