Variation and change and historical linguistics

The deepening engagement of sociolinguistics and variation studies with

language change has brought these fields into close contact with the traditional

discipline of historical linguistics. Initially, these were complementary

approaches to diachronic questions. Historical linguistics focused on real-time

data, used written evidence (necessarily, since no sound recordings of speech

existed prior to the late nineteenth century), and dealt with completed changes

across large-scale time spans€ – change across centuries and millennia. The

focus on completed changes typically implied invariant, categorical models

and descriptions. Variationist studies introduced the use of apparent-time evidence,

focused on speech, and dealt with changes in progress, over shorter

time spans€– decades and generations€– using quantitative models and descriptions.

But more recently, this neat division of labour and focus has eroded.

Variationist analyses have been conducted of documentary evidence from

times long past. Quantitative approaches have been brought to bear on completed

long-term changes using written materials to study their time course and

the variation that occurred while they were in progress. Sociolinguistic models

of the mechanisms of change have illuminated historical questions. Whereas

traditional historical linguistics sought, in effect, to use the past to explain the

present, the addition of the variationist perspective to diachronic research has

also, in Labov’s words (1994:€9), made it possible to ‘use the present to explain

the past’.

Particularly noteworthy examples of the fruitfulness of this fusion of variation

and diachrony have occurred in research on historical syntax, such as

the previously mentioned work of Kroch on English, and other studies on languages

as diverse as Yiddish (Santorini 1993), Greek (Ann Taylor 1994), and

Portuguese (Tarallo 1996). Phonological change is less amenable to this kind

of approach, because of the limitations of orthographic evidence; nevertheless,

some fruitful work has been undertaken, such as Toon’s study of ‘the politics

of early Old English sound change’ (1983).

The principal impact of variation research on historical linguistics, however,

may be less methodological and empirical, and more theoretical. The

variationist perspective has had little impact on the comparative method and

the reconstruction of proto-languages, but understanding that variation is an

essential way station in the course of change has substantial implications for

evaluating the plausibility of the changes that are postulated and their social

settings. For example, change-in-progress studies show that the period of variation

can last for a long time, and multiple changes may be underway simultaneously.

This has implications for reconstructing the sequencing of events (e.g.

chain shifts). Contemporary variables sometimes exhibit lexical conditioning

or irregularity, with implications for the regularity of sound change. Studies of

the sociolinguistic types of change have implications for evaluating prior language

contact. The transmission of lexical items, for example, implies borrowing

as a primary mechanism of language change; hence in England after the

Norman conquest, the huge inventory of French loanwords in English implies

that native speakers of English were borrowing from French, while the paucity

of phonological and syntactic effects suggests that Norman accents had very

little impact on the English of descendants of the conquerors who underwent

language shift. By comparison, English in India, which shows phonological

characteristics common in Indian languages, such as retroflex consonants, and

syntactic phenomena such as an invariant tag question ‘isn’t it’, is a case in

which the main mechanism of change was imposition.

A significant consequence of the variationist perspective in historical studies

has been the development of quantified models of change. Yang (2001) is a

noteworthy example of this trend. Treating syntactic change as the product of

the interaction between the distribution of syntactic structures in the input and

the choices that child language learners face in their construction of a mental

grammar, Yang proposes a probabilistic model of grammar competition that

drives change forward along an S-shaped time course. The model crucially

depends on variation:€child language learners do not construct a single, static,

invariant grammar to account for all the facts they encounter; rather, they entertain

multiple alternatives, and select among them probabilistically.

Another significant contribution of variationist studies to historical linguistics

is the refined view they permit of the stages of change. Conventional historical

studies, relying on reconstruction from fragmentary evidence, typically

account only for the endpoints of a change; a diachronic statement like x → y

tells us that an early form x is realised centuries later as y, but gives no perspective

on what happens during the intervening years. But synchronic studies of

changes in progress make it possible to investigate triggering events and onsets

of change (the actuation phase), and subsequent expansion of the innovation

(the implementation phase).

Actuation appears to involve both social and linguistic factors; thus Labov

(2010) attributes the original generalised tensing of /æ/ in the Inland North of

American English to a social event:€ the early nineteenth-century mixture in

north central New York state of speakers coming from several different dialect

regions (including New England and southern New York), during the construction

of the Erie Canal. These source dialects had different contexts for /æ/

tensing. The new communities that emerged from this mixture koinéised these

conflicting patterns by tensing /æ/ in all contexts. The completion of the Erie

Canal provided the pathway to settlement of the Upper Midwest, disseminating

the new vowel phonology across a wide area. The tensed /æ/ vowel subsequently

raised, vacating the low front corner of the vowel space; this provided a

linguistic trigger for the fronting of the other low vowels. 

Analysing Variation in English- Edited by Warren Maguire and April McMahon 

Cambridge University Press, New York

Chapter 8 Variation and change- Gregory R. Guy

8.4- Variation and change and historical linguistics- Pages 193,194