This is a study of the geographical distribution in Indiana and the social distribution in a few Indiana counties of several hundred lexical items taken from the Linguistic Atlas work-sheets. The material was gathered in 1957 and 1958 by means of a questionnaire distributed using a variation of the correspondence method described and shown to be valid by Alva L. Davis in his Ph. D. dissertation "A Word Atlas of the Great Lakes Region" (University of Michigan, 1948). The purpose of this study was to discover what the primary material shows about the northern and southern boundaries of the Midland dialect area in Indiana, and to show what it reveals about the effect of age and education on vocabulary.The study is based on 263 questionnaires consisting of 147 checklists like those used by Davis in his dissertation. The informant was asked to circle the word or expression in each checklist which he would use to express the idea defined in that semantic unit.The study makes frequency counts of the recurring lexical items by a methodology developed by Charles L. Houck and recorded in his "A Computerized Statistical Methodology for Linguistic Geography: A Pilot Study" [Folia Linguistica, I (1967), 80-95] and in his "A Statistical and Computerized Methodology for Analyzing Dialect Materials" (Ph.D dissertation, University of Iowa, 1969). Houck's programs, designed for the IBM 7044, 32K core computer, are adapted in this study to the IBM 360-40, 331K core computer.The first three chapters of this dissertation describe the problem and the method, review related studies, and survey Indiana settlement history. Chapter IV shows the geographical distribution of items in 133 of the checklists, only those which contain items the regional classification of which could be discovered in former studies. Chapter V is a record of the distribution by age and education among the informants from eastern central Indiana of the items in 23 of the checklists. The Appendix contains a sample questionnaire, maps showing the geographical distribution of the items in 50 checklists, and sample computer programs and read-outs.The conclusions in this study conflict with Davis' "A Word Atlas of the Great Lakes Region" in 50 instances out of 96 checklists which appear on both his questionnaire and the one used in the present study. These conclusions suggest that considerable change in vocabulary occurred in the decade between Davis' study and the time the material was gathered for this study.The limited analysis of the distribution of lexical items based on age and education shows little that is surprising. The older informants tend to have more alternate terms for a specific meaning than do the younger ones. The less well educated informants are generally made up of the older ones; therefore, the discovery that the less education, the more variety of vocabulary is likely insignificant.Generally, this study indicates that dialect boundaries among Northern, Midland, and Southern Regions on the East Coast which other studies have shown to extend westward-are blurring considerably in Indiana.

Recent discussions on African-American Vernacular English (AAVE) have focused on the linguistic divergence and convergence hypotheses. Some linguists (Ash and Myhill 1986; Bailey and Maynor 1987, 1989; Graff, Labov, and Harris 1986; Labov 1983, 1987; Labov and Harris 1986; Luthin 1987; Myhill and Harris 1986; Thomas 1989) claim that AAVE is diverging from White Vernacular English (WVE) on a national level. However, other linguists (Butters 1987, 1988, 1989; Vaughn-Cooke 1986, 1987; Wolfram 1987) have challenged the divergence hypothesis, and have argued that AAVE is actually converging with WVE. They point out that the data in most of the studies supporting the divergence hypothesis were incomparable and manifested age-grading. In addition, these studies investigated only a few linguistic features. Most importantly, most of these studies lack the time-depth data which are essential to investigate language change.This study analyzed the time-depth data of speech samples from thirty-two African-American subjects, sixteen from 1980 and sixteen from 1993, in Muncie, Indiana. The subjects were both males and females, equally divided into young adult and elderly speakers. The analysis of the study focused on twenty-three syntactic and five phonological features.The results from the study have found no innovative features, either syntactic or phonological, in the speech of Muncie AAVE subjects. More importantly, the findings of the study, based on the time-depth data, have shown that Muncie AAVE was not divergent with WVE, but convergent with it, at least from 1980 to 1993. Thus, the findings of the study do not support the divergence hypothesis. / Department of English