Abstract
In order to make use of novel experiences and knowledge to guide our future behavior, we must keep large amounts of information accessible for retrieval. The memory system that stores this information needs to be flexible in order to rapidly incorporate incoming information, but also requires that memory traces are stable and resistant to forgetting over a long period of time. More than a century of research on this topic has led to the insight that memory is not a unitary system but rather depends on a complex interaction between multiple memory systems that are subserved by different neural networks. The Complementary Learning Systems (CLS) model, derived from the standard theory of systems level memory consolidation, posits that a novel memory is initially encoded by the hippocampal memory system, and subsequently undergoes a shift to the neocortical memory system. This process serves to stabilize and integrate novel information into persistent long-term memory without causing catastrophic interference to existing knowledge. This chapter introduces the concept of multiple memory systems, which brain structures are engaged, and how they interact through the process of consolidation. Memory consolidation takes place both at the microscopic level, in the form of changes in synaptic strength, and at the whole-brain systems level, where it refers to a representational shift between different neural networks. We focus mainly on the systems level, and discuss how the hippocampal (episodic/contextual) and neocortical (semantic/content) memory systems interact during specific stages of memory formation and consolidation. Of particular relevance for the entrenchment of linguistic information is the idea that memory consolidation processes support qualitative changes in the representation of acquired information. For instance, consolidation is thought to underlie the generalization of overlapping information from multiple episodic memories into abstract representations, and the integration of novel information with existing knowledge. We will touch on the role of sleep in offline memory consolidation, which has been shown to boost memory stabilization as well as generalization and integration. Linguistic information, just like other memory representations, is thought to be acquired and retained in our memory system through the process of memory consolidation. The final section of the chapter focuses on recent findings related to linguistic memory consolidation, from word-form to semantics and syntax. These findings broadly support and add refinements to the CLS model of memory consolidation.
