Evolutionary Pressures Promoting Complexity in Navigation and Communication Dimitar Kazakov and Mark Bartlett Department of Computer Science, University of York, United Kingdom {dimitar.kazakov|mark.bartlett}@york.ac.uk Abstract This article presents results from simulations studying the hypothesis that mechanisms for landmark-based navigation could have served as preadaptations for compositional lan- guage. It is argued that sharing directions would significantly have helped bridge the gap between general and language- specific cognitive faculties. A number of different levels of navigational and communicative abilities are considered, re- sulting in a range of possible evolutionary paths. The se- lective pressures for, resp. against, increased complexity in either faculty are then evaluated for a range of environ- ments. The study aims specifically to identify whether there is a viable evolutionary path leading to compositional lan- guage, and if so, under what circumstances. The results show that environmental conditions can render a step towards more complex communication either desirable or harmul, and sug- gest that very specific initial conditions and changes in the en- vironment, resp. the ecological niche occupied, would have been needed to select for compositional language. Subject to these conditions, a (proto)language using order, but no hierarchical structure could evolve. This represents a mid- dle ground, which brings closer hypotheses about syntax that have so far appeared difficult to reconcile. Introduction This article builds on our previous studies of the origins of language, in which we used simulations to study naviga- tion as a potential preadaptation for compositional language, and compositional language as a way of sharing information about resources. We have also studied the factors and mech- anisms that make sharing an evolutionary viable strategy, and identified specific conditions, which make information sharing advantageous, when compared to sharing in kind or selfish behaviour (Kazakov and Bartlett, 2002; Bartlett and Kazakov, 2003; Kazakov and Bartlett, 2004; Bartlett and Kazakov, 2005; Kazakov and Bartlett, 2005). In the course of the last decade, we have put forward a number of ideas which were later advanced by some of the leading proponents in this debate. This includes: (1) the suggestion that landmark-based navigation and parsing re- quire the same type of abstract computation, and therefore the two could be homologous, with landmark navigation serving as a preadaptation for compositional language (com- pare Kazakov and Bartlett (2002) with Hauser’s comments (Wade, 2003)); (2) the idea that language initially served to direct other individuals to valuable resources (cf. Kazakov and Bartlett (2002) and Bickerton (2009)); (3) the idea that language may have spread as a form of kin selection (cf. Kazakov and Bartlett (2002); Bartlett and Kazakov (2003), and Fitch (2004)); (4) the finding that environmental factors, such as the type of diet our ancestors had, could have played a critical role to promote language (cf. Kazakov and Bartlett (2005) and Bickerton (2009)). We have also demonstrated in simulations how the processes promoting language on the basis of the above mechanisms would reinforce each other to make the emerging faculty of language a lot more likely to remain and spread in the population (see Bartlett and Kaza- kov (2003) for the first experimental results and Kazakov (2010) for a more in-depth analysis of the interplay between the processes). The existence of such interplay does not imply that the above mentioned hypotheses stand or fall together, only that evidence for one would increase the estimated benefits of the others as promoters of language. Similarly, the proposed mechanisms need not be exclusive. For instance, replac- ing or complementing kin selection with reciprocal altru- ism (Slocombe et al., 2010) or sexual selection (Gomes and Boesch, 2009) as alternative mechanisms making communi- cation an evolutionary viable strategy would not affect the validity of the remaining aspects of our studies. The hypothesis that sharing of directions is behind the first use of compositional language is appealing as it implies a much reduced gap between general cognitive faculties and language-specific faculties, for the following reasons. With respect to syntax, the planner used for optimal (shortest- path) navigation can directly double up as a parser for this task. Also, indexical signs, rather than fully symbolic ones, are sufficient to describe a landmark, and the sign itself could be either spoken or gestural. Combining gesture with speech could allow all speakers to develop a shared, open vocabulary in a relatively simple way, as shown in computer simulations (Steels, 1999). With respect to such an essential