How do readers interpret ambiguous characters in novel Chinese compound words?

<p>Most Chinese words are compounds comprising multiple constituent characters. Research has shown that compounds are routinely broken down into their constituents during word recognition, but it is unknown how readers deal with characters that have multiple possible meanings (e.g., 花 can mean both flower and to spend, depending on the compound in which it appears) and that are highly prevalent in Chinese. The present study investigated how a computational model (Study 1) and human readers (Study 2) compute the meanings of novel Chinese compounds that contain semantically ambiguous characters. We found that both model and human readers prioritise characters that have a consistent meaningful function when making sense of novel compounds, and a further study revealed that semantically-consistent characters serve as anchors to help readers determine the contextually appropriate meaning for ambiguous characters. These findings suggest that the stored knowledge of the consistency of form–meaning mappings shapes the comprehension of novel words. We conclude by arguing that, despite its apparent ambiguity, the Chinese writing system may be more systematic and learnable than it appears.</p>