Human languages vary in many ways but also show striking cross-linguistic universals. Why do these universals exist? Recent theoretical results demonstrate that Bayesian learners transmitting language to each other through iterated learning will converge on a distribution of languages that depends only on their prior biases about language and the quantity of data transmitted at each point; the structure of the world being communicated about plays no role (Griffiths & Kalish, , ). We revisit these findings and show that when (...) certain assumptions about the relationship between language and the world are abandoned, learners will converge to languages that depend on the structure of the world as well as their prior biases. These theoretical results are supported with a series of experiments showing that when human learners acquire language through iterated learning, the ultimate structure of those languages is shaped by the structure of the meanings to be communicated. (shrink)

We examined 11-month-olds’ tendency to generalize properties to category members, an ability that may contribute to the inductive reasoning abilities observed in later developmental periods. Across 3 experiments, we tested 11-month-olds’ (N= 113) generalization of properties within the cat and dog categories. In each experiment, infants were familiarized to animal-sound pairings (i.e., dog-barking; cat-meowing), and tested on this association and the generalization of the sound property to new members of the familiarized categories. After familiarization with a single exemplar, 11-month-olds generalized (...) the sound to new category members that were both highly similar and less similar to the familiarized animal (Expt. 1). When familiarized with mismatched animal-sound pairings (Expt. 2; i.e., dog-meowing; cat-barking), 11-month-olds did not learn or generalize the sound properties, suggesting that infants have pre-existing expectations about the links between the characteristic sound properties and the animal categories. When familiarized with unfamiliar sound-animal pairings (Expt. 3; i.e., dog-unfamiliar sound), 11-month-olds linked the animals with the novel sounds but did not generalize to new category members. Taken together, these findings highlight the conditions under which young infants generalize properties from one exemplar to other category members. (shrink)

It is well-established that toddlers can correctly select a novel referent from an ambiguous array in response to a novel label. There is also a growing consensus that robust word learning requires repeated label-object encounters. However, the effect of the context in which a novel object is encountered is less well-understood. We present two embodied neural network replications of recent empirical tasks, which demonstrated that the context in which a target object is encountered is fundamental to referent selection and word (...) learning. Our model offers an explicit account of the bottom-up associative and embodied mechanisms which could support children’s early word learning and emphasises the importance of viewing behaviour as the interaction of learning at multiple timescales. (shrink)