Accommodation across borders

The authors analysed the speech of a Scottish English speaker who conducted sociolinguistic interviews along the Scottish–English border; this is a significant linguistic boundary between the vastly different varieties of Scottish English and (northern) English English. 

• In particular, the Eyemouth variety is rich in Scottish features, while Carlisle speakers use considerably more northern English features.

In their investigation, the authors selected a range of phonological, discoursal and lexical features. 

• They looked at how the interviewer accommodated her speech to older and younger speakers from the two localities that she interviewed, by comparing the distribution of variants in her speech with the patterns in the communities.

The interviewer accommodated to the speech of her interviewees by responding to their social characteristics: national identity (Scottish/English) and age (older/younger).

• With traditional variants that are on their way out, she showed convergence with the community patterns, using them generally more with older speakers than with younger speakers.

• With forms that are typically perceived as Scottish and traditional, such as f(r)ae as from, she made use of them far more frequently when speaking with older Eyemouth informants than with other speakers, to the extent that she overshot the actual high/low usage by the corresponding groups.

Thus, the interviewer's pattern of accommodation is strongly influenced by the perceived social categories of the informants, which are indexed by the forms in question.

The speaker did not exhibit accommodation in all linguistic features, but the selection of features that show accommodation is not random. The authors found that the main factor behind whether a feature is accommodated to is its stability in the community:

• Accommodation is unlikely for forms that are stable and near-categorical, such as coda-r (presence of /r/ in car in Scottish English vs absence in English English), where the speaker maintained a high percentage of realised coda-r regardless of her interlocutor.

• Accommodation is much more likely for variables that are undergoing change, such as the (stereotypically) traditional Scottish use of ken to mean know, which appears to be dying out in the younger generation. In this case, the speaker frequently used ken with older Eyemouth informants, but not at all with younger Eyemouth speakers or Carlisle speakers.

Contrary to previous suggestions by other linguists investigating dialect contact, it does not seem to matter whether the linguistic variable is a phonological, discoursal or lexical one.