About

Margaret E. L. Renwick is an Associate Research Professor and the Director of Undergraduate Studies in the Department of Cognitive Science at Johns Hopkins University. Her research interests include sociophonetics, laboratory phonology, phonetics, speech acoustics, and linguistics. She focuses on how speakers manipulate fine phonetic detail in patterned ways and how listeners extract multiple layers of linguistic and social meaning from speech signals. Her work incorporates variation into models of spoken language to explore the nature of phonological contrast, the origins and realization of phonological patterns, regional accents of US English, and linguistic change across generations. Renwick has a background that includes a PhD from Cornell University in Cognitive Science. She has previously held a tenured position in Linguistics at the University of Georgia and was a member of the Phonetics Laboratory at the University of Oxford. Her research on Southern accents has been featured in prominent publications such as the Washington Post and the Wall Street Journal. She is actively involved in teaching and mentoring graduate students, offering courses on dialects of English, acoustic phonetics, and sociophonetics, and she has co-edited works on the phonetics and phonology of contrast.

Research topics

• Computer Science
• Linguistics
• Natural Language Processing
• Geography
• Sociology
• History
• Acoustics
• Meteorology
• Speech recognition
• Psychology
• Physics
• Astrophysics
• Philosophy
• Demography
• Archaeology
• Mathematics

Selected publications

• Reconsidering the mid-vowel system of Parisian French:

  Isogloss Open Journal of Romance Linguistics · 2025-07-16

  articleOpen accessSenior author

  The French mid vowel system exhibits a complex marginal phonological contrast. The distributions of three pairs of vowels (/e, ɛ/; /o, ɔ/; /ø, œ/) are neither completely contrastive nor allophonic, and their heights vary, leading to overlapping phonetic realizations. This paper explores the phonetic and phonological factors that determine mid vowel height in Parisian French. We especially explore the relative contributions of phonological grammar, including the tension between contrast maintenance and positional neutralization based on syllable structure alongside the phonetic factor of vowel duration. Results are formalized in Maximum Entropy Grammar (Goldwater & Johnson 2003) with constraint scaling (Coetzee & Kawahara 2013) to model the interplay of phonetics and phonology. We find that speakers of Parisian French do tend to realize these vowels more faithfully than in varieties in which the loi de position is consistently productive, but realizations are still highly variable, and although an effect of duration was observed, we find that vowel length may not play a role in helping speakers distinguish high- and low-mid vowels.

  PublisherOA PDFDOI

• Statistics in Phonetics

  Annual Review of Statistics and Its Application · 2024-10-01 · 4 citations

  articleOpen access

  Phonetics is the scientific field concerned with the study of how speech is produced, heard, and perceived. It abounds with data, such as acoustic speech recordings, neuroimaging data, and articulatory data. In this article, we provide an introduction to different areas of phonetics (acoustic phonetics, sociophonetics, speech perception, articulatory phonetics, speech inversion, sound change, and speech technology), an overview of the statistical methods for analyzing their data, and an introduction to the signal processing methods commonly applied to speech recordings. A major transition in the statistical modeling of phonetic data has been the shift from fixed effects to random effects regression models, the modeling of curve data (for instance, via generalized additive mixed models or functional data analysis methods), and the use of Bayesian methods. This shift has been driven in part by the increased focus on large speech corpora in phonetics, which has arisen from machine learning methods such as forced alignment. We conclude by identifying opportunities for future research.

  PublisherDOI

• Phonological Feature Detection for US English using the Phonet Library

  2024-09-01 · 1 citations

  articleSenior author
  PublisherDOI

• 1. Language Change at the Intersections of Movement, Economy, and Orientation

  Publication of the American Dialect Society · 2024-12-01 · 3 citations

  articleOpen access

  A growing amount of research on North American English varieties shows that many sociolinguistic variables peaked among the Baby Boomer Generation (born approximately 1946-64).By "peak," we mean that a linguistic phenomenon in one community has reached its point of greatest intensity and is maximally divergent from realizations of that same variable in most other North American communities.Similarly, we also find studies that describe a "cliff" beginning among Generation X (born approximately 1965-82) such that after generations of relative stability, local dialect features are lost.Nesbitt (2021, 359) observes that "the Baby Boomer-Gen X transition appears to be a pivotal transition throughout the country [and] appears to be of great importance to North American English dialects in general."In this volume, we present studies that illuminate the nature of these changes.The eight studies contained here discuss changes in phonological, grammatical, and lexical variation among White Americans, Black Americans, Irish Catholics, and Yats from communities in New York, North Carolina, Georgia, Louisiana, Pennsylvania, Missouri, and California.More examples can be found in other phenomena, communities, and regions, and we lament that we could not include them all here.However, this volume represents the first themed discussion of such phenomena, and we look forward to further discussion around these topics.We are not proposing that all communities underwent linguistic change in the Baby Boomer to Gen X transition, and we purposely sought out cases that go against that pattern.We feel that including such counterexamples is useful because they circumscribe not the variable context ( la Poplack and Tagliamonte 1989, 60) but rather the predictor context, or help define the range of influence any one predictor variable has on a linguistic variable.They also highlight communities that go against or resist trends in mainstream North American English.It is worth exploring why such communities did or did not change together with many other communities, to deepen our understanding of the kind of language-external factors that drive language change.

