Dissertations, Theses, and Capstone Projects

Date of Degree

9-2026

Document Type

Doctoral Dissertation

Degree Name

Doctor of Philosophy

Program

Linguistics

Advisor

Michael Newman

Committee Members

Cecelia Cutler

Bill Haddican

Christine Mallinson

Subject Categories

Anthropological Linguistics and Sociolinguistics | Phonetics and Phonology

Keywords

Baltimore, low back vowels, African Americans' English, North American English, race/ethnicity

Abstract

This dissertation is concerned with the low back vowels of Baltimore English: what low back systems occur, and how these systems have changed among Black and white Baltimoreans. I draw on two sources of data: 48 interviews that I conducted with natives of Baltimore and its inner suburbs, and 18 oral history interviews that were conducted from 1978 to 1981. In Chapter 3, I argue for the inclusion of the pre-rhotic sets START and NORTH in typologies of English low back vowels. With these sets included, several previously unattested low back systems are found in Baltimore. These are united by the independence of NORTH from THOUGHT, which sets them apart from other English low back systems. Instead of NORTH, START represents pre-rhotic THOUGHT in Baltimore. This leads to an account for the lesser degree of THOUGHT-raising in Baltimore: in other cities, THOUGHT raises as a generalization of NORTH raising to high mid position, but in Baltimore, THOUGHT only raises as high as START raises. Evidence is then presented that suggests a regional difference in the identity of START and NORTH: in the Southeastern region, pre-rhotic THOUGHT is START, whereas in the North, pre-rhotic THOUGHT is NORTH. I address the back vowel chain shift before /r/ and other mergers about the back pre-rhotic sets, arguing that the various patterns of chain shifting and merger observed in North American English trace back to a single origin: the backing of START.

The focus of Chapter 4 is the loss of the PALM vowel in Baltimore. For white speakers, it was primarily lost via LOT-PALM merger. Though MANOVA results initially suggest retention of a distinct PALM into the present century, this proves to be misleading. The youngest white speakers with a distinct PALM vowel were born in the 1960s. For Black speakers, the PALM vowel was primarily lost via PALM THOUGHT merger. However, this merger appears highly irregular, with different speakers using different vowels in the same words and the same speaker using different vowels in the same word across elicitation tasks. Similar irregularity appears among white speakers, and in all cases appears to be largely confined to the elicitation. I argue that this irregularity occurs because speakers with interracial social networks have access to two norms for the treatment with PALM—merger with LOT or merger with THOUGHT—and vary in which norm they adopt in the elicitation. Finally, I address the mechanism of merger. I argue that PALM THOUGHT merger is a merger by approximation, whereas LOT-PALM merger is a merger by expansion. This latter merger is triggered by the restoration of post-vocalic /r/ among white Baltimoreans, which weakens the vowel quality contrast between the LOT and PALM vowels. Finally, Chapter 5 investigates LOT THOUGHT merger, which appears among the youngest participants in the sample. Both Black and white Baltimoreans are moving towards merger, with the merger appearing most advanced among graduates from elite city public schools. I argue that the merger is led by externally oriented and upwardly mobile Baltimoreans. This argument is corroborated by three case studies. The most externally oriented speaker is most advanced in LOT-THOUGHT merger and in the Low Back Merger Shift and is less advanced in local features such as fronted GOOSE and GOAT. Conversely, speakers who are less upwardly mobile and more locally oriented are more conservative both in the Low Back Merger Shift and in the retention of features associated with Baltimore English. In this chapter I also provide an account for the maintenance of raised THOUGHT. Structurally, I argue that the retention of raised START favors the retention of raised THOUGHT. Socially,

I argue that stereotype avoidance in Baltimore is focused not on raised THOUGHT but on unrounded THOUGHT, a highly stigmatized variant found among Black Baltimoreans. This work points to the importance of /r/ in shaping developments among the low back vowels: non-rhoticity bolsters the LOT-PALM distinction, and shifts in pre-rhotic environments reshape the low back system when they generalize to other environments. It also shows structural changes can have butterfly effects on the linguistic system, as when the restoration of post-vocalic /r/ leads to LOT-PALM merger. Conversely, changes whose immediate cause is structural may have an indirect social motivation if they occur as a knock-on effect of a socially motivated change. Finally, the results here support prior findings on the role of social networks in the spread of linguistic features across racial lines and on the role of place and class identity in the maintenance of local features.

Download

Share

COinS