For the past fifty years, sociolinguistic studies of linguistic change have focused mainly on phonological variables, but recently some attention has been paid to other features, particularly discourse features used by younger speakers that may change within a relatively brief period. Such features have included expletive like (Romaine & Lange, 1991), quotatives (Tagliamonte & D'Arcy, 2004), and intensifiers (Stenström, 2000). This article deals with the appearance of an unusual intensifier in the speech of adolescents in Glasgow, Scotland, as illustrated in (1).
Recordings of Glasgow adolescents in the past seven years have shown that the use of pure is well established in contexts such as (1) and used frequently by a range of speakers. This article will examine the significance of this development in possibly illuminating a process of linguistic change, although it is too early to tell what the lasting effects will be.
DATA AND METHODS
In 2003 Jane Stuart-Smith recorded 36 Glasgow working-class adolescents in same-sex dyads talking to each other for approximately half an hour, with no investigator present, with equal numbers of girls and boys 10–11 years old, 12–13 years old, and 14–15 years old. This was a follow-up to an earlier project in which eight middle-class and eight working-class Glasgow adolescents aged 13–14 were recorded under similar conditions in 1997, as part of an investigation into language change (Stuart-Smith, 1999). In 2004 the two younger groups were recorded again under similar circumstances.1
Claire Timmins, who was in charge of recording the adolescents for all three samples, commented on the 2004 recordings: “My opinion of these recordings is that the children are bored with the task a second time round” (Personal communication, 9/29/2005). This may explain the lower frequency of all intensifiers in these recordings.
Total words used by Glasgow adolescents by year and gender (in 1,000s).
The older adolescents produce more speech than the younger ones and in each age group the girls speak more than the boys, but even the youngest boys produce nearly 11,000 words. Neither the age nor gender differences in the total number of words are statistically significant.2
The statistical measure used on all frequency tests is the Mann-Whitney nonparametric test.
There are few signs that the adolescents were constrained by the fact that the recordings were made on the school premises and would be heard by adults. The frequency with which some of the adolescents used taboo expressions suggests that they did not feel inhibited by the recording situation.3
There are 205 examples of the word fuck, 57 of shit(e), 35 of cunt (usually with reference to a male), 15 of bastard, and 11 of arse. There are also 41 references to farting.
One result of this lively interaction is that the speakers often used emphatic forms, of which the most frequent is pure. There are 1,054 examples of the word in the complete corpus, a frequency of 6.7 per 1,000 words. This is vastly more frequent than is recorded anywhere else.4
In the British National Corpus pure does not reach the minimum frequency of 0.16 per 1,000 words to be included in the rank frequency list for spoken English (Leech, Rayson, & Wilson, 2001:144).
The earliest citations in the OED for pure both as an adjective and as an adverb date from 1297, but the adverbial use is said to be ‘slang’ or ‘colloq.’ by the 18th century. It is given as ‘dial.’ (especially in the U.S.) and the latest citations are from William Faulkner in 1932 and 1942.
Figure 2 gives the frequency of use of pure by year and gender. It can be seen that pure is used much more frequently by girls than by boys and this is statistically significant at the .05 level in 1997 and at the .01 level in 2003. The gender difference just fails to meet significance in 2004 but this is the result of the extremely high use (22.7 per 1,000 words) of pure by one 14-year-old boy. This is the highest frequency of any individual in the whole set. Without this outlier, the gender difference would again be significant at the .05 level. Figure 2 also shows a substantial increase in the use of pure from 1997 to 2003 and then a falling off in 2004. Although there are noticeable differences in the frequency for the three age groups in 2003, the use of pure by the 10–11-year-olds correlates with that of the 12–13-year-olds at the .05 level (Pearson = .635), and the use by the latter group correlates with that of the 14–15-year-olds at the .01 level (Pearson = .886).
Use of pure by year and gender (per 1,000 words).
