Homo Academicus & East Timor: Boundary-Maintenance and Erasure in the Social Sciences

Julian Torelli, McMaster University
Gregory Hooks, McMaster University

Did the social sciences ignore and erase Indonesia’s genocidal occupation of East Timor (1975-1999)? Did the social sciences display reflexivity in recognizing these atrocities and self-consciously adapt conceptual frameworks to shed light on them? Our research takes advantage of a large and comprehensive corpus of social science articles to document erasure and reflexivity. Erasure and reflexivity are prominent concerns, but they are difficult to examine empirically. As erasure refers to people, places, and topics that are overlooked, it can only be inferred by absences. As reflexivity, in Bourdieu’s framework, refers to overcoming “social and intellectual unconscious,” empirical documentation must consider both absence and presence of concepts and conceptualizations. We mobile artificial intelligence techniques (specifically, Latent Dirichlet Allocation (LDA) to identify topics in a large corpus of English-language social science articles (approximately 600,000). This corpus includes all articles indexed by JSTOR (1800 – 2013, metadata and full-text) in: anthropology, sociology, political science, and criminal justice. Taking advantage of this large and comprehensive corpus, we provide an empirically rich documentation of erasure and reflexivity. We filtered social science articles for mentions of East Timor (fewer than 1%) – and of this relatively small set, mentions of the atrocities that occurred there (fewer than half of the article discussing East Timor). Mobilizing AI and machine learning tools, we examine the pattern of topics found in social science articles that discuss East Timor. Despite the tumultuous and tragic changes that occurred in East Timor, we found a great deal of continuity in the topics that were examined in social science articles. Change does occur; but this change appears to be disciplinary boundary maintenance, not reflexivity. Machine learning and artificial intelligence (AI) can build compelling models of historical change; these tools can also document change and stasis in social scientific accounts of society.

See paper

 Presented in Session 29. Reflexivity and Complexity in Social Science History