---
title: "A Sociophonetic Analysis of Racial Bias in — Scott 2026"
description: "Commercial speech recognizers are known to transcribe African American speakers less accurately."
authors: ["Michael Scott", "Siyu Liang", "Alicia Wassink", "Gina-Anne Levow"]
year: 2026
venue: "Proceedings of the Fifteenth Language Resources and Evaluation Conference"
source: "https://aclanthology.org/2026.lrec-1.376/"
doi: "10.63317/4qkgjt7n9ehb"
licence: "CC BY-NC 4.0 (ELRA, LREC proceedings)"
licence_url: "https://creativecommons.org/licenses/by-nc/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/papers/scott-2026-sociophonetic-analysis-racial-bias-commercial/"
---
[Elamur Research](https://md.research.elamur.ai/) · [this page in HTML](https://research.elamur.ai/free/papers/scott-2026-sociophonetic-analysis-racial-bias-commercial/)

[home](https://md.research.elamur.ai/) / [papers](https://md.research.elamur.ai/free/papers/) / [2026](https://md.research.elamur.ai/free/papers/2026/) / A Sociophonetic Analysis of Racial Bias in Commercial ASR Systems Using the Pacific Northwest English Corpus

# A Sociophonetic Analysis of Racial Bias in Commercial ASR Systems Using the Pacific Northwest English Corpus

**Authors:** Michael Scott, Siyu Liang, Alicia Wassink, Gina-Anne Levow · **Year:** 2026 · **Venue:** Proceedings of the Fifteenth Language Resources and Evaluation Conference · **Licence:** [CC BY-NC 4.0 (ELRA, LREC proceedings)](https://creativecommons.org/licenses/by-nc/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.lrec-1.376/) · [PDF](https://aclanthology.org/2026.lrec-1.376.pdf) · [DOI 10.63317/4qkgjt7n9ehb](https://doi.org/10.63317/4qkgjt7n9ehb) · [OpenAlex](https://openalex.org/W7160027011)

**Citations:** 0 citations (openalex-snapshot · 2026-10-14)

## Summary
Commercial speech recognizers are known to transcribe African American speakers less accurately. Which pronunciation traits drive these gaps, however, has received little systematic study. The authors combine ASR evaluation with sociophonetic annotation from the Pacific Northwest English (PNWE) corpus to address this. They use the Word List task, in which speakers read target words inside a fixed carrier frame. The data come from 16 speakers in four communities: African American, Caucasian American, ChicanX and Yakama. Eleven dialect-related variables were marked in Praat TextGrids and cross-referenced with the output of four commercial ASR systems. Alongside word error rate, the study introduces a heuristic Phonetic Error Rate (PER) that links misrecognitions to particular phonological features. Errors concentrate around vowel quality. Speakers who resist the low-back merger or show non-mainstream pre-nasal /æ/ patterns are misrecognized more often, and African American speakers are hit hardest in every system tested. The authors conclude that acoustic modeling of dialectal phonetics, rather than vocabulary or grammar, is the main source of bias. They also position PNWE as a resource for fairness benchmarking.

## Key points
- The full PNWE corpus (Wassink, 2007) has 112 speakers from five ethnic groups in Washington State (68 Caucasian American, 19 Asian American, 12 ChicanX, 8 Yakama, 5 African American).
- The evaluation subset has 16 speakers (AA 3, CA 4, CX 4, YA 5), each producing about 750 Word List utterances. The Asian American group was left out because its error rates closely matched the Caucasian American group.
- Transcripts were produced with CLOx (a Microsoft-ASR-based tool), then hand-corrected in Praat TextGrids. A regional extension of the CMU Pronouncing Dictionary, PNWEdict, handles local pronunciation variants.
- The new Phonetic Error Rate (PER) metric connects errors to eleven sociophonetic variables. Low-back merger resistance and pre-nasal /æ/ variation show the strongest association with errors for African American speakers.
- The study builds on Koenecke et al. (2020), which reported average WER of 35% for African American versus 19% for Caucasian American speakers across commercial systems.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

Fields: [Linguistics & languages](https://md.research.elamur.ai/free/shelves/linguistics/)

