MEI AND DOU IN CHINESE: A TALE OF TWO QUANTIFIERS
This study addresses two outstanding puzzles about the two well-known quantifiers mei and dou in Chinese: ( = 1 \* roman i) the indefinite/definite asymmetry when mei leads the subject NP: dou is not needed when there is an indefinite or a reflexive object within the scope of mei and ( = 2 \* roman ii) the subject/object asymmetry: when mei leads the subject NP, its distribution is restricted, depending on the type of the objects, and, by contrast, when it leads the object NP, its distribution is much freer. We propose a unified account for these puzzles. We argue that ( = 1 \* roman i) the indefinite/definite asymmetry can be explained away if we assume that mei is a distributive quantifier with a portmanteau semantic structure, i.e., that it is a standard universal quantifier plus a matching function; ( = 2 \* roman ii) mei can be domain-shifted into a distributive determiner to satisfy interpretability, and this explains the subject/object asymmetry and ( = 3 \* roman iii) the other intricate facts which arise when mei and dou co-occur can be explained by the Principle of Economy (cf. Reinhart (2006)), which regulates type-shifting.
Key words: distributive quantification, determiners, quantifiers, 'mei', 'dou'