Kanzi, a bonobo, has been learning English since infancy. I examine a corpus of instructions given to Kanzi at age 8 (Savage-Rumbaugh et al 1993), as a source of evidence concerning his syntactic abilities. Although Kanzi has learned the meaning of many English words, the corpus suggests that Kanzi cannot interpret NP-coordination structures like “Give [[the water] and [the doggie]] to Rose”. I claim that this reflects an inability to form hierarchical phrase-structural representations of sentence structure: NP-coordination is one of the simplest structures that cannot be interpreted without some notion of constituency, because an interlocutor has to recognize that “water” and “doggie” jointly form a single argument of “give”. Kanzi’s failure to interpret NP-coordination can therefore taken to be indicative of a general inability to generate hierarchical representations of sentence structure. I compare this finding to the literature on the acquisition of NP-coordination by human infants. Humans also initially struggle to understand NP-coordination structures (e.g. Gertner & Fisher 2012), but unlike Kanzi, their comprehension quickly improves.
Rob Truswell is Chancellor’s Fellow in the school of Philosophy, Psychology, and Language Science at the University of Edinburgh, and adjunct professor of Linguistics at the University of Ottawa. His main interest is in pairing minimal theories of syntactic competence with rich theories of nonsyntactic influences on syntactic phenomena. He has published on a range of topics in the syntax/semantics interface, diachronic syntax/semantics, and language evolution, including the monograph “Events, Phrases, and Questions” (OUP, 2011) and edited volumes “Syntax and its Limits” (OUP, 2013) and “Micro-change and Macro-change in Diachronic Syntax” (OUP, 2017).