In this chapter, we’ll look at sociolinguistics. This is where linguistics meets sociology. So the topics here have to do with language in society. The classic areas of study in sociolinguistics have to do with how language varies from region to region and from social group to social group. Another very important topic, especially in recent years, is language and gender: how men and women speak differently.
Here I’m going to look at just three areas:
Accents of English
The social stratification of speech
Negation in non-standard English
The first topic is pretty self-explanatory. Whole books can be and have been written on phonetic and phonological variation in English, and I’ll just provide a very sketchy overview, picking up on a couple of points I mentioned in Chapters 1 and 2. Under the second topic, I’ll introduce the foundational work in sociolinguistics by William Labov, concentrating on his study of social variation in the English of New York City, and Peter Trudgill’s later study of Norwich English. Finally, I’ll briefly discuss non-standard forms of negation in English and explain why these are not, as is sometimes claimed, in any sense ‘illogical’.
*
Everyone is aware of how accents vary from place to place. Most people who have lived in England for a few years can spot a ‘Northern’ accent, for example. And of course ‘Northern’ covers a multitude of sins: Liverpool and Manchester, two great northern cities separated by 35 miles and one of the bitterest football rivalries on the planet, have dramatically distinct forms of local English, and no inhabitant of either city would thank you for taking them to be from the other one. Then there are Leeds, Newcastle and Sheffield, which are just the main urban varieties. Yorkshire alone boasts seemingly innumerable local varieties, and the speech of Northumberland north of Newcastle is remarkable in a number of ways.
And that’s just the North. We’re all more or less familiar with the West Country ‘burr’, the Welsh ‘lilt’, the Black Country … er, whatever, and of course good old Cockney. Then there’s Scotland, then there’s Ireland (and it’s a tin ear that can’t distinguish Glaswegian from the Edinburgh accent, or an Ulsterman from a Dubliner).
It’s not just Britain, of course. Americans can easily spot a New Englander (go on YouTube and find a speech by the late President John F. Kennedy if you want to hear an authentic Boston accent), a New Yorker (more on these later) and of course the Southern drawl. In the American West, there’s less variation (more space, fewer people and English hasn’t been there all that long), but a fascinating Southern Californian variety has emerged in Valspeak (check out ‘Valley Girl’ by Frank and Moon Unit Zappa). Then there’s Canada, and then there’s the Southern Hemisphere, mate.
But this is all just anecdote: we want patterns. Coming back to the basic North-South distinction in England, there are two main differences in the vowel system which are common to all the ‘Northern’ as opposed to all the ‘Southern’ varieties. First, Northern accents don’t have the /ʌ/ vowel. In these varieties, put and putt, book and buck and look and luck are all pronounced with the same vowel /ʊ/.
The second difference has to do with the vowel, mostly spelled <a>, in words like grass, dance, laugh and pass. In the South, this is /ɑ:/, the same vowel as in father and car. In the North, the vowel is short /æ/, although usually pronounced somewhat lower than /æ/and more like an IPA [a]. So the vowels in grass and gas are the same in these varieties.
Systematic studies of regional variation make it possible to plot these trends on a map. An isogloss is a line we draw on a map to separate the two variants. Figure 8.1 shows the isogloss for the /ʊ/ vs /ʌ/ vowel of luck, etc.
Figure 8.1 Isogloss for the /ʊ/ vs /ʌ/ variants in sun, Etc.
The isogloss for /ɑ:/ vs /a/ in grass runs close to the one for the sun vowel, hence the widespread impression that these variants characterise ‘Northern’ vs ‘Southern’ English accents.
More generally, we can divide all of the accents of English the world over into three main groups. One group of accents is basically Scots, extending across to Ulster in the northern part of the island of Ireland. These accents have very different long vowels and diphthongs to the rest of the English-speaking world, and in fact make the ‘long’ vs ‘short’ distinction in a very different way from a phonetic perspective.
In the second group, we have most of the accents of England, Wales (in large areas of which English is a fairly new arrival), the Southern Hemisphere, New England, New York and many varieties of the American South. What these varieties all have in common is that they fail to pronounce the <r> in words like car. In these varieties, <r> is pronounced only at the beginning of a syllable, as in red or Fred. This is of course familiar in Standard British English, but for example Bostonians are often lampooned in the US for saying ‘pahk the cah in Hahvahd Yahd’, where the spelling tries to indicate the r-less pronunciation here; similarly, the Big Apple is sometimes referred to as ‘New Yawk’.
