Relations in a chemistry text

[John Bateman]

Here's an attempt....

I think that what makes texts of this kind (register) quite
complex for a straight RST analysis is that they
employ significant and irreducible amounts of grammatical
metaphor (Halliday, 1994, _Introduction to Functional
Grammar_, Chapter 10).. The work that this does
in science writing is absolutely crucial to the function
of scientific texts as they have developed over the
past few hundred years. See:, e.g., Halliday/Martin,
1993, _Writing science: literacy and discursive
power_, Falmer Press. This also needs to bring in
Martin's (1992) _English Text_, Benjamins views
of conjunctive relations because much of what
would be in a simple text a rhetorical relationship
is expressed within clauses as processes rather
than across clauses as discourse connectives.

This is the cause of some of the complexities in the
following...

 > (d) The data used to construct Fig 5.7. gives a straight line when we
plot the pressure against 1/volume (Fig 5.8).

 > (e) The linear plot implies that, for a fixed amount of gas at constant
temperature, pressure is inversely proportional to volume, V:

 > (f) this relation is known as Boyle's law.

The evidence relationship between the linear plot and the
inverse proportionality is actually buried in the "implies that", so
there is an evidence (linear plot, pressure inversely proportional),
or maybe an interpretation relation is better. Given the
text's explanation and its position of power, there is
not really much of a motivation for evidence I think... I.e.,:

interpretation:
  nuc: linear plot
  sat: inverse proportion

that is, the linear plot is related to a frame of ideas drawn from
pressures and volumes and given an interpretation in that
light.

But we are down to the interesting issue of whether we interpret
the text or what the text is about. Of course, the fact that there
is a linear relationship is evidence for making certain conclusions
in the process of scientific discovery... but is this also how the
text is being structured? How about:

d: circumstance: (nucleus: d1,   satellite: d2)
e: interpretation: (nucleus:  e1, satellite: e2)

So we have (d:nucleus) "data giving a straight line" and
(e: nucleus)  "there is a linear plot". Which means
that there is basically a restatement, since a
straight line is a linear plot, or perhaps even
a summary, although we have the "N must be
more than one unit" constraint. But here again
I think I see evidence that we have different
levels of analysis to tease apart: the picking
up and repackaging of  "give a straight line
on a  graph" as "a linear plot" is absolutely
typical of the packing (condenstation) operations discussed in
Halliday/Martin above and is crucial to the
development of scientific text. This textual
cohesive reference should probably not be
RST-ized at all (although if we insisted on doing
so then it might well be a summary). So if we
absorb this into the flow of the discourse (i.e.,
treat the text as if it were saying instead of
"the linear plot implies that" rather "which implies that"),
then the former satellite of the interpretation
within the summary becomes
instead a direct interpretation satellite to (d).

And then comes the naming... This is I presume of
e2?, pressure inversely proporitional to volume, and
I suppose has to be an elaboration, just telling
us a bit more about the relation, i.e., that it is
called something.

Summing up:

d: circumstance: (nucleus: d1,   satellite: d2)
e: interpretation: (nucleus: d, satellite: e2-f)
e2-f: elaboration: (nucleus: e2, satellite: f)

Which makes the text very much about one of interpreting
data, graphing data and looking at the results, and
seeing if they make sense. Incidentally it is about
Boyle's law.

And this clearly supports the text as an example of the chemistry
text book teaching students to take this approach to
data rather than a chemistry book for experts or
a history of chemistry book.

As for (a) and (b), this is a projection (Halliday, 1994) and is
not an RST relationship at all. (a) just sources (b). There may
be texts which single out the projecting element rather than
the projected element as "nucleus", but this probably isn't
one of them.


Food for thought?
John B.