The emergence of linguistic structure in an online iterated learning task

Previous research by Kirby et al. has found that strikingly compositional language systems can be developed in the laboratory via iterated learning of an artificial language. However, our reanalysis of the data indicates that while iterated learning prompts an increase in language compositionality, the increase is followed by an apparent decrease. This decrease in compositionality is inexplicable, and seems to arise from chance events in a small dataset (four transmission chains). The current study thus investigates the iterated emergence of language structure on a larger scale using Amazon Mechanical Turk, encompassing twenty-four independent chains of learners over ten generations. This richer dataset provides further evidence that iterated learning causes languages to become more compositional, although the trend levels off before the 10th generation. Moreover, analysis of the data (and reanalysis of Kirby et al.) reveals that systematic units arise along some meaning dimensions before others, giving insight into the biases of learners