A study tested the validity of a theory of count/mass syntax in word learning. The theory proposes that children infer one of two procedures, depending on whether the referent is an object or a non-solid substance. Subjects were 36 2-year-olds, divided according to three experimental conditions. All were taught a novel word with reference to either an object or a non-solid substance, then asked to select another referent of the new word, given two choices. In the neutral syntax condition, the syntax used indicated that the new word was a noun but not whether it was a count or mass noun. In the matching syntax condition, objects were labeled with count nouns and corresponding determiners and quantifiers. In the conflicting syntax condition, mass syntax was used in object trials and count syntax was used in substance trials. Results indicate that the children were unable to use syntactic distinctions to support the two procedures for inferring meaning, contradicting the position that children learn ontological distinctions through the syntactic distinction. Further, it is concluded that some 2-year-olds do use syntax to influence their inferences about new words when the referents are non-solid substances, but not when they are solid objects. (MSE)