How is prejudice learned? And how does it influence our behavior?

Humans learn about other people in multiple ways (Amodio, 2019), and multiple learning mechanisms contribute to the formation of prejudice (Amodio & Cikara, 2021). We can consciously remember specific people, places, and events (episodic memory) or recall all kinds of general knowledge (semantic memory). We store conceptual knowledge about people and groups (semantic memory), affective reactions to threat or reward (Pavlovian learning), approach/avoidance behavioral action preferences (instrumental learning), and automatic behavioral responses (habits). Each of these depends on different underlying neural processes, and in everyday life, they usually work together to guide our thoughts, emotions, and behaviors. Some of these memory systems function explicitly, subject to our conscious awareness, whereas others function with little conscious awareness and may be delpoyed automatically. To understand these effects, we have developed a memory systems model of social cognition and attitudes (Amodio, 2019; Amodio & Ratner, 2011).

Research in the Amodio Lab examines the role of different memory systems in the formation and expression of implicit bias. We have shown that implicit bias may be supported by semantic memory, Pavlovian learning, or instrumental learning (Amodio & Ratner, 2011). Importantly, these memory systems influence our thoughts and actions in different ways. For example, semantic associations predict stereotype-based judgments, whereas Pavlovian associations relate to intergroup anxiety and social distance (Amodio et al., 2003; Amodio & Devine, 2006). Moreover, we found that intergroup anxiety selectively amplifies the activation of implicit prejudice but not implicit stereotypes (Amodio & Hamilton, 2012).

More recently, we have examined the role of instrumental learning in the formation of prejudice. Instrumental learning supports the formation of reward associations through action and feedback, such as in reward reinforcement learning paradigms, and it addresses how we learn about people through direct social interaction. Research in the lab shows that social stereotypes influence how we form preferences toward group members in direct interactions, via instrumental learning (Schultner, Stillerman, et al., 2024), and that preferences formed toward individual group members through instrumental learning generalize to group-level prejudiced attitudes (Hackel et al., 2022). Furthermore, we show that this kind of interaction-based prejudice easily spread to others via social learning: when observers who are naive to any group stereotypes  merely view the interaction behaviors of a prejudiced person, they acquire prejudice in their own personal beliefs and express it in their future interaction behaviors (Schultner, Lindström, et al., 2024). Much of our current work examines the role of instrumental reinforcement learning processes in different aspects of prejudice and stereotyping.

Broadly, we believe that a consideration of the learning mechanisms underlying prejudice and stereotyping is essential for understanding intergroup bias, its expression in behavior, and its reduction through predicting how such biases influence behavior and how they may be effectively reduced.