2.4 Back vowels

The parallel frontward shift of the back vowels has been found in English dialects of the United States other than that of the American South. Over the past few decades, linguistic research has revealed that back vowel fronting, particularly of /u/, may be an increasing dialect pattern in areas of the American West and Midwest.

Two decades ago, Hinton et al. (1987) conducted a real time study of Bay Area California vowels with data collected in the 1920’s, 1950’s, and (by the authors) in 1986. Analysis of back vowels showed that the percentage of speakers with fronted high back vowels had increased by a very small amount between the 1920’s and 1950’s data, from approximately 22% to approximately 29%. However, the data collected in 1986 showed that 70% of young Californians had fronted /u/ vowels in environments other than preceding /l/. This study suggests that, by 2010, /u/ fronting should be well established in the San Francisco Bay Area and most likely, in areas nearby as well. Evidence for this assertion can be found in, for example, Fought (1999), who mentioned that she has found that Anglo residents of the Los Angeles area demonstrate significant /u/ fronting in interviews she has conducted in the area. However, the primary goal of the research shown in Fought (1999) was to determine if the Chicano minority was also participating in the wider change. She examined the back vowel patterns of 32 speakers and found that fronting of /u/ among Chicanos exhibited a complex pattern that was dependent on the social categories of sex, class, and whether or not the person was affiliated with a local gang. Non -gang affiliated speakers tended to have the most fronted /u/ vowels, demonstrating that the gang identity may be partly expressed through the rejection of participation in wider (Anglo) norms.

Elsewhere in the West, as part of a larger project at Portland State University (the PDS, or Portland Dialect Survey), Ward (2003) looked at back vowel fronting in Portland, Oregon. His study analyzed data from 18 speakers of the area and found pervasive /u/ fronting in the area, particularly with younger speakers, with almost no correlation with gender or social class. Somewhat oddly, young adults had more fronting than teens (who still showed far more than older speakers). Though age was shown to be the best predictor of fronted /u/, the results showed that females fronted slightly more than males and that working class fronted slightly more than middle class. Ward’s study reinforces the idea that the dialect of the American West exhibits a pattern of back vowel fronting, particularly with the back vowel /u/.

Ash (1996) additionally reports that some speakers in the Midwest actually participate in /u/ fronting as well, even when they otherwise exhibit traits associated with the Northern Cities Shift (NCS), which originally counted “absence of change” in the back vowel /u/ as indicative of NCS (4). The findings were therefore surprising and suggest that /u/ fronting may occur in conjunction with other features of the NCS. The article also cites research done by Luthin (1987) which shows back vowel fronting in San Francisco, California, further supporting the idea of back vowel fronting in the American West.

Recent research has shown that there may in fact be a difference between the execution of /u/ fronting by Southern and non-southern /u/-fronters. Koops (2010) analyzed fronted /u/ vowels as spoken by Houstonians. He created a rating system based on the front vowel shifts of the SVS and monophthongization of (ay) and, out of a greater sampling of 42 speakers, ten speakers were selected who participated in the vowel shift the least (presumably not at all) and ten were selected who participated the most. The mean ages for Southerners and non-Southerners (as defined by participation in the SVS) were 54 and 19, respectively. While both groups demonstrated /u/ fronting, the formant trajectories of the two groups were significantly different. In simplified terms, the non-Southern group’s fronted /u/ contains a backed offglide (resulting in a diphthong similar to [ɪ^u]), while the Southern group’s /u/ remains relatively steady in a fronted position (similar to [y]).

Taken together, these studies contribute to the decision to exclude investigation of back vowel movement as one of the features strongly indicating a Southern accent. The above research demonstrates that, while back vowel fronting is undoubtedly a feature of Southern speech, the phenomenon is obviously not exclusive to that geographic region. While Koops (2010) shows that the types of /u/ fronting occurring in the South and elsewhere may not be the same, it is difficult to determine whether listeners would make this distinction. Though this feature would certainly be able to contribute to the current study, keeping the study at a feasible level is vital and necessitates trimming variables. Perhaps the forward movement of the back vowels can be included in future work to test if listeners associate all types of /u/ fronting with the South, no matter the reality, particularly if listeners are told they are listening to a Southerner. Given the above research, however, it is not likely that a fronted back vowel alone would be enough to obtain a percept of a Southern accent.

The study can be simplified by only looking at the two front vowel pairs in the shift. In addition, work such as that done by Fridland, Bartlett, and Kreuz (2004) suggests that the front shifts may indeed be much more salient for listeners than the back vowel movement. In fact, they write “midfront vowel shifts are generally more socially identifying than the shifts affecting the back vowels” (2004, 13). In this work, the researchers conducted a study testing listeners’ perceptions of the SVS to investigate to what level listeners are attuned to formant changes in vowels and, presumably, the ability to assign a social value to the sound. Fridland et al. had two speakers, one male and one female, read a word list. The speakers, from Memphis, Tennessee, had a small amount of Southern shifting. From this natural (as opposed to artificially synthesized) data, they were able to acoustically manipulate the vowels, resulting in a shifted and non-shifted token for each speaker for each of six vowels. Reactions were elicited from 141 listeners from the Memphis area. In rating the degree of Southernness of the variants, pairs of words were presented to listeners. For each pair, one word contained a vowel with a more advanced version of the SVS than the other. The listeners circled the word they considered more Southern and rated how different the two stimulus words sounded from each other. The researchers found that Memphians were better able to recognize those front vowel shifts that were most prevalent in their own speech, i.e. the mid vowels were more salient than high vowels. By contrast, the back vowel shifts did not show the same pattern; even though speakers in the Memphis area participate in this sub-shift, they do not associate fronted back vowels as readily with sounding Southern. Apropos to the studies discussed above, they observe that “it certainly appears that /uw/ and /U/ fronting are quietly spreading through North American dialects with little social significance attached” (13). Further, for those front vowel shifts in which the Memphians participated the most, they were also better able to label (as more Southern) finer distinctions in formant movements toward a more advanced SVS. In Fridland, Bartlett, and Kreuz (2005, 370), which summarizes the results from the 2004 article, they conclude: “These results suggest that speakers’ perceptions are sensitive to community norms.”