Tuesday 18 February 2020

The "Remarkably Honest" Diagram Of Matthiessen and Bateman

Fawcett (2010: 131-2):
It is highly illuminating to compare a recent description of the version of Halliday's Nigel grammmar [sic] that was implemented in the computer with the original account of a generative systemic functional grammar in Halliday (1969/81). … Consider the diagram shown in Figure 9, which is taken directly from Matthiessen & Bateman (1991:109). 
Here Matthiessen and Bateman provide a visual representation of the output from their generator, following as far as they can the IFG layout. It looks, at first sight, very like Figure 7 in Section 7.2, and so also like one of the many representations of clauses given in IFG.
It is also a remarkably honest diagram, in the sense that it has not been adapted to make it more like the standard IFG analysis than the output from the computer really is (with one vital proviso, which we shall come to shortly). Yet the fact is, as we shall now see, that it is unlike the IFG diagrams in one vital respect. So let us compare Figure 9 with the IFG-style analysis of a clause in Figure 7.
Two differences that are not important for our present purposes are (1) that Figure 9 includes the 'tag' isn't it, and (2) that it contains no line for the INFORMATION structure shown in Figure 7 as "Given" and "New". (The reason for this omission is probably that their version of the Penman generator does not generate intonation.)

Blogger Comments:

[1] Here Fawcett begins a comparison of a very early statement of SFL Theory (Halliday 1969) with a later application of SFL Theory to the programming of computers to generate texts (Matthiessen & Bateman 1991). As a form of argumentation, it is an instance of the red herring logical fallacy, a fallacy of relevance, because the adaptation of a linguistic theory to the domain of computer capabilities is irrelevant to the validity of a linguistic theory as a model of human language.

[2] Here Fawcett introduces the possibility that Matthiessen & Bateman are otherwise dishonest, since their diagram (Figure 9) is not only "honest", but "remarkably" so.

[3] To be clear, one reason why Figure 9 does not include a 'line for the INFORMATION structure' is that INFORMATION is not a system of the clause. The reason why Figure 7 does include a 'line for the INFORMATION structure' is that Figure 7 was devised by Fawcett, and Fawcett does not realise that INFORMATION is not a system of the clause.

No comments:

Post a Comment