Communication, symbolic communication, and language: reply to Seidenberg and Petitto.

Seidenberg and Petitto's (1987) assertion that Kanzi and Mulika's lexigram usage is not representational is evaluated by contrasting their abilities with Nim's. Kanzi and Mulika's data indicate that they (a) comprehend spoken English words; (b) can identify lexigram symbols when they hear these words; (c) can comprehend lexigram usage; (d) can use lexigrams when referents are absent and can, if asked, lead someone to the referent; and (e) that all these skills were acquired through observation, not conditioning. Nim evidenced no comprehension of signs and could not use signs when referents were absent. He was forced to sign and encouraged to imitate his teachers. Seidenberg and Petitto's negative experiences with Nim apparently led them to overgeneralize to all other apes, regardless of species, modality, or training history. Consequently, they unjustifiably disregard important components of Kanzi and Mulika's comprehension data which demonstrate that their lexical knowledge could not have been acquired in an instrumental fashion.