The frequency at which a particular variant/linguistic variable occurs may change over time, but the conditioning factors are stable.
In which case, for language change to occur at that locus of variability, something needs to happen to disrupt the transmission of those conditioning factors to the next generation.
I’m trying to unpack the diachronic consequences of long term, stable multilingual communities with balanced bilinguals making up the population. For where I am in Southern New Guinea, I feel I should read up of this, about the social networks of hunter-gatherer societies… Because there must be some effect about interaction frequency and social networks in explaining language change?
A few months ago a few of us Wellspringers and CoEDL students at the Australian National University were having lunch, and talking about the topic of my last blog post: the overlaps and differences between sociolinguistics and variationism. I was somewhat glad to see that quite a few of us had conflated the two, or were still unsure about how the two are different (this was particularly the case if you didn’t deal with variationism in your research directly). I tried to throw in my two-cents about why we have this conflation of the two things, and mumbled something about Labov. So I think it’s time to tick off that “Write About Labov in relation to Sociolinguistics and Variationism” box on my to-do list.
I struggled a little bit on how to start this section, because introductory textbooks to sociolinguistics all talk about, in-depth, the famous Martha’s Vineyard study by William Labov (1). (If you want a really quick and dirty overview, checkout here.) As for the man himself, Wikipedia already has a pretty good overview, so it’ll be a bit redundant for me to write out more again… Having said that, there’s no way around extrapolating on the Martha’s Vineyard study if we’re to understand how sociolinguistics and variationism get conflated. So even if you know the Martha’s Vineyard story, bear with me, I’ll try and make it worth your while.
Once upon a time, there was a concept called “free variation”. I think this term is still in use, still with the original meaning of “speakers freely alternate between various forms” (you definitely see it used in descriptive grammars). Along comes Bill Labov and shows, empirically, that there are sometimes social motivations for why speakers alternate between forms. Why do some speakers on Martha’s Vineyard pronounce their vowels one way, and others another way? That’s because those speakers who are more oriented towards main land American culture pronounce their vowels more main-landy, while those who are more island oriented tend to pronounce their vowels like the vernacular dialect.
Drawing on social motivations to explain variation was a break through at this time in modern linguistic history. But another thing that was revolutionary about Labov’s study was that it used quantitative and qualitative data collection and analysis.
These quantitative methods that Labov started, and were further developed by linguists such as David Sankoff, Gregory Guy, Shana Poplack etc etc did NOT necessarily look at the social dimensions of language. In fact, the Labovian method was used to study variation in a numeric and systematic way. So to simplify it:
Assumption of variation + Quantitative method part = variationism
Assumption of variation + social explanations + social dimensions of language = sociolinguistics
As I talked about in this post, sociolinguistics is a pretty broad term with a range of interests that cross-over with anthropological linguistics and linguistic anthropology. But “Labovian sociolinguistics” (which is what I meant in said post when I said “sociolinguistics”) has that emphasis on quantitative methods, which sets it apart from the other socio-culturally oriented linguistic things.
People say the same thing in different ways.
The theory was that there was free variation if the grammar couldn’t give an explanation for the variation.
Labov said that was rubbish, and that social and cultural factors can explain variation.
Labov used a method that sought patterns and correlations between who and when these variations occurred.
The explanation of why variation occurred, was a social reason.
So this conflation of “sociolinguistics” and “variationism” makes sense historically, so to speak. It’s kind of an interesting case of how meaning drift can happen, if you aren’t paying attention to the particulars of how things develop, I guess.
Of courser I would lie, and my first meaningful post of 2016 is neither about PRAAT or micro variation, but about the definition/scope of sociolinguistics (again) and variationism. But I guess if I’m going to talk about microvariation, it’s kind of necessary to clear this ground about what is generally meant by variation and variationism in linguistics, and the close relationship it has with sociolinguistics (read: Labov. If you don’t know what Labov is, he’s a godfather of linguistics, and we’ll do a proper post about him later. I guess it’s kind of inevitable).
So what is the relationship between sociolinguistics and variationism? From the title you can tell that they are not the same thing. But their origins come from (essentially) the same place, so they can be confounded.
When I started off in my PhD two years ago (argh!) I began by reading a book by Sali Tagliamonte titled Variationist Sociolinguistics: Change, Observation, Interpretation (2012). Since my topic is about studying “variation” within a population of speakers, this made sense. What I mean by variation is something like “how do people talk about the same thing in different ways”. A famous example often used in text books is the alternation between word final [n] or [ŋ] in English speakers: for example pronouncing a word laughin’ for the first case, laughing for the second case. (2) Is there a reason why some people, sometimes, pronounce words with the [n], not the “proper” [ŋ]? Studying this kind of alternation in speech is what “studying variation” means.
So why do some people say the same thing differently? The big break through in the twentieth century for the science of language was that these differences are in part due to where the speaker is from, their background, identity and other ‘social things’ (what in the field is called “socially conditioned variation”). The [ŋ] vs [n] is a well-known dialect difference in England, which is also sensitive to speech style (e.g. polite and careful vs. rapid and casual) (see Trudgill 1974).
I loved reading Tagliamonte’s textbook. I had taken a subject in my undergraduate level degree called Introduction to Sociolinguistics or something like that, so I was vaguely familiar with the things Tagliamonte was talking about.. but not in this kind of depth. But it took me a while after reading this text book’s methods section, and reading papers such as the one provided by a member of my supervisory panel, that variationism and sociolinguistics are not the same thing. I had, somehow, conflated variationism as a theory and method with frameworks and approaches inherent in sociolinguistics. There is plenty of overlap, but actually, this is another one of those things where the origins and history of sociolinguistics as a field have (probably) helped conflate the two things.
Let’s articulate this here:
Variationism: Research into what patterns and correlations can be found in regard to how people vary their speech.
Sociolinguistics: Research into the social meaning of language use, with an emphasis on seeking patterns and correlations from a largish data sample.
Tagliamonte’s book is titled Variationist Sociolinguistics. That is, she is talking about sociolinguistic research that uses the variationist framework and method. But variation doesn’t have to be socially conditioned. In fact, it’s very often conditioned by the grammar of a language. Going back to that [ŋ] vs [n] difference from before, it turns out that there are grammatical reasons why people vary their pronunciation between [ŋ] vs [n]. According to Houston (1985) nouns are much more likely to be pronounced with the [ŋ] rather than [n], while verbs are more likely to be pronounced with [n].
This conflation was a real n00b mistake, I’m aware. Somehow I had thought that variationism and sociolinguistics were the same thing because, you know, Labov.
I think my next post is going to have to be about William Labov, and how much his work has shaped the field of sociolinguistics, both theoretically, but also methodologically. There are plenty of things out there that talk about his theoretical contributions to linguistics, but I find that very few have explicitly highlighted his methodological contribution (probably because it’s so self evident). This is the second post I’ve had to shove aside the history of sociolinguistics and Labov, so I guess we’ll cross that bridge next post.
(1) Walker and Meyerhoff (2014) have recently writtenabout how it may be that the variable of “individual” is very often the most important factor in explaining the pattern of variation found within speech. That is, a person’s personal language history and their individual quirks are able to best explain patterns of variation found in their speech, rather than things like class, ethnicity, etc., especially when looking at small language’s communities.