Language variation and algorithmic bias: understanding algorithmic bias in British English automatic speech recognition
Open accessThe DOI matches its Crossref record; 32 of the 32 references it cites were found in a registry or an index.
Abstract
All language is characterised by variation which language users employ to construct complex social identities and express social meaning. Like other machine learning technologies, speech and language technologies (re)produce structural oppression when they perform worse for marginalised language communities. Using knowledge and theories from sociolinguistics, I explore why commercial automatic speech recognition systems and other language technologies perform significantly worse for already marginalised populations, such as second-language speakers and speakers of stigmatised varieties of English in the British Isles. Situating language technologies within the broader scholarship around algorithmic bias, consider the allocative and representational harms they can cause even (and perhaps especially) in systems which do not exhibit predictive bias, narrowly defined as differential performance between groups. This raises the question whether addressing or “fixing” this “bias” is actually always equivalent to mitigating the harms algorithmic systems can cause, in particular to marginalised communities.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Mapping the Margins: Intersectionality, Identity Politics, and Violence against Women of Color
Crenshaw, K. (1991) · Stanford Law Review · doi:10.2307/1229039 · Cited by 32,792
Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective
Haraway, D. (1988) · Feminist Studies · doi:10.2307/3178066 · Cited by 18,364
On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988
Gebru, T., Morgenstern, J., Vecchione, B., Vaughan, J. W., Wallach, H., Daumé, H., & Crawford, K. (2018) · arXiv · doi:10.48550/arxiv.1803.09010 · Cited by 2,474
Gebru, T., Morgenstern, J., Vecchione, B., Vaughan, J. W., Wallach, H., III, H. D., & Crawford, K. (2021) · Communications of the ACM · doi:10.1145/3458723 · Cited by 2,469 · cited 1×
“This includes where, how, when, why, about whom or what, by whom and for whom the dataset was compiled and, where relevant, who annotated it [52].”
BackgroundVariation and the indexical field 1
Eckert, P. (2008) · Journal of Sociolinguistics · doi:10.1111/j.1467-9841.2008.00374.x · Cited by 2,046
Unsettling race and language: Toward a raciolinguistic perspective
Rosa, J., & Flores, N. (2017) · Language in Society · doi:10.1017/s0047404517000562 · Cited by 1,675
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
Bender, E. M., & Friedman, B. (2018) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00041 · Cited by 854 · cited 1×
“Understanding datasets as infrastructure and pri-oritising good documentation is a first step to mitigating harms of biased systems as it allows us to anticipate them [15, 71, 99].”
BackgroundMitigating bias in algorithmic hiring: evaluating claims and practices
Raghavan, M., Barocas, S., Kleinberg, J., & Levy, K. (2020) · Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency · doi:10.1145/3351095.3372828 · Cited by 845
The Risk of Racial Bias in Hate Speech Detection
Sap, M., Card, D., Gabriel, S., Choi, Y., & Smith, N. A. (2019) · Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/p19-1163 · Cited by 817 · cited 4×
“They also show that White annotators are more likely than Black annotators to label non-toxic AAE tweets a “toxic” [106].”
BackgroundInfluentialThe State and Fate of Linguistic Diversity and Inclusion in the NLP World
Joshi, P., Santy, S., Budhiraja, A., Bali, K., & Choudhury, M. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.560 · Cited by 766
Racial 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
Measuring and Mitigating Unintended Bias in Text Classification
Dixon, L., Li, J., Sorensen, J., Thain, N., & Vasserman, L. (2018) · Proceedings of the 2018 AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3278721.3278729 · Cited by 674
Where fairness fails: data, algorithms, and the limits of antidiscrimination discourse
Hoffmann, A. L. (2019) · Information, Communication & Society · doi:10.1080/1369118x.2019.1573912 · Cited by 535 · cited 1×
“While the recent interest algorithmic oppression (or “bias”) in machine learning, including SLT applications, is a step in the right direction, there are fundamental limits to these discourses of bias and fairness [20, 56, 64].”
BackgroundEthical and social risks of harm from Language Models
Weidinger, L., Mellor, J., Rauh, M., Griffin, C., Uesato, J., Huang, P.-S., Cheng, M., Glaese, M., Balle, B., Kasirzadeh, A., et al. (2021) · arXiv · doi:10.48550/arxiv.2112.04359 · Cited by 477
Persistent Anti-Muslim Bias in Large Language Models
Abid, A., Farooqi, M., & Zou, J. (2021) · Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3461702.3462624 · Cited by 476 · cited 2×
“For example, the islamophobic predictions by GPT-3 documented in [1] are both harmful to any Muslim users who see the offensive output and feed into (and, of course, result from) much larger islamphobic discourses harming all Muslims.”
BackgroundRickford, J. R., & King, S. (2016) · Language · doi:10.1353/lan.2016.0078 · Cited by 374
Algorithmic Colonization of Africa
Birhane, A. (2020) · SCRIPTed: A Journal of Law, Technology & Society · doi:10.2966/scrip.170220.389 · Cited by 349 · cited 1×
“As I discuss below, it may well be that some communities do not want or need these technologies and in any case would like to be actively involved in their creation [22, 23, 65].”
