Language Evolution and Computation Bibliography

The problem of how young learners acquire the meaning of words is fundamental to language development and cognition. A host of computational models exist which demonstrate various mechanisms in which words and their meanings can be transferred between a teacher and learner. However these models often assume that the learner can easily distinguish between the referents of words, and do not show if the learning mechanisms still function when there is perceptual ambiguity about the referent of a word. This paper presents two models that acquire meaning-word mappings in a continuous semantic space. The first model is a cross-situational learning model in which the learner induces word-meaning mappings through statistical learning from repeated exposures. The second model is a social model, in which the learner and teacher engage in a dyadic learning interaction to transfer word-meaning mappings. We show how cross-situational learning, despite there being no information to the learner as to the exact referent of a word during learning, still can learn successfully. However, social learning outperforms cross-situational strategies both in speed of acquisition and performance. The results suggest that cross-situational learning is efficient for situations where referential ambiguity is limited, but in more complex situations social learning is the more optimal strategy.

There is a growing body of research on multi-agent systems bootstrapping a communication system. Most studies are based on simulation, but recently there has been an increased interest in the properties and formal analysis of these systems. Although very interesting and promising results have been obtained in these studies, they always rely on major simplifications. For example, although much larger populations are considered than was the case in most earlier work, previous work assumes the possibility of meaning transfer. With meaning transfer, two agents always exactly know what they are talking about. This is hardly ever the case in actual communication systems, as noise corrupts the agents' perception and transfer of meaning. In this paper we first consider what happens when relaxing the meaning-transfer assumption, and propose a cross-situational learning scheme that allows a population of agents to still bootstrap a common lexicon under this condition. We empirically show the validity of the scheme and thereby improve on the results reported in (Smith, 2003) and (Vogt and Coumans, 2003) in which no satisfactory solution was found. It is not our aim to reduce the importance of previous work, instead we are excited by recent results and hope to stimulate further research by pointing towards some new challenges.

Color categories enjoy a special status among human perceptual categories as they exhibit a remarkable cross-cultural similarity. Many scholars have explained this universal character as being the result of an innate representation or an innate developmental program which all humans share. We will critically assess the available evidence, which is at best controversial, and we will suggest an alternative account for the universality of color categories based on linguistic transmission constrained by universal biases. We introduce a computational model to test our hypothesis and present results. These show that indeed the cultural acquisition of color categories together with mild constraints on the perception and categorical representation result in categories that have a distribution similar to human color categories.

This article proposes a number of models to examine through which mechanisms a population of autonomous agents could arrive at a repertoire of perceptually grounded categories that is sufficiently shared to allow successful communication. The models are inspired by the main approaches to human categorisation being discussed in the literature: nativism, empiricism, and culturalism. Colour is taken as a case study. Although we take no stance on which position is to be accepted as final truth with respect to human categorisation and naming, we do point to theoretical constraints that make each position more or less likely and we make clear suggestions on what the best engineering solution would be. Specifically, we argue that the collective choice of a shared repertoire must integrate multiple constraints, including constraints coming from communication.

In this paper we present our work on developing a shared repertoire of action categories through imitation. A population of robotic agents invents and shares a repertoire of actions by engaging in imitative interactions. We present an experimental set-up which enables us to investigate what properties agents should have in order to achieve this. Among these properties are: being able to determine the other's actions from visual observation and doing incremental unsupervised categorisation of actions.

Human colour perception is continuous, but humans categorise the colour continuum and often label the resulting colour categories. The debate on whether colour categorisation is an individual process, or whether it is embedded in genetic constraints has not been settled yet. Further- more, as colour categories have colour names, it is claimed that language could have an influence on the categorisation. This paper reports on agent-based simulations that test the validity of dirent theories, and uncovers the weak and strong points of each. We conclude, from experi- ments using AI techniques, that colour categorisation is most likely to be cultural process.

Humans perceive a continuous colour spectrum, but divide the spectrum into colour categories in order to reason and communicate about colour. There is an ongoing debate on whether these colour categories necessary for language communication are universal or culture-specific, whether these categories are genetically determined or learned, and whether there is a causal influence of language on colour category acquisition or not. The dissertation presents a number of models, each examining one of these outstanding issues. The models draw on techniques from multi-agent systems, machine learning and evolutionary programming. After considering the behaviour of each model, we conclude in favour of a cultural specificity of language categories and argue that learning under the influence of language is the most plausible explanation for their acquisition.

This paper investigates the formation of color categories and color naming in a population of agents. The agents perceive and categorize color stimuli, and try to communicate about these perceived stim- uli. While doing so they adapt their internal representations to be more successful at conveying color meaning in future interactions. The agents have no access to global information or to the representa- tions of other agents; they only exchange word forms. The factors driving the population coherence are the shared environment and the interactions. The experiments show how agents can form a coherent lexicon of color terms and -particularly- how a coherent color categorization emerges through these linguistic interactions. The results are interpreted in the light of theories describing and explaining universal tendencies in human color categorization and color naming. At the same time, the experiments confirm aspects of the theories of Luc Steels who views language as a complex dynamic system, arising from selforganization and cultural interactions.