The autonomous computation of morpho-phonological structure in reading

Is morphological decomposition automatic? To address this question, we examine whether Hebrew readers decompose morphologically complex words when reading is not required, in the Stroop task. Morphological decomposition is assessed using two markers. One marker examines whether color-naming is modulated by morphologically complex words generated from color roots. For example, we compare words generated from the Hebrew root of “blue” displayed in either blue or an incongruent color. The second marker examines whether color-naming is sensitive to root phonotactics. Here we compare color-naming with words whose (color-unrelated) roots are either phonologically illicit (e.g., ssm) or well-formed (e.g., smm). Results suggest that morphological decomposition proceeds even when reading is discouraged, but unlike previous research with intentional reading tasks, Stroop-like conditions do not allow for a detailed representation of the root’s internal structure.