(Cat)egory mistake: the invalidity of animal shelter behavior assessments

Abstract

Animal shelters face diverse challenges, which often necessitate making life-or-death decisions for animals in their care. One strategy used to determine whether admitting, adoption, or euthanasia is appropriate is to assess an animal’s in-shelter or pre-admission behavior to infer its “personality.” Shelters do this because potential adopters are often interested in knowing an animal’s personality as it provides information about whether the animal will fit in their home. However, shelter behavior assessments are a broad topic. To narrow focus, I explore a relatively novel development: feline behavior assessments. These assessments suggest that shelter workers can make a valid inference from in-shelter behavior to long-term, consistent personality. I argue that assessments do not, and might not be able to, validly infer personality in shelters; I utilize recent philosophical work by Kaiser & Müller (Biol Philos 36:1–25, 2021) on personality to do so. I build on their work by showing how shelter assessments do not meet their criteria and fall victim to epistemic bias in privileging some behaviors. Because feline assessments fail to meet philosophically robust definitions of animal personality and have methodological biases, these assessments do not provide valid insight into a shelter animal’s personality and should not be used.