ABSTRACT This research will use a telephone survey of North American dialects of English to trace the development of linguistic changes in progress on a broad geographic scale. Telephone interviews offer a rapid and inexpensive method of sampling broad geographic areas with a sampling density proportionate to the variation in the data. The survey will establish a network of native speakers who are available for repeated interviews to obtain both phonetic and phonological information on a wide range of variables. The data obtained will be used to investigate general questions in the theory of language change, as well as provide national maps of the regional features of English. Current research on English dialects shows continued change and increasing diversification of broad regional patterns through chain shifts, mergers, lexical splits and the vocalization of liquids. A number of principals have been proposed to account for the geographic spread and transmission of these changes across generations, including specific quantitative models for areal diffusion. Efforts to confirm these models have been limited so far to sociolinguistic studies of small areas; the regional data that has been gathered is dispersed over time and space in ways that inhibit comparison. Furthermore, the traditional methods of dialectology do not gather the information on perception and categorization of speech needed to investigate the mechanism of phonological change, and the traditional focus on rural communities does not provide the data needed to investigate the correlation between community size and the advancement of change. Results of a pilot survey of the North Central States show considerable success in recruiting a number of subjects and in obtaining phonetic and phonological information that shows clear regional differentiation of sound changes in progress. Current methods of instrumental analyses of telephone recordings provide data of sufficient quantity and internal reliability to map individual vowel systems. An extended survey of the Middle Western United States will be conducted from the eastern boundary of Ohio westward to the Continental Divide and from the Ohio River northward to the Canadian border. The survey will gather data that will trace the progress and mechanism of the Northern Cities vowel shift, its relation to the southern vowel shift, the merger of long and short open /0/, ongoing mergers before liquids and nasals, and the lexical distribution of the tensing of short a and o in those regions where splits have occurred in the past. The sample will be representative of both geography and the population size of communities. The basic unit of sampling will be areas of influence of the largest urbanized centers as established by newspaper circulation.