## Cite

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{scott2026sociophonetic,
  title = {{A Sociophonetic Analysis of Racial Bias in Commercial ASR Systems Using the Pacific Northwest English Corpus}},
  author = {Scott, Michael and Liang, Siyu and Wassink, Alicia and Levow, Gina-Anne},
  booktitle = {Proceedings of the Fifteenth Language Resources and Evaluation Conference},
  year = {2026},
  doi = {10.63317/4qkgjt7n9ehb},
  url = {https://doi.org/10.63317/4qkgjt7n9ehb}
}
```

- [← previous A Recipe for Adapting Multilingual Embedders to OCR-Error Robustness and Historical Texts](https://md.research.elamur.ai/free/papers/michail-2026-recipe-adapting-multilingual-embedders-ocr/)
- [next → A Subword Embedding Approach for Variation Detection in Luxembourgish User Comments](https://md.research.elamur.ai/free/papers/lutgen-2026-subword-embedding-approach-variation-detection/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/papers/2026/)

## References

31 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Amazon. 2022. Amazon Transcribe. [find](https://scholar.google.com/scholar?q=Amazon%2E+2022%2E+Amazon+Transcribe%2E)
2. Apple. 2022. Apple Speech API. [find](https://scholar.google.com/scholar?q=Apple%2E+2022%2E+Apple+Speech+API%2E)
3. Paul Boersma. 2007. Praat: doing phonetics by computer. [http://www.praat](http://www.praat). org/. [↗](http://www.praat)
4. Kalvin Chang, Yi-Hui Chou, Jiatong Shi, Hsuan-Ming Chen, Nicole Holliday, Odette Scharenborg, and David R. Mortensen. 2024. Self-supervised Speech Representations Still Struggle with African American Vernacular English. ArXiv:2408.14262 [cs]. [↗](https://arxiv.org/abs/2408.14262)
5. Jay Cunningham, Su Lin Blodgett, Michael Madaio, Hal Daumé Iii, Christina Harrington, and Hanna Wallach. 2024. Understanding the Impacts of Language Technologies’ Performance Disparities on African American Language Speakers. In Findings of the Association for Computational Linguistics: ACL 2024, pages 12826–12833, Bangkok, Thailand. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Jay+Cunningham%2C+Su+Lin+Blodgett%2C+Michael+Madaio%2C+Hal+Daum%C3%A9+Iii%2C+Christina+Harrington%2C+and+Hanna+Wallach%2E+2024%2E+Understanding+the+Impacts+of+Language+Technologies%E2%80%99+Performance+Disparities+on+African+Am)
6. Rachel Dorn. 2019. Dialect-Specific Models for Automatic Speech Recognition of African American Vernacular English. In Proceedings of the Student Research Workshop Associated with RANLP 2019, pages 16–20, Varna, Bulgaria. IN-COMA Ltd. [find](https://scholar.google.com/scholar?q=Rachel+Dorn%2E+2019%2E+Dialect%2DSpecific+Models+for+Automatic+Speech+Recognition+of+African+American+Vernacular+English%2E+In+Proceedings+of+the+Student+Research+Workshop+Associated+with+RANLP+2019%2C+pages+16)
7. Google. 2022. Google Cloud Speech-to-Text. [find](https://scholar.google.com/scholar?q=Google%2E+2022%2E+Google+Cloud+Speech%2Dto%2DText%2E)
8. Lisa J. Green. 2002. African American English: A Linguistic Introduction. Cambridge University Press, Cambridge. [find](https://scholar.google.com/scholar?q=Lisa+J%2E+Green%2E+2002%2E+African+American+English%3A+A+Linguistic+Introduction%2E+Cambridge+University+Press%2C+Cambridge%2E)
9. Ke Hu, Antoine Bruguier, Tara N. Sainath, Rohit Prabhavalkar, and Golan Pundak. 2019. Phoneme-Based Contextualization for Cross-Lingual Speech Recognition in End-to-End Models. ArXiv:1906.09292 [cs]. [↗](https://arxiv.org/abs/1906.09292)
