Postersession 3
Poster #: 105
Topic: Speech and language (incl. deficits)
Friday, Sep 11, 2015
1st floor

Lexical MMN effects are too late: automatic lexical ERP enhancement in the oddball paradigm is already present in P1

Yury Shtyrov1 & Maria Lenzen2

1Center of Functionally Integrative Neuroscience, Aarhus University, Aarhus, Denmark
2Aachen, Germany

Previous MMN research made a strong case for automatic and rapid access of word memory traces in the brain, seen as a specific enhancement of MMNs to meaningful words at ~150ms. Recent evidence using non-MMN paradigms has, however, suggested an even earlier lexical access reflected in a word-specific enhancement of P1-type responses at ~60ms (MacGregor et al, 2012). This raises the question of why previous MMN studies missed early lexical effects. Two principle possibilities exist. One (in line with the MacGregor results) is that repetitive stimulus presentation leads to habituation of P1 amplitudes, obliterating any neurolinguistic effects. The other lies with signal-processing procedures: whereas MMN analysis typically employs 20-40Hz low-pass filter, P1 studies use a much higher range: a high-pass of >10 and low-pass of >70Hz.

To test this, we recorded EEG responses to a matched set of words and pseudowords in a typical lexical oddball paradigm known to elicit a robust word MMN enhancement, and subjected the data to P1-optimised analysis: bandpass filtered of 10-100Hz and ICA artifact correction. We found that, for all response types – standards, deviants and subtraction ERPs – P1 responses, peaking at ~30ms after stimulus recognition points, showed a significant amplitude increase for real words over acoustically matched pseudowords. The results suggest that the well-known lexical MMN enhancement is a secondary process preceded by first-pass lexical access already in the P1 range. With appropriate analysis techniques, MMN paradigms are well-suited to capturing both of these early neurolexical processes.