Morphological segmentation and lexical processing of nouns and verbs: evidence from Segment Shifting in Italian

Some psycholinguistic models (e.g., Taft, 1994) emphasize the role of morphological structure in lexical access and describe lexical processing as the result of the activation of individual morphemic constituents. Distributional data suggest that the morpheme access process is sensitive to the differential tendency of affixes to also appear as pseudoaffixes, i.e. orthographic patterns that do not play a morphological role and may act as confounding factors in morphological analysis.
The probability that letter patterns of Italian are processed as morphemes during lexical access was tested in two segment shifting tasks (Feldman, 1991). This technique can be used to highlight the internal structure of words by asking participants to shift a final letter segment from a source word (e.g., “-en” in “garden”) to an isolated word (e.g., “hard”) and produce a new output (e.g., “harden”, see Feldman, 1991).
Using final segments with varying degrees of morphological (and orthographic) confusability, we test the effectiveness of morphological parsing and the mechanisms of (pseudo-) suffix shifting with or without (pseudo-)morphological structure. The results support the hypothesis that verbs are more likely to activate morphemic components. The reliability of “suffix consistency” in predicting the shifting is discussed in terms of an interplay between sublexical and lexical information during processing.