The role of the root morpheme in mediating word production in Hebrew

Abstract

Using the picture–word interference paradigm with auditorily presented distractors, we investigated the effect of morphological processes on producing words, as distinct from phonological and semantic effects. The study was performed in Hebrew, a language with a nonconcatenated morphology, and focused on the derivational root morpheme, which is distributed nonsequentially throughout the word. The semantic, morphological, and phonological relatedness between the auditory distractor word and the picture name was manipulated, as well as the Stimulus Onset Asynchrony (SOA) between the distractor and picture presentation. Compared with an unrelated distractor, the results revealed a facilitatory morphological effect within the time window of SOAs from −200 to 300 ms. This effect was distinct from the semantic and phonological effects in the time course and/or direction. These results extended both the experimental evidence for the role of morphology in language production to a language with a nonconcatenated morphological structure, and previous findings in word perception in Hebrew to the realm of word production.

Keywords:

• Speech production
• Picture–word interference
• Morphology

Acknowledgements

This research was supported by a grant from the Israeli Science Foundation (#179/09) to Avital Deutsch. We thank Ram Frost for helpful comments on the manuscript, and Roni Pener-Tessler for her extensive help and assistance in running the experiments.

Notes

¹In fact there are a few specific cases where the word-pattern does not definitively determine the word class of adjectives or participles. This complication is connected with the linguistic definition of word class, which is beyond the scope of the present discussion.

²Since not all pairs met the criteria in the first round, we had to change some of the pairs and conduct additional rounds of ratings. We used the criterion of having at least 10 participants rate each pair of words.

³The evaluation of the semantic relation between the targets and the morphologically related pairs using the same procedure described in Section “Methods” revealed a mean semantic relation of 4.78, with a fairly high variance of 2.34 and a range of 1.6–6.5.

⁴The issue of representation at the word level, as well as the question of at what phase of lexical access written stimuli are translated into a phonological representation, belongs to a different, although related, issue—namely, models of printed-word recognition. Admittedly, there is ample evidence from written-word recognition in Hebrew to support the idea that orthographic units are mapped into phonological representations at very early phases of lexical access in reading (e.g., Frost, 1998). These claims of early phonological decoding of orthographic information can easily accommodate the notion that the root morpheme that mediates lexical access in written word perception is represented phonologically.