10. IBM. 2022. IBM Watson Speech to Text. [find](https://scholar.google.com/scholar?q=IBM%2E+2022%2E+IBM+Watson+Speech+to+Text%2E)
11. Alexander Johnson, Natarajan Balaji Shankar, Mari Ostendorf, and Abeer Alwan. 2024. An exploratory study on dialect density estimation for children and adult’s African American English. The Journal of the Acoustical Society of America, 155(4):2836–2848. [↗](https://doi.org/10.1121/10.0025771)
12. Allison Koenecke, Andrew Nam, Emily Lake, Joe Nudell, Minnie Quartey, Zion Mengesha, Connor Toups, John R. Rickford, Dan Jurafsky, and Sharad Goel. 2020. Racial disparities in automated speech recognition. Proceedings of the National Academy of Sciences, 117(14):7684– 7689. Publisher: Proceedings of the National Academy of Sciences. [↗](https://doi.org/10.1073/pnas.1915768117)
13. Changye Li, Trevor Cohen, and Serguei Pakhomov. 2024. Reexamining Racial Disparities in Automatic Speech Recognition Performance: The Role of Confounding by Provenance. ArXiv:2407.13982 [cs]. [↗](https://arxiv.org/abs/2407.13982)
14. Joshua L Martin and Kelly Elizabeth Wright. 2023. Bias in Automatic Speech Recognition: The Case of African American Language. Applied Linguistics, 44(4):613–630. [↗](https://doi.org/10.1093/applin/amac066)
15. Microsoft. 2022. Azure AI Speech. [find](https://scholar.google.com/scholar?q=Microsoft%2E+2022%2E+Azure+AI+Speech%2E)
16. NIST. 2021. SCTK, the NIST scoring toolkit. Original-date: 2016-05-03T19:00:47Z. [find](https://scholar.google.com/scholar?q=NIST%2E+2021%2E+SCTK%2C+the+NIST+scoring+toolkit%2E+Original%2Ddate%3A+2016%2D05%2D03T19%3A00%3A47Z%2E)
17. Rachael Tatman. 2017. Gender and Dialect Bias in YouTube’s Automatic Captions. In Proceedings of the First ACL Workshop on Ethics in Natural Language Processing, pages 53–59, Valencia, Spain. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Rachael+Tatman%2E+2017%2E+Gender+and+Dialect+Bias+in+YouTube%E2%80%99s+Automatic+Captions%2E+In+Proceedings+of+the+First+ACL+Workshop+on+Ethics+in+Natural+Language+Processing%2C+pages+53%E2%80%9359%2C+Valencia%2C+Spain%2E+Associat)
18. Rachael Tatman and Conner Kasten. 2017. Effects of Talker Dialect, Gender & Race on Accuracy of Bing Speech and YouTube Automatic Captions. pages 934–938. [find](https://scholar.google.com/scholar?q=Rachael+Tatman+and+Conner+Kasten%2E+2017%2E+Effects+of+Talker+Dialect%2C+Gender+%26+Race+on+Accuracy+of+Bing+Speech+and+YouTube+Automatic+Captions%2E+pages+934%E2%80%93938%2E)
19. Erik R. Thomas. 2007. Phonological and Phonetic Characteristics of African American Vernacular English. Language and Linguistics Compass, 1(5):450–475. [↗](https://doi.org/10.1111/j.1749-818x.2007.00029.x)
20. Siyin Wang, Chao-Han Huck Yang, Ji Wu, and Chao Zhang. 2024. Can Whisper perform speech-based in-context learning? ArXiv:2309.07081 [eess] version: 2. [↗](https://arxiv.org/abs/2309.07081)
21. Alicia Beckford Wassink, Cady Gansen, and Isabel Bartholomew. 2022. Uneven success: automatic speech recognition and ethnicity-related dialects. Speech Communication, 140:50–70. [↗](https://doi.org/10.1016/j.specom.2022.03.009)
22. Alicia Beckford Wassink, Rob Squizzero, Campion Fellin, and David Nichols. 2020. CLOx. [find](https://scholar.google.com/scholar?q=Alicia+Beckford+Wassink%2C+Rob+Squizzero%2C+Campion+Fellin%2C+and+David+Nichols%2E+2020%2E+CLOx%2E)
