Low Back Merger Research Articles

Vowel trajectories of African Americans in Georgia, USA

Within the United States, dialectal variation is often characterized by vowel shifts: systematic differences in vowels' relative qualities and vowel-inherent dynamics. The African American Vowel Shift (AAVS), in particular, includes raising and fronting of front lax /ɪ ɛ æ/, among other features. The more recent pan-regional Low Back Merger Shift (LBMS), by contrast, includes lowering and backing of the same vowels. We evaluate these two shifts in an audio corpus of over 40 Black speakers from the Southern state of Georgia. Speakers, born between the 1930s and 2000, represent five demographic generations. Normalized formant values (F1,F2) from five temporal points per token are input to Generalized Additive Mixed Models (GAMMs). We test for significant changes in vowels' trajectories across time by fitting Year of Birth as a continuous smooth term. Additionally, we use linear mixed-effects modeling to test for raising versus lowering on the (F2–F1) front-vowel diagonal, across generations. Evidence from GAMMs and linear modeling indicates raised positions of /ɪ ɛ æ/ among older generations (1950s–1980s), followed by significant retraction from 1990–2000. These acoustic results are consistent with strengthening of the AAVS in the third quarter of the 20th Century, followed by a rapid transition to the pan-regional LBMS.

Generational Phases: Toward the Low-Back Merger in Cooperstown, New York

This paper reports on a new sociolinguistic sample of Cooperstown, a village in rural central New York. Previous research suggested Cooperstown was losing the Northern Cities Shift (NCS) and acquiring the low back merger via koineization as a result of dialect contact among locally-born children of parents from other regions. The new data shows abrupt retreat from NCS patterns between the Baby Boom generation and Generation X. A “phase transition” pattern is observed in progress toward the low back merger: Millennial women are the first to describe low back minimal pairs as merged, despite no appreciable difference between Millennials and Generation X in production of the low back vowels. No evidence is found to support the hypothesis that koineization is responsible for these changes; it appears that Cooperstown is subject to the same trend away from NCS documented in many other communities, subject to many of the same constraints.

Phonological mergers have systemic phonetic consequences:palm, trees, and the Low Back Merger Shift

AbstractThis paper provides a unified phonologically motivated explanation for the movement oftrap,dress, andkitfollowing the low-back merger in North American English (i.e., the Canadian Shift, California Shift, Low Back Merger Shift, Third Shift, etc.). The explanation puts forth that the three-way merger oflot,palm, andthoughtresults in the loss of the [+Front] feature specification fortrap, opening the door for dispersion focalization to pulltraptoward the low central region of the vowel space. Analogy then prompts all other [−Peripheral] vowels, includingstrutandfoot, to centralize. Crucial to this explanation is that the low-back merger includespalm, not justlotandthought. Evidence for this requirement is presented in a phonetic analysis of older speakers from conservative Victoria, British Columbia. The explanation presented here reconciles an earlier proposal (Roeder & Gardner, 2013) with Fruehwald's (2017) observation that parallel movement requires a shared feature specification.

Structure, Chronology, and Local Social Meaning of a Supra-Local Vowel Shift: Emergence of the Low-Back-Merger Shift in New England

AbstractThe Low-Back-Merger Shift (LBMS) is a major North American vowel chain shift spreading across many disparate dialect regions. In this field-based study, we examine the speech of fifty-nine White Western Massachusetts speakers, aged 18–89. Using diagnostics in Becker (2019) and Boberg (2019b), we find the LBMS emerging at the expense of the Northern Cities Shift (Labov, Yaeger, & Steiner, 1972) and traditional New England features (Boberg, 2001; Kurath, 1939; Nagy & Roberts, 2004). In Becker's LBMS model (2019:9), the low-back merger (lot-thought) triggers front-vowel shifts. Our results suggest that local social meaning can sometimes override this chronology such that the front-vowel shifts occur before the low-back merger, even as the overall configuration comes to match Becker's predictions. Sociosymbolic meaning associated with the older New England system has led to a different temporal ordering of LBMS components, thus providing new theoretical and empirical insights into the mechanisms by which supralocal patterns are adopted.

The Foot of the Lake

In 2013, Dinkin reported an unexpectedly sharp dialect boundary in northern New York between the communities of Ogdensburg and Canton in St. Lawrence County: Ogdensburg exhibited the Northern Cities Vowel Shift (NCS) and very little evidence of the low back merger, while Canton showed low back merger nearing completion and no NCS. This article investigates the nature of this dialect boundary via new sociolinguistic interview data from eight neighboring communities: four along the St. Lawrence River and four 25 miles south of it. An east-west division is observed in merger incidence: the four communities to the west, including Ogdensburg, show relatively robust lot-thought distinction, though apparent-time trends toward merger exist; east of Ogdensburg, the merger is much more advanced. A similar sharp boundary may hold for the NCS raising of trap (though the data are spottier due to the NCS’s obsolescence). The geographical sharpness of this boundary suggests that it is not due merely to socioeconomic differences between communities. It may be due to historical patterns of transportation: in the nineteenth century, Ogdensburg was the easternmost navigable point of the upper St. Lawrence River, meaning communities east of Ogdensburg were not directly accessible to the Great Lakes shipping network.

