Semantic influences on visual processing of compounds and pseudo-compounds: a masked priming study in Swedish

Abstract Semantic transparency has been shown to influence the processing of morphologically complex words, yet its role in early visual word recognition remains debated. The present study reports two masked priming analyses using Swedish noun–noun compounds. First, we compared priming effects from compounds, pseudo-compounds, and orthographic control words to assess the relative contributions of form-and-meaning overlap versus form-only overlap in early lexical processing. Second, focusing exclusively on compounds, we examined whether graded differences in constituent transparency modulated priming magnitude. The results revealed robust facilitative priming from true compounds, whereas pseudo-compounds and orthographic controls produced no reliable effects. Within the compound condition, priming effects were not modulated by degree of constituent transparency, operationalised as a continuous measure of semantic relatedness. Thus, while morphological relatedness reliably facilitated lexical decisions, graded semantic transparency among true compounds did not influence priming under masked conditions. We argue that, overall, these findings challenge accounts positing a semantically blind morpho-orthographic decomposition stage and instead support word-based, connectionist approaches in which masked priming reflects experience-dependent sensitivity to reliable form-meaning mappings.