Running head : LOGICAL WORD LEARNING 1 Logical word learning : The case of kinship

In this paper, we propose a framework for conceptual development through the lens of program induction. We implement this framework to model the acquisition of kinship term concepts, resulting in the first formal learning model for kinship acquisition. We demonstrate that our model can learn any kinship system consistent with its learning data using cross-linguistic data from English, Pukapuka and Turkish. More importantly, the behavioral patterns observed in children learning kinship terms, under-extension and over-generalization, fall out naturally from our learning model. We conducted interviews to simulate realistic learning environments and demonstrate that the characteristic-to-defining shift is an epiphenomenon of our learning model in naturalistic contexts. We use model simulations to discuss the influence of simplicity and learning environment on the order of acquisition of kinship terms, positing novel predictions for the learning trajectories of kinship terms under different conceptual architectures for learning inter-related systems. We conclude the paper with a discussion of how this model framework generalizes beyond kinship terms and the limitations of our model.