Productivity and priming: Morphemic decomposition in Arabic![Sample our Language & Literature journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5941/TOC-Subject-Sample-2021-Language-Literature.jpg"})
Abstract
Word formation in Arabic involves the interleaving of two abstract morphemes—a root consisting exclusively of consonants and conveying semantic meaning, and a word pattern comprised primarily of vowels and conveying phonological and morpho-syntactic information. In masked and cross-modal priming experiments, we probed the processing relationship between these two morphemes during word recognition by examining the roles of word pattern and root productivity (family size) in producing word-pattern priming in Arabic deverbal nouns. Co-varying word pattern and root productivity in a 2×2 design, we found that priming was determined entirely by the productivity of the root. Even very productive word patterns did not prime if they appeared in the context of an unproductive root. This pattern of results, which is identical in cross-modal and masked priming, indicates the importance of the root in driving the on-line decomposition of Arabic surface forms into their constituent morphemes.
Acknowledgements
This work was supported by the Medical Research Council UK (U.1055.04.002.00001.01). The authors would like to thank Abdallah Megbli, Headmaster of the High School of Tataouine, Tunisia, for providing access to the participants who took part in the study. We thank Ian Nimmo-Smith, and other members of the Language Group at the MRC Cognition and Brain Sciences Unit, for their help. We are also thankful to the anonymous reviewers for helpful comments on earlier versions of this work.
Notes
1Note that our experience of masked priming using Arabic script shows that even at SOAs of 80 ms the prime still seems strongly masked and not detectable by the participants (Boudelaa & Marslen-Wilson, Citation2005).
2Family size will normally correlate with both lemma and word frequency. In all of the studies reporting family size effects, this is in the context of regression analyses where the effects of frequency have been partialled out.
3Note that this study used a different definition of productivity, defined as the continuing use of the affix in novel word formation. This is not the same as, though correlated with, a simple type count.
4Note that the final ~un at the end of each word is the indefinite article which is part of the phonological representation of the word but has no corresponding graphemes in the written form.
6Note that what we are reporting in this and the next experiment are partial effects of significant predictors. This means that the effects of a given predictor are evaluated given (a) that other predictors are in the model and (b) that they are held constant.
7Given the results reported by Martin et al. (2005) for Hebrew, in separate analyses we also assessed the effects of root family size broken down into transparent and opaque family members. There were significant facilitatory effects on overall RT of target transparent family size (β=–0.02908; t(4816)=–2.77; p<.006) and of the transparent/opaque ratio for the target (β=–0.07273; t(4816)= –3.33; p<.001). In both cases, neither there was effect on RT associated with the prime root family size properties, nor was there any sign of Martin et al.'s (2005) finding for Hebrew that opaque family members were associated with an increase in overall RT. In the analyses of the effects of these variables on the amount of priming, evaluated in terms of interactions with Prime Type, we see small facilitatory effects of both prime and target opaque family size (respectively, β=0.01720; t(4816)=2.22; p<.03; β=0.01293; t(4816)=1.935; p=.05). There was also an effect for the target ratio (β=–0.07273; t(4816)=3.329; p=.001) but not the prime ratio.