The third set of accents pronounces these <r>s. So car is pronounced /kar/, roughly. This is the case in ‘Standard’ American English (if there is such a thing), basically all American varieties except for New England, New York and parts of the South and all of Ireland south of Ulster, as well as in Canadian English. It’s also true of the varieties of the West Country and Norfolk in England, as well as parts of Lancashire (Blackburn and Burnley, for example) and a small part of the East Riding of Yorkshire.
Of course, accents aren’t only regional: we’re all aware of what ‘talking posh’ means, and are often inadvertently snobbish ourselves about ‘local’ varieties. In Britain, Received Pronunciation (RP) refers to a particular variety of English which lacks regional connotations, this is roughly the ‘conservative’ variety of Southern British English I described in Chapter 1. The funny thing about RP is that hardly anyone speaks it anymore. It very roughly corresponds to what people call ‘The Queen’s English’, ‘BBC English’ or ‘Oxford English’. In fact, Her Majesty speaks a particularly conservative form of RP (characterised by her pronunciation of words like off with /ɔ:/ rather than /ɒ/), normal enough for an aristocrat born in 1926, but hardly representative of contemporary English. And you can listen to the BBC for a long while without hearing real RP these days. As for Oxford, I couldn’t possibly comment, but you don’t hear much RP in Cambridge. If you want to hear a good example of RP, listen to Julie Andrews in The Sound of Music.
Most middle-class people in Britain nowadays speak a mildly regionalised approximation to an innovative form of RP, but the original is all but dead. Working-class people tend to use more regionally diverse forms of English. There’s an approximate correlation between social class and the degree of regional differentiation.
This is where the serious sociolinguistics comes in. Thanks to the work of Labov and his followers, we now know that the situation is more complicated and interesting. To see this, I’ll first try to summarise the main elements of Labov’s study of New York City speech, before returning to the UK to look at Trudgill’s study of Norwich English.
Labov recognised that certain variant pronunciations have social value: they are recognised as markers of the speech of a particular social class. From a purely linguistic perspective, the ascription of social value to a given pronunciation is completely arbitrary. The <r> business described above illustrates this well. In Britain, the r-less (or non-rhotic) pronunciation is the prestige form, with the r-ful ones (the rhotic ones) being considered hokey, regional, working class, low prestige. In the US, it’s the other way round: in New York City in particular, the nonrhotic variety is low prestige, and the rhotic one high prestige (Boston is a bit more complicated, as the example of JFK suggests).
Labov studied the incidence of rhoticity as a sociolinguistic variable, marking a correlation between accent and social class in New York. The conception of the study was brilliantly simple. Labov chose three Manhattan department stores, Saks Fifth Avenue, Macy’s and S. Klein. These stores are known for being very upmarket, middle class and somewhat downmarket, respectively. On entering each store, Labov checked what was on sale on the fourth floor, and then he asked a shop assistant where he could find whatever it was, in this way eliciting the words ‘fourth floor’. So then he could check whether the assistant said /fɔrθ flɔr/ or /fɔ:θ flɔ:/ (or, in New York, /fɔəθ flɔə/). He also pretended not to hear the assistants’ first answers, getting them to repeat the words more carefully. This allowed him to distinguish spontaneous and more careful speech.
The results were impressive. The incidence of rhoticity (i.e. rhotic or nonrhotic pronunciation) closely reflected the social cachet of the department stores. The assistants at Saks had the highest incidence of the rhotic pronunciation (/fɔrθ flɔr/), those at S. Klein the least. Another interesting result was that the Macy’s assistants showed the biggest difference between spontaneous and careful speech, indicating an awareness of the social value of the rhotic pronunciation.
The department-store results showed awareness of overt prestige, the desire to sound ‘classy’ and align oneself with the speech of the dominant group in society. Labov also discovered the phenomenon he called hypercorrection, which is the overuse of a prestige form. This, he observed, was commonest at Macy’s, the ‘middle’ store. In fact, Labov went a step further and pointed out that it was middle-class women who showed the highest incidence of hypercorrection and therefore the greatest sensitivity to overt prestige.
Still another innovation was the concept of covert prestige. This arises when people want to distinguish themselves from the ‘dominant group’, and so use seemingly low-prestige forms in order to show a distinct social or regional identity. The high incidence of nonrhotic forms at S. Klein (the downmarket department store) shows covert prestige. Middle- and working-class males are most likely to use covert-prestige forms, Labov observed. So Labov was the first person to demonstrate a clear correlation between social class, gender and accent, indicating the social stratification of speech.
