Whether the animals studied by Pepperberg, Herman, and Savage-Rumbaugh can learn prepositions remains an open and important question, though the dolphins' use of demonstratives and conjuctions suggests an impressive competence for at least some closed-class items. It is unlikely that human-like brain structure can explain the capacities of these animals. Their capacities may instead be built upon the cognitive mechanisms favored by social complexity.In their thoughtful, stimulating commentaries, Pepperberg (1999), Herman and Uyeyama (1999), and Shanker, Savage-Rumbaugh, and Taylor (1999) have addressed a number of important issues. I will here focus on just three: closed-class items, the relations of brain structure and social complexity to linguistic capacity, and the importance of continued research on the linguistic capacities of nonhuman animals.
Closed-Class ItemsI am pleased to hear that Pepperberg (1999) has proposed to teach Alex closed-class items, although I am disappointed that she has not yet done so. I disagree that Alex's success in learning in versus on (or above vs. below) would be trivial. Like Pepperberg, I strongly suspect that Alex possesses the necessary spatial concepts. The question is whether he can map those concepts onto lexical items and whether he can use those lexical items compositionally, in comprehension and production (to interpret a question like "What color hide in bowl?" or to respond to the question "Where yellow hide?" with "In bowl"). To date, Alex has succeeded in the mapping and compositional use of concepts related to object class, property, and number. It remains to be seen whether he can map and use the geometric and relational concepts relevant to prepositions.Herman and Uyeyama (1999) present persuasive evidence that the dolphins have learned two closed-class items-the demonstrative (pronoun) THAT and the conjunction AND. I am especially impressed by their capacity to use demonstratives in the comprehension of relational commands (e.g., THAT FETCH THAT). The dolphins' mastery of LEFT and RIGHT is very striking, as these terms encode abstract, egocentric reference points. However, they are not closed-class items, but rather belong to the open-class set of nouns (at least in English).' Whether Correspondence concerning this article should be sent to E. Kako,