Monica Tamariz

Compression and communication in the cultural evolution of linguistic structure

Language exhibits striking systematic structure. Words are composed of combinations of reusable sounds, and those words in turn are combined to form complex sentences. These properties make language unique among natural communication systems and enable our species to convey an open-ended set of messages. We provide a cultural evolutionary account of the origins of this structure. We show, using simulations of rational learners and laboratory experiments, that structure arises from a trade-off between pressures for compressibility (imposed during learning) and expressivity (imposed during communication). We further demonstrate that the relative strength of these two pressures can be varied in different social contexts, leading to novel predictions about the emergence of structured behaviour in the wild.

Culture: Copying, Compression, and Conventionality.

Through cultural transmission, repeated learning by new individuals transforms cultural information, which tends to become increasingly compressible (Kirby, Cornish, & Smith, ; Smith, Tamariz, & Kirby, ). Existing diffusion chain studies include in their design two processes that could be responsible for this tendency: learning (storing patterns in memory) and reproducing (producing the patterns again). This paper manipulates the presence of learning in a simple iterated drawing design experiment. We find that learning seems to be the causal factor behind the increase in compressibility observed in the transmitted information, while reproducing is a source of random heritable innovations. Only a theory invoking these two aspects of cultural learning will be able to explain human cultures fundamental balance between stability and innovation.

The cultural evolution of language

Human language has unusual structural properties that enable open-ended communication. In recent years, researchers have begun to appeal to cultural evolution to explain the emergence of these structural properties. A particularly fruitful approach to this kind of explanation has been the use of laboratory experiments. These typically involve participants learning and interacting using artificially constructed communication systems. By observing the evolution of these systems in the lab, researchers have been able to build a bridge between individual cognition and population-wide emergent structure. We review these advances, and show how cultural evolution has been used to explain the origins of structure in linguistic signals, and in the mapping between signals and meanings.

Cultural selection drives the evolution of human communication systems

Human communication systems evolve culturally, but the evolutionary mechanisms that drive this evolution are not well understood. Against a baseline that communication variants spread in a population following neutral evolutionary dynamics (also known as drift models), we tested the role of two cultural selection models: coordination- and content-biased. We constructed a parametrized mixed probabilistic model of the spread of communicative variants in four 8-person laboratory micro-societies engaged in a simple communication game. We found that selectionist models, working in combination, explain the majority of the empirical data. The best-fitting parameter setting includes an egocentric bias and a content bias, suggesting that participants retained their own previously used communicative variants unless they encountered a superior (content-biased) variant, in which case it was adopted. This novel pattern of results suggests that (i) a theory of the cultural evolution of human communication systems must integrate selectionist models and (ii) human communication systems are functionally adaptive complex systems.

Regularity in Mappings Between Signals and Meanings

We combine information theory and cross-situational learning to develop a novel metric for quantifying the degree of regularity in the mappings between signals and meanings that can be inferred from exposure to language in context. We illustrate this metric using the results of two artificial language learning experiments, which show that learners are sensitive, with a high level of individual variation, to systematic regularities in the input. Analysing language using this measure of regularity allows us to explore in detail how language learning and language use can both generate linguistic variation, leading to language change, and potentially complexify language structure, leading to qualitative language evolution.

Completion norms for 112 Spanish sentences

Sentence completion norms are a valuable resource for researchers interested in studying the effects of context on word recognition processes. Norms for 112 Spanish sentences were compiled with the use of experimental software accessed over the World-Wide Web. Several measures summarizing the distribution of responses for each sentence are reported, including Schwanenflugel’s (1986) multiple-production measure of sentence constraint strength, the type-token ratio, and the information-theoretic measure of redundancy. The complete set of completion norms is available at http://www.ling.ed.ac.uk/ ~monica/spanish_completion_norms.html.

The role of practice and literacy in the evolution of linguistic structure

Recent iterated language learning studies have shown that artificial languages evolve over the generations towards regularity. This trend has been explained as a reflection of the learners’ biases. We test whether this learning bias for regularity is affected by culturally acquired knowledge, specifically by familiarity and literacy. The results of non-iterated learning experiments with miniature artificial musical and spoken languages suggest that familiarity helps us learn and reproduce the signals of a language, but literacy is required for regularities to be faithfully replicated. This in turn indicates that, by modifying human learning biases, literacy may play a role in the evolution of linguistic structure.

The Interactive Origin of Iconicity

We investigate the emergence of iconicity, specifically a bouba-kiki effect in miniature artificial languages under different functional constraints: when the languages are reproduced and when they are used communicatively. We ran transmission chains of (a) participant dyads who played an interactive communicative game and (b) individual participants who played a matched learning game. An analysis of the languages over six generations in an iterated learning experiment revealed that in the Communication condition, but not in the Reproduction condition, words for spiky shapes tend to be rated by naive judges as more spiky than the words for round shapes. This suggests that iconicity may not only be the outcome of innovations introduced by individuals, but, crucially, the result of interlocutor negotiation of new communicative conventions. We interpret our results as an illustration of cultural evolution by random mutation and selection (as opposed to by guided variation).

Experiments and Simulations Can Inform Evolutionary Theories of the Cultural Evolution of Language

Language is a complex adaptive system supported by humans. With evidence and illustrative examples from recent computer simulation and experimental work, this chapter defends that much linguistic structure can be explained as emergent phenomena. Cultural processes such as social transmission to new generations or the patterns of speaker interaction, operating in large populations over many generations give rise to structure at different levels, from categories of phonemes to compositional structure. This position shifts the burden of explanation of linguistic structure from a biologically evolved mental organ to more tractable cultural processes, which are being investigated with a host of innovative empirical methods.

The Effects of Generation Turnover and Interlocutor Negotiation on Linguistic Structure

Cultural transmission has a key impact on the evolution of linguistic structure (Kirby, Cornish & Smith, 2008). During transmission, new individuals learn the conventions of the language through observation and participation in communicative interactions in context. We present a series of experiments which disentangle the role of generation turnover, or new minds coming into the linguistic community (e.g. Kirby et al. 2008) from that of negotiation of shared form-meaning conventions during interlocutor interaction (studied in graphical, but not linguistic, systems by e.g. Fay et al. 2007, Galantucci 2005). In the experimental task, two interlocutors played a communicative cooperative game using an artificial miniature language. Players were trained on a random language, which they then used to ask each other for specific objects; they scored a point for each successful interaction. The language produced by one (randomly selected) participant was then used as training data for the following training-and-usage round. We manipulated generation turnover: in dyads, two participants played six training-and-usage rounds; in chains, six different participant pairs played one round each. We also manipulated negotiation. In the negotiation condition there were two human players; in the no-negotiation condition, one of the players was replaced with a simulated computer agent who had perfect memory (i.e. it always used the training language without errors, and never offered or adopted suggestions during usage). The training languages in this condition came from the human player. We examined the effects of generation turnover and negotiation on the systematic structure of the resulting languages (measured using the technique described in Kirby et al., 2008) and found a significant effect of both factors. Structure increased over the six rounds in chains, but not in dyads. Looking at chains only, the level of structure at the final round was markedly higher when there was negotiation between two interlocutors than when a single participant