Jianhua Mei

What fixes the reference of proper names and kind terms remains contested. Experimental semantics has revived debates between descriptivism and causal‐historical accounts and revealed complex patterns. This article reviews the field from early cross‐cultural findings to recent developments, including renewed attention to Jonah‐style cases, extensions from names to natural kind terms and other predicates, and methodological disputes like alleged “red herrings” and the new meaning hypothesis that prompted refinements. Looking ahead, we propose broadening targets (artifacts, social kinds, disease terms, etc.) and integrating richer methods (Socratic dialogue, online processing techniques like eye‐tracking, and corpus‐based analysis, etc.) to clarify referential mechanisms.

One of the central findings in experimental semantics over the past two decades is the systematic difference in how East Asians and Westerners judge the reference of proper names, particularly in the well-known Gödel-style thought experiment. This large-scale study aims to evaluate the robustness of the alleged cross-cultural style of semantics across scenarios involving other prominent thought experiments, such as the Jonah case and the Twin Earth case. Through five online replication studies, we sampled over 2000 Mandarin Chinese speakers to investigate their responses to various tasks involving judgments about the reference of proper names, natural kind terms, and other predicates, such as terms for tools and games. These results were then compared with those from English-speaking participants as reported in previous studies. Our findings reveal that the contemporary Gödel case peculiarly stands out in its ability to elicit robust cultural differences. However, the Jonah-style case and the Twin Earth thought experiment do not consistently produce culturally distinct responses. Notably, when popular experimental setups modelled on classical thought experiments are extended to fictitious predicates for tools and games, the response patterns from both Chinese and English-speaking participants resemble their responses to proper names and natural kind terms. This finding deepens concerns about the validity of such designs in testing theories of reference. The study concludes that claims about cross-cultural style of semantics must be carefully constrained, and that the future development of experimental semantics depends on addressing key methodological challenges.