My main research interests focus upon different aspects of the analysis of natural language, and, therefore, the academic areas with which I am most closely involved include Philosophy, Theoretical Linguistics, English Literature, and Speech Recognition Technology. The following paragraphs briefly summarise some of my work in these fields.
My research in the Philosophy of Language and Theoretical Linguistics has recently focused upon the relationship between mathematics and linguistic theory during the 20th century. Sometimes I have approached this topic from an historical perspective (for instance, exploring Leonard Bloomfield's (1887-1949) knowledge of mathematics) while, at other times, I have concentrated upon analysing distinctive mathematical techniques that are present in contemporary linguistic theories (for instance, reconsidering the so-called recursive components). In addition, I am intrigued by the problem of language death, and I have written at length about Haida, an endangered language that is spoken on the Pacific Northwest Coast of North America.
I am also interested in the relationship between grammatical theories and contemporaneous literary texts from various historical periods. For instance, 18th century grammarians such as Robert Lowth (1710-1787) and Lindley Murray (1745-1826) produced grammar textbooks which exerted a profound influence upon literary perceptions during the 18th and 19th centuries, especially concerning notions of grammaticality, and it is revealing to explore the manner in which different authors (such as Christopher Smart, William Cobbett, and William Hazlitt) responded to, and negotiated with, formulations concerning grammatical correctness.
As a Research Associate in the Engineering Department, I have been involved in developing Automatic Speech Recognition (ASR) systems. I work on various problems, such as the difficult tasks of identifying `Structural MetaData' in input speech signals, and building Statistical Language Models for agglutinative languages. I helped to implement an effective morphological decomposition scheme for the Cambridge Arabic ASR system as part of the Global Autonomous Language Exploitation (GALE) project. I currently work on the Natural Speech Technology programme, seeking to develop ways of improving the grammaticality and acceptability of ASR output.