Is morpho-orthographic decomposition purely orthographic? Evidence from masked priming in the same–different task

Abstract

Two experiments used the cross-case same–different task to test whether the orthographically driven morphological decomposition effects that have been found in the lexical decision task are obligatory. Experiment 1 replicated the manipulation used by Duñabeitia, Perea, and Carreiras (2007), testing transposed-letter (TL) priming effects spanning the boundary between the affix and the stem. In contrast to their finding observed with the lexical decision task, TL priming effect did not vanish with polymorphemic or pseudomorphemic words. Experiment 2 used the manipulation used by Rastle, Davis, and New (2004), comparing the effects of polymorphemic affixed words (e.g., walker), pseudo-affixed words (e.g., corner), and nonaffixed monomorphemic words (e.g., brothel) in target word recognition. Unlike the results observed in the original lexical decision study, equal priming effects were observed with all three types of words. These results suggest that the presence of an orthographically defined subunit (affix) is not sufficient to drive morphological decomposition processes.

Keywords:

• Letter transpositions
• Masked priming

Acknowledgements

This research was supported by grants, PSI2009-08889 and CONSOLIDER-INGENIO2010 CSD2008-00048 from the Spanish Government, and the Australian Research Council Discovery Project grant (DP0877084). The authors are grateful to Jukka Hyönä and to two anonymous reviewers for their helpful comments on earlier drafts.

Notes

¹In any case, it should be noted that the significance levels corresponding to the F ratios over items in the present experiments essentially mimicked the reported significance levels of the F ratios over participants.

²As seen, suffixed words were responded to faster than prefixed words in the “same” and “different” response analysis. A similar effect was found in Duñabeitia et al. (2007). A priori, we have no theoretical reason for this effect. We believe that this effect might be due to uncontrolled factors that could have potentially enhanced processing of suffixed words over prefixed words, but we have no clear explanation for this result. In any case, considering that the critical manipulations were carried out within-items, we believe that slight processing differences between the different target sets are not critical for the ultimate aim of the present study (especially if these effects are mainly evident in the “different” response analysis).

³The same complete set of items had been previously used in a Spanish two-choice lexical decision masked priming experiment, showing the classical “corner” effect (Diependaele, Duñabeitia, Morris, & Keuleers, 2010).

⁴It should be noted that there is a current debate on the issue of semantic transparency and how it affects masked morphological priming in the lexical decision task (see Feldman et al., 2009). Some recent data show that the magnitude of the priming effects observed for opaquely related pairs like corner–CORN is lower than the magnitude of the priming effects observed for transparently related pairs like walker–WALK. However, although relevant, this issue is not the focus of the present study. Rather, the critical point for our purposes is whether the presence of an orthographically defined morphological unit triggers a morphological decomposition process (i.e., whether the amount of priming would differ between pseudo-affixed words like in corner–CORN and merely orthographically related pairs such as in brothel–BROTH.