2. Dimensions of Variance and Contrast in the Low Back Merger and the Low-Back-Merger Shift

2. Dimensions of Variance and Contrast in the Low Back Merger and the Low-Back-Merger Shift

Dutchified English in an Ohio Beachy Amish-Mennonite Community

This article compares the English spoken by Beachy Amish-Mennonites just outside the Holmes County, OH, plain community with the English of non-plain neighbors. Although there is no previous linguistic research about the Beachys in this region, previous research on the dialects of English spoken by Amish communities suggests that differences from regional standards are influenced by deliberate border maintenance, primacy of language in expressing ethnic identity, and interference from Pennsylvania German, known to speakers as “Dutchified English.” However, little research has examined the English of the Beachys, who, compared to the Amish, occupy a less conservative position between separation from and engagement with broader society. The variables I focus on are initial th-stopping, final obstruent devoicing, and the low back vowel merger. The first two variables are documented for Amish English and various German bilingual communities. Th-stopping is also socially salient and is associated with the more religiously orthodox Amish churches in Holmes County. I collected production and perception data from ten speakers in the Beachy congregation and five non-Anabaptist speakers from the same region, and I present here my findings of the use of th-stopping and final consonant devoicing, both of which are attested, although neither are used categorically, and the low back merger, which is attested categorically in all non-Anabaptists but in only one Beachy speaker. The perceptual data has theoretical implications for the role of community identity and border maintenance between non-plain neighbors and the heavily Amish Holmes County community.

Patterns of the Mainstream Sound Change in a Liminal Region: Low Back Merger in Washington DC

This study investigates the low back vowel merger (lot-thought merger) of African American and European American speakers in Washington DC. The study aims to follow up with the previous investigation by Labov et al. (2006) on low back merger in DC, and investigate whether African American speakers in DC are participating in the merger, one of the majority-led sound changes. The study also examines the difference in low back vowel production between African Americans from a particular quadrant of the city, namely, Southeast (SE), and those from elsewhere in DC. Data are taken from forty sociolinguistic interviews with natives of DC. The vowels were analyzed acoustically, taking F1 and F2 measurements at three different points of a vowel. Both the F1 and the F2 dimensions of the low back vowels and the degree of overlap were examined. The degree of overlap was measured by calculating the Pillai score for each speaker, with duration and following environment taken into account. Results indicate that DC is conservatively participating in the merger, with European American speakers exhibiting higher degrees of overlap than African American speakers. An age effect is also observed, but only among African American speakers who are not from SE. The study further argues that the participation in the mainstream sound change by some African Americans can also be analyzed as convergence to the local white norm, and that this convergence is more likely to be carried out by African American speakers who are not from SE, a section of the city in which the contact between African American speakers and European American speakers is minimal.

Production and perception of the low vowels in American English

Two vowel features are especially pivotal in defining contemporary U.S. dialects: The merger of the low back vowels and the variable realizations of the low front vowel. Several scholars (Bigham 2010, Gordon 2005, Labov, Ash and Boberg 2006) suggest a relationship between the low front and low back vowels such that /æ/ raising and subsequent fronting of /ɑ/ in the North inhibits the tendency toward low back merger. However, little work examines the robustness of this “structural linkage” or whether a similar relationship obtains across different regional varieties. Further, little work examines whether differences in production correlate with differences in regional perception patterns. Here, we compare the low vowel system across U.S. regional dialects and also consider interrelationships (i.e., correlations) between low vowel categories using data from eight fieldsites across the U.S. We then look at what these speakers’ perceptions of category shift between the low front and low back vowel tells us about this relationship in production. Our results suggest both that the regional vowel shifts create significantly divergent low vowel systems across regions but also that /æ/ and /ɑ/ indeed show a structural linkage across regions, while /æ/ and /ɔ/ and /ɑ/ and /ɔ/ show no such relationship.