23. Walter A. Wolfram. 1969. A Sociolinguistic Description of Detroit Negro Speech. Urban Language Series, No. 5. Technical Report 5, Publications Section, Center for Applied Linguistics, 1717 Massachusetts Avenue, N. ERIC Number: ED028431. [find](https://scholar.google.com/scholar?q=Walter+A%2E+Wolfram%2E+1969%2E+A+Sociolinguistic+Description+of+Detroit+Negro+Speech%2E+Urban+Language+Series%2C+No%2E+5%2E+Technical+Report+5%2C+Publications+Section%2C+Center+for+Applied+Linguistics%2C+1717+Massachuset)
24. Yu Zhang, Wei Han, James Qin, Yongqiang Wang, Ankur Bapna, Zhehuai Chen, Nanxin Chen, Bo Li, Vera Axelrod, Gary Wang, Zhong Meng, Ke Hu, Andrew Rosenberg, Rohit Prabhavalkar, Daniel S. Park, Parisa Haghani, Jason Riesa, Ginger Perng, Hagen Soltau, Trevor Strohman, Bhuvana Ramabhadran, Tara Sainath, Pedro Moreno, Chung-Cheng Chiu, Johan Schalkwyk, Françoise Beaufays, and Yonghui Wu. 2023. Google USM: Scaling Automatic Speech Recognition Beyond 100 Languages. ArXiv:2303.01037 [cs]. 9. [↗](https://arxiv.org/abs/2303.01037)
25. Language Resource References Ardila, Rosana and Branson, Megan and Davis, Kelly and Henretty, Michael and Kohler, Michael and Meyer, Josh and Morais, Reuben and Saunders, Lindsay and Tyers, Francis M. and Weber, Gregor. 2020. Common Voice: A Massively-Multilingual Speech Corpus. Mozilla. Mozilla Foundation. Available at https://commonvoice.mozilla.org/. [↗](https://doi.org/10.48550/arxiv.1912.06670) [our card](https://md.research.elamur.ai/free/papers/ardila-2019-common-voice-massively-multilingual-speech/)
26. Carnegie Mellon University. 2015. The Carnegie Mellon Pronouncing Dictionary (CMUdict). Carnegie Mellon University. Carnegie Mellon University, 0.7b. Available at [https://github.com/cmusphinx/cmudict](https://github.com/cmusphinx/cmudict). [↗](https://github.com/cmusphinx/cmudict)
27. Kendall, Tyler and Farrington, Charlie. 2018. Corpus of Regional African American Language. University of Oregon. The Online Resources for African American Language Project. Available at [https://oraal.uoregon.edu/coraal/](https://oraal.uoregon.edu/coraal/). [↗](https://oraal.uoregon.edu/coraal/)
28. Meyer, Josh and Rauchenstein, Lindy and Eisenberg, Joshua D. and Howell, Nicholas. 2020. Artie Bias Corpus: An Open Dataset for Detecting Demographic Bias in Speech Applications. Artie Inc. Artie. Available at [https://github.com/artie-inc/artie-bias-corpus](https://github.com/artie-inc/artie-bias-corpus). [↗](https://github.com/artie-inc/artie-bias-corpus)
29. Olatunji, Tobi and Afonja, Tejumade and Yadavalli, Aditya and Emezue, Chris Chinenye and Singh, Sahib and Dossou, Bonaventure F. P. and Osuchukwu, Joanne and Osei, Salomey and Tonja, Atnafu Lambebo and Etori, Naome and Mbataku, Clinton. 2023. AfriSpeech-200: Pan-African Accented Speech Dataset for Clinical and General Domain ASR. Intron Health and collaborating institutions. Intron Health. 200 hours of Pan-African English speech from 120 indigenous accents across 13 countries. [↗](https://doi.org/10.1162/tacl_a_00627)
30. Wassink, Alicia Beckford. 2007. Pacific Northwest English Corpus. University of Washington Sociolinguistics Laboratory. University of Washington, Department of Linguistics. Funded by the National Science Foundation. Available at [https://zeos.ling.washington.edu/PNWEnglish/index.php](https://zeos.ling.washington.edu/PNWEnglish/index.php). [↗](https://zeos.ling.washington.edu/PNWEnglish/index.php)
31. Weinberger, Steven. 2013. Speech Accent Archive. George Mason University. George Mason University. Available at [http://accent.gmu.edu/](http://accent.gmu.edu/). [↗](http://accent.gmu.edu/)