The use of pure by these adolescents must be seen in the context of a phenomenon first reported in Macaulay (1991) and later elaborated in Macaulay (1995, 2002, 2005), namely, the extremely low frequency of derived adverbs in -ly and also the intensifier very in recordings of Scottish working-class speakers. It has to be emphasized that this absence of inflected adverbs is not caused by the use of uninflected adjectives as adverbs, as in it's real funny. There is only this one example of an uninflected adverb in the whole adolescent corpus other than the two that will be discussed later. There are only 410 inflected adverbs in the corpus, a frequency of 2.6 per 1,000 words. This contrasts with the frequency of 11.9 for inflected adverbs used by middle-class adults in Glasgow (Macaulay, 2005:113).6
In the list of the 1,000 most frequent words in the Bergen Corpus of London Teenage Language (COLT), there are 15 derived adverbs in -ly with a frequency of 7.4 per 1,000 words. Since there will be other derived adverbs, the overall frequency will be higher than this.
In the British National Corpus very occurs with a frequency of 2.4 per 1,000 words in the rank frequency list for spoken English (Leech et al., 2001:144) and with a frequency of 1.5 per 1,000 words in the conversational English sample (Leech et al., 2001:222).
Ito and Tagliamonte (2003:266) list 11 intensifiers that occur with adjectival heads in their York corpus, the most frequent being very (38%), really (30%), and so (10%). In contrast to their findings in York, really is not used frequently with adjectival heads by the Glasgow adolescents. There are only 11 examples of really in this function, a frequency of 0.07 per 1,000 words.8
In the COLT corpus really occurs with a frequency of 3.6 per 1,000 words (calculated from the list of the 1,000 most common words in the COLT corpus).
The proportion of the most frequent intensifiers used with adjectival heads is shown in Figure 3. Figure 3 shows that there has been a dramatic reduction in the use of dead since 1997, while at the same time there has been a gradual increase in the use of so. This latter development is similar to the favored use of so that Tagliamonte and Roberts (2005) found in the American television program Friends over a period of eight years. However, it is the high frequency of pure as an amplifier or booster that stands out in the Glasgow materials.9
There are also 44 examples of quite, a frequency of 0.28 per 1,000 words. This compares with a frequency of 2.49 per 1,000 words used by the middle-class adolescents in 1997. The working-class adolescents use quite generally as a positive amplifier rather than as a hedge.
Main boosters used by Glasgow adolescents by year.
In the 1997 recordings pure is used predominantly with negative affect, as in the examples in (3a–c).
Only 12% of the adjectival examples with pure in 1997 have a positive aspect as in (3d); the others are negative. In contrast, 74% of the examples with dead have positive or neutral affect. The situation has changed by 2003, since 42% of the examples of pure with adjectives have positive or neutral affect, although dead continues to be used more (70%) with positive or neutral effect. What has happened is that the use of pure has been extended to more positive contexts, as in (4).
In the 2004 sample, the use of pure with adjectives of positive or neutral aspect is 28%, while the proportion of such adjectives with dead has gone down to 53%. Thus over a period of seven years, what we find is that the use of both pure and dead has broadened to more general use with adjectives. There are over a hundred different adjectives used with pure. Most of them are evaluative in Dixon's (1982) categories. Only 5% represent color and 6% a physical property. Hunston and Thompson (2000:5) give the following definition: “Evaluation is the broad cover term for the expression of the speaker's or writer's attitude or stance towards, viewpoint on, or feelings about the entities or propositions that he or she is talking about.” It is clear that the Glasgow adolescents use pure mainly for evaluative purposes, either positive or, more often, negative.
The majority of adjectives (89%) are predicatives (e.g., he's pure lovely) and only 11% are in attributive position (e.g., makes a pure bad noise don't it?). This is consistent with the pattern Ito and Tagliamonte (2003:271–273) found for very and really in the York sample, though the preference for predicative position is much stronger in the Glasgow sample. Ito and Tagliamonte suggested that “use of intensifiers with predicative adjectives could be taken as evidence for a later stage in the delexicalization process” (2003:271). This point will be taken up later. First, it is necessary to look at other uses of pure by the Glasgow adolescents.
Use with adjectival heads accounts for only 34% of the examples of pure in the Glasgow adolescent corpus. In 1997, 22% of the examples of pure were with nouns; in 2003, 16% were with nouns; and in 2004, 21% were with nouns. Examples are shown in (5).
