Race, Classification & AI

I work on questions of racial classification, colonialism and technologies of state violence from the 18th to the 21st centuries in Germany, France, Japan and the US.

My dissertation, “Making the Master Race: Germany, Japan, and the Rise and Fall of Racial States,” mapped how states translate scientific, political and folk ideas about race into practice in the realms of citizenship and welfare policies in authoritarian regimes. My current book project, “Racial Vision: Failed Projects of Human Difference,” carries forward my concern with race, technology and epistemology to study how scientists and other political actors translate their racial projects and perceptions into facial recognition technologies in ways that enshrine and naturalize incompatible forms of racial categorization.

My work has been published in Qualitative Sociology, the American Journal of Sociology, and several other venues. It has among others won best article awards from the ASA, the American Journal of Sociology, the Eastern Sociological Association, the Society for the Advancement of Socio-Economics, and the École des Hautes Études en Sciences Sociales.

My methodological interests span qualitative and computational methods: original-language archival and text analysis; interviews; translation; and computer vision. My research has been supported by Oxford University, the Japan Foundation, the NOMIS foundation, the Max Weber Foundation, the Leibniz Association, the HONJO Japanese American Association, the American Sociological Association and through various internal grants at NYU, Harvard, the Social Science Center Berlin and CUNY Queens College.

If you expand the menu on the right, you will find details on research & links to papers. Thanks for visiting!

  • “Racial Vision: Failed Projects of Human Difference” is a cultural historical investigation into the pre-histories of racialized facial recognition technologies. The qualitative and computational investigation of racial recognition begins with an analysis of 18th century French doctors studying “racial metamorphoses” and ends with 21st century Chinese computer vision engineers trying to remedy racial bias in machine learning datasets.

  • How does racial classification emerge in contexts where it previously did not exist? And how does racialization happen within Whiteness and within Asianness?

    "Horror Vacui: Racial Misalignment, Symbolic Repair, and Imperial Legitimation in German National Socialist Portrait Photography." American Journal of Sociology 129 (2): 313-383.

    *Charles Tilly Award, Comparative Historical Section of the American Sociological Association (2025)

    *Social Sciences Research Award, CUNY Queens College, 2025

    *2024 Roger V. Gould Prize, American Journal of Sociology

    *Political Sociology Article/Chapter Award, American Sociological Association (2024)

    *Global and Transnational Section Best Article Award (Honorable Mention), American Sociological Association (2024)

    *Clifford Geertz Prize in the Sociology of Culture (Honorable Mention), American Sociological Association (2024)

    Link here. If you do not have institutional access, please feel free to email me for a preprint at Anna.skarpelis ((at)) qc.cuny.edu. A totally free preprint (without accurate pagination) is available here: https://osf.io/preprints/socarxiv/8vk9r.

    Paper 2, R&R with American Sociological Review.

  • Things that look the same are often rather different. Hidden variation is extremely consequential for our understanding of social mechanisms and processes. Three of my papers look at things that appear the same on the surface, but disguise fundamentally different interests, technologies, and processes.

    “Race’s Many Metamorphoses: Racial Vision from Skin Tone Studies to Artificial Intelligence”

    “Race in Parentheses: Historical Legacies in the Production of Racial Absence”

    “What is it like to be a Nazi? Racial Vision and Scientific Selves in German Portrait Photographic Practice” (2020, Link to Preprint)

  • My interest in computational sociology is twofold: First, in establishing meaning in images and videos. Second, in devising methods to better capture perception across the board of sociological methods (ethnography, experiments, vision, etc.). I learned about Computer Vision at Harvard and MIT, and about art historical ways of seeing at Eikones.

    “From Archives to Algorithms: Distance, Evidence, and Inference” (forthcoming). Sociologica

    “The Moral Pixel” (2022, Working Paper)

    “Erasing Hitler” (2022, Working Paper)

    “Consider the Scallop (Scallop Gallop)” (2020) asks fundamental questions about the nature of perception. My first virtual reality exhibition, it was shown at the MIT Museum Studio between December 2020 - April 2021, as part of a group exhibition called “Total Internal Reflection.”

    “Untitled” (2020, ongoing), collaborative project with Fiona Rose Greenland. We track the development of “untitled” works of art in art price and visual content over the 20th century.

  • My main interests in qualitative work center on questions of classification, epistemology, and translation.

    “Life on File: Archival Epistemology and Theory” (2020) provides a framework for sociological analysis in and through archives. The article is part of a special issue “Archival Work as Qualitative Sociology” in the Journal Qualitative Sociology. It has also been translated into Portuguese.

    “Race in Translation” (2021-) is an ongoing project with Elizabeth Onasch on how social scientists who work on race & ethnicity navigate the fraught linguistic terrain of translation.

    “Hounded by Nuance, or: Freeing the B*tch from the Iron Cage” (2020, Working Paper).

  • I enjoy reviewing books, doing public policy work and writing handbook chapters. I have written on minimum wage regulation, the politics of demographic change, and the rise of fascism.

    Please see the publications section on my CV for links to the full texts of these publications.