Abstract words, due to the absence of material referents, are more difficult to learn and process than concrete ones (the concreteness effect). Here, we have investigated this effect by assessing the outcomes of concrete and abstract word learning provided by meaningful textual or graphical contexts. Five tasks including Free recall, Recognition, Lexical decision, Definition, and multiple-choice Semantic judgment tasks were used to assess the success of the acquisition of newly learnt words at lexical and semantic levels. Within-group analysis revealed the concreteness effect in both experimental groups. However, it was more pronounced after a graphical than a textual presentation (in three tasks vs one) supporting the crucial role of the non-verbal (imagery) system in learning of concrete words. Between-group analysis showed more accurate Recall and Definition of concrete nouns as well as better performance of both types of words in terms of Definition quality in the graphical group in comparison with the textual one. However, participants from the textual group recognised novel abstract words better than the other learners. Interestingly, only one between-group difference that was found for abstract words in the Definition task (definition quality) reached significance after the Bonferroni corrections for multiple comparisons. The results show that (1) concrete and abstract word processing may have partly distinct cognitive mechanisms, and (2) visual associations may play a crucial role in the semantic acquisition, especially for concrete words.
