Christian M.I.M. Matthiessen
The Hong Kong Polytechnic University
Like many other linguistic phenomena, or indeed more generally semiotic ones, register — or more helpfully, register variation — has proved to be “slippery”. This is partly because semiotic phenomena are inherently hard to pin down: they exist (or unfold) as semiotic phenomena, of course, but at the same time are also (enacted) as social phenomena, (embodied) as biological phenomena, and (ultimately manifested) as physical phenomena. And within their own order of phenomena, while they are located stratally, they derive their value from their stratal neighbourhood, and (crucially) they are extended somewhere along the cline of instantiation. The slippery nature of register is reflected in the history of the term in Systemic Functional Linguistics. Taken from Reid (1956) by Halliday and his colleagues in the 1960s (e.g. Halliday, McIntosh & Strevens 1964; Gregory 1967), the term register meant registerial variation — functional variation in language according to context of use; it reflected the nature of language as an adaptive system. However, in J.R. Martin’s work, the term slipped stratally from language into context, and it came to stand for the contextual variables implicated in functional variation, i.e. field, tenor and mode. Martin has documented this terminological slippage very clearly and carefully (e.g. 1992). So terminologically, we now have two distinct (but related) uses of the term register— its original and still current sense of functional variation, and its later sense as (roughly) situation type. But the phenomenon of functional variation — register variation — is still recognized, regardless of the terminology (cf. Martin 2010). In what follows I use register in its original sense of functional variation.
The slippery nature of the phenomenon of register variation makes me think of one of M.A.K. Halliday’s technical terms, viz. “semantic slime”. He had in mind, in the first instance, the semantic slime that accompanies terms as they slide from everyday use to technical-scientific use; but perhaps we need to recognize such slime also when terms such as register slither from one theoretical area to another. (And of course, “register” is also used with other senses in linguistics, as in phonetics.)
Fortunately, SFL comes with a theoretical principle, and method, that enables us to deal with slippery phenomena like register. This is Halliday’s trinocular vision (spelled out by him and others in many places, e.g. Halliday 1978, 1979, 1996; Halliday & Matthiessen 2004). The principle is simple but very powerful: since the systemic functional theory of the “architecture” of language is relational in nature (rather than modular), and is based on intersecting semiotic dimensions like the hierarchy of stratification (cf. Matthiessen 2007), the cline of instantiation and the spectrum of metafunctions, we can shunt along these dimensions (cf. Halliday 1961, on shunting — borrowed from the railways) and adopt different observer points, viewing phenomena trinocularly. Halliday (e.g. 1978) worked this out for the hierarchy of stratification: any phenomenon can be viewed “from above” (from a higher stratum), “from below” (from a lower stratum) and “from roundabout” (from its own stratum, its own primary location). Register variation is semantic variation in the first instance, so its primary location is the stratum of semantics. Consequently, when we view it “from above”, we look at it from the point of view of context, when we view it “from below”, we look at it from the point of view of lexicogrammar (and by further steps, phonology, and then phonetics, or graphology, and then graphetics).
Now, I think that Halliday’s trinocular vision can be applied to all semiotic dimensions, not only to the hierarchy of stratification, where it was first applied. For instance, locally within a stratum, we can use it to move up and down the rank scale as we adopt different views on some particular phenomenon. While I think this is by now a well-known extension of the use of trinocular vision, I believe it is still helpful to view register trinocularly in terms of all the relevant semiotic dimensions, and this is what I propose to focus on in my talk:
· global semiotic dimensions:
o the hierarchy of stratification: register viewed from above — contextual variables and values; register viewed from below — lexicogrammatical realizations (and lower-stratal ones as well); register viewed from roundabout — registers as “meanings at risk”;
o the cline of instantiation: register viewed from above — from the point of view of the overall meaning potential: registers as subpotentials [with potentially distinct probabilities of instantiation]; register viewed from below — from the point of view of instances of this potential, i.e. texts as flow of meaning: registers as particular patterns (in context of situation), possibly emergent as new adaptations detectable at first as text types [with new relative frequencies]; register viewed from roundabout — the point of view of the mid-region of the cline of instantiation, between potential and instance: registers as systems of semantic strategies adapted to institutional settings (as in Halliday 1973, and Patten 1988).
o the spectrum of metafunction: all metafunctions (and their contextual correlates) are equally involved in the characterization of register, but we still benefit from viewing registers horizontally as it were — ideationally (logically, experientially), interpersonally and textually.
