Abstract:

High-frequency syllables are spoken faster than low-frequency ones in spoken word production, which is termed the syllable frequency effect. Compared with alphabetic languages, syllables’ frequency might have different processing mechanisms in Mandarin Chinese with fewer syllables and no resyllabification in connected speech. We review theories and studies about the syllable frequency effect in spoken word production. Specifically, we review the syllable frequency effect across languages, comparing Mandarin Chinese as a non-alphabetic language with alphabetic languages such as English, Dutch, and French.
In the article, we first distinguish different definitions for syllabic neighbors in Chinese and alphabetic languages, and introduce two distinct measures of syllable frequency. The first is the token measure of syllable frequency, which refers to the accumulated word frequency of all syllabic neighbors. The second is the type measure of syllable frequency, which refers to the number of syllabic neighbors. A few studies in alphabetic languages found that the token syllable frequency plays an inhibitory role in lexical selection due to the competition among activated words at the lexical level, while other studies reported that the type syllable frequency plays a facilitative role in spoken word production. Mixed findings indicated that different measures of syllable frequency influence spoken word production distinctly.
In alphabetic languages, studies have observed facilitatory, inhibitory, or the null effect of syllable frequency. Most studies confirmed that the effect arises at phonetic encoding in spoken word production. The direction of the syllable frequency effect depends on the balance of the inhibitory effect during lexical selection and the facilitation effect during phonetic encoding. According to the findings of syllable frequency effects in alphabetic languages, three hypotheses have been proposed to explain the facilitation effect, including the mental syllabary theory, the mixed model, and the dual-route model. All models consistently assume that syllable-sized motor programs can be retrieved from the mental syllabary during phonetic encoding, but they have different explanations for the facilitation effect of syllable frequency. The debate concerns the different processing mechanisms among high frequency syllables, low frequency syllables and novel syllables, which are stored in the mental syllabary or assembled online.
According to the proximate units principle, syllables are retrieved during phonological encoding in Chinese speech production. Using a picture-word interference task, studies revealed that syllable frequency facilitates phonological encoding or phonetic encoding in Chinese speech production. One possibility is that high frequency syllables are retrieved faster than low frequency ones during phonological encoding, resulting in a facilitation effect. Another possibility is that as a tonal language, the syllable-sized motor programs in Chinese are prepared through syllable-to-tone association in phonetic encoding. Speech is one of the most practiced motor behaviors, and practice leads to the storage of motor programs. Thus, the programs of high frequency syllables are retrieved faster than those of low frequency ones. Studies also showed an inhibitory effect in word reading-aloud tasks. We propose a theoretical framework to interpret the cognitive mechanisms underlying syllable frequency effects (including facilitatory and inhibitory) in Chinese speech production.
So far, only a few studies have distinguished the effects of token syllable frequency and type syllable frequency, but it remains unknown about the role of two measures of syllable frequency in speech production. Future studies should clarify the cognitive mechanism of syllable frequency effects underlying different measures. Whether there is a mental syllabary in Chinese speakers is still controversial, and there is little evidence of Chinese syllables’ positional frequency effect in spoken word production. It is necessary to elucidate the mechanism of the facilitation effect of syllable frequency in Chinese speech production, and investigate the processing stages and neural mechanisms of Chinese syllable frequency effects through a variety of techniques and paradigms.

Key words: speech production, syllable frequency effect, cross-language, mental syllabary, syllabic neighbours