My research is focused around three main themes:
1. Formulaic language
I am interested in the role that lexically specific units (fixed phrases and frames with slots that can be filled with novel lexical material) play in first language acquisition as well as adult production and comprehension. In a series of studies based on various child corpora, I show that children's grammatical knowledge is best captured by means of such units, that the units can be derived from the input that children are exposed to, and that children become syntactically productive as they generalize over lexically specific units which share both formal and semantic characteristics. In another project (funded by the AHRC), which combined corpus and experimental methods, I argue that this also applies to complex syntactic constructions such as questions with long distance dependencies. This is important because most syntacticians believe that such constructions require complex syntactic machinery built on linguistically specific innate representations.
In more recent work, I have argued that adult informal conversational speech is also based on lexically specific units and the psychological reality, or otherwise, of corpus-based measures of collocation strength. My current research in this area explores collocations as an area of particular difficulty for second language learners and the role of collocational information in acquiring word meaning.
2. The nature of linguistic generalizations
A number of studies, including some of my own, have argued that syntactic productivity relies on stitching together stored chunks. I have also shown that learners don't necessarily represent generalizations that are demonstrably present in their language. This tension between linguists' descriptions (driven by principles of economy) and what speakers know about the grammar of their language (formulas and low-scope patterns) raises some interesting questions: How do speakers manage to behave as if they have extracted a generalization when in fact they haven't? Even more puzzlingly, how did the patterns come into being, how do they survive, and what is their ontological status? This, and the research on individual differences described below, has led me to the realization that languages belong to communities, not individuals: that is to say, individual speakers “own” only parts of their language and a new interest in the social and cultural processes that shape language, and how they interact with cognitive processes – something which I would like to pursue in future research.
3. Individual differences in linguistic knowledge
Most linguists assume, either implicitly or explicitly, that all native speakers of a language converge on (more or less) the same grammar. In a series of studies I demonstrated that is not the case: there are, in fact, considerable individual differences in speakers' knowledge of various areas of grammar, including quantifiers, passives, subordination, and some aspects of inflectional morphology. Many, though not all, of these differences are education-related. Those that are all show a characteristic pattern: highly educated speakers perform at or near ceiling, while less educated speakers show vast individual differences, with performance ranging from ceiling to chance (and sometimes below chance).
While most of my early work was devoted to demonstrating the reality of individual differences, in more recent projects I focus on exploring their causes. I have also expanded my interests to comparing native speakers and adult L2 learners, and examining differences in grammatical knowledge in the context of speakers' knowledge of vocabulary and collocational patterns.