Abstract

Almost all types of learning involve, to some degree, the ability to encode regularities across time and space. Two decades ago, statistical learning (SL) was proposed as a powerful domain-general mechanism for processing a wide range of regularities. However, because of its rather narrow focus, SL research has largely failed to deliver on the wide-reaching promise of SL as a theoretical construct. This is mainly due to SL being investigated largely a separate ability, isolated from other aspects of cognition.

This theme issue fosters a transition to studying SL as an integral part of different cognitive systems, taking into consideration complementary perspectives from neurobiology, computation, development and evolutionary studies. This collection of work shows that SL is not simply learning to accurately represent the regularities of the environment. Rather it is a product of the complex interaction between environmental statistics, the neurocomputational principles of the cognitive systems in which learning takes place, and pre-existing biases due to previous experience and/or architectural constraints of the brain. This new perspective will enable SL to impact a broad range of theories related to language, vision, audition, memory and social behaviour.

Copyright Notice (courtesy of David Plaut): The documents distributed here have been provided as a means to ensure timely dissemination of scholarly and technical work on a noncommercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.