2. Consonants

Below is a chart of the consonants of Toisan. There are 19 consonant phonemes, all of which may serve as syllable onsets.

In addition to the 18 consonants on the chart, Toisan has a voiceless lateral fricative: /¬/.

2.1 Consonant Onsets

Margaret’s Toisan allows all 19 consonants to serve as onsets. All syllables must begin with one of these consonants. On the tape, they are words (1) – (19). Below, you may click on the hyperlinked number to hear the sounds. The first fifteen occur as onsets before /i/. Tones vary as indicated in the transcription.

Sound (14) sounds almost like a variant of English [s]; to me it seemed too strident to be a postalveolar fricative. Margaret assured me that the tip of her tongue was on her bottom teeth when making this sound and that the point of contact was behind the alveolar ridge. I do not know of any IPA diacritic to indicate that a sound is more or less strident, but I feel the only proper symbol for this sound is [˛].

Also, sound (9) appears more palatal than velar. It probably is somewhere in between. In Toisan, when velars precede front vowels, they are fronted slightly. This phenomenon occurs in English too, as in [kHç]‘caw’ vs. [k+Hi] ‘key’.

I selected four additional words to appear before [A]:

Because Toisan is considered to be a Cantonese dialect, it is important that I show the sounds (16) and (17) placed before [A] rather than [i]. In my source, [˛] and [˛]-affricates were treated as allophones of [ß] and [ß]-affricates, occurring before [i] and high glides. Even in some dialects of modern Cantonese, [˛] has a restricted distribution before high vowels and glides. By showing [t˛, t˛H] appearing before [A], I hope to provide evidence that these sounds are non-alternating phonemes.

2.2 Consonant Codas

Although any Toisan consonant can fill the syllable onset, the number which can fill the coda is rather limited. Only the unreleased stops [p|, t|, k|] and nasals [m, n, N] may fill the coda. A syllable does have the option of having an open coda. All these possibilities are listed below, following the segments [?a]. The tones vary as transcribed.