A meta-analysis of letter–sound integration: Assimilation and accommodation in the superior temporal gyrus

Abstract

Despite being a relatively new cultural phenomenon, the ability to perform letter–sound integration is readily acquired even though it has not had time to evolve in the brain. Leading theories of how the brain accommodates literacy acquisition include the neural recycling hypothesis and the assimilation–accommodation hypothesis. The neural recycling hypothesis proposes that a new cultural skill is developed by “invading” preexisting neural structures to support a similar cognitive function, while the assimilation–accommodation hypothesis holds that a new cognitive skill relies on direct invocation of preexisting systems (assimilation) and adds brain areas based on task requirements (accommodation). Both theories agree that letter–sound integration may be achieved by reusing pre-existing functionally similar neural bases, but differ in their proposals of how this occurs. We examined the evidence for each hypothesis by systematically comparing the similarities and differences between letter–sound integration and two other types of preexisting and functionally similar audiovisual (AV) processes, namely object–sound and speech–sound integration, by performing an activation likelihood estimation (ALE) meta-analysis. All three types of AV integration recruited the left posterior superior temporal gyrus (STG), while speech–sound integration additionally activated the bilateral middle STG and letter–sound integration directly invoked the AV areas involved in speech–sound integration. These findings suggest that letter–sound integration may reuse the STG for speech–sound and object–sound integration through an assimilation–accommodation mechanism.