BackgroundGender 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
Documenting Large Webtext Corpora: A Case Study on the Colossal Clean Crawled Corpus
Dodge, J., Sap, M., Marasović, A., Agnew, W., Ilharco, G., Groeneveld, D., Mitchell, M., & Gardner, M. (2021) · Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2021.emnlp-main.98 · Cited by 306 · cited 2×
“very direct way by “parroting” biases in the training data [16], for example islamophobic content [1, 43, 86].”
BackgroundHutchinson, B., Smart, A., Hanna, A., Denton, R., Greer, C., Kjartansson, O., Barnes, P., & Mitchell, M. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445918 · Cited by 279
Hargittai, E., & Shaw, A. (2015) · Information, Communication & Society · doi:10.1080/1369118x.2014.957711 · Cited by 257 · cited 1×
“The majority of Wikipedia editors are men [6, 58] due to (perceived) skills gaps [58], lack of leisure time and perceptions of the existing (misogynistic or antagonistic) culture on Wikipedia.”
BackgroundPredictive biases in natural language processing models: A conceptual framework and overview
Shah, D. S., Schwartz, H. A., & Hovy, D. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.468 · Cited by 241 · cited 2×
“shown in the case study of British English ASR below, SLTs can exhibit “predictive bias” [109], producing systematically higher error rates for some, usually marginalised, groups (e.”
BackgroundMoving beyond “algorithmic bias is a data problem”
Hooker, S. (2021) · Patterns · doi:10.1016/j.patter.2021.100241 · Cited by 181 · cited 4×
“3As [66] points out, the pervasiveness of “skewed” distributions in the real world, is one of the reasons why careful model development is crucial.”
BackgroundInfluentialDecolonising Speech and Language Technology
Bird, S. (2020) · Proceedings of the 28th International Conference on Computational Linguistics · doi:10.18653/v1/2020.coling-main.313 · Cited by 151
Language and discrimination: Generating meaning, perceiving identities, and discriminating outcomes
Craft, J. T., Wright, K. E., Weissler, R. E., & Queen, R. M. (2020) · Annual Review of Linguistics · doi:10.1146/annurev-linguistics-011718-011659 · Cited by 147 · cited 3×
“[5] conclude that women should avoid creaky voice to avoid discrimination and similar advice is often given to anyone who doesn’t speak the “standard variety”[35].”
BackgroundMethodUnderstanding and Mitigating Worker Biases in the Crowdsourced Collection of Subjective Judgments
Hube, C., Fetahu, B., & Gadiraju, U. (2019) · Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3290605.3300637 · Cited by 129 · cited 1×
“Other research on crowd-sourcing also highlights the importance of balancing annotators from diverse backgrounds and accounting for variation in values and opinions [12, 69].”
Background(2020) · doi:10.1145/3386296.3386301 · Cited by 107
A Framework for Understanding Sources of Harm throughout the Machine Learning Life Cycle
Suresh, H., & Guttag, J. (2021) · Equity and Access in Algorithms, Mechanisms, and Optimization · doi:10.1145/3465416.3483305 · Cited by 82
A Framework for Understanding Sources of Harm throughout the Machine Learning Life Cycle
Suresh, H., & Guttag, J. V. (2019) · arXiv · doi:10.48550/arxiv.1901.10002 · Cited by 82 · cited 6×
“the origins [116], types [109] and consequences [13] of algorithmic bias have been proposed.”
BackgroundInfluentialTo Live in Their Utopia: Why Algorithmic Systems Create Absurd Outcomes
Alkhatib, A. (2021) · Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3411764.3445740 · Cited by 81 · cited 2×
“As [4] notes, those who canmake themselves “legible””
Background
Cited by
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
“However, when technologies with performance disparities are used in consequential domains — such as in job application videos — degraded service can not only stigmatize users but also lead to other types of harm, such as allocative harms [140].”
BackgroundWorking at the Intersection of Race, Disability and Accessibility
Harrington, C. N., Desai, A., Lewis, A., Moharana, S., Ross, A. S., & Mankoff, J. (2023) · The 25th International ACM SIGACCESS Conference on Computers and Accessibility · doi:10.1145/3597638.3608389 · Cited by 53
“research point to the advantages for digital inclusion [54, 105], and highlight the potential of neglecting such perspectives.”
BackgroundMeryam, R., & Ali, F. (2026) · Convergence: The International Journal of Research into New Media Technologies · doi:10.1177/13548565261431143
Metadata
- Authors
- Markl, N.
- Venue
- 2022 ACM Conference on Fairness Accountability and Transparency
- Publisher
- ACM
- Published
- 2022-06-20
- Pages
- 521-534
- OpenAlex
- W4283161703
- Cited by
- 78
- References
- 32
- FWCI
- 10.79
- Type
- conference paper
- Language
- en
- Open access
- green
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.