Machine Learning, Abduction, and Computational Ethnography

Abstract

Abstract This chapter shows how interpretive work can reveal sociological discoveries in analyses using machine learning (ML). It proposes integrating ethnographic and quantitative research around a common concern with abduction. The chapter’s main section develops this computational ethnography approach across a series of examples, including computational analyses of historical and contemporary corpora, classic and recent network-analytic studies, and some of the newest ML techniques that are geared toward detecting sources of inequality. Despite technical and substantive differences, the studies’ empirical designs involve settings that resemble fields, in the ethnographic sense, more than representative samples and that reveal local meaning and locally meaningful actions. The chapter argues that this era’s challenges and opportunities call on researchers to cultivate and integrate technical skills, critical reflection on technological infrastructures, and sensitivity to surprising observations in light of existing theories.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 2 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.