Recent neurophysiological evidence suggests that a hierarchical neural network of low-to-high level processing subserves written language comprehension. While a considerable amount of research has identified distinct regions and stages of processing, the relations between them and to this hierarchical model remain unclear. Magnetoencephalography (MEG) is a technique frequently employed in such investigations; however, no studies have sought to test whether the conventional method of reconstructing currents at the source of the magnetic field is best suited for such across-subject designs. The present study details the results of three MEG experiments addressing these issues. Neuronal populations supporting responses to low-level orthographic properties were housed posteriorly near the primary visual cortex. More anterior regions along the fusiform gyrus encoded higher-level processes and became active ~80ms later. A functional localiser of these early letter-specific responses was developed for the production of functional regions of interest in future studies. Previously established response components were successfully grouped based on proximity to the localiser, which characterised location, latency and functional sensitivity. Unconventional anatomically constrained signed minimum norm estimates of MEG data were most sensitive to the primary experimental manipulation, suggesting that the conventional unsigned unconstrained method is sub-optimal for studying written word processing.

Speech is an inherently noisy and ambiguous signal. To fluently derive meaning, a listener must integrate contextual information to guide interpretations of the sensory input. Although many studies have demonstrated the influence of prior context on speech perception, the neural mechanismssupportingtheintegrationofsubsequentcontextremainunknown.UsingMEGtorecordfromhumanauditorycortex,weanalyzed responses to spoken words with a varyingly ambiguous onset phoneme, the identity of which is later disambiguated at the lexical uniqueness point. Fifty participants (both male and female) were recruited across two MEG experiments. Our findings suggest that primary auditory cortex is sensitive to phonological ambiguity very early during processing at just 50 ms after onset. Subphonemic detail is preserved in auditory cortex over long timescales and re-evoked at subsequent phoneme positions. Commitments to phonological categories occur in parallel, resolving on the shorter timescale of ϳ450 ms. These findings provide evidence that future input determines the perception of earlier speech sounds by maintaining sensory features until they can be integrated with top-down lexical information.

No abstract