We report here our studied on the infl uence of the lexical frequency of Russian words on the latency and amplitude of mismatch negativity (MMN) in auditory event-related potentials (ERP). ERP were recorded in a multideviant oddball paradigm by presenting different-frequency Russian words and pseudowords. These experiments showed that the pattern of intrinsic MMN differed signifi cantly between words with different lexical frequencies (p = 0.001) – the higher the frequency, the greater the amplitude and the shorter the latent
period of the intrinsic MMN of the words. It is suggested that the greater amplitude and shorter latency of MMN for high-frequency words as compared with the pattern of MMN for low-frequency words is due to activation of memory traces for these words, these being stored in the cerebral cortex as distributed neuron populations. The suggestion that there is superfast access to lexical information during speech perception is confi rmed, with access being possible 100–200 msec after presentation of a word. The ratio of MMN
amplitudes for different pseudowords was somewhat reminiscent of data on MMN for words (analogs of high-frequency words produced higher-amplitude responses, while analogs of low-frequency words produced
weaker responses, with no signifi cant difference between low- and intermediate-frequency analogs), though MMN amplitudes for pseudowords were signifi cantly greater and latent periods were signifi cantly
longer. Increases in the amplitude and latency of MMN to pseudowords as compared with MMN to words is associated with later and uncertain recognition of rarely encountered low-frequency words and completely unfamiliar stimuli, which are later classifi ed as signals of a different category.
Original languageEnglish
Pages (from-to)1043-1050
Number of pages8
JournalNeuroscience and Behavioral Physiology
Volume47
Issue number9
StatePublished - 2017

    Scopus subject areas

  • Neuroscience(all)

    Research areas

  • event-related potentials, word frequency, speech perception, mismatch negativity.

ID: 9569897