The examples of pure with a countable noun occur after the indefinite article with the exception of examples (5c) and (5d). Examples (5c) and (5d) may be instances of scope expansion (from DP-internal to DP-external),11
I am grateful to Elizabeth Traugott for this suggestion.
Examples (5e) and (5f) suggest an adjectival interpretation of the noun. There are 16 examples of crap and 28 examples of shit(e).12
In Scotland shite and shit are in free variation, though shite is the more common.
This process may also be apparent in a number of examples with compound nouns or noun phrases that convert them, as it were, into adjectives, as shown in (6).
In (6a) the sense of pickle brain is ‘stupid, idiotic’ and thus is a descriptive term rather than a referential one. There are more than 50 nouns that occur with pure. Most of them are evaluative in the way that the adjectives are. This can be interpreted as an increase in subjectification.
Sometimes pure occurs in the sense of very with other premodifiers:
In addition to its occurrence with adjectives, adverbs, and nouns, pure is used frequently with verbs, as can be seen in the examples in (8).
In cases such as the examples in (8), the verbs show degrees of activity, and pure has the effect of indicating a high degree. There is some similarity here to the use of really, but pure does not have the same scope ambiguity as really. For example, in contrast to (8b), I think she really fancies David is ambiguous between the reading ‘I think it is true that she fancies David’ and ‘I think she fancies David to a high degree.’ There is no such ambiguity in (8b), which can only have the latter interpretation.
There are, however, examples of verbs where this interpretation is not possible, as shown in (9).
All of the verbs in (9) are achievement verbs in the sense of Vendler (1967) and cannot be varied to a greater or lesser degree. In these examples, pure has instead the effect of highlighting the event, much in the same way as a pseudo-cleft (e.g., what happens is that they fall off aw the time). In the examples in (9) there is a somewhat negative implication. There are other examples of pure with verbs where the negative sense is even stronger.
Without the context it may be hard to see that these are, in fact, all complaints of one kind or another. The effect of pure is that of an adverse evaluative comment. In (10a) the speaker is complaining about John's behavior and in (10c) the speaker is upset that a boy went out with Marie (I was pure gutted). The effect of pure is to focus attention on the action.
Sometimes pure can be used in quick succession in very different contexts, as in (11).
The occurrence of pure in a range of contexts is thus similar in many ways to the use of just by the adolescents. After pure, just is the most frequent adverb in the corpus, at 6.6 per 1,000 words.13
This is exactly the same frequency as found in the COLT sample (calculated from the list of the 1,000 most frequent words in the COLT corpus).
As demonstrated by Lee (1987), just has a range of meanings, although it is not always easy to distinguish them. In Macaulay (1991:130–132), four uses were identified and all four occur in the adolescent conversations. The first is ‘exactly’ (the drawer just beside me), which is not found with verbs; 7% of the examples, mostly with nouns and adjectives, are of this type. The next is ‘recency’ (this chip's just broken off); 11% of the occurrences of just with verbs are of this kind. The third meaning is ‘only’ (I was just kidding on); 26% of the verbal examples are of this kind. The most common meaning is ‘simply’ (we're just good at everything); 63% of the examples are of this kind. The percentages are only approximate because the meanings overlap. It may be helpful to quote Lee's conclusion at this point:
Although it is possible to identify different categories of meaning for just in different utterances, these categories are linked to each other in intricate ways. Examples can be found in which two (or even more) meanings combine, so that one type of meaning overlays and shades into another. Borderline cases can be identified where it is difficult to decide to which category a particular case should be assigned. (Lee, 1987:395)
Lee's remarks apply equally well to the use of pure by the Glasgow adolescents where the meaning is not always obvious. It may be worth pointing out that Lee is describing the use of just by adults in the standard language. The flexibility that the adolescents display in their use of pure thus has its parallel elsewhere.
The verbal contexts in (8–10) account for a quarter of the occurrences of pure; 13% after an auxiliary and 12% before a simple verb. However, it is not always the case that pure follows the auxiliary. Compare (12c) with (12d).