One very important result which emerges from the kind of sociolinguistic work that Labov’s New York City study initiated is that sound changes can be observed going on, in real time. Indeed, by comparing the incidence of rhoticity across different age groups in New York (in 1962), Labov was able to show that rhoticity was not a social marker for the older generation at the time but was for the younger generations. This is an example of a linguistic variable taking on social value; Labov demonstrated how speakers may be conscious of such social value.
A similar study was carried out in Britain by Peter Trudgill in the early 1970s. He looked at sociolinguistic variation and social stratification in the speech of the city of Norwich, in East Anglia. One of the variables he looked at is actually found in non-standard varieties of English all over England (and elsewhere in the English-speaking world). This is the pronunciation of the ending -ing in words like walking, talking and so on, as what’s usually written -in’ (IPA /ɪn/ as opposed to /ɪŋ/. Here the apostrophe indicates the ‘missing’ <g>, but in fact the two sounds are different: an alveolar nasal in the non-standard varieties but a velar nasal in the more standard varieties). He also looked at the presence or absence of the 3Sg present -s ending (as in She go to the shop or She goes to the shop). The -in forms and those without the -s ending are low-prestige forms, of course.
Trudgill’s main findings were similar to what Labov found in his New York study: class was more of a determiner of non-standard usage than gender, though women in all social classes were more likely to use the overt prestige forms, i.e. the standard ones, especially in careful speech. Men, on the other hand, showed evidence of covert prestige, in that they over-reported their non-standard usage, while women over-reported their standard usage. One of Trudgill’s general conclusions was that women are more susceptible to overt prestige than men, while men are more susceptible to covert prestige. Like Labov, Trudgill found that the differences between men’s and women’s usage of the standard forms were greatest in the lower middle class and the upper working class, and in careful speech. So the same forces of social stratification seem to be at work with different formal variables in both New York and Norwich. Since the 1970s, the same results have been replicated in many places and in many languages.
In the previous chapter, we took the notion of change for granted, and effectively treated things like the Germanic Consonant Shifts as though they were instantaneous changes. Of course, this was justified, since they took place thousands of years ago and we know nothing in detail about the circumstances under which they happened (all we really know is that they must have happened). But sociolinguists have shown us how changes actually take place in real time and the role of sociological factors in inhibiting or facilitating them. Today’s variation is tomorrow’s change, looked at in retrospect.
Sociolinguistic variation doesn’t only affect phonetics and phonology. Morphological variation exists, and certain forms are regarded, again arbitrarily, as high or low prestige. For example, in Standard English, the -s ending on verbs only shows up in the 3sg of the present tense; however, there are many varieties where this form has a different distribution. In many non-standard Midlands varieties, the -s ending appears on all persons in the present tense, and so we find forms like I thinks, they sings, etc. Some varieties in the West Country and East Anglia don’t have this ending at all, so we have he think, she sing, etc, as we saw above in connection with Trudgill’s Norwich study. Finally, in parts of the North of England, as well as Scotland and Ulster, there’s the Northern Subject Rule, which requires a singular verb form with a non-pronoun subject NP, singular or plural and the standard form of agreement (i.e. no -s in the plural) with an adjacent pronoun subject. This rule extends to the forms of the verb to be, so it’s not just about the -s ending but about the system of subject-verb agreement in general, as examples like (1) show:
Them eggs is cracked, so they are.
All of these variants are highly regional and low prestige, but every single one of them is a coherent linguistic system, every bit as much as Standard English.
It’s sometimes said that working-class language or non-standard forms are somehow intrinsically inferior to the standard language, on the basis of some notion of ‘logic’ or ‘clarity’ or ‘precision’. From a linguistic point of view, there’s absolutely no reason to think this whatsoever. Non-standard forms of English are just that: not the standard. It’s best not to use them when taking tea at Buckingham Palace, or in a job interview or (more debatably) on the BBC, but there’s nothing intrinsically ‘wrong’ with them. To believe that non-standard varieties are inherently inferior to the standard is to dress up social snobbery with pseudo-linguistics.
A good example of this comes from how negation works in non-standard English (in many parts of the English-speaking world, including both the US and the UK and elsewhere). In non-standard English, sentences like (2) express various kinds of negation:
a. I ain’t done nothing.
b. I can’t stand it no more.
c. I ain’t never been there.
d. She don’t like no-one.
