Re< mock linguistic play

Geoffrey Williams (geoffrey.williams who-is-at english.su.edu.au)
12 Dec 1995 15:55:05 U

Gordon,
Thanks for your further comments on the Bristol study in relation to my own
work. I agree that the time sampling of talk during book-reading is an
important factor. What this study attempted to do, in conscious contrast, was
to gather data for intact occasions of reading and talk, defined theoretically
in terms of context of situation following Halliday, Hasan and others. A
particularly important resource was Hasan's description of contextual
variables in 'The conception of context in text', recently published in Fries
and Gregory, Discourse in society (Ablex).

If one were to be able talk about possible differences in semiotic mediation
implicated in the ontogenesis of literacies in the two sets of social
locations, it seemed important to be able to relate a primary unit of
linguistic analysis (text) to some theorization of the features of the
occasions of language use (context). From a specific sense of text understood
as a semantic unit, it was possible to move to more delicate semantic units,
specifically messages and their realization in clauses, and then to discuss
variation in terms of frequently occurring patterns of specific features of
linguistic interaction. It was also possible to discuss subtle variations in
the situation type in different social locations realized by the linguistic
paterning. A highly specific example of this was the agentive roles taken up
by mothers and children. (Contrary to a common stereotype, in this data it
turned out that children in the two social locations did not differ in the
extent to which they initiated interaction during j b-r.) The agentive roles
of the mothers was a particularly complex issue.

(Hasan has recently discussed the problem of a unit of analysis in semiotic
mediation, and possibilities for using the systemic linguistic sense of text
for describing variable forms of semiotic mediation in an article in Language
Sciences 14(4), p489-528, contra suggestions for the use of Bakhtin's speech
genre.)

Another difference between the studies was in the conceptualization of
locations in social class relations. In this study they were described in
terms of Bernstein's analysis of sc relations, particularly boundary strengths
between labour categories and their effects, communicative practices often
associated with labour categories in relation to different fields (economic,
symbolic control and cultural fields). I think there are significant
implications for interpreting both the prior structuring conditions for, and
effects of, semantic variation.

You mention frequency (extent) of reading as an important factor in the
Bristol study. Interestingly in the Sydney data, the participants from the
"lower working class" and "middle class" groups both read extensively. Using
the linguistic unit of message as a measure of extent, rather than time, I
found that the number of messages read in object texts was not significantly
different in the two social locations. (In fact the lwc group read more
extensively in raw terms.)

I will resist giving a brief summary of other findings since I don't think
they make much sense outside the theoretical paradigms. (You'll recall that
the original comparative comment was made in a private post to Angel, in
relation to her observations about relative academic literacy success amongst
Hong Kong students from different sc backgrounds.) I find it particularly
difficult to write about differences in 'quality' of talk about text.

A last, quick comment in this over-long post. It is perhaps not surprising
that the Maintown/Gateway mothers seemed to act like teachers since my
understanding is that many (?most) of them were in graduate classes for
further professional training as teachers. But it also seems that recent
pedagogical discussion of j b-r may be projected back into family discourses,
regulating home interaction around text even where there is no direct link
with the teaching profession. Obviously, many handbooks describing
'desirable' family reading practices and K school class practices are based on
observations about interaction during j b-r in "middle-class" families, often
I think the families of "middle-class" academics. What kind of 'natural'
partnership is envisaged here? I was stunned to read the comment in one such
handbook that 'For Marie-Ellen storybook reading is almost an extension of
breast-feeding her little girl'.

Geoff Williams