What variationist sociolinguistics is, and whether it has anything to do with writing

According to Gordon’s (2013) book on Labov, there are three tenets of the “Labovian paradigm,” aka “a variationist approach” to language:

1. Variation is inherent to linguistic structure.
2. A socially realistic linguistics offers valuable insights to the study of language.
3. Quantitative methods can reveal patterns where casual observation sees only chaos.

This is useful to me for a couple of reasons. First of all, when I started the project I am working on (both the dissertation project and the larger ‘project’ of conceptualizing deviation from SWE in writing), I had absorbed a little sociolinguistic theory by osmosis and wanted to call what I was doing “a variationist approach” to L2 writing. While I am very interested in variation, it isn't appropriate to call what I'm doing “variationist” because a) I am not a linguist, so strictly speaking I am not studying linguistic structure, even though I might be interested in contributing methodological tools for people whose interest is in studying features of ‘new’ varieties of English and b) I am not using quantitative methods to look for patterns in usage based on social variables.

As I have read Coulmas and Lillis, I have been wondering where variation fits into their work. I wonder the same about Milroy and Milroy. “Variation” to most sociolinguists has to do with differences in pronunciation across social contexts, and also sometimes words, grammar, etc.

I'm increasingly convinced that Horner et al’s ‘translingual approach’ to writing, which collapses all variation/deviation from SWE into the category of ‘language difference,’ is extremely fruitful for studying readers’ reactions. Making everything fair game for acceptability is a really good way to go, since ‘acceptability’ is actually in large part what underwrites social and institutional concepts of good writing, standard English, and even to some extent the very notion of “a language.”

Yet I do think Labov’s theory, or variationist sociolinguistics,  has something to offer a program of research that is (as mine is) primarily interested in writing. Gordon says that for Labov, “the fundamental question driving sociolinguistics” is “understanding why anyone says anything.” I appreciate how open-ended this leaves things in terms of methodology and what to focus on, even if Labov ultimately went a very specific direction. Coulmas, too, calls sociolinguistics “the study of language choice”.

Gordon points out that the “third wave” of sociolinguistic variationist research focuses more on “what language does” and “its social and cultural functioning” as well as “how social meaning is constructed from linguistic resources.” The focus of this ‘wave’ tends to be, according to Gordon, on what was called “style” by early sociolinguists -- that is, “intra-speaker variation.”

I need to back up a little here. If we're looking still at how sociolinguistic theory/approaches can influence the study of writing, the  introduction of style looks scary at first, because it sounds like something that is already included in ‘register,’ or ‘genre' -- pretty well-covered territory in traditional studies of writing. But remember, ‘style’ here is referring to individual choices that individual language users make.

Even though ultimately my own study is more in the realm of attitude (or really, ‘ideology’), I think we finally arrive at a very robust and worthwhile place here. The sociolinguistics of writing, whatever else it comprises, is going to be a place where we look carefully at the choices writers make, and their reasons for making those choices (whether they be related to traditional sociolinguistic variables like class, race, gender, etc, or not), in terms of social and cultural factors and meanings. It will be about "understanding why anyone writes anything." And of course, for my purposes, the ‘uptake’ of lexicogrammatical features of texts by readers (i.e. teachers, editors, etc) is a prime site for examining these sociocultural factors and meanings.