Do episodic memory representations mediate masked priming effects in the lexical decision task? A confrontation of the prospective and retrospective accounts

Date: 27 November 2015

Venue: UAntwerp, Stadscampus, Promotiezaal Grauwzusters - Lange Sint-Annastraat 7 - 2000 Antwerp

Time: 5:00 PM - 7:30 PM

PhD candidate: Lien Van Abbenyen

Principal investigator: Dominiek Sandra

Short description: Phd defence Lien Van Abbenyen - Faculty of Arts



Abstract

Masked priming is a frequently used paradigm in psycholinguistics but there are two radically different accounts on the source of masked priming effects. The classic or prospective account, proposed by Forster and Davis (1984), situates that source in the mental lexicon, while the retrospective account proposed by Bodner and Masson (1997, 2003, 2014) situates it in episodic memory.

In three separate series of experiments, we have tried to find out which memory system is the true source of masked priming effects in lexical processing tasks, more particularly, in lexical decision tasks (LDT). Our chief goal was to find out whether masked priming effects in LDTs are (in part) mediated by episodic representations. This was done by investigating whether priming in a LDT (partially) depends on two consecutive processes, namely (a) the creation of an episodic trace by the masked prime and (b) the recruitment of this episodic trace during subsequent processing, as claimed by the retrospective account.

The overall conclusion that can be drawn from this doctoral dissertation is that our results are most in line with the prospective account of masked priming. Our empirical research strongly indicates that masked primes do create episodic traces, both in an ERT and a LDT, but that these episodic representations are only recruited and used when the experimental task explicitly requires the activation of episodic memory. As a result, episodic representations only have an effect on subsequent target processing when the experimental task explicitly requires the use of these episodic representations (e.g., an ERT). As the use of an episodic representation cannot help participants decide whether a stimulus is a word or not (i.e., participants must simply wait whether or not a lexical representation is accessed), no masked episodic priming effects are found in a LDT. This outcome is in line with the prospective account.



Contact email: lien.vanabbenyen@uantwerpen.be