You may have missed my comment that the 'natural/conventional' debate is out of date... by 6 decades, since MAKH showed how the phono/grammar relation is not ‘arbitrary’ at the ranks of intonation and rhythm, and often not at syllable rank. The IFG intro briefly acknowledges that old debate but proceeds to ignore it, preferring the terms ‘congruent/incongruent’ for grammar/semantics relations.
Blogger Comments:
[1] As previously explained, this misunderstands Halliday on the conventional (arbitrary) relation between lexicogrammar and phonology. The relation obtains between grammatical forms and their phonological realisations. Clearly, although participants and processes are naturally distinguished grammatically as nominal and verbal groups, nominal and verbal groups are not naturally distinguished phonologically by their realisations in intonation, rhythm and articulation. According to Halliday & Matthiessen (2014: 11), the natural relation of prosodic phonology is to semantics, not grammar.
Importantly, it is the natural relation between semantics and grammatical form that makes it possible to interpret grammatical form in terms of the meanings they realise. If there were a natural relation between grammatical form and phonology, it would be possible to interpret phonology in terms of the grammatical form it realises.
[2] This is misleading, because the IFG introduction (Halliday 1985 & 1994: xvii-xix) does not acknowledge any "debate"; it merely sets out what Halliday means by a natural relation between semantics and grammar.
[3] This is very misleading indeed. Halliday (1985 & 1994: xvii-xix) does not "proceed to ignore" the natural/conventional distinction and does not "prefer" the terms 'congruent/incongruent' for grammar/semantic relations. As can be seen by reading what Halliday actually wrote here, Halliday begins by explaining what he means by a 'natural' relation between semantics and grammar, and then explains how grammatical metaphor exploits this natural relation.
Importantly, the relation between semantics and grammar is realisation, the 'symbol' sub-type of intensive (elaborating) identification, with semantics as Value and grammar as Token. This is the relation between all adjacent strata, whether the relation is natural or conventional, or congruent or incongruent.
