Kees Sommer

We propose that self-domestication in humans, correlating with increased activity of neurochemicals associated with prosociality, resulted in a heightened sensitivity to cross-modal associations. Evidence supporting this hypothesis is presented. The specific connection between prosociality and cross-modal associations is explored through a behavioural experiment, manipulating and measuring participants’ prosociality and testing their sensitivity to cross-modal associations. The resulting data… Show more

We propose that self-domestication in humans, correlating with increased activity of neurochemicals associated with prosociality, resulted in a heightened sensitivity to cross-modal associations. Evidence supporting this hypothesis is presented. The specific connection between prosociality and cross-modal associations is explored through a behavioural experiment, manipulating and measuring participants’ prosociality and testing their sensitivity to cross-modal associations. The resulting data show a significant effect of measured prosociality on sensitivity to cross-modal associations. Show less

A key communicative resource available to sign language users is the use of space to distinguish between referents and to express relationships between them. While spatial tracking of referents is found in co-speech gesture, evidence from young sign languages suggests that systematic use of spatial devices emerges gradually. We present an investigation of the cultural transmission of spatial devices using a novel experimental method, drawing on artificial sign language experiments and the… Show more

A key communicative resource available to sign language users is the use of space to distinguish between referents and to express relationships between them. While spatial tracking of referents is found in co-speech gesture, evidence from young sign languages suggests that systematic use of spatial devices emerges gradually. We present an investigation of the cultural transmission of spatial devices using a novel experimental method, drawing on artificial sign language experiments and the cultural evolution of stories. Participants first watched a video of a story depicted in pantomime, and were asked to record their own video, which was then shown to the next person in one of the 9 chains of 6 generations each. We find that participants identified characters using a mixture of lexical labels and spatial strategies, typically shifting their body orientation when embodying different characters. Over generations, the use of body shifting increased across chains. Show less

Abstract
The mainstream approach in NLP is to train systems on large amounts of data. Such passive learning contrasts with the waylanguage is learnt by humans. Human language is acquired within communities, it is culturally transmitted and changesdynamically. These evolutionary mechanisms have been extensively studied in the field of Language Evolution. Despitelimited prior interaction between fields, such mechanisms are now increasingly incorporated into NLP systems. Suchmodels have the… Show more

Abstract
The mainstream approach in NLP is to train systems on large amounts of data. Such passive learning contrasts with the waylanguage is learnt by humans. Human language is acquired within communities, it is culturally transmitted and changesdynamically. These evolutionary mechanisms have been extensively studied in the field of Language Evolution. Despitelimited prior interaction between fields, such mechanisms are now increasingly incorporated into NLP systems. Suchmodels have the potential to both study the evolution of language in multi-agent simulations with state-of-the-art (deep)learning systems in more naturalistic settings and improve NLP systems by having language emerge organically. Weexamine how findings from a model by Havrylov & Titov (2017) compare to those from traditional Language Evolutionmodels and quantify the emerging compositionality using an existing Language Evolution method (Tamariz, 2011). Thisapproach reveals novel insights into the generated data, the applied methodology and the nature of compositionality. Show less