The topic of language evolution is characterised by the scarcity of records, but also by a large flow of research produced within multiple subtopics and perspectives. Over the past few decades, significant advancement has been made on the geographical and temporal origins of language, while current work is rather devoted to the underpinnings of language, in brain, genes, body, and culture of humans. Much of this literature is polarized over the crucial dichotomy of nativism versus emergentism. Our state of affairs report also confirms a high degree of speculation, albeit with a decrease for modelling. To tackle the speculation and the large research flow, we propose a more impersonal kind of review, focused on the topic’s questions rather than on particular accounts. Another observation is that novel perspectives are on the rise. One of these highlights the importance of perceptual cognition, often dubbed ‘embodiment,’ in the earlier evolution of language. In following this lead, we adapted a previous experiment which had investigated the correspondence between certain perceptual features of events, and different grammatical orders arising as participants acted out those events. That design made a perfect basis for us to put in an additional variable, namely the contrast between body-based communication (gestures), and more disembodied communication (symbol matching). Albeit tentative, the results of this pilot experiment reveal a greater effect of the embodiment variable on the grammatical preferences, which we see as inviting further exploration of embodied cognition in language evolution.