“I don’t Think These Devices are Very Culturally Sensitive.”—Impact of Automated Speech Recognition Errors on African Americans
Open accessAbstract
Automated speech recognition (ASR) converts language into text and is used across a variety of applications to assist us in everyday life, from powering virtual assistants, natural language conversations, to enabling dictation services. While recent work suggests that there are racial disparities in the performance of ASR systems for speakers of African American Vernacular English, little is known about the psychological and experiential effects of these failures paper provides a detailed examination of the behavioral and psychological consequences of ASR voice errors and the difficulty African American users have with getting their intents recognized. The results demonstrate that ASR failures have a negative, detrimental impact on African American users. Specifically, African Americans feel othered when using technology powered by ASR—errors surface thoughts about identity, namely about race and geographic location—leaving them feeling that the technology was not made for them. As a result, African Americans accommodate their speech to have better success with the technology. We incorporate the insights and lessons learned from sociolinguistics in our suggestions for linguistically responsive ways to build more inclusive voice systems that consider African American users’ needs, attitudes, and speech patterns. Our findings suggest that the use of a diary study can enable researchers to best understand the experiences and needs of communities who are often misunderstood by ASR. We argue this methodological framework could enable researchers who are concerned with fairness in AI to better capture the needs of all speakers who are traditionally misheard by voice-activated, artificially intelligent (voice-AI) digital systems.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
A study of automated speech recognition showed that African American users had difficulty controlling voice-activated technologies unless they accommodated their speech patterns.
“For example, a study of automated speech recognition showed that African American users had difficulty controlling voice-activated technologies unless they accommodated their speech patterns (Mengesha et al. 2021).”
From Epistemic Injustice in Generative AI (2024) · cites this paper as Mengesha et al. 2021
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Dissecting racial bias in an algorithm used to manage the health of populations
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019) · Science · doi:10.1126/science.aax2342 · Cited by 7,137 · see citing works
Harrell, S. P. (2000) · American Journal of Orthopsychiatry · doi:10.1037/h0087722 · Cited by 1,589 · see citing works
Language (Technology) is Power: A Critical Survey of “Bias” in NLP
Blodgett, S. L., Barocas, S., Daumé III, H., & Wallach, H. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.485 · Cited by 1,050 · cited 3× · see citing works
“…techniques for measuring or mitigating ‘bias’ are poorly matched to their motivations and do not engage with the relevant literature outside of NLP” (Blodgett et al., 2020, p. 1), illuminating the need for speech recognition researchers to look to literature in fields where the relationship…”
BackgroundMethodInfluentialRacial disparities in automated speech recognition
Koenecke, A., Nam, A., Lake, E., Nudell, J., Quartey, M., Mengesha, Z., Toups, C., Rickford, J. R., Jurafsky, D., & Goel, S. (2020) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1915768117 · Cited by 765 · see citing works
Self-citationRickford, J. R., & King, S. (2016) · Language · doi:10.1353/lan.2016.0078 · Cited by 374 · see citing works
Gender and Dialect Bias in YouTube’s Automatic Captions
Tatman, R. (2017) · Proceedings of the First ACL Workshop on Ethics in Natural Language Processing · doi:10.18653/v1/w17-1606 · Cited by 338 · see citing works
Discrimination through Optimization
Ali, M., Sapiezynski, P., Bogen, M., Korolova, A., Mislove, A., & Rieke, A. (2019) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3359301 · Cited by 268 · see citing works
Reactions to racial discrimination: Emotional stress and help-seeking behaviors.
Carter, R. T., & Forsyth, J. (2010) · Psychological Trauma: Theory, Research, Practice, and Policy · doi:10.1037/a0020102 · Cited by 190 · see citing works
Discrimination through optimization: How Facebook's ad delivery can lead to skewed outcomes
Ali, M., Sapiezynski, P., Bogen, M., Korolova, A., Mislove, A., & Rieke, A. (2019) · arXiv · doi:10.48550/arxiv.1904.02095 · Cited by 72 · see citing works
Flagged first, then by citations
Research works printed but not linked: 22
- Baugh J. (2003). Linguistic Profiling, In A. Ball, S. Makoni, G. Smitherman, and A. K. Spears (Eds.), Black Linguistics: Language, Society and Politics in Africa and the Americas (pp. 167–180). New York, NY: Routledge.