In (12a) and (12b) the sense suggests that it might have been more appropriate for pure to occur before the modal. Since there are so few examples it is difficult to be sure about the scope of the adverb. In (12d), for example, the use of pure may be parallel to the use of really in standard varieties. In that case, there would be two possible interpretations: (a) ‘you can sing very well’ and (b) ‘it is really the case that you can sing.’
There are also some other contexts that occur less frequently, as in the examples in (13), in which pure occurs directly before an adverbial or prepositional phrase.
Very occasionally pure occurs in quotatives, as in the examples in (14).
One of the quotatives used by the Glasgow adolescents is be like that, as in I was like that “On you go” (Macaulay, 2001:9).
The wide range of contexts in which pure occurs resembles that of really (Stenström, 1987), except that pure rarely occurs in initial or final position. There is only one example of each.
There are too few examples of the kinds illustrated in (13)–(15) to make the situation clear, but they show how the use of pure has been extended from the other contexts in which it occurs more frequently.
DISCUSSION
As has frequently been pointed out, intensifying adverbs are an unstable category (Bolinger, 1972; Ito & Tagliamonte, 2003; Labov, 1984; Parkington, 1993; Stoffel, 1901). Bolinger said that they “afford a picture of fevered invention and competition that would be hard to come by elsewhere, for in their nature they are unsettled” (1972:18). Parkington (1993:181–183) described the process by which a word such as very has lost its original meaning of ‘truly’ or ‘genuinely’ and can collocate with a wide range of modifiers. However, he also pointed out that as intensifiers such as very and utterly develop their new function in addition to losing their earlier semantic qualities they also “seem today to have a more limited syntactic range than in the past” (Parkington, 1993:190). The use of pure by Glasgow adolescents shows a very different pattern since it has developed a much wider range of syntactic contexts.
Haspelmath (2004:18) suggested as a possible universal of language change, the notion of The survival of the frequent: “When a grammatical distinction is given up, it is the more frequent category that survives.” A corollary of this might be that the absence of a frequent form may lead to innovation. In the situation with the Glasgow adolescents, it could be argued that the absence of the most common intensifier very and the low frequency with which boosters in -ly occur left a vacuum in which there were no standard degree words to perform this function. Since adolescents are often assumed to “exaggerate rather than modulate” (Paradis, 2000:147), it is to be expected that they will need some way in which to emphasize items in their speech. Peters (1994:271) pointed out that “it is a well-known fact that among degree adverbs, it is the booster class which has the highest degree of fluctuation.” He also claimed that “boosters frequently function as symbols of group identification” (1994:271). The Glasgow adolescents appear to have chosen a booster that might function effectively as a sign of group identification. This may be an example of what Keller (1994) called the effect of “the invisible hand,” that is, an explanation that depends upon its plausibility and cogency. In the absence of comparative studies it is impossible to be sure that this is an innovation originated by the Glasgow adolescents, but there is no published evidence of a similar use of pure elsewhere. In her study of degree modifiers, Paradis (2000) cited no examples of pure, nor is this usage found in the study of London teenage talk by Stenström et al. (2002). There is also no evidence of pure being used in this way by the Glasgow adults in the 1997 recordings (Macaulay, 2002) or in the later 2004 recordings.15
I did not notice it in my interviews with Glasgow 10-year-olds and 15-year-olds in 1973, but I was not looking for such items, and the interviews were a very different kind of speech event from the later conversations.
Stoffel, in his study of intensives and down-toners, said that most intensives “are adverbs derived from adjectives expressing absolute qualities, i.e. such as do not admit of variation, as, for example, pure, full, very” (1901:1). He pointed out that most of these intensives “have in course of time come to mean merely a high degree of a quality” (1901:1). He observed that “frequent use is apt to weaken the sense of a word” (1901:1) and that “the process is always going on” (1901:2). Stoffel cited examples from Middle English and Early Modern English showing the use of pure as an adverb in the senses of (1) ‘very’, (2) ‘completely, quite’, and (3) ‘merely, exclusively, only’ (1901:14–15). All these senses can be found in the examples of pure used adverbially by the Glasgow adolescents, but there is no evidence that the adolescents' use represents historical continuity from the earlier situation.
