Andrew Radford - Linguistics An Introduction [Second Edition]

Table 5 Deletion of [t] and [d] in English

Followed by a consonant

Followed by a vowel

% deletion in

% deletion in % deletion in

% deletion

Language variety

non -ed clusters -ed clusters

non -ed clusters in -ed clusters

Standard American English

66

36

12

3

White working-class American 67

23

19

3

English

Black working-class American 97

76

72

34

English

Puerto Rican working-class

93

78

63

23

English

56

sounds

Table 6 A hypothetical implicational scale

easy to delete

hard to delete

!

non -ed clusters -ed clusters

non -ed clus-

-ed clusters

followed by a

followed by a

ters followed

followed by

Language variety consonant

consonant

by a vowel

a vowel

Dialect A

+

+

+

+

Dialect B

+

+

+

−

Dialect C

+

+

−

−

Dialect D

+

−

−

−

(29)

If a particular dialect deletes final [t] and [d] in a specific linguistic environment, then the same dialect will delete [t] and [d] in all environments that more readily allow for deletion.

In Dialect B, for instance, the most unlikely environment that allows consonant deletion is in non -ed clusters followed by a vowel. This implies that it is possible to delete consonants in all environments to the left of this one on the grid. In the actual study reported above, we do not find deletion occurring always or never in a particular environment; rather we see different frequencies of deletion. For such a case, then, it is necessary to replace (29) with the implicational statement in (30): (30)

If a particular dialect deletes final [t] and [d] with a certain frequency in a specific linguistic environment, then it will delete final [t] and [d] with

a greater frequency in all environments that more readily allow for deletion.

The statement in (30) is true of table 5 because in each row the figures increase as we move from right to left.

To summarise, we can see that variability in language is not free and random but is characterised by what William Labov has called ‘orderly heterogeneity’ – a set of social, interactional and linguistic factors which have complex effects on the linguistic forms found within a speech community.

Variation and language change

Finally, here, we introduce the vital role that variation plays in lan-

guage change, the subject of the next section. If a sound changes in a particular community, this implies the existence of sound variation as an intervening stage in the process of change. A change from an old form to a new one necessarily involves a stage where both the old and new forms coexist, not only in the speech of the community as a whole, but also in the speech of individuals. You do not go to bed one night with an old sound and wake up the next morning with a new

Sound variation

57

sound having completely replaced the old one! The coexistence of old and new forms leads, of course, to language variation.

In order to introduce briefly the intimate relationship between language variation and language change, we present here further research carried out by William Labov (see the main introduction). He had noticed that in New York some people pronounced the ‘r’ following vowels in words such as car and park and others did not. He proposed that the New York speech community was changing from being

‘r’-less (or non-rhotic, see section 2, p. 37) to being ‘r’-ful (rhotic), and in order to investigate how this change was spreading throughout the community, he carried out an unusual but rather simple investigation. He visited three department stores, one middle-class, expensive store (Saks), one inexpensive store (Klein) and one in between (Macy’s) and asked as many assistants as he could find the whereabouts of a product he knew to be on the fourth floor of each store. The expected answer

‘fourth floor’ was, of course, carefully chosen, as it contains two examples of the

‘r’ he was looking for: in fourth the ‘r’ occurs before a consonant, and in floor it occurs at the end of the word. Having received the answer ‘fourth floor’, Labov pretended that he hadn’t heard properly, asking the assistant to repeat. He thereby doubled the size of his data set and introduced a further variable into the study, as the assistants’ second replies could be regarded as ‘emphatic’ or ‘careful.’ Having posed his question, which required the answer ‘fourth floor’ to over 250 assistants, he was able to compare the use of (r) across a number of speaker characteristics, contextual styles and linguistic environments, such as those in table 7.

Some of Labov’s results appear in figure 22.

As might be expected from our earlier discussion, Labov found that the

different social, contextual and linguistic factors had varying effects on the use of (r). He found, for example, that assistants in Saks were more likely to use

[r] than those in the other stores; younger people were more likely to use [r]

than older; [r] was more likely to be used in the emphatic second reply; and [r]

Table 7 Social, contextual and linguistic variables from Labov’s study of (r) in New York department stores

characteristics of the shop assistants

store (upper-middle-class, lower-middle-class, working-class)

job within store (floorwalker, till operator, shelf filler, etc.)

floor within store (higher floors sell more expensive products)

sex

ethnicity

age

contextual characteristics

first reply given versus emphatic reply given after Labov had pretended not to hear linguistic environment

(r) before a consonant versus (r) in word-final position

58

sounds

70

60

50

40

30

20

10

0

% of assistants using [r] some or all of the time

Saks

Macy's

S Klein

Department store

Figure 22a Percentage of department store assistants using [r] by store (from Labov 1972: 51). Reprinted with permission of the University of Pennsylvania Press.

90

80

70

60

e 50

tim

40

30

20

10

% of assistants using [r] some or all of the

0

old

middle aged

young

Age

Figure 22b Percentage of Saks department store assistants using [r] by age

(based on Labov 1972: 59). Reprinted with permission of the University of

Pennsylvania Press.

was more likely in the word floor than in fourth. Particularly important for our discussion of the role of variation in language change, however, is the fact that every stage in the advancing change to [r] could be found in the speech of some of the assistants. Some used virtually no [r] at all, others – who were obviously further ahead in the change – used [r] all the time, but most used it some of the time but not on every occasion. The study thus provided Labov with a convenient

snapshot of the progress of this change through the speech of individuals, particular groups and the whole New York speech community (exercise 5).

Sound variation

59

60

50

40

30

% use of [r] 20

10

0

fourth

floor

FOUR

TH

FLOOR

casual and EMPHA

TIC pronunciations

Figure 22c Casual and emphatic pronunciations of [r] in New York

department stores (based on Labov 1972: 66). Reprinted with permission of the University of Pennsylvania Press.

Exercises

1.

If we are able to shift our speech so readily, why do you think that