We use all the techniques we can get our hands on to study these areas, depending on what method might provide the right type of answer given a certain problem. That is, we are not ‘methodological imperialists’. Beyond typical behavioral paradigms, the majority of the physiological work uses magnetoencephalography (MEG), although we also use electroencephalography (EEG) and functional magnetic resonance imaging (fMRI). We have used PET and TMS as well, although these are not techniques we commonly employ. Unifying themes in the lab include (i) what is the 'parts list' for speech perception/lexical representation/processing, both from the linguistic and from the neurobiological points of view, i.e. what are the primitives and the elementary operations; and (ii) how do temporal mechanisms encode or represent information in the auditory cortex and form the basis for speech perception and auditory cognition (latency-based codes, oscillatory activity, phase, etc.)? Are here temporal primitives? Our experiments use stimuli of varying complexity and ecological relevance, ranging from pure tones to FMs to ripples to syllables to words to connected speech.
One principle guiding some of the work is based on the assumption that the problem in speech perception is to execute the transformations from ‘vibration in the ear to abstraction in the head.’ Speech perception and successful lexical access means making contact with the internal mental representations that have specific characteristics. Given this perspective, two further topics that are increasingly investigated in the lab concern multi-sensory (audio-visual) speech perception as well as lexical processing. Questions regarding the nature and processing of lexical representations are investigated using a combination of behavioral and MEG studies.
Some of the topics on the brain basis of speech & language that we worry about are discussed more or less informally on a new blog: Talking Brains