Where does the imageability effect in working memory come from?

Working memory is influenced by several linguistic long-term memory factors. One of these factors is the imageability effect, in which concrete or high imageability words (e.g. cat, tree) are better recalled as compared to abstract or low imageability words (e.g. phase, doubt). The precise origin of this effect is however a matter of debate. In this study, we assessed the hypothesis that the occurrence of the imageability effect in WM tasks is conditioned to the implementation of semantic elaborative processes. Thirty-nine young healthy adult participants were invited to encode and immediately recall six-items lists composed of either high or low imageability words. In half of the trials, participants had to perform a secondary interfering task involving an animal classification judgement task aimed at preventing the implementation of semantic elaborative processes on the memoranda. We found a very strong impact of the imageability dimension and the interfering task (BF10 > 100) on recall accuracy but there was no reliable evidence for a reduction of the imageability effect in the interfering condition (BF10 = .833). Our results suggest that semantic elaborative processes are not a necessary condition for the observation of imageability effects in WM.