RB_ERP_cogneuro.htmlTEXTHTML .Ze Illusions electrified: ERP priming effects are induced by illusory words

"Illusions electrified": ERP priming effects induced by illusory words

Eva Ludowig1, Michael Niedeggen1, Martin Heil2, Bettina Rolke3 & Catherine L. Harris4

1Institute of Experimental Psychology II, Heinrich-Heine-University of Duesseldorf, Germany

2Department of Psychology, University of Hull, UK

3Psychological Institute, University of Tuebingen, Germany

4Department of Psychology, Boston University, USA


The perception of an illusory word can be induced by a partial repetition blindness: If two words (W1, W2) presented within a rapid serial visual presentation (RSVP) share the same string of letters ("CREEP" — "SHEEP"), these letters will be inhibited, thus generating a free floating fragment (FR1, "SH"). FR1 is likely to be linked to a subsequently presented fragment (FR2, "IFT") if both combine to a meaningful word (Illusion: "SHIFT").

Using event related brain potentials (ERPs), we examined whether the illusion is generated at a lexical level of word processing. RSVP streams containing two meaningful words (W1, W2) and a word fragment (FR2) were presented to the subjects (n=8). A preceding prime word was semantically unrelated, or related either to the real word (W2), or the illusion.

The probability of naming W2, but not the illusion, was at about 31% for unrelated, and increased to 55% for related primes. The corresponding probabilities for naming the illusion, but not W2, were at about 44% and 67%, respectively. If W2 was named, the semantic priming effect (unrelated — related) was reflected in an N400 effect peaking at 600 ms at right-frontal leads. Most importantly, an N400 effect was also evoked by illusory words but its latency was shortened (420 ms), and its topographical distribution differed (left-posterior leads).

Our preliminary ERP results support the notion that illusory words are already available at lexical level but the differences in timing and topography suggest that priming of illusions facilitates word-form processing rather than higher semantic associations.