- Blodgett S. L. Green L. O'Connor B. (2016). “ Natural Language Processing,“ in EMNLP 2016: Conference on Empirical Methods, Austin, TX. Available at: https://www.aclweb.org/portal/content/emnlp-2016-conference-empirical-methods-natural-language-processing.
- Brody G. H. Chen Y.-F. Murry V. M. Ge X. Simons R. L. Gibbons F. X. (2006). Perceived Discrimination and the Adjustment of African American Youths: A Five-Year Longitudinal Analysis with Contextual Moderation Effects. Child. Development 77 (5), 1170– 1189. 10.1111/j.1467-8624.2006.00927.x 16999791 · DOI · Paper page
- Buolamwini J. Gebru T. (2018). “ Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification,” in Conference on fairness, accountability and transparency, New York, NY, February 23 and 24, 2018 (PMLR), 77– 91.
- Carter S. Mankoff J. (2005). “ When Participants Do the Capturing: the Role of media in Diary Studies,” in Proceedings of the SIGCHI conference on Human factors in computing systems, Portland, Oregon, USA, April 2-7, 2005, 899– 908.
- Clark R. Anderson N. B. Clark V. R. Williams D. R. (1999). Racism as a Stressor for African Americans: A Biopsychosocial Model. Am. Psychol. 54 (10), 805– 816. 10.1037/0003-066x.54.10.805 10540593 · DOI · Paper page
- Coughlin S. S. Smith S. A. (2017). Community-based Participatory Research to Promote Healthy Diet and Nutrition and Prevent and Control Obesity Among African-Americans: A Literature Review. J. Racial Ethnic Health Disparities 4 (2), 259– 268. 10.1007/s40615-016-0225-0 PMC5055839 27059053 · DOI
- Ellen Barnidge E. Elizabeth A. Baker E. A. Freda Motton F. Frank Rose F. Teresa Fitzgerald T. (2010). A Participatory Method to Identify Root Determinants of Health: The Heart of the Matter. Prog. Community Health partnerships: Res. Educ. Action. 4 (1), 55– 63. 10.1353/cpr.0.0105 PMC2866451 20364079 · DOI
- Grogger J. (2011). Speech Patterns and Racial Wage Inequality. J. Hum. Resour. 46 (1), 1– 25. 10.1353/jhr.2011.0017 · DOI
- Jones T. Kalbfeld J. R. Hancock R. Clark R. (2019). Testifying while Black: An Experimental Study of Court Reporter Accuracy in Transcription of African American English. Language 95 (2), e216– e252. 10.1353/lan.2019.0042 · DOI
- King S. (2020). From African American Vernacular English to African American Language: Rethinking the Study of Race and Language in African Americans' Speech. Annu. Rev. Linguist. 6, 285– 300. 10.1146/annurev-linguistics-011619-030556 · DOI
- Nelson A. (2002). Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. J. Natl. Med. Assoc. 94 (8), 666– 668. 10.17226/12875 12152921 PMC2594273 · DOI
- Purnell T. Idsardi W. Baugh J. (1999). Perceptual and Phonetic Experiments on American English Dialect Identification. J. Lang. Soc. Psychol. 18 (1), 10– 30. 10.1177/0261927x99018001002 · DOI
- Redmiles E. M. Kross S. Mazurek M. L. (2017). “ Where Is the Digital divide? a Survey of Security, Privacy, and Socioeconomics,” in Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems, Denver, CO, May 6–11, 2017, 931– 936.
- Rickford J. R. (1999b). The Ebonics Controversy in My Backyard: a Sociolinguist's Experiences and Reflections. J. Sociolinguistics 3 (2), 267– 266. 10.1111/1467-9481.00076 · DOI
- Seaton E. K. Caldwell C. H. Sellers R. M. Jackson J. S. (2008). The Prevalence of Perceived Discrimination Among African American and Caribbean Black Youth. Dev. Psychol. 44 (5), 1288– 1297. 10.1037/a0012747 18793063 PMC2556985 · DOI · Paper page
- Su J. Duh K. Carreras X. (2016). “ November)Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing,” in Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, Austin, TX, November 1–5, 2016.
