Mergers have been a much-researched topic in sociolinguistics (e.g., Baranowski 2013, Thomas and Hay 2005, Hall-Lew 2013), including in pre-lateral contexts (Bowie 2000, Faber & Di Paolo 1995, Thomas 2001). However, aside from Thomas and Bailey’s (1992) study on “competing mergers” in Texas, little work has been done on the interaction among several mergers involving some of the same phonemes and occurring in the same contexts when they coexist in a given community. Even less research has addressed the role of perception in competing-merger contexts.
This study examines the status of mergers among /ul/, /ol/, and /Ul/ in Youngstown, Ohio. Although at least three forms of the merger have been cited in this community—/ul/-/Ul/, /ol/-/Ul/, and /ul/-/ol/-/Ul/—not all speakers in the community are merged, and those who are merged do not all merge the same phonemes. Using acoustic analyses and multiple-forced-choice perception task results from 26 speakers from the Youngstown area ranging from ages 9-81, this project examines 1) whether these mergers are all progressing in production and/or perception (and at the same rate) in this region and whether production directly correlates with perception, 2) whether they are motivated by the same internal linguistic forces, 3) how the presence of multiple patterns of pre-/l/ merger affects both the realizations and progression of these variants in the community, and 4) whether there is evidence to suggest that alternative patterns of distinction are maintained in cases of qualitatively merged vowels.
Acoustic analyses of F1 and F2 measured 25% into the vowel-liquid sequence, as well as multivariate statistical analyses, suggest that the /ol/-/Ul/ merger is progressing in apparent time, mainly with respect to F1, while the /ul/-/Ul/ merger is remaining relatively steady. Additionally, the /ul/-/Ul/ merger appears to exhibit features unique to Youngstown in that it is realized more closely to /ul/, unlike what has been typically described of this merger throughout the United States (Labov, Ash and Boberg 2006). Triple mergers, on the other hand, are so scarce in the community that generalizations about the merger’s progression cannot be made. However, those who are triple merged realize the merger closer to /Ul/ than those merged only between /ol/ and /Ul/ or /ul/ and /Ul/. Initial analysis of perception data suggests that production does not directly correlate with perception, perhaps as a result of exposure to multiple patterns of merger. Though this region does not show a simultaneous progression of the three “competing mergers,” it does exhibit considerable inter-speaker variation that, though puzzling, allows for another angle from which to examine sound change.
"Multiple Mergers: Production and Perception of Three Pre-/l/ Mergers in Youngstown, Ohio,"
University of Pennsylvania Working Papers in Linguistics:
2, Article 2.
Available at: http://repository.upenn.edu/pwpl/vol21/iss2/2