Word-object and action-object learning in children aged 30 to 48 months appears to develop at a similar time scale and adheres to similar attentional constraints. However, children below 36 months show different patterns of learning word-object and action-object associations when this information is presented in a bimodal context (Eiteljoerge et al., 2019b). Here, we investigated 12- and 24-month-olds’ word-object and action-object learning when this information is presented in a unimodal context. Forty 12- and 24-month-olds were presented with two novel objects that were either first associated with a novel label (word learning task) and then later with a novel action (action learning task) or vice versa. In subsequent yoked test phases, children either heard one of the novel labels or saw a hand performing one of the actions presented with the two objects on screen while we measured their target looking. Generalized linear mixed models indicate that 12-month-olds learned action-object associations but not word-object associations and 24-month-olds learned neither word- nor action-object associations. These results extend previous findings (Eiteljoerge et al., 2019b) and, together, suggest that children appear to learn action-object associations early in development while struggling with learning word-object associations in certain contexts until 2 years of age.

Human agents draw a variety of inferences effortlessly, spontaneously, and with remarkable efficiency – as though these inferences were a reflexive response of their cognitive apparatus. Furthermore, these inferences are drawn with reference to a large body of background knowledge. This remarkable human ability seems paradoxical given the complexity of reasoning reported by researchers in artificial intelligence. It also poses a challenge for cognitive science and computational neuroscience: How can a system of simple and slow neuronlike elements represent a large body of systemic knowledge and perform a range of inferences with such speed? We describe a computational model that takes a step toward addressing the cognitive science challenge and resolving the artificial intelligence paradox. We show how a connectionist network can encode millions of facts and rules involving n-ary predicates and variables and perform a class of inferences in a few hundred milliseconds. Efficient reasoning requires the rapid representation and propagation of dynamic bindings. Our model (which we refer to as SHRUTI) achieves this by representing (1) dynamic bindings as the synchronous firing of appropriate nodes, (2) rules as interconnection patterns that direct the propagation of rhythmic activity, and (3) long-term facts as temporal pattern-matching subnetworks. The model is consistent with recent neurophysiological evidence that synchronous activity occurs in the brain and may play a representational role in neural information processing. The model also makes specific psychologically significant predictions about the nature of reflexive reasoning. It identifies constraints on the form of rules that may participate in such reasoning and relates the capacity of the working memory underlying reflexive reasoning to biological parameters such as the lowest frequency at which nodes can sustain synchronous oscillations and the coarseness of synchronization.