- Topaz M. Schaffer A. Lai K. H. Korach Z. T. Einbinder J. Zhou L. (2018). Medical Malpractice Trends: Errors in Automated Speech Recognition. J. Med. Syst. 42 (8), 153– 154. 10.1007/s10916-018-1011-9 29987660 · DOI
- Vaughn L. M. Jacquez F. Lindquist-Grantz R. Parsons A. Melink K. (2017). Immigrants as Research Partners: a Review of Immigrants in Community-Based Participatory Research (CBPR). J. Immigrant Minor. Health 19 (6), 1457– 1468. 10.1007/s10903-016-0474-3 27491305 · DOI
- Vetulani Z. Marcinak J. Obrębski J. Vetulani G. Dabrowski A. Kubis M. (2010). “ Zasoby językowe I Technologie Przetwarzania Tekstu. POLINT-112-SMS Jako Przykład Aplikacji Z Zakresu Bezpieczeństwa Publicznego (in Polish),” in Language Resources and Text Processing Technologies. POLINT-112-SMS as Example of Homeland Security Oriented Application (Poznań: Adam Mickiewicz University Press)
- Williams F. Whitehead J. L. Miller L. M. (1971). Ethnic Stereotyping and Judgments of Children's Speech. Speech Monogr. 38 (3), 166– 170. 10.1080/03637757109375706 · DOI
- Winnick M. (2012). “ dScout,” in Ethnographic Praxis in Industry Conference Proceedings, Savannah College of Art and Design, October 14–17, 2012 (Oxford, United Kingdom: Blackwell Publishing Ltd), 378. 10.1111/j.1559-8918.2012.00058.x · DOI
Cited by
Sorted by citations
Sociotechnical Harms of Algorithmic Systems: Scoping a Taxonomy for Harm Reduction
Shelby, R., Rismani, S., Henne, K., Moon, A., Rostamzadeh, N., Nicholas, P., Yilla-Akbari, N., Gallegos, J., Smart, A., Garcia, E., et al. (2023) · Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3600211.3604673 · Cited by 279 · see citing works
Communication accommodation theory: Past accomplishments, current trends, and future prospects
Giles, H., Edwards, A. L., & Walther, J. B. (2023) · Language Sciences · doi:10.1016/j.langsci.2023.101571 · Cited by 152 · see citing works
Kurian, N. (2025) · Contemporary Issues in Early Childhood · doi:10.1177/14639491231206004 · Cited by 117 · see citing works
Epistemic Injustice in Generative AI
Kay, J., Kasirzadeh, A., & Mohamed, S. (2024) · Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1609/aies.v7i1.31671 · Cited by 76 · see citing works
The Role of Inclusion, Control, and Ownership in Workplace AI-Mediated Communication
Kadoma, K., Aubin Le Quere, M., Fu, X. J., Munsch, C., Metaxa, D., & Naaman, M. (2024) · Proceedings of the CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642650 · Cited by 24 · see citing works
Challenges in Automatic Speech Recognition for Adults with Cognitive Impairment
Cohn, M., Lanzi, A., Ishihara, Y., Chuah, C.-N., Zellou, G., & Weakley, A. (2026) · Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3772318.3791581 · Cited by 4 · see citing works
“First, ASR errors meaningfully shape users’ interactions with voice-based AgeTech systems, impacting trust [12], satisfaction [61], and adoption [47].”
- Authors
- Mengesha, Z., Heldreth, C., Lahav, M., Sublewski, J., & Tuennerman, E.
- Affiliations
- Google (United States) (US)
- Corresponding author
- Mengesha, Zion
- Venue
- Frontiers in Artificial Intelligence
- Publisher
- Frontiers Media SA
- Published
- 2021-11-26
- Volume
- 4
- Pages
- 725911-725911
- ISSN
- 2624-8212 (ISSN-L)
- PMC
- PMC8664002
- OpenAlex
- W3215236363
- Cited by
- 140
- References
- 9
- Citations by year
- 2026: 43 · 2025: 38 · 2024: 24 · 2023: 29 · 2022: 6
- FWCI
- 17.01
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Frontiers in Artificial Intelligence (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Social Sciences › Safety Research › Ethics and Social Impacts of AI
- Social Sciences › Linguistics and Language › Linguistic Variation and Morphology
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Keywords
- speech recognition, African American English, othering, identity, speech accommodation, voice assistants, sociolinguistics, algorithmic fairness, user experience, diary study
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
