Language, Community and Identity Event

Last Friday, I popped down to London for the Language, Community and Identity event at the British Academy. There was a very good line up of speakers, including Professor Wendy Bennett, Professor Jenny Cheshire, Professor Dennis Preston, and Dr Devyani Sharma. It was a packed house with nary a free seat in the room, not surprising given the academic pedigree of the speakers, all of whom have done fantastic work in the field of language and gender, perceptual dialectology, historical sociolinguistics and so on.

Perhaps the most surprising thing, though, was that it was chaired by Professor Peter Trudgill, someone who hasn’t really popped up on my radar as a ‘language and identity’ research. Indeed, as he acknowledged in his opening address, he has in the past been accused of being a bit of an anti-identity campaigner. He took some time, however, to say that he has never doubted the importance of identity in terms of language variation, but rather that he couldn’t see how identity could be the principal reason why a particular linguistic variable would change over time and that other forces must be in action. In any event, he didn’t spend a great deal of time on this point and instead the event was focused on dealing with three main questions:

1. How do we research identity and community through language?
2. What role do identity and community play in language change?
3. How do different parts of language construct identities?

For each question, each of the four panel members gave a short overview of their thoughts on the topic, before the floor was opened up to the audience and questions were taken. Because of lack of space, I’m not going to go over everything that was said (and I’ve surely forgotten stuff…), but there were some interesting things brought up, including the issue of bringing together the perception/production cycle. Related to this point, Preston made a really good observation along the lines of ‘what’s the point of doing identity if there’s no-one there to hear you talk?’. Researchers in the UK and the US have been making strides in dealing with how production and perception are linked, and recently there’s been work that’s integrated identity within this framework (like Erez Levon’s work I’ve discussed elsewhere on the blog), but it’s a slow process and I think that more work can be done here, especially within work that adopts a more ethnographic approach (mine included).

The other part that was interesting was how different levels of language contributed to identity construction. Most sociolinguistic research focuses on phonetic/phonological variation, but the interface between syntax/phonology/discourse is still a really under-investigated area of work (Katie Drager’s work on discourse marker ‘like’ is a good example of what can be done). I would have liked the panel to have spent a bit more time on this area since it’s an interesting one and it was good to get viewpoints from different people working within syntax, discourse, and phonetics. I think that speakers and listeners are perhaps more attuned to changes in syntax (since it can be so marked) than they are for variation in discourse or phonology, although that raises a whole bunch of other questions on what makes a variant ‘noticeable’. It is frequency of use, recency of use, salience of use (and how does a variant become salient) or some sort of combination of these (thanks to Norma Mendoza-Denton for these points!)?

Beyond the event itself, it was great to catch up with colleagues I hadn’t seen in a while. Working in an English Department, it gets a bit lonely not having other sociolinguists to talk to and I always like seeing how people are getting on. It also cheered me up before I got back on the train to do another batch of marking…

– The Social Linguist

The Early History of Sociolinguistics in Glasgow

The field of sociolinguistics is relatively new (at least in the ‘quantitative sociolinguistic’ format that William Labov developed through his work on Martha’s Vineyard), and Glasgow was one of the first places outside of the United States where Labov’s techniques and methods were applied in a large-scale sociolinguistic investigation (the other was Norwich in an influential study carried out by Peter Trudgill). The man who was key in developing the field of sociolinguistic enquiry in Glasgow was a researcher named Ronald Macaulay, and his book Language, Social Class and Education: A Glasgow Study (1977) was a milestone in sociolinguistic research in Scotland. Given that the tools and techniques of phonological analysis which we take for granted in contemporary sociolinguistic research (stuff like PRAAT, for example) were unavailable to Macaulay, the robustness of his findings are perhaps all the more remarkable, and his work really was a pioneering study in sociolinguistics. The work was based on preliminary research Macaulay had carried out with Gavin Trevelyan for a Social Science Research Council report on language, education and employment in Glasgow, working with teachers, employers and communities across the city. This report formed the basis of Language, Social Class and Education: A Glasgow Study, which also introduced more qualitative analysis of attitudes towards Glaswegian.

In terms of his analysis, Macaulay was interested primarily in how linguistic variation correlated with social class. With Glasgow being such a class-conscious city, the idea that how you spoke depended on which class you were in (marked through things like where you were from and what you did for a job) was an accepted social ideology, but it had never been tested empirically before Macaulay’s work. The study focused on five main phonological variables (Macaulay 1977: 27):

1. The vowel in words like hit, kill, and risk
2. The vowel in words like school, book, full, and fool (yup, Scottish English doesn’t contrast the final two words)
3. The vowel in words like hat, sad, and back 
4. The diphthong in words like now, down, and house
5. The glottal plosive as an alternative to /t/ in words like butter and get

Now, obviously Macaulay’s work came out well before John Wells’ work on lexical sets, so he didn’t use them as a descriptor for the variables he was looking at. It was also before we knew more about the phonological structure of Scottish English (Macaulay 1977: 29), so in his analysis of the TRAP/BATH/PALM set¹, he only included those words which belonged to the TRAP set, rather than the BATH/PALM set. Through an auditory analysis of these variables (using a point-scale format and giving points to variants depending on whether they were ‘more’ or ‘less’ Glaswegian), and by categorising his speakers according to the Registrar-General’s classification of occupations (Macaulay 1977: 18), he was able to show how different social classes used different forms of the variable. For example, for variable 1 (the vowel sounds in hill), the highest rated variant was [ʌ] (scored at 5 points) and the lowest rated variant was [ɪ] (scored at 1 point), with a number of intermediary variants. So, the higher the mean index, the more likely speakers were to produce  [ʌ]-like tokens, and the lower the mean index, the more likely speakers were to produce [ɪ]-like tokens. The analysis showed that ‘speakers from the lower social classes were more likely to use a vowel that is more retracted and lowered than speakers from the higher social class groups’ (Macaulay 1977: 31). A relationship between linguistic variation and social class was found for the rest of the variables, ultimately giving a stronger empirical basis to folk-ideologies about language use in Glasgow.

But where Macaulay’s work really shines is in his discussion of the qualitative data, particularly the interviews with the teachers and employers. It is in this section of his book where interviewees make clear their attitudes and ideologies about Glaswegian and Glasgow. It was perhaps the following quote (Macaulay 1977: 94) which helped me pursue my own line of research:

“The accent of the lowest state of Glaswegian is the ugliest one can encounter, but that is partly because it is associated with the unwashed and the violent.”

It’s important to say that this isn’t Macaulay’s opinion on Glaswegian, but rather that of a university lecturer (quell surprise!), and when I read it, it hit home just how deeply-entrenched the negative stereotypes of Glasgow as a criminal, dirty and violent city are and how influential this image can be on people’s perception of Glaswegians.

Macaulay’s work was a massive inspiration to me during my own research, and I was lucky enough to meet him a few times at various conferences over the years, and it was this book in particular which was one catalyst to me pursuing my PhD thesis topic. I’m also hugely chuffed that he’ll be contributing a chapter on the history of sociolinguistic research in Scotland to an edited volume I’m currently working on, so watch this space!

– The Social Linguist

1. Paul Johnston argues in his chapter ‘Regional Variation’ (in Edinburgh History of the Scots Language), the lexical sets which are used for English English don’t work for Scottish English because the distribution is different. For example, while TRAP, BATH and PALM are the lexical sets for American English and English English, these lexical sets all have the same vowel in Scottish English /a/. Johnston’s keyword for this vowel is CAT instead. Obviously, Paul’s work wasn’t around when Macaulay was doing his research, so he had to make do with what was available.