  PublisherOA PDFDOI

• Evaluating Italian Vowel Variation with the Recurrent Neural Network Phonet

  2024-09-01

  articleSenior author
  PublisherDOI

• Vowel Harmony in Romance Languages

  Oxford University Press eBooks · 2024-10-22

  book-chapterOpen accessSenior author

  Abstract This chapter presents the main features of vowel harmony (VH) patterns in Romance. Virtually all Romance harmonies are unidirectional—that is, they are right-to-left or left-to-right processes. Furthermore, in most Romance vowel harmonies the stress-bearing vowel is either the target (most usually assimilating to the final vowel, a process known as ‘metaphony’ in Romance linguistics) or the trigger; even when this is not the case, the boundaries of the process are prosodically defined. In some VH processes, there can be intervening transparent vowels between the target and the trigger. Most harmonies displayed by Romance varieties involve vowel height, though there are some instances of rounding or backness assimilation in Swiss Lombard, Valencian Catalan, and Romanian, as well as cases of centralization (or laxing) in Canadian French, Cantabrian, and central Sicilian, and complete vowel copy in Italo-Romance.

  PublisherOA PDFDOI

• Statistics in Phonetics

  arXiv (Cornell University) · 2024-04-11 · 1 citations

  preprintOpen access

  Phonetics is the scientific field concerned with the study of how speech is produced, heard and perceived. It abounds with data, such as acoustic speech recordings, neuroimaging data, or articulatory data. In this paper, we provide an introduction to different areas of phonetics (acoustic phonetics, sociophonetics, speech perception, articulatory phonetics, speech inversion, sound change, and speech technology), an overview of the statistical methods for analyzing their data, and an introduction to the signal processing methods commonly applied to speech recordings. A major transition in the statistical modeling of phonetic data has been the shift from fixed effects to random effects regression models, the modeling of curve data (for instance via GAMMs or FDA methods), and the use of Bayesian methods. This shift has been driven in part by the increased focus on large speech corpora in phonetics, which has been driven by machine learning methods such as forced alignment. We conclude by identifying opportunities for future research.

  PublisherOA PDFDOI

• Robustness and Complexity in Italian Mid Vowel Contrasts

  Languages · 2024-04-18 · 1 citations

  articleOpen access1st authorCorresponding

  Accounts of phonological contrast traditionally invoke a binary distinction between unpredictable lexically stored phonemes and contextually predictable allophones, whose patterning reveals speakers’ knowledge about their native language. This paper explores the complexity of contrasts among Italian mid vowels from a multifaceted perspective considering the lexicon, linguistic structure, usage, and regional variety. The Italian mid vowels are marginally contrastive due to a scarcity of minimal pairs alongside variation in phonetic realization. The analysis considers corpus data, which indicate that the marginal contrasts among front vowels vs. back vowels are driven by different sources and forces. Functional loads are low; while front /e ɛ/ have the weakest lexical contrast among all Italian vowels, back /o ɔ/ are separated by somewhat more minimal pairs. Among stressed front vowels, height is predicted by syllable structure and is context-dependent in some Italian varieties. Meanwhile, the height of back mid vowels is predicted by lexical frequency, in line with expectations of phonetic reduction in high-frequency contexts. For both front and back vowels, the phonetic factor of duration predicts vowel height, especially in closed syllables, suggesting its use for contrast enhancement. The results have implications for a proposed formalization of Italian mid vowel variation.

  PublisherOA PDFDOI

• Comparing Kaldi-Based Pipeline Elpis and Whisper for Čakavian Transcription

  2024-01-01

  articleOpen access
  PublisherOA PDFDOI

• 5. Demographic Change, Migration, and the African American Vowel System in Georgia

  Publication of the American Dialect Society · 2024-12-01 · 1 citations

  article

  The theme of this volume is generational trends in language change, with an emphasis on the Baby Boomer to Gen X transition as an important period of time. Many studies have found sharp changes in vowel systems in communities after the Baby Boomer Generation on the East Coast (Thiel and Dinkin 2020) and in the Midwest (Nesbitt 2021). Though we know a great deal about the geographic consistency of the Baby Boomer to Gen X transition, in all of these cases, the focus has been on White speakers and regional varieties. Besides the general importance of including a breadth of social backgrounds in theorizing, drawing solely from White regional varieties particularly causes problems when extending generational trends to varieties of African American Language (AAL). For some time, AAL was thought by linguists to be relatively uniform across regions (Wolfram 2007; King 2020), due to the finding of a relatively consistent set of features. Challenges to the idea of homogeneity in vowel systems arose in the early twenty-first century, culminating in a series of studies looking at regional differences between Black and White speakers’ vowel systems (Yaeger-Dror and Thomas 2010). Findings showed that varieties of AAL differed across regions in terms of adoption of novel sound changes, and further work has found there is a great deal of heterogeneity in the adoption of regional sound changes within communities as well (King 2018). This level of variability makes it especially important to include varieties of AAL in the theorization of generational shift across the United States. To help remedy this issue, we use data from Georgia AAL to look for any generational trends that may appear and to compare them with patterns found elsewhere.

  PublisherDOI

Frequent coauthors

• Joseph A. Stanley

  11 shared

• Rachel M. Olsen

  University of Georgia

  8 shared

• John Coleman

  Children's Health Ireland at Crumlin

  7 shared

• Lori Lamel

  Laboratoire Interdisciplinaire des Sciences du Numérique

  6 shared

• Ioana Vasilescu

  Laboratoire Interdisciplinaire des Sciences du Numérique

  6 shared

• Camille Dutrey

  Laboratoire de Phonétique et Phonologie

  4 shared

• Michael L. Olsen

  University of Georgia

  4 shared

• Michele Gubian

  Ludwig-Maximilians-Universität München

  3 shared