One way of looking at the situation is to treat it as evidence of a process of grammaticalization. Grammaticalization is a term that has been used with a variety of meanings (e.g., Bybee, Perkins, & Pagliuca, 1994; Heine, Claudi, & Hünnemeyer, 1991; Hopper & Traugott, 1993, 2003; Ramat & Hopper, 1998; Traugott & Heine 1991). Common to most approaches is the notion that increasing use of linguistic forms can lead to changes in meaning and function, what Heine and Kuteva call decategorialization (2002:379). According to Traugott and Dasher (2002:81), grammaticalization should be “conceived as the change whereby lexical material in highly constrained pragmatic and morphosyntactic contexts is assigned functional category status.” This is what appears to have happened with pure in the speech of the Glasgow working-class adolescents. Hopper and Traugott stated (1993:121) that “[i]n divergence existing forms take on new meaning in certain contexts, while retaining old meanings in other contexts.”16
Haspelmath (2004:33–34) pointed out that older meanings can coexist with new uses.
Himmelmann (2004:31–32) argued that grammaticalization proceeds by a process of context expansion. He described three stages. In the first stage, the class of elements that the item is in construction with “may be expanded.” Secondly, “the larger syntactic context in which the construction at hand is used may change. Thirdly, and most importantly, the semantic and pragmatic contexts in which the construction is used is expanded.”
It is not possible to trace the development of pure in the speech of the Glasgow adolescents in the way that Traugott (1988, 1990) and Nevalainen (1994) have examined the grammaticalization of just as an adverb. One reason is that we have no chronological record. By the time of the 1997 recordings the Glasgow adolescents are already using pure in a wide range of functions.17
This is true only of the girls, but the boys produced very few narratives, and their conversations were less varied than those of the girls (Macaulay, 2005:30). This may account for their very limited use of pure.
Inflected adverbs in -ly and intensifiers such as very and really were not favored in the discourse style of the Glasgow adolescents. One alternative at an earlier period was the form dead, but for some reason it lost ground to pure. Pure was available both as an adjective and in its standard adverbial use with colors, as in your teeth are pure yellow. As dead lost its appeal as a “booster” (Paradis, 2000:150) in construction with adjectives, pure increased in frequency in that function and was extended to indicate positive as well as negative affect. This was an increase in subjectification. In addition, the class of adjectives with which pure could be used was expanded. Haspelmath (1999:586) suggested that “frequent exposure” to an innovation is the mechanism that alters speakers' grammars after the initial acquisition stage, and pure is remarkably frequent in the recordings. This expansion of the use of pure is an example of Traugott's (1990:499) “first tendency” by which the meaning of an item based in the externally described situation is extended to meanings based in internal evaluations.
The next stage extends the syntactic context in which pure could be used. The model for this could have been just, which is equally frequent in the speech of the adolescents. This extended the context for the use of pure to include verbs, where the sense was of “to a high degree.” This meaning was not one that would compete with just in verbal contexts. The contrast can be seen in several examples with the same or similar verbs as shown in (16).
In (16a) the speaker is pointing out that something had happened a moment before, while (16b) is a complaint. (16c) is an example of the exclusionary use of just equivalent to “that is all she does,” while in (16d) pure reinforces the notion of “all the time.” In (16e) just functions as a minimizer in contrast to the use of pure as a booster in (16f).
König (1991:3) described the role of what he calls “focus particles” that “interact with the focused part of the sentence they occur in.” This seems to be the pragmatic function of pure in some examples given by the Glasgow adolescents, as in (17).
In (17a) the sense seems to be “Is it actually the case that she is standing at the window?” The examples (17b) and (17c) are like pseudo-clefts, that is, the meaning is something like “What he always tries to do is give me money” (cf. the examples in (9)). Example (17d) shows pure actually used in a cleft sentence. All these examples illustrate the use of pure as a focus particle in König's sense. In examples such as these pure has achieved a state of grammaticalization (or delexification) whereby the meaning of ‘unadulterated’ has been bleached out and its function is mainly pragmatic.
There are also examples where just and pure occur together, always in that order, as in (18).
