Research conducted by the Linguistic research unit covers four key areas:
Contrastive linguistics is based on rigorous, systematic comparison of the linguistic characteristics of two or more languages from both a theoretical (for better understanding of the differences and similarities between the languages studied) and a practical perspective (implications for language learning, translation, bilingual lexicography, etc.) These studies make use of large bilingual or multilingual corpora containing a wide variety of texts, ranging from newspapers and novels to scientific texts and parliamentary debates.
Research on linguistic variation looks at the various factors that determine how speakers use language. These factors may be related to the traits of the speaker, his or her objectives, the communication context, etc.
These variation phenomena are studied at the most elementary levels of linguistic organisation (phonology, morphology), and at the most complex (prosody, discourse). Since variation phenomena are particularly prevalent in spoken language and in informal contexts, the study of linguistic variation most commonly involves recording spontaneous verbal interactions and analysing corpora of transcriptions of these verbal interactions.
Linguists at the Institute for Language and Communication (ILC) have developed a large corpus of oral interactions in French and have acquired expertise in building, processing and managing oral corpora. This empirical research on the diverse ways in which language is actually used helps to develop theoretical models of linguistic variation and change and examines the relationship between linguistic norms and observed usage, which in turns raises important societal questions relating to both linguistic policy and the knowledge required for successful oral communication.
Linguistics applied to language learning and teaching links the theoretical analysis of the process of foreign and second-language acquisition and the teaching practices and tools that can be used to optimise this process. Research focuses on the factors that influence the learning process: both learner-related, such as age or mother tongue, and situation-related, such as the learning task or context (home environment, classroom environment, immersion). For analysis purposes, the emphasis is on using authentic data (oral or written) produced by learners, known as learner corpora. By comparing this data with native-speaker data, those aspects of language acquisition that present problems for learners can be identified, and a range of teaching tools best suited to their needs proposed. These applications are incorporated into, for example, reference tools (learner dictionaries, grammars, terminology databases, etc.), CALL tools (computer-assisted language learning) and corpus consultation interfaces to help users write in a foreign language.
The objective of natural language processing is to develop software capable of generating or analysing documents in natural language (for the purpose of text generation, speech synthesis, information retrieval, machine translation, spellchecking, etc.). These applications usually rely on standard linguistic descriptions and corpus-based language models.
UCL's research in this field focuses mainly on information search and extraction, two topics that led to the creation of a spin-off in 2012: Knowbel Early Tracks. Other fields of research are also explored, such as assessment of text difficulty for the purpose of automatic text simplification, opinion detection, or style variations for speech synthesis.