Xu F. & Griffiths T. (2011) Probabilistic models of cognitive development: Towards a rational constructivist approach to the study of learning and development. Cognition 120(3): 299–301.
Probabilistic models of cognitive development: Towards a rational constructivist approach to the study of learning and development.
Cognition 120(3): 299–301.
Excerpt: The papers that appear in this special issue bring together researchers working on probabilistic models of cognition with developmental psychologists, to consider how “rational constructivism” could shed light on some of the challenges of understanding cognitive development. Our goal in collecting these papers together is to illustrate that this new approach to the study of cognitive and language development has already shown a lot of promise – both computational modeling and empirical work have opened up new directions for research, and have contributed to theoretical and empirical advances in understanding learning and inference from infancy to adulthood. The rational constructivist view embodies two key ideas: one is the commitment that the learning mechanisms that best characterize learning and development from infants to adults are a set of rational, inferential, and statistical mechanisms that underlies probabilistic models of cognition. The application of these domain-general mechanisms may give rise to domain-specific knowledge. The second is to call into question both the nativist characterization of innate conceptual primitives (e.g., is object or agent an innate concept?), and the empiricist’s characterization of a newborn infant with nothing but perceptual primitives and associative learning mechanisms. It is an open question how best to think about the initial state of a human learner. Perhaps in addition to a set of perceptual (proto-conceptual?) primitives, the infant also has the capacity to represent variables, to track individuals, to form categories and higher-order units through statistical analyses, and maybe even the representational capacity for logical operators such as and/or/all/some – these capacities enable the infant to acquire more complex concepts and new learning biases. As such, this view departs from the traditional Piagetian view of development in at least two ways – development does not progress through stages, driven by qualitative changes in the child’s logical capacities, and development does not start with sensory-motor primitives and a lack of differentiation between the child and the world. Instead, the construction of new concepts and new learning biases is driven by rational inferential learning processes. At the moment, there is by no means any consensus on these issues. With further empirical and computational work, a more detailed explication will emerge.
We will upload a full textversion shortly.