· local semiotic dimensions:
o the hierarchy of rank: any given register will extend across all of the relevant ranks of the semantic rank scale, but we still need to shunt along this hierarchy, viewing texts as instantiating particular registers (and so the registers themselves) both from the highest rank and the lowest rank, making sure that they meet in the middle. (The question whether there is one general semantic rank scale, comparable to the general rank scales in lexicogrammar and phonology is a very interesting one, but a tough one to answer since it depends on extremely extensive analysis of texts from a vast number of different registers, and nobody has even come close in any framework. I suspect it will turn out that the semantic rank scale needs to be differentiated both metafunctionally and registerially. And it is also important to note that rank is as it were experientially biased; the other metafunctional modes have complementary models for dealing with “composition” — including the logical one of complexing, which is what Bill Mann, Sandy Thompson and I focussed on when the developer RST [Rhetorical Structure Theory].)
o the hierarchy of axis: this hierarchy has only two “values”; but any given register can be viewed from above in terms of systemic organization [paradigmatic axis] and from below in terms of structural organization [syntagmatic axis].
If we re-view register trinocularly along the lines that I have suggested, have we covered everything there that needs to be said about register? The short answer is obviously no.
On the one hand, phenomenologically, we also need to consider the ordered typology of systems operating in different phenomenal domains into account, viewing register not only semiotically, but also social, biologically and physically. For example, we need to take account of the role of register in the complex relationship between social hierarchies and the division of labour, noting the way that register variation and dialect variation intersect. (Many of the semiotic upheavals that we witness today can be related to the physical technology of the Internet — i.e. in the first instance (but not only!) to the rapid changes in the channel of the mode variable of context. The ramifications are extensive, just as when the printing press was introduced as another new channel technology.)
On the other hand, to address and take account of all the insightful observations that have been made about register (and also the potentially misguided ones), we need to go meta — we need to find or create a framework of observations that transcends SFL since a great deal of very valuable work on register has been done and is being done outside SFL, as will be easy to see once the new pioneering journal Register Studies has been launched in 2019).
On the third hand (semiotically, we are not at all limited to our two biological hands — and even this may change Frankenstenially within the biological order of systems if Yuval Noah Harari is on the right track with his vision of Homo Deus, his “history” of tomorrow — which I would call a forecast), there are quite a few phenomena that have yet to be interpreted consensually — like ideology (see Lukin 2019) and individuation (discussed by various contributors to Bednarek & Martin 2010). Ideology and individuation are two of the issues in the exploration of language in context that I pointed to in a talk at ALSFAL XIV hosted by BUAP in Puebla (8-12 October, 2018): “Issues: ideology, individuation, institutions; intervention; impact; implementation”. Time permitting, I will try to touch on these issues since I think they are all central to the theme of this Halliday-Hasan Forum, “Register: New Questions and Possibilities”. One central area of impact is education, and fortunately Kazuhiro Teruya will deal with this area in his talk.
This talk is part of my attempt over the years to contribute to our understanding and investigation of register — e.g. Matthiessen (1993, 2015). The focus on register in this third Forum is very timely, also in view of the launch next year of the new journal Register Studies that I mentioned above. The first issue will include interviews with scholars dealing with register in different traditions and from different point of departure, and my contribution represents SFL; a later issue will include an interview of me about register, conducted by Wang Bo and Helen Ma.