There is general consensus among usage-based researchers that the development of linguistic structure is driven by domain-general processes, but these processes are not always explained in light of psychological research on cognition. Chapter 3 provides a systematic overview of the various cognitive processes that are involved in language use and explains, in general terms, how grammar, usage and cognition are related. It is argued that language use involves a unconscious decision-making process that is determined by cognitive factors from three general domains: (1) social cognition (e.g., joint attention, common ground), (2) conceptualization (e.g., figure-ground, metaphor) and (3) memory-related processes (e.g., automatization, priming). The various processes can reinforce each other but can also be in competition. Of particular importance is the competition between other-oriented processes of social cognition and self-oriented processes of memory and activation spreading. One general advantage of the network approach is that it provides a natural explanation for the effects of frequency on usage and development.

The final chapter provides a short summary of the main results and highlights the major differences between the dynamic network approach and other theoretical frameworks.

Every construction has a particular ecological location in the grammar network that is defined by its relationship to other constructions in the system. Since the relationships between constructions are similar to those between lexemes, Chapter 10 begins with a short discussion of psycholinguistic research on the mental lexicon, which is commonly analyzed as an activation network (Dell 1986). There is abundant evidence that lexical access is influenced by several interacting factors including frequency, priming, similarity and neighborhood density, or family size. Considering research on sentences processing, L1 acquisition and language change, the chapter argues that the availability, or accessibility, of constructions is influenced by the same factors as lexical access, that is, by frequency, priming, similarity and neighborhood density, supporting the hypothesis that lexemes and constructions are organized in similar ways. Specifically, the chapter argues that grammar includes “construction families” that influence the use and the development of syntactic patterns (Diessel and Tomasello 2005; Wells et al. 2009).

Chapter 2 provides a background on the use of network models in different scientific disciplines and introduces the general architecture of the grammar network. The proposed network model has two levels of analysis: a lower level, at which linguistic signs, notably constructions, are defined by three different types of associations, or relations: (1) symbolic relations connecting form and meaning, (2) sequential relations connecting linguistic elements in sequence and (3) taxonomic relations connecting linguistic patterns at different levels of abstraction. Together the three relations define the basic units of speech, i.e., lexemes and constructions. Every unit constitutes a (local) network shaped by language use, but these networks also serve as nodes of a higher-level network that involves three other types of relations: (4) lexical relations connecting lexemes with similar or contrastive forms and meanings, (5) constructional relations connecting constructions at the same level of abstraction and (6) and filler-slot relations connecting particular lexemes with constructional schemas.