Before saying any more about the examples in (2), let’s take a quick detour into how languages express negation. The key point is that, while in propositional logic, as we saw in Chapter 5, two negatives make a positive (i.e. ¬¬p ↔ p), in natural languages two negatives very often just make an emphatic negative. Or even a non-emphatic one. French is a very good example of this. In French, if you want to negate a sentence, you have to put in two negative words, one before a finite verb and one after it:
(3)
Je ne bois pas. (‘I don’t drink’)
Of course, we can’t possibly say that each of ne and pas are equivalent to logical ¬; if we did, then (3) would be a positive sentence (and Standard French would seemingly have no way of expressing simple negation).
French has other negative words, as in (4):
a. Je n’ai rien fait. (‘I haven’t done anything’)
b. Je ne le supporte plus. (‘I can’t stand it anymore’)
c. Je n’ai jamais été là. (‘I have never been there’)
d. Elle n’aime personne. (‘She doesn’t like anyone’)
On their own, rien means ‘nothing’, plus means ‘no more’, jamais means ‘never’ and personne means ‘no-one’. So all the sentences in (4) have two markers of negation too, but, as their English translations show, each expresses only a single negation.
There’s nothing ‘illogical’ here (how could the language of Descartes, Voltaire and Rousseau be illogical?). Many languages express negation in the way French does, and we call this negative concord. All it amounts to is that, when you want to express a negative, you negate several bits of the sentence at once. Propositional logic (a ‘language’ invented by logicians and philosophers) and some languages, like Standard English, don’t do this: instead each negative expression ‘counts’, and so two negatives do indeed make a positive, in these languages.
Now, purists often claim that the varieties of English illustrated in (2), which we can now say differ from Standard English in showing negative concord, are ‘illogical’ or – worse – that the people who say such things are ‘illogical’ or incapable of thinking clearly. This is simply not true. In fact, you might have noticed that the sentences in (2) exactly parallel the French sentences in (4). Is French therefore ‘illogical’? Were Descartes, Voltaire and Rousseau incapable of thinking clearly? Of course, the notion is utterly absurd, and shows a very imperfect understanding of language and languages – in this case, the widespread phenomenon of negative concord – as well as a naïve, and quite unjustified, willingness to assimilate natural languages to propositional logic.
But it can be worse: social prejudice masquerading as pseudo-linguistics can actually disguise racial prejudice. William Labov, once again, has made very significant contributions to revealing these prejudices. In particular, he pioneered the study of what is usually known as African American Vernacular English (AAVE), sometimes called ‘Ebonics’. This variety of English, spoken by much of the African American population of the US, also has negative concord. In a classic article from 1972, Labov made a number of striking observations concerning negative concord, and other aspects of the syntax of negation, in AAVE. Labov observed sentences like those in (5):
a. Well, wasn’t much I couldn’t do.
b. Down here nobody don’t know about no club.
In (5a) negative concord goes ‘across clauses’: here wasn’t is in the main clause and couldn’t is in a separate (relative) clause, but they are interpreted as one (so the sentence means ‘There wasn’t much I could do’). In (5b) the subject participates in negative concord (the examples in (2) don’t have this). These two kinds of negative concord can combine to give spectacular examples of the following type:
a. It ain’t no cat can’t get in no coop.
b. When it rained, nobody don’t know it didn’t.
Each of these examples actually contains a single logical negation (‘No cat can get into a coop’; ‘… nobody knew it did’), the further expressions of formal negation just being the negative-concord system of this variety. Many a purist would probably have a hard time figuring out what these sentences mean, but kids in the urban centers of the US would probably not have any problems.
The examples above are not emphatic. AAVE has ways of emphasising negation, as in:
a. She ain’t in no seventh grade.
b. But not my physical structure can’t walk through that wall.
c. Ain’t nobody on the block go to school.
d. Don’t so many people do it.
Again, spectacular stuff, and hardly the product of unclear, illogical or imprecise thought processes. All of this clearly shows that AAVE is a distinct grammatical system from Standard English (and from the many non-standard white varieties), and that it may have undergone, or be undergoing, certain changes in how negative concord works which do not affect other varieties. And of course this distinct variety has social value since it is associated with a particular ethnic group. So here Labov showed a case of social stratification, correlated with ethnicity, involving a syntactic feature.
*
As I said at the beginning of this chapter, there’s much more to sociolinguistics than what we have seen here. But nonetheless we’ve looked at a couple of the major areas: accents and dialects, evidence for the social stratification of speech and language change in real time and evidence that non-standard forms are entirely coherent linguistic systems, every bit as ‘logical’ as standard varieties.
As with all of the other chapters, this has been an attempt to give a taste of what is involved. It should be clear, at least, that sociolinguistics can do much to counter unjustified social and racial prejudice when this is manifested through pseudo-linguistics.
