Recursion is a powerful technique in computer science. If wecan express a solution to problem X in terms of solutionsto smaller instances of the same problem X—and wecan solve X directly for the “smallest”inputs—then we can solve X for all inputs. There aremany examples. We can sort an array A by sorting the lefthalf of A and the right half of A andmerging the results together; 1-element arrays are by definition alreadysorted. (That’s merge sort.)

Morphosyntax describes the form and function of grammatical constructions in the world’s languages. The form of constructions includes both syntactic structure and relevant morphology. The function of constructions includes both information content (semantics) and information packaging of the content. The same semantic content can be packaged in different ways. The approach in this textbook is crosslinguistic and empirical: we compare grammatical constructions across languages and describe patterns of variation, universals constraining variation, and diachronic processes that give rise to the variation. Crosslinguistic comparison is done using crosslinguistically valid concepts (comparative concepts). Crosslinguistic constructions are defined as all grammatical forms expressing a particular function. Strategies are crosslinguistically defined formal means for expressing a function. The analysis of grammatical structure in a particular language is the categorization of constructions in the language by their form and their function. Language-particular analysis of constructions and crosslinguistic analysis of constructions can be united via the function of the construction.

In the first chapter, we learnt that the basic units of a Construction Grammar analysis are FORM-MEANING pairings of varying degrees of schematicity. In this chapter we will see that for most Construction Grammarians, constructions are not just descriptive tools for linguistic analysis. They also maintain that constructions are in fact the basic unit of our mental grammars. This obviously raises the question of how people in general, and children in particular, acquire constructions. The majority of constructionist approaches answer this question by claiming that people acquire constructions through actual language use and with the help of general cognitive processes. These approaches are therefore known as 'usage-based'. In this chapter, we will explore a Usage-based Construction Grammar account of language acquisition, survey the types of data sources used in such approaches and discuss how we have to refine our definition of constructions in light of the results of usage-based studies.

In the last chapter, we explored word constructions and the basic phrasal constructions that they appear in. In addition to that, we saw that English has a great number of schematic and substantive idioms that can best be described as constructions. In the present chapter, we continue this approach and investigate how syntactic phenomena, such as argument structure (which tells us what happened) and its interaction with active and passive voice (which represent different vantage points from which to construe events), as well as tense and aspect (when and how something happened), can be analysed within Usage-based Construction Grammar. Moreover, we also look at abstract constructions for the various clause types (e.g., declaratives, interrogatives and imperatives, all of which basically express speakers’ illocutions). Finally, we also look at how Information Structure constructions can be used to structure information in a way that is most beneficial for a specific hearer in a discourse context.