Effects of syllable preparation and syllable frequency in speech production: Further evidence for syllabic units at a post-lexical level

Abstract

In the current paper, we asked at what level in the speech planning process speakers retrieve stored syllables. There is evidence that syllable structure plays an essential role in the phonological encoding of words (e.g., online syllabification and phonological word formation). There is also evidence that syllables are retrieved as whole units. However, findings that clearly pinpoint these effects to specific levels in speech planning are scarce. We used a naming variant of the implicit priming paradigm to contrast voice onset latencies for frequency-manipulated disyllabic Dutch pseudo-words. While prior implicit priming studies only manipulated the item's form and/or syllable structure overlap we introduced syllable frequency as an additional factor. If the preparation effect for syllables obtained in the implicit priming paradigm proceeds beyond phonological planning, i.e., includes the retrieval of stored syllables, then the preparation effect should differ for high- and low frequency syllables. The findings reported here confirm this prediction: Low-frequency syllables benefit significantly more from the preparation than high-frequency syllables. Our findings support the notion of a mental syllabary at a post-lexical level, between the levels of phonological and phonetic encoding.

Keywords:

• Phonological/phonetic encoding
• Mental syllabary
• Implicit priming
• Syllable frequency

Notes

¹Dots indicate syllable boundaries.

²The claim that phonological codes are not pre-syllabified is, in part, a language-specific claim. For a language like Mandarin Chinese, which has a small set of syllables and limited resyllabification processes, the story might be different (see Chen, Chen, & Dell, 2002).

³Much discussion has been given to the results of the apparent syllable priming effect in French (Ferrand et al., 1997, Experiment 5). However, Brand et al.'s (2003) failure to replicate the Ferrand effects suggests that this should not be taken as strong evidence for a syllable priming effect (see also Evinck, 1997; and for a review Schiller et al., 2002). The only study that showed a clear syllable priming effect was a study by Chen et al. (2003) conducted in Mandarin Chinese. Mandarin Chinese compared to the other languages under investigation consists of a low number of syllables. Mandarin Chinese has (not counting tone) a syllable inventory of 400 different syllables whereas languages as Dutch and English have more than 12,000 different syllables. Additionally, syllables in Mandarin Chinese are not resyllabified in connected speech. Thus, in languages with far less syllables that are not resyllabified the storage of syllables might be different. It might be the case that the syllable structure is in fact stored within the word-form that is retrieved from the mental lexicon. This would explain why a significant syllable priming effect could be found in Mandarin Chinese but not in Indo-European languages (see also O'Sheagdha, Chen, Shen, & Schuster, 2004). The issue of cross-linguistic differences has to be further investigated.

⁴For one experiment, Meyer (1991) reports that sets with open initial syllables (CV) that share only those two initial segments produced preparation effects that were equivalent to effects produced for sets with closed syllables (CVC) that shared three initial segments. This result was surprising because a pure segmental length effect would predict larger preparation effects in the CVC sets since they comprise one more shared segment. This finding supports the possibility of syllabic effects that are independent of segmental length. However, contrary to this result, Roelofs (1996), Experiment 6) showed that the size of the preparation effect depends on the length of the shared syllable in terms of number of segments.

⁵Note that the idea of a mental syllabary is also in principle compatible with Dell's syllabified phonological code, the idea of syllabified word-forms itself does not deny the existence of a syllabary storing phonetic syllables.

⁶The four high-frequency syllables that were used to construct the second syllables in the disyllabic pseudo-words were among the first percentile of the most high-frequency Dutch syllables; for their frequency values see notes of Appendix A. We opted for these very high-frequency syllables because these syllables are most likely to be stored within the syllabary. The retrieval of those high-frequency syllables should be fast and least error-prone. Furthermore, we decided to have frequency-constant second syllable in order to keep this condition equal across items.

⁷Since the selection of one item determined the selection of three remaining items within one quartet, items cannot be considered a random factor in this design.