Since April 2015 is the sesquicentennial of the end of the Civil War, now is a particularly appropriate time to review the progress of the Corpus of American Civil War Letters (CACWL) project and to suggest directions it might go in the future. Since 2007, we have located and collected images of nearly 11,000 letters and transcribed over 9,000 of these, totaling well over four million words. Of the transcribed letters, just over 6,000 were written by southerners (490 individual letter writers), a corpus extensive enough to begin identifying and describing what features were distinctively Southern in 19th-century American English. We have already mapped many of these features that are especially common in southern letters, for example, fixing to, howdy, past tense/past participle hope ‘helped’, qualifier tolerable, intensifier mighty, pronoun hit, and the noun heap. By way of comparison, we also have a somewhat smaller but rapidly growing collection of 3,000 transcribed letters written by individuals from northern states, and variant features from these letters are also being mapped. The work at present is very preliminary; there are thousands of additional letters to be collected and transcribed, particularly from northern states and from states west of the Mississippi. However, by mapping variants from letters that have already been transcribed, we can begin to get a better understanding of regional differences, as well as how regional features spread westward in the decades before the Civil War. We can also begin to obtain some sense of how American English in general, and particularly its regional dialects, may have changed since the mid 19th century. This article presents a preview of a number of those findings.

Distinctive regional varieties of English have only recently emerged in parts of North America, including Wisconsin, where differences appear to be increasing today. We present an experiment in which listeners heard two short samples each from three Wisconsin regions and three other dialect areas. For each area, one sample was recorded pre-1970 and another recorded post-2010. Regional stereotypes were excluded. In a situation of new and still-emerging regional varieties, we expected listeners to be able to more accurately identify recent speech samples versus old samples from Wisconsin. Listeners proved better at recognizing speakers from Wisconsin in newer over older recordings. This complicates previous discussions of dialect awareness, in particular ‘enregisterment,’ with our listeners able to identify Wisconsin speech even in the absence of salient, known dialect features.

The complex series of vowel changes known as the Northern Cities Shift has been extensively documented over the last four decades across the broad territory of the Inland North dialect region. Little is known, however, about the origins of the shift, and there remain open questions about where the changes began and which vowel initiated the process. This paper examines such questions by analyzing the speech of several people born in the late 19th and early 20th centuries using archival recordings of oral history interviews. Drawing on acoustic data we identify what appear to be early stages of the Northern Cities Shift in some individual speakers though many in the sample give no evidence of participating in the changes. We consider the implications of these findings for accounts of how the shift began with particular focus on Labov’s (2010) proposal.