Linguistics is a broad interdisciplinary discipline with research in many different areas concerning the nature, processing, development and evolution of language. The following areas are represented within Linguistics at the Centre for Languages and Literature (CLL):
Phonetic variation is concerned with the diversity of speech sounds and their pronunciation which can be found in various languages and dialects, for different genders, ages, status of health and moods. Variation in pronunciation also occurs for second language learners who can be influenced by the sound system of the learner's first language. Phonetic variation occurs in vowels, consonants and prosody, e.g. melodic patterns (cf. Prosody). Sound variation and similarities among languages is also studied within the field of Linguistic typology.
- Studies of Swedish dialects carried out at our department deal with typical pronunciations of a phoneme within a specific dialect or a comparison of the pronunciation of a particular phoneme in different Swedish dialects, e.g. r-sound variation or diphthongs.
- Variation in pronunciation based on second language learning is studied with the goal of increasing the awareness in teachers of Swedish as a second language. Phonetic analyses of pronunciation difficulties - foreign accents - and comparisons of the Swedish sound system with those of the first languages of the students may provide tools for teachers to help individual students improve their pronunciation.
Simulekt: Simulation of Swedish prosodic dialect types
Funktions- och produktionsbaserad modellering av svensk prosodi
Exotic vowels in Swedish: an articulographic study of palatal vowels [VOKART]
Prosody deals with languages' melodic and rhythmic characteristics as well as their function in communication. Research areas at our department include:
- Phonological and acoustic modeling: A prosody model for Swedish (the "Lund Model") has been developed, which describes how word-level prosody (Accent I, Accent II) and sentence intonation are realized in different dialect types. Recently, the model was refined by a project using the Swedia2000 database (recordings from 100 dialects). A current project studies discourse-related meanings of sentence intonation – for example, how we highlight information in different contexts – and builds a new, quantitative model for Central and Southern Swedish (see Speech Technology).
- Neurocognitive processing: Grammar-prosody interactions and their processing in the brain are studied using neurophysiological methods (see Neurolinguistics).
- Prosodic typology: Prosodic research focuses on Swedish dialects (see 1) as well as Asian languages such as Kammu (spoken in Northern Laos) and Formosan languages (Taiwan) (see Linguistic Typology).
Function- and production-based modeling of Swedish prosody
Images of tones: fMRI-studies on the processing of prosody in the human brain
Information structure, prosody and the origins of Formosan relativization
Within the field of phonetics, speech production deals with how we use our respiratory system and vocal tract in speech, e.g. how we move our lips and tongue to articulate a certain vowel or consonant. Speech production is studied in a number of research projects within some of the following fields:
- Articulation of Swedish vowels is the study of how speakers of different Swedish dialects move their lips and tongue to pronounce vowels. We also look at dialectal differences, including South Swedish diphthongisation and the so called Viby-i in Stockholm and Gothenburg dialects.
- Development of motor speech control can be investigated by comparing speech movements in children of various ages with those of adults. One method is to study the variation in speech movements between several repetitions of the same utterance to see if we improve our speech motor control as we grow up.
Language and speech technology is about how computers can be programmed to process, understand and produce human language. The department conducts research in the following areas:
- text-to-speech conversion: how an arbitrary text can be converted to speech and read aloud by a computer. Research is carried out here both on how to convert written text into phonetic representations and on how to produce sounds in the form of spoken language based on these representations.
- dialectally colored synthesis: together with the Prosody field research is conducted on how to vary the text-to-speech pronunciation so that it can mimic different dialects within the Swedish-speaking area.
- software for automatic speech analys: development of methods and software for systematically analyzing speech databases.