David Rose continued his third reply to Erich Steiner on sysfling on 16 May 2018 at 11:47 as follows:
The discourse semantic model attempts to resolve this contradiction, by treating the discourse/grammar relation as stratal abstraction, rather than instantiation. That is, the discourse semantic stratum consists of feature/structure systems that are instantiated in the discourse patterns of texts. These discourse semantic patterns are realised inter-stratally as grammatical patterns, that instantiate grammatical systems.
Blogger Comments:
[1] See the previous post for why the said contradiction only arises from Rose's misunderstandings of instantiation and cohesion. This clause thus begins the logical fallacy known as a straw man argument:
A straw man argument is one that misrepresents a position in order to make it appear weaker than it actually is, refutes this misrepresentation of the position, and then concludes that the real position has been refuted. This, of course, is a fallacy, because the position that has been claimed to be refuted is different to that which has actually been refuted; the real target of the argument is untouched by it.
[2] To be clear, treating discourse and grammar as two strata of symbolic abstraction is reinterpreting the word 'discourse' to mean 'semantics'. A complicating factor here is that the stratum of discourse semantics (Martin 1992) was not theorised with an understanding of strata as different levels of symbolic abstraction. Instead, Martin misinterpreted all strata as interacting modules of meaning. Martin (1992: 390, 392, 488):
Each of the presentations of linguistic text forming resources considered above adopted a modular perspective. As far as English Text is concerned this has two main dimensions: stratification and, within strata, metafunction. …
The modularity imposed by stratification is also an important issue. Discourse systems generate structures which in principle cut across grammatical and phonological ones. …
In this chapter a brief sketch of some of the ways in which discourse semantics interacts with lexicogrammar and phonology has been presented. The problem addressed is a fundamental concern of modular models of semiosis — namely, once modules are distinguished, how do they interface? What is the nature of the conversation among components?
[3] There is a sense in which the relation between grammar and discourse is one of instantiation, though not the sense understood by Rose. The relation between grammatical systems and the grammar of a text is instantiation, and for Halliday (2008: 78), text and discourse are two views on the same phenomenon:
[5] A most serious shortcoming of the discourse* semantic model (Martin 1992), in this regard, is that its system networks do not specify how choices in discourse semantic systems are realised as choices in grammatical systems. Moreover, attempts to do so would expose deficiencies in the theorising. Two examples will illustrate this fact. On the one hand, the logical discourse semantic system of CONJUNCTION does not recognise the three most general types of expansion: elaboration, extension and enhancement; because of this, most, if not all, realisation relations between discourse semantics an grammar are incongruent, whether metaphorical or not. On the other hand, the logical discourse semantic system of CONJUNCTION does not include the major logical semantic system of projection, and so cannot account for projection relations anywhere at the level grammar.
* Martin's interpersonal discourse semantics is a general exception, because it is largely a rebranding of genuinely semantic systems and structures, devised by others, not Martin.
I do make a difference between these two; but it is a difference in point of view, between different angles of vision on the phenomena, not in the phenomena themselves. So we can use either to define the other:
[4] A most serious shortcoming of the discourse semantic model (Martin 1992), in this regard, is that its system networks do not specify how choices in discourse semantic systems are realised as discourse semantic structures. Moreover, attempts to do so would expose deficiencies in the theorising of structure. For example, on the one hand, discourse* semantic units are not structures, since they have no internal organisation; and on the other hand, the relations between Martin's discourse* semantic units are not structural, since they are rebrandings of Halliday's (non-structural) cohesive relations. Moreover, the type of structure Martin interprets them to be, co-variate, was later acknowledged by its source, Lemke (1988: 159), as not being a type of structure after all. For more detailed critiques of discourse semantic structure, see here.
- 'discourse' is text that is being viewed in its sociocultural context, while
- 'text' is discourse that is being viewed as a process of language.
[5] A most serious shortcoming of the discourse* semantic model (Martin 1992), in this regard, is that its system networks do not specify how choices in discourse semantic systems are realised as choices in grammatical systems. Moreover, attempts to do so would expose deficiencies in the theorising. Two examples will illustrate this fact. On the one hand, the logical discourse semantic system of CONJUNCTION does not recognise the three most general types of expansion: elaboration, extension and enhancement; because of this, most, if not all, realisation relations between discourse semantics an grammar are incongruent, whether metaphorical or not. On the other hand, the logical discourse semantic system of CONJUNCTION does not include the major logical semantic system of projection, and so cannot account for projection relations anywhere at the level grammar.
* Martin's interpersonal discourse semantics is a general exception, because it is largely a rebranding of genuinely semantic systems and structures, devised by others, not Martin.