The Participation of a Northern New Jersey Korean American Community in Local and National Language Variation

This study investigates the variable patterning of two English phonemes, /ɔ/ and /æ/, in the speech of 24 second- and 1.5-generation Korean Americans from Bergen County, New Jersey, which borders New York City and contains the most densely populated Korean American community in the United States. Detailed acoustic phonetic analysis reveals that in general Bergen County Korean Americans produce a raised /ɔ/ nucleus, in accordance with New York City English, and tense /æ/ in prenasal environments, in accordance with General American English. Positive correlations between speaker age and the frontness of /ɔ/ and distance in vowel space between /ɔ/ and /ɑ/ suggest a change in apparent time toward fronting of /ɔ/ and shortening of the distance between two vowels (and hence possibly movement toward a low back merger). Speaker sex patterns are also apparent, with female speakers showing a closer distance and more overlap between /ɔ/ and /ɑ/ than male speakers, and females showing a clearer split between tense /æ/ in prenasal environments and lax /æ/ elsewhere. In general, the patterning of variation for /æ/ among Bergen County Korean Americans shows an alignment to the general American norm, while the patterning for /ɔ/ suggests that Korean Americans are adopting the local norm. Future research will be needed to corroborate these patterns across a wider range of speaker ages.

Phonological Transfer as a Forerunner of Merger in Upstate New York

Herold (1990) discusses three mechanisms by which phonemic merger can take place: expansion, approximation, and transfer. A fourth possibility Herold touches on but does not explore might be called phonological transfer: as in (lexical) transfer, words move abruptly from one phonemic class to another; but rather than one lexeme at a time being transferred, all words of a particular phonological class move simultaneously. This paper provides evidence that phonological transfer is playing a role in the movement toward merger of /o/ (as in lot) and /oh/ (as in thought) in Upstate New York. Words containing (olF)—i.e., historical /o/ followed by /l/ plus a labiovelar, as in golf and revolve—are produced with /oh/ rather than /o/ in 74 percent of tokens; this use of /oh/ is increasing in apparent time. Many speakers using /oh/ in (olF) words have an otherwise clear phonemic distinction between /o/ and /oh/; however, the geographic distribution of this phonological transfer is correlated with other indices of progress toward the low back merger. This indicates that phonological transfer can be regarded here as an early sign of merger in progress, and that a single merger can proceed by two mechanisms simultaneously (here, approximation and phonological transfer).

The Michigan Upper Peninsula English Vowel System in Finnish American Communities in Marquette County

This article investigates the acoustic properties of the American English vowel spaces of 65 Finnish Americans from Marquette County in Michigan’s Upper Peninsula (UP). The author compares the general system of these UP vowel spaces to substrate Finnish and neighboring English vowel systems and examines the subsystems of these UP vowel spaces along the sociolinguistic variables of language dominance, age, sex, and educational attainment to ascertain whether the internal system dynamics show any substrate or exogenous characteristics. The apparent-time evidence in this study indicates a potential change in progress in this UP speech community, with younger speakers using variants typically associated with neighboring Canada (e.g., low-back merger of /ᵅ, ᵓ/, the lowering and retraction of /ᴵ, ε, æ/, the raising of /aᴵ, aᶷ/ nuclei before voiceless obstruents). Consequently, Michigan’s UP more closely aligns with the Canadian English’s regional boundary than the proposed “residual territory” as defined by Labov, Ash, and Boberg.

Phonetics as a complement to phonology in the Canadian Shift

Previous accounts of the Canadian Shift have interpreted this diachronic change in vowel pronunciation as a purely phonetic consequence of the low back LOT-THOUGHT vowel merger; however, such an analysis does not transparently explain the strong connection between the (phonological) low back merger and the subsequent (phonetic) retraction of the TRAP vowel in the acoustic vowel space. This paper addresses this issue by presenting an analysis of the shift that combines the approaches of Modified Contrastive Specification theory and the Contrastive Hierarchy—two phonological frameworks—with phonetic insights from Vowel Dispersion-Focalization theory. We propose that the catalyst of the Canadian Shift is a three-way vowel merger, in combination with a simultaneous change in the underlying feature specifications of the TRAP vowel. This results in a phonology that allows for the TRAP and DRESS vowels to succumb to the influence of the phonetic principles of dispersion and focalization. This hypothesis is illustrated by comparison of data from 59 speakers in Thunder Bay, Ontario, and Industrial Cape Breton, Nova Scotia. Our analysis predicts that a Canadian Shift-type phonetic change will occur in any North American dialect of English where the PALM-LOT-THOUGHT merger occurs, unless an intervening phonological change alters systemic contrasts.

