CNL (Early) Lunch

Doug Saddy (U. Potsdam)

Timing is Everything

Tuesday March 19th, 11:00am, 3416 Marie Mount Hall

     CreRecordings of the electro-magnetic fields generated in the cortex
provide important information about the time course and timing of cortical
events underlying higher order cognitive processes. In this short talk I will
review some critical shortfalls associated with the traditional
averaging approach to analysing such data which undermine the temporal sensitivity of the ERP/EEG recordings. I will sketch out an alternative approach that relies on recent advances in non-linear Complex Systems techniques.
Complex systems approaches do not require the assumption of linearity or
ergodicity associated with conventional averaging, provide a new set of tools for assessing time dependent effects and provide a range of alternatives to
the problem of base line determination. I will exemplify the benefits of
non-linear analyses by presenting recent results from two ERP sentence
processing paradigms (time permitting). One set of data focuses on
modulation within the first 120 ms and is interpreted as reflecting
parsing confidence. ERP effects in such early time windows are generally
associated with stimulus detection and attention. The finding of systematic
variation in this time window in response to higher order processing information
provides important new evidence regarding predictive aspects of the time
course of language processing. A second new result comes from studies of
the processing of Polarity constructions. In a series of ERP experiments we
have found differential processing for positive and negative polarity items
as well as structure sensitive influences from non-licensing operators. In
the standard averaging approach the Evoked components associated with the
appearance or lack of intrusive operators appear to be nested with
respect to amplitude and duration while non-linear analyses reveal different
temporal displacements for these effects. These results demonstrate
that the traditional averaging approach can hide contrastive behaviour in the
earliest time windows and obscure temporal distinctions. The adoption
of a Complex Systems approach provides a more sensitive and robust suite of analytic tools and makes assumptions about the nature of cortical
processing that are in-line with our current understanding of the nature of
cortical computation.