Mutual intelligibility of Chinese dialects: Predicting cross-dialect word intelligibility from lexical and phonological similarity

This paper aims to predict mutual intelligibility (defined here as cross-dialectal word recognition) between 15 Chinese dialects from lexical and phonological distance measures. Distances were measured on the stimulus materials used in the experiment. Their predictive power was compared with earlier similar distance measures based on large word lists. Predictors based on just the stimulus materials used afford the better prediction. Segmental Levenshtein distance was the strongest predictor, outperforming both lexical and tonal similarity measures.