Abstract
In this paper, we consider the syntactic properties of languages emerged inreferential games, using unsupervised grammar induction (UGI) techniquesoriginally designed to analyse natural language. We show that the consideredUGI techniques are appropriate to analyse emergent languages and we then studyif the languages that emerge in a typical referential game setup exhibitsyntactic structure, and to what extent this depends on the maximum messagelength and number of symbols that the agents are allowed to use. Ourexperiments demonstrate that a certain message length and vocabulary size arerequired for structure to emerge, but they also illustrate that moresophisticated game scenarios are required to obtain syntactic properties moreakin to those observed in human language. We argue that UGI techniques shouldbe part of the standard toolkit for analysing emergent languages and release acomprehensive library to facilitate such analysis for future researchers.