Ethics

[dhicks]

The "ethics" section does not include any work by ethicists, other humanities scholars, or any members of groups who are especially harmfully impacted by applied data science — such as residents of neighborhoods policed using predictive policing algorithms.

O'Neil, Wallach, and Patil are all mathematicians/computer scientists. Wheeler is an anthropologist. I can't find a list of members of the ASA Committee; but this press release suggests that committee members are selected from the ASA membership, i.e., statisticians.

Last fall, O'Neil wrote an opinion piece that was published in the New York Times. The piece was sharply criticized by STS scholars and ethicists for ignoring the existence of longstanding fields of study devoted to exactly the topics that O'Neil claimed academia was neglecting.

In addition, all of these authors speak as data science agents — people who develop and use data science systems and models — rather than data science subjects — people whose lives are governed by data science systems and models. Compared to data science agents, data science subjects are much more likely to be poor, disabled, queer, and people of color. In the context of a small class taught in the statistics department at an elite university, the overall effect is that data science subjects are effectively excluded from the conversation. This reproduces the systemic, intersectional problems that data ethics is supposed to criticize.

[hadley]

Would you like to suggest some readings?

[dhicks]

I'm not really an expert on the topic. But I'm the closest thing my organization has at the moment to a "data science ethicist," so I've been paying some attention to several of the issues in this space. There are also several bibliographies of "data ethics" and "critical algorithm studies" in circulation. I got a few of the readings below from this one: https://socialmediacollective.org/reading-lists/critical-algorithm-studies/

Here are 6 papers or book chapters that will probably be accessible to data science/statistics/computer science students without much background in the humanities.

• Noble, Safiya. 2012. “Missed Connections: What Search Engines Say about Women. Bitch magazine, 12(4): 37-41. https://safiyaunoble.files.wordpress.com/2012/03/54_search_engines.pdf

• Winner, Langdon. “Do Artifacts Have Politics?” Daedalus 109, no. 1 (1980): 121–36. http://www.jstor.org/stable/20024652

• Leonelli, Sabina. “Locating Ethics in Data Science: Responsibility and Accountability in Global and Distributed Knowledge Production Systems.” Phil. Trans. R. Soc. A 374, no. 2083 (December 28, 2016): 20160122. https://doi.org/10.1098/rsta.2016.0122.

• Nissenbaum, Helen. 2001. “How Computer Systems Embody Values,” IEEE Computer, 120, 118-119. https://www.nyu.edu/projects/nissenbaum/papers/embodyvalues.pdf

• Vallor, Shannon. "Surveillance and the Examined Life." Chapter 7 of Technology and the Virtues: A Philosophical Guide to a Future Worth Wanting. New York, NY: Oxford University Press, 2016.

• Citron, Danielle Keats, and Frank A. Pasquale. 2014. “The Scored Society: Due Process for Automated Predictions.” Washington Law Review 89. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2376209

This list includes authors in philosophy, law, and the humanistic side of information science. At least one author (Noble) is a woman of color; 5 of the 7 authors are women. I think all of these papers speak either from a critical perspective, looking over the shoulder of data science agents, or explicitly from the perspective of data science subjects.

Read in the order given, the papers start with a concrete example of the problem (Noble); give three general perspectives /conceptual frameworks from Winner, Leonelli, and Nissenbaum; then come back to more concrete question of what individuals (Vallor) and institutions (Citron and Pasquale) should do about these issues.