Predictions of Entropy Reduction Theory on Chinese Relative Clauses
Borui Zhang · The Knowledge Bank (The Ohio State University) · 2018
Processing difficulty in Chinese has been a challenge for theories of sentence parsing due to its lacking morphology and structural ambiguity. I use entropy reduction (ER) theory, based on a minimalist analysis of relative clauses (RCs), and tree-banks-based probabilistic grammar to make predictions on Chinese RCs (with various amounts of temporary ambiguity). The predictions match the results from the previous human experiments on Chinese. This provides supporting evidence to the reliability of ER theory and highlight the ability of the theory to deal with ambiguous sentences with the right grammar and analysis.