On the role of social factors in the loss of phonemic distinctions

The article tests the generalisation of the curvilinear hypothesis and the tendency of females to lead linguistic change in vocalic mergers on the basis of two mergers currently in progress in Charleston, South Carolina: the low-back merger and thepin–penmerger. It is based on sociolinguistic interviews with 100 informants, aged 8–90, covering the socioeconomic spectrum of the city. The speech of 90 of the informants is analysed acoustically; it is supplemented with minimal-pair tests and word list reading. F1/F2 measurements and minimal-pair test results are subjected to a series of multiple linear regression analyses, with social class, gender, age and style as independent variables. While the low-back merger is a change from below showing a female lead and a curvilinear effect of social class, thepin–penmerger shows a decreasing monotonic relationship with social class and no female advantage. The difference is argued to be due to the two mergers being at different levels of social awareness.

Chain Shifting and Centralization in California Vowels: An Acoustic Analysis

This article documents the acoustic properties of the vowels of young adults from various regions of California, with special attention to a chain shift that lowers short front vowels /I/ (as in kit), /ϵ/ (as in dress), and /æ/ (as in trap). It also tracks the centralization of /ow/ (as in goat). Quantitative analysis of subjects' formants shows a consistently lower set of front vowels and a consistently centralized /ow/ relative to baseline comparisons drawn from Labov, Ash, and boberg's 2006 Atlas of North American English, suggesting a recent diachronic emergence of both phenomena. A small gender effect is observed, suggesting women are further advanced than men in the chain shift. These findings suggest that the chain shift is not clearly a pull-chain brought about by the low-back merger of /Ɔ/ (as in lot) and /ɑ/ (as in thought) and may have occurred independently as a push-train.

Weakening resistance: Progress toward the low back merger in New York State

AbstractThis paper examines the status of the low back caught-cot merger in Upstate New York. Most of this region is subject to the Northern Cities Shift (NCS) and therefore, according to Labov, Ash, and Boberg (2006), ostensibly “resists” the spread of this merger. It is found that the phonology of this region is indeed trending toward the merger in apparent time, in terms of both phonetic distance between the two phonemes and speakers' explicit judgments. It is argued that the fronting of the cot vowel in the NCS region is not sufficient to withstand the spread of the merger because fronting a low vowel is a “reversible” sound change (Labov, 2010). It is further argued that the expansion of a merger to new communities may take place indirectly, through launching a sound change in the direction of merger rather than causing merger to take place immediately in the new community.

Low-Back Merger and Phonemicization of /ɒ/ in Modern General American English

Low-Back Merger and Phonemicization of /ɒ/ in Modern General American English

The Canadian Shift in Toronto

AbstractThis study provides the first wide-scale, apparent time, instrumental description of the Canadian Shift in mainstream Toronto English. In contrast with some previous findings, the Toronto data suggest that for the last 70 years or more the shift has not affected the high front lax vowel (i). We observe that the movement of the nonhigh front lax vowels (ε) and (æ) involves both lowering and retraction in Toronto English, although retraction is the primary direction of more recent change and the shift appears to be slowing down. Our findings also suggest that continued retraction of the vowel resulting from the low back merger is involved in the final stage of the shift. We do not find evidence of a chain shift but instead propose that a parallel shift is occurring and make reference to Vowel Dispersion Theory in our discussion.

ACCOMMODATION TO THE LOCAL MAJORITY NORM BY HMONG AMERICANS IN THE TWIN CITIES, MINNESOTA

This study investigates selected English sounds spoken by immigrants in a region of the United States undergoing vowel space restructuring. The vowels of Hmong Americans in the Minneapolis–St. Paul area are compared to those of local white middle-class speakers. Results of this comparison suggest that young white speakers produce a fronted bat vowel and a lowered and backed bet. Males reduce the mean bot-bought spectral distance, leaving the impression of an apparent merger. Younger Hmong Americans who arrived in the United States at a young age accommodate to local norms for bat fronting and raising but not to the apparent low-back merger, regardless of speaker age, sex, generation, or social mobility.

Improved representation of variance in measures of vowel merger.

Current measures of vowel merger, such as the Euclidean distance between averages, have only been able to capture some of the variability between two given vowel clusters. Reliance on averages obscures the amount of variability within a given vowel class, while other techniques, such as calculating distance between minimal pairs, rely on few tokens per speaker. Hay et al. (2006) introduced an alternative approach that accounts for the variability between two vowel clusters, taking formant values as input, rather than averages. The measure is the Pillai-Bartlett statistic [Baayen (2008)], an output of a multivariate analysis of variance (MANOVA), which represents the proportion of one variance that can be predicted by another variance. A higher Pillai value indicates a lower degree of overlap between two vowel clusters in F1/F2 space. Since the value is derived from a MANOVA, Pillais can account for known internal factors influencing the production of merger, such as phonological environment, thereby reducing the need to obtain minimal pair lists. This talk argues for using Pillais as measures of merger by comparing results from low back merger in California English [Hall-Lew (2009)] with the analysis of front vowel merger by Hay et al. (2006).