This is a preprint of a manuscript submitted as a doctoral dissertation to the Department of Psychology at New York University. The dissertation is in stapled form, where a broad introduction and discussion explain the theoretical trajectory of research published elsewhere (peer-reviewed journals, and preprint form). Each chapter references these external manuscripts for the reader. ABSTRACT: In order to efficiently navigate our social world, humans sort one another along dimensions and categories intended to reflect the structure of human behavior. Popular models of social perception generally theorize a relatively fixed detection process to identify functionally and adaptively significant social attributes (e.g., warmth, competence, anger, race). However, recent research suggests considerable malleability in social perception, which is not adequately accounted for by current models. Here I argue that a number of social perception phenomena may be parsimoniously explained not by a set of fixed detectors but by a domain-general account of spreading activation between social concepts through an associative network. Specifically, I propose that, similar to other forms of non-social inference, perceivers form a knowledge structure of what social concepts exist in the world (e.g., frequent speakers are ‘extroverted’) and how those concepts associate with one another (e.g., ‘extroverted’ people are often ‘kind’ and ‘male’), and they then use this structure to make inferences (e.g., 'this kind male is likely extroverted'). Although quite simple, this perspective provides rich predictions that may describe many facets of the social perception process, such as how perceptions vary in their initial formation from cues, automaticity and temporal dynamics, variance within and between perceivers and contexts, and dimensional and categorical structure. This perspective also helps integrate theory of social perception, bridging both perceptual classes (e.g., traits, social categories, and emotion) and their contexts (e.g., face impressions, person knowledge, and group stereotyping).To provide an initial test of this perspective, I examine how social perceptions (e.g., warmth, extroversion) correlate with one another along the lines of their conceptual associations (e.g., ‘are warm people likely to be extroverted?’). In Chapter 1, we demonstrate that face-based trait impressions color one another in to the extent they are conceptually associated. Faces perceived to possess one personality trait (e.g., trustworthiness) elicited additional trait impressions (e.g., creativity) to the extent perceivers conceptually associated the traits (e.g. ‘trustworthy people are often creative’). Chapter 2 extends the findings of Chapter 1 across contexts of social cognition, where the same conceptual structuring of trait impressions emerged across the domains of face impressions, familiar person knowledge, and group stereotype content. Lastly, in Chapter 3, I apply this perspective to the domains of emotion recognition and social categorization. Survey, mouse-tracking, and neuroimaging analyses showed categories apparent in a face (e.g., ‘male’) facilitated or impaired perceptions and neural representations of other categories (e.g., ‘black’) to be in accordance with their conceptual associations. Together, these findings provide evidence for a domain-general account of social perception, which assumes only basic semantic-processing principles, accounts for a number of social perception phenomena, and generates several new theoretical predictions. Overall, this research demonstrates that the perceptions and dimensions which emerge in social perception are bound to perceivers’ conceptual representations of the social world.
Great ape communication as contextual social inference: a computational modeling perspective
