SIMULATING THE COEVOLUTION OF LANGUAGE AND JOINT ATTENTION

Abstract

Joint attention, as a set of socio-cognitive abilities to establish common ground in general communicative activities (Tomasello et al., 2005), is essential for social behaviors such as cooperation and language, yet the reason why humans, compared with nonhuman primates, possess an exclusively high level of joint attention remains unknown. Comparative evidence between humans and contemporary animals implies that a high level of joint attention is a humanunique prerequisite for communication and language (Tomasello, 2008), but such conclusions overlook the limitations of the comparative approach and fail to address questions such as whether the initial level of joint attention in humans must be high, and if not, what factors induce the level difference. We propose an evolutionary explanation for the level difference of joint attention. We suggest that via exaptation, joint attention was adopted to comprehend linguistic utterances encoding simple, frequent events taking place in the immediate environment of conversations. Once it starts to assist linguistic comprehension, joint attention triggered the acquisition of preliminary linguistic knowledge. With its level being correlated with linguistic comprehension, according to the socio-biological explanation (Hurford, 1989), communicative success could help adjust the level of joint attention among language users. This scenario indicates that the initial level of joint attention in humans was not very high, and that the level difference could result from a coevolution (Thompson, 1994) with language. It also shows that during language origin comprehension underwent a transition from relying on nonlinguistic information to relying on linguistic information, which led to displacement (Hockett, 1960) in language.