One's purpose in conducting a language survey can have an effect on the types of data that are collected and the types of analyses that are used (Simons 1977c). WORDSURV was written primarily as a tool for the dialect surveyor who is concerned with gaining a present-day picture of the language situation in a given area. This contrasts with the historical linguist, who is interested in the changes languages have undergone in relation to each other through time. WORDSURV (in its COMPASS module) implements an approximation of the comparative method used by historical linguists, so it should be of some benefit to them. However, the routines in WORDSURV by themselves are inadequate for rigorous historical study. Yet they are sufficient for the rough types of analysis used by the synchronic surveyor. Thus this chapter centers on application for this latter type of study.
It may appear strange, in light of this introduction, to make the statement that a major goal of word list analysis for the synchronic surveyor is to gain a picture of the genetic relationships between languages. But it is through an understanding of genetic relationships that one can really understand the synchronic situation.
With an understanding of the genetic relationships between dialects, the field linguist can do a number of things. He can make initial hypotheses of patterns of intelligibility, which can then be refined through direct testing. He can estimate the degree of difficulty the speakers of various dialects might have in learning to understand a particular dialect. He can estimate the feasibility of implementing computer-assisted adaptations of written materials between dialects. And finally, he can gain an idea of the historical interactions of the languages which can supplement his sociolinguistic observations.
Wimbish, John S. 1989. WORDSURV: A Program for Analyzing Language Survey Word Lists, page 54. Occasional Publications in Academic Computing, number 13. Dallas, TX: Summer Institute of Linguistics.