Bob Hodge On Understanding What Grammar Is

Bob Hodge wrote to sys-func on 8 Jul 2024, at 17:50:

Thanks for this great intervention, which raises so many important issues about the architectures of SFL, as proposed by Halliday and can be hypothesised now.

Two points:

1. I don't think Halliday had entirely settled views on grammar in his social phase. 'Grammaticality' seems to have a multimodal scope, yet 'grammar' often is restricted to verbal language. I'm not sure how much that difference should be insisted on.
2. Whatever it is, it seems to be something that comes in late in the evolutionary process. A kind of emergent structure. If it is something which comes in earlier, then it should be different in some key respects from what we understand as grammar drawing on adult language.

Whatever we come to think of it, I suggest that all linguists, SFL included, should be less sure that we understand what grammar is

Blogger Comments:

[1] To be clear, Halliday was quite definite on grammar, in the SFL sense, being restricted to language, a decade after his retirement (Halliday 1996; Halliday & Matthiessen 1999). Moreover, it is the only interpretation that is consistent with his theory, given his definition of grammar.

[2] Cf. Halliday (2002[1996]: 388):

We could locate grammatico-semantic systems within the framework of an evolutionary typology of systems, as in Figure 1. In this frame, semiotic systems appear as systems of a fourth order of complexity, in that they are at once physical and biological and social and semiotic. Within semiotic systems, those with a grammar in them are more complex than those without.

[3] To be clear, SFL Theory is formulated in the tradition of the 'immanent' orientation to meaning. This means that its epistemological stance is that meaning does not transcend semiotic systems — a position that is supported by the experimental findings of Quantum Physics. This means that the meaning of 'grammar' — "what grammar is" — does not transcend semiotic systems. This means that "what grammar is" is a construal of a semiotic system, such as a linguistic theory. This means that each linguistic theory determines for itself "what grammar is". This, in turn, depends on the assumptions on which the theory is based. The validity of a theory and its assumptions are a separate matter.

It is the theory which decides what we can observe.

— Albert Einstein

Ed McDonald On Halliday's Grammar-Centrism

Edward McDonald continued sys-func on 8 Jul 2024, at 14:44:

Ok, so this is not "linguistic imperialism" but it does seem like "grammar-centrism". It strikes me that, though I'm not aware if anyone has ever made this argument, that both (adult) language and music could equally be seen as "outgrowths" of protolanguage, particularly if we look at spoken language and singing. Given that music lacks the line of arbitrariness, i.e. that musical meanings are in the first instance indexical, not symbolic, the argument for the priority of grammar seems less convincing in this instance.

Any multimodalists like to weigh in here about other semiotic systems? Do they all depend on the prior existence of grammar? It would seem to me that systems depending on iconicity don't necessarily.

Blogger Comments:

[1] To be clear, Halliday's view that language is the only semiotic system with a grammar is merely the view that is consistent with his Systemic Functional Theory and the assumptions on which is based.

[2] To be clear, song lyrics are language, as are musical notation and theory. The question of whether music itself requires its makers to have language, depends on how music is defined. If music is narrowly defined as restricted to humans, then music is defined as requiring its makers to have language. However, if music is more broadly defined to include the structured use of rhythm and pitch variation in other species, most notably birdsong and whalesong, then music is not defined as requiring its makers to have language. This still leaves open the question of whether music is social (value) or semiotic (symbolic value), in terms of Halliday's linear taxonomy of complex systems.

[3] Given that no evidence is given for musical meanings, or a natural indexical relation between such meanings and their expression, this is an instance of the bare assertion (ipse dixit) fallacy.

[4] Lest this be misunderstood, the question is whether the users of non-linguistic semiotic systems have a grammar. Halliday (2002[1996]: 389): 

On the other hand, human adults also develop numerous non-linguistic semiotic systems: forms of ritual, art forms, and the like; these have no grammar of their own, but they are parasitic on natural language – their meaning potential derives from the fact that those who use them already have a grammar.

[5] To be clear, one way to assess whether a semiotic system requires its users to have a grammar is to look for such a system in species without a grammar. To date, no other species have been demonstrated represent their experience iconically by drawing pictures, and the earliest evidence of human cave painting long postdates the estimated emergence of language.

Ed McDonald On Halliday's Linguistic Imperialism

Edward McDonald wrote to sys-func on 8 Jul 2024, at 08:49:

We've often had this debate before, and presumably if music has no "grammar" it has no "semantics" either. I would agree, in the sense that those two terms are traditionally applied to language and so come with all sorts of associations and presuppositions. … 

But as a semiotic system - by definition (IMFFHO) - must have the two mutually defining strata of what in general terms I like to call interpretation and expression: specifically for music, what I call "phonotactics" and "e/motion" - see my 2021 paper in Language, Context, Text, then it presumably has a "grammar" of sorts, in the sense of distinctive patterns with significant interpretations. In a separate step, In relation to music, I also depart from orthodoxy in characterising the expression stratum of phonotactics as purely textural (= textual), with the other metafunctions coming in only at the stratum of e/motion, whether interactional (= interpersonal) or figurative (= ideational) (I haven't found the need to distinguish between experiential and logical). …

Halliday "privileging" language in this way for me invokes the spectre of what, following Horst Ruthrof, I call "linguistic imperialism"; and the fact that Social Semiotics (again IMFFHO) is shot through with linguistic imperialism for me explains why Ruqaiya for one didn't "believe in" multimodality, in the sense of a model of same that wasn't simply parasitic on the theory of language.

Blogger Comments:

[1] To be clear, here McDonald is commenting on Halliday's view (2002[1996]: 388-9) that language is the only semiotic system that has a grammar, in his sense of the term.

[2] This misunderstands Halliday. On Halliday's model, if music is a social semiotic system, then it has content and expression, but its content is not stratified into semantics and lexicogrammar. For Cléirigh, on the other hand, in terms of Halliday's linear taxonomy of complex systems, music is social, not semiotic. That is, it carries value, but not symbolic value. In terms of Edelman's neuroscientific theory, music selects perceptual categorisations of positive or negative value, and such value systems play an important rôle in the experience of emotions.

[3] Again, this misunderstands Halliday. For Halliday, the "two mutually defining strata" are the content and expression planes, following Hjelmslev. This is a prerequisite of semiotic systems, but it does not entail that content is stratified to include a grammar, in Halliday's sense, as 'an entirely abstract semiotic construct that emerges between the content and the expression levels of the original, sign-based primary semiotic system' (ibid.).

[4] To be clear, the problem with reconstruing 'content' as 'interpretation' is that the latter is the perspective of the listener only.

[5] This misunderstands the metafunctions. The metafunctions are types of meanings, and so are a dimension of the content plane only. For example, phonological systems are not metafunctional; they realise metafunctional meanings of the content plane.

[6] As demonstrated above, McDonald's accusation of "linguistic imperialism" is entirely unwarranted, since derives only from his own misunderstandings of Halliday's theorising.