Abstract
The intervocalic voicing and gemination induced by the suffixation of the diminutive suffix a in Taiwan Southern Min have long been observed. For example, the phonetic form of tik a ‘bamboo’ is [tik ga]. In this study we conducted two experiments to examine the status of onset consonants of the suffixes –a and –e: how the native speakers treat the intervocalic consonants between mono-syllable stems and the following suffix a and e. The first experiment is called ‘syllable-flip-flop’ test, in which the experimenter asked the subjects to reverse the order of two syllables in bisyllabic words and observed how they syllabify it. The second experiment was ‘concept formation’ test in which the subjects in six different conditions were trained to respond positively to the target sound in the onset position of the second syllables in the learning session to examine whether the native speakers thought that there was an onset consonant preceding the suffix a and e in the test session. The results of Experiment 1 showed that the subjects inclined to treat the suffix –a and –e as onsetless, whereas in Experiment 2 we found that the subjects’ tendencies were still equivocal: the subjects of [l] and [g] conditions tended to treat the suffix –a and –e with an onset; subjects’ answering tendencies to the test items in [b], [m] and [n] conditions were still vague; the subjects’ answering pattern in zero-onset condition was to treat the second syllables as onsetless. The interpretation of the results from these two experiments may be that the subjects were aware of the existence of the onsets in the surface form in the initial positions of the suffixes –a and –e; however, they didn’t consider this kind of onsets as underlying. On the other hand, some responses in both experiments showed that a few words containing the suffixes –a and –e have been changed: as time goes on, some spread onsets before the suffixes –a and –e will be fossilized becoming and underlying, especially the [l] sound.