research

[title redacted] - under review

There is more than one prima facie plausible statistical criterion for evaluating whether classification algorithms are fair with respect to marginalized groups. These criteria are nevertheless incompatible with one another when applied under real world conditions. Given their mutual incompatibility, there is widespread debate over which one fairness criterion is correct. I argue, against the contemporary trend in this debate, that there is not one fairness criterion which ought to be used to the exclusion of others. Instead, multiple fairness criteria each correspond to different right-making features of decision-making involving socially salient groups. As a result, we should recognize the satisfaction of multiple criteria as separate pro tanto duties. My proposal both vindicates the considerations in favor of each individual criterion and solves a problem of triviality for monistic views about algorithmic fairness metrics.

[title redacted] - under review

I argue that domestic animals can have adaptive preferences. Adaptive preferences, as commonly understood in the feminist literature, are preferences which result from an entity’s experiences with oppression and are in violation of an entity’s basic welfare. This paper argues that feminist accounts of adaptive preferences can likewise be used to explain circumstances in which domestic animals face oppression and subsequently seem to express a preference for their continued subordination. Furthermore, the adaptive preferences of oppressed animals are often used to justify ongoing animal oppression. Contrary to these claims, I draw on feminist literature to show that animals’ apparent preferences are insufficient to justify systems or practices when we have reason to believe that these preferences are adaptive and thus caused by those same systems or practices.

The Moral Limits of Speculation

Intimate interpersonal speculation is the practice of hypothesizing about sensitive features of other people’s lives where the participants lack the requisite information to form a definite conclusion. In this paper, I define and distinguish speculation from related social phenomena, such as gossip, and give an account of how speculation can sometimes wrong its subjects. In some cases, I hold that speculation wrongs its subjects through violating their right to informational privacy. An intriguing upshot of my view is that it raises a challenge for existing conceptions of informational privacy, which tend to rely upon a person’s limited right to control the information others have about them. In my view, acts of speculation can violate the subject’s privacy even when they involve only public information about a person or even only information that has been knowingly and freely provided. As a result, accounts of privacy which place constraints upon a person’s access to information are inadequate. I provide an account of privacy which fills this explanatory gap. Ultimately, I argue that when speculation wrongs a person through invading their privacy, such speculation violates privacy because it manifests undue attention upon private features of their life.