ABSTRACT
Research on the moral foreign language effect (MFLE) shows that using a foreign language can shift moral judgements, and sometimes this shift can reverse. The present study tested whether a reverse MFLE would appear in a self-driving-car dilemma that usually favours protecting the larger group (pedestrians). In Experiment 1 (N = 266; L1 Chinese vs. L2 English, between-subjects), a canonical sacrificial dilemma was used for benchmarking before introducing the self-driving-car dilemma; a reverse MFLE was observed in the latter (fewer utilitarian choices in L2 than in L1). Experiment 2 (N = 361) replicated the reverse MFLE in the self-driving-car dilemma and showed moderation by self-reported driving frequency: the reverse effect was clear among participants who drive less often and weaker or absent among those who drive more often. These findings extend MFLE research beyond the footbridge paradigm, reducing over-reliance on a single dilemma, and—together with prior work—strengthen the conclusion that using a foreign language can dampen fast, default processes (System 1) responses, which can lead to shifts, including reversals, in moral choice patterns.