The use of just is always exclusionary in this context, while pure can be a focus particle, as in (18a)–(18c) or have the meaning of ‘to a high degree,’ as in (18d). Finally, there are the more limited examples of the kinds in (12)–(15), wherein the use of pure seems to have been extended to a variety of other contexts.
The only support for any claim that pure might have developed in this sequential manner comes from the significant gender difference. In 1997 the boys did not use pure in construction with verbs, whereas the girls did with a very high frequency (32%). In 2003, the proportion of pure used in construction with verbs by the older girls (12–13-year-olds, and 14–15-year-olds) is 24% compared with 14% for the boys at the same age. In 2004 there were fewer examples with verbs (perhaps for the reason given in endnote 2), but they were three times more frequent in the conversations between girls than in the boys' conversations. Since the girls have presumably been the leaders in introducing pure, it would not be surprising that they should lead in the extension of its use to other functions, but the evidence from the 2003 and 2004 recordings shows that the pattern has been adopted by boys. This is true also of the other uses of pure illustrated in (12)–(15). The proportion of these miscellaneous uses by the older girls in 2003 is 15%, compared with only 7.5% for the older boys. There are not enough examples in the 2004 recordings to draw any conclusions. The extremely high use by one 14-year-old boy in 2004 (22.7 per 1,000 words) suggests that by this date pure has become established in all its functions for boys as well as girls in the adolescent community.
CONCLUSION
The use of pure as a complex intensifier seems to have developed in the speech of these Glasgow working-class adolescents without an obvious external model. There may have been outside influences but there is no obvious source. There is a hint in the conversations as to how changes may evolve in a somewhat conscious way. Two 13-year-old boys had this exchange early in their conversation:
There are no other references to haggard anywhere else in the 2003 recordings so it does not appear to be a term that has caught on, and there are no examples in the 2004 recordings. This is not evidence for the origin of pure, but it suggests that the adolescents are aware of innovations that arise within their own speech community. It also suggests that innovations may arise from within a tightly linked community, contrary to the position taken by the Milroys (Milroy & Milroy, 1985).
The increase in the uses of pure in the six years since 1997 combined with other changes suggests that the Glasgow working-class adolescents have developed a set of norms for their speech community that owes little to adult or outside influence.18
The 1997 recordings included adults (aged 40 years and over) and adolescents, with equal numbers of middle-class and working-class speakers at each age. The largest number of significant statistical differences were between adults and adolescents (Macaulay, 2005:158). Unfortunately, there are no recordings of younger adults that might indicate when (or if) the adolescents begin to adopt adult norms.
Keller's invisible hand seems to be getting a lot of exercise.
Given the paucity of information about adolescent speech, it is impossible to say whether the use of pure by Glasgow adolescents is a unique development in this community and it is even less clear whether this is an age-graded form that will disappear as the adolescents move into the adult world. The drop in frequency in 2004 may indicate that it is losing popularity. On the other hand, the lower frequency in 2004 may be because the adolescents are less enthusiastic in their conversations (see endnote 2). The future of pure in the school and in the wider community of Glasgow will require, as usual, further research. What is undeniable is that for this adolescent speech community (essentially one school in one district of Glasgow) during the period 1997–2004 pure is a powerful and important feature of their linguistic resources.
As Eckert (2001:382) has observed, “adolescents enter into a kind of time warp—or a cultural sink” and only a few adults (e.g., Cheshire, 1982; Eckert, 1989, 2000; Fought, 1999) have been brave enough to make a serious attempt to explore the use of language employed in this cultural sink. The recordings made by Jane Stuart-Smith for very different purposes have revealed a glimpse of the ways in which adolescents create their own norms and assert their identities. The technique of recording unsupervised conversations of this kind provides samples of adolescent speech that reveal patterns that might be hard to obtain from interviews conducted by adults. The comparability of the samples permits quantitative analysis of frequently occurring discourse features. More investigation of this type in addition to the kind of ethnographic studies carried out by Eckert (1989, 2000) might make a significant contribution, not only to the study of linguistic variation, but also to the functioning of “the invisible hand” (Keller, 1994) in the processes of linguistic change.
