Reversal of near-merger

Guy Bailey, Jan Tillery & Tom Wikle

Sat. 11:00-12:40 A

The work of Labov and DiPaolo and Faber on near mergers provides strong empirical support for the idea that mergers are irreversible by linguistic means. Labov (1994) is able to show that a number of mergers which were thought to have been reversed were actually just near mergers, while DiPaolo and Faber (1990) demonstrate how phonetic contrasts among vowels can sometimes be maintained even when F1/F2 distinctions are lost. While we now have firm evidence for the idea that mergers cannot be reversed by linguistic means, we have little evidence on the nonlinguistic means by which a merger can be unmerged.

This paper explores the nonlinguistic means by which mergers can be reversed by examining a reversal in progress in Oklahoma -- the reversal of the merger of /'/ and /^/ before nasals (or the pen/pin merger). Brown (1991) has shown that the pen/pin merger began to expand in the American South during the last quarter of the 19th century and went virtually to completion by the end of the 1930s. Data from a Survey of Oklahoma Dialects (SOD) shows that the pen/pin merger had gone virtually to completion by 1930 in Oklahoma too, but the data also suggests that over the last 50 years the distinction has begun to expand at the expense of the merger. In other words, SOD suggests that in Oklahoma we have the reversal of a merger in progress.

Our analysis of SOD data suggests that the convergence of a number of nonlinguistic factors probably account for this reversal. First, over the last 50 years Oklahoma has undergone dramatic changes in population. The establishment of military bases during WW II and escalating oil prices and the Sunbelt phenomena during the 1970s and 80s brought large numbers of people from other areas to Oklahoma. These new arrivals are significantly more likely than native Oklahomans to preserve the pen/pin distinction. Second, many of the new arrivals settled in the upper-middle class suburbs of Oklahoma City and Tulsa that serve as foci for the diffusion of linguistic innovations. Thus, the pen/pin distinction was well-positioned to expand along with other innovations. Finally, the fact that the pen/pin merger is a partial merger, that spelling is a reliable guide to the distinction, and that some teachers explicitly teach the distinction may have made the merger particularly reliable to reversal.