Research
Social categories help people navigate social life by carving the complexity of humanity into neat (sometimes, too neat) categories. They allow people to notice and explain social regularities, predict the behavior of others, and communicate about the social world.
Where do these social category representations come from? I focus on childhood, when many of our most foundational social categories are acquired. What learning mechanisms and input explain how children learn social category representations? How are social category representations transmitted from generation to generation, and how do they come to stabilize in collective cognition?
Stereotypes present a particularly interesting case for examining these questions. Stereotypes are representations that do not fully capture the reality of a social group, but nonetheless emerge and take collective hold, leading to undesirable social outcomes. Why are stereotypes so persistent and stable, despite their seeming inaccuracy and problematic social effects? How could rational learning mechanisms possibly lead to the emergence of stereotypes? And how do stereotypes contribute to the very differences and disparities they purport to index in the first place?
Current research
Reasoning about social categories in the context of social structures
I’m interested in how social category representations and social structures emerge and stabilize together over time, including how social category representations are acquired, how those representations support social structures, and how those social structures feed back into the acquisition of social category representations.
Tracing this loop can reveal how pernicious social representations and structures become mutually entrenched, and how the cycle might be altered to support equitable futures.
How do children reason about social differences and disparities?
- Children tend to attribute social differences and disparities to inherent differences between groups, but they are capable of structural reasoning, i.e., attributing social differences and disparities to social structures, starting around age 6 (Zhang, Sullivan, Markman, & Roberts, 2023, Child Dev Persp).
What are the consequences of structural reasoning?
- Structural reasoning leads to less group-based generalization, less justification of disparities, and more structural interventions, compared to inherent reasoning about a group’s biology or culture (Zhang, Liu, & Markman, 2024, Proc Cog Sci; in prep).
How to encourage structural thinking?
- The strategies often used by White parents in conversations with children about race lead children to inherent rather than structural thinking; structural explanations are more effective in fostering structural thinking (Zhang, Sullivan, Markman, Dev Sci).
How do children evaluate social structures?
- Children expect teachers to hold gender stereotypes and endorse classroom practices based on such gender stereotypes (Zhang, Asaba, & Leonard, Comms Psych).
How do social structures in turn shape children’s social category representations?
- Rational learning mechanisms may lead to stereotypes in social environments where statistical evidence is skewed by social processes (Zhang, Ho, & Rhodes, in prep).
Learning about social categories from language
I’m particularly interested in how information about social categories can be gleaned from subtle cues in language, going beyond or even subverting the intentions of speakers. Such effects would result in stable collective beliefs about social categories without explicit individual intention being necessary.
How are abstract beliefs about categories learned through language?
- People can learn abstract conceptual beliefs about categories from generic language, through language interpretation, pragmatic reasoning, and world knowledge (Zhang, Malaviya, Leslie, Rhodes, & Ho, 2025, Proc Cog Sci; in prep).
How are stereotypes learned through language?
- Although often said with egalitarian intention, directional comparisons like “Girls are just as good as boys at math” can transmit stereotypes to children through their syntax, which implies one group (e.g., boys) sets the standard (Chestnut, Zhang, & Markman, 2025, Dev Psych).