Speaking to the Heart or the Mind? A Dual‐Process Model of Bilingual Prosocial Behavior in Human and GPT Synthetic Participants

This research examines how the language of a charitable appeal influences bilingual consumers' prosocial intention and behavior. Drawing on the dual‐process model of sympathy and deliberation, we propose that native‐ and nonnative‐language appeals jointly activate emotive and cognitive routes to prosocial behavior. Two human experiments and two GPT‐based simulations test this account among Mandarin‐English bilingual consumers. Results show that Mandarin appeals elicited stronger sympathy and prosocial behavior through an affective route (Mandarin → ↑ sympathy → ↑ prosocial behavior), whereas English appeals operated through a cognitive route in which perceived self‐other overlap offsets an otherwise lower level of sympathy, increasing prosocial behavior (English → ↑ self‐other overlap → ↑ sympathy → ↑ prosocial behavior). Two GPT simulations reproduced the direction of every mediation path. The findings contribute to bilingual persuasion research by showing that language choice is not just a matter of translation but a strategic lever that influences which psychological pathway an appeal activates. For marketers and nonprofits in linguistically diverse markets, our findings offer a practical framework for aligning language choice with the pathway most likely to drive donor response. The dual‐process model may also extend to other persuasion contexts, such as health, sustainability, and public policy communication, where the choice between a native and a nonnative language can similarly determine whether an affective or cognitive pathway is activated.