Book Title: Computational Criminology: AI Applications in Forensic Science and Criminal Justice
Editors: Dr. Xavier Louis, Dr. Surbhi Girdhar, Ms. Aswathi Chandran Nair, Mr. Ravi Kumar, and Ms. Nandini Katare
Chapter: 4
DOI: https://doi.org/10.59646/704/4
Author: Shruthik Sharma
Abstract
The most consequential disciplinary divide in contemporary criminology is no longer between qualitative and quantitative research, between micro and macro analysis, or between left and right realism. It runs, instead, between two modes of inquiry that have developed in increasing isolation from one another: a tradition of formal criminological theory that has accumulated, over more than a century, a sophisticated vocabulary for describing the causal architecture of criminal behaviour, and a tradition of computational analysis that has accumulated, over the last three decades, a sophisticated technical vocabulary for detecting patterns in large administrative and digital datasets.[1]These two intellectual cultures share a subject matter but speak distinct languages, address distinct audiences, and reward distinct competencies. The risk, increasingly evident in the empirical literature, is that they will continue to drift apart producing, on one side, theoretical frameworks insufficiently disciplined by data, and on the other, predictive models insufficiently disciplined by theory. This chapter advances the central thesis of Part I: that the maturation of computational criminology depends on the systematic integration of these two traditions, not on the displacement of one by the other. The argument has both a defensive and a constructive dimension. Defensively, the chapter articulates the reasons why purely data-driven, theory-free approaches to criminal justice prediction are epistemically inadequate, ethically hazardous, and operationally unstable.
[1]David Garland, The Culture of Control: Crime and Social Order in Contemporary Society (University of Chicago Press, 2001) 4.