Eastern Fujianese (Fuzhou)

The government seat for the Fuzhou prefecture as well as the whole of Fujian province lies in the Fuzhou metropolitan district, and the speech here represents the prestige accent for the Eastern Fujianese language, also known as Mindong (閩東語) or Eastern Min. The specific variety spoken in Fuzhou is also commonly known as Fuzhounese. In a wider sense, the similar accents and dialects around the Fuzhou metro area are also often lumped in with the term Fuzhounese. Fuzhounese is also spoken on the island of Matsu, administrated by the Republic of China (Taiwan).

Fuzhounese speakers have migrated far and wide, with large concentrations in Sarawak, Malaysia as well as in metro areas of North America, especially the eastern coast of the United States. In Southeast Asia, Fuzhounese is often called by its Hokkien/Minnan (Southern Fujianese/Southern Min) name Hokchiu, while in the United States, Fuzhounese immigrants often refer to their language as Fujianese when speaking in English, though it is important to keep in mind that Fujian is home to many different languages, and so on Lexxify the Mindong language will be referred to consistently as Eastern Fujianese, and in this description the Fuzhou variety of Eatern Fujianese will also be referred to as Fuzhounese.

As with all Sinitic (Chinese) languages and varieties on Lexxify Hub, we have provided a Lexx Rom transcription system that is as consistent as practically possible across all of Sinitic.

Initials

Notes on Initials
/ts/, /tsʰ/, /s/ may be palatalized for some speakers before high vowels /y/ and /i/. For many speakers these allophones are in free variation.

For some speakers, /n/ and /l/ are either merged into /l/ or exist in free variation.

Initial Assimilation

Initial consonants can undergo changes according to the phonetic nature of the sound in front of them. There are two different environments that trigger this change:

1. Post-Vocalic
(after a vowel or a weak disappearing glottal stop coda <-h>:)

<b> and <p> soften to <vh> [β]

<d>, <t>, and <s> assimilate to <l> [l ~ ɾ]

<z> and <c> soften to <zz> [z~ʒ]

<g>, <k>, and <h> disappear, resulting in no initial consonant (not even a glottal stop)

2. Post-Nasal
(After a nasal coda <-ng>:)

<b> and <p> are nasalized into the corresponding bilabial nasal <m>

<d>, <t>, <s>, and <l> are nasalized into the corresponding alveolar nasal <n>

<z> and <c> soften to <zz> [z~ʒ]

<g>, <k>, <h>, and the null initial nasalize into the velar nasal <ng>

The nasal coda <-ng> can subsequently morph its place of articulation to match the following initial consonant, resulting in <-mm->, <-nn->, and <-nzz-> sequences. This optional assimilation is not indicated in the Lexx Rom romanization or Zhuyin Fuhao.

The glottal stop coda <-t> blocks any change to the following initial consonant. However, after blocking the following consonant from changing, this glottal stop may subsequently drop out.

Rimes

1. Stable Rimes

These rimes do not exhibit major changes depending on the tone/position. Theoretically, some descriptions of Fuzhounese have proposed schemas to include minor shifts in vowel timbre for every single Fuzhounese rime, but this was generally based on impressionistic aural perception, and has not been confirmed upon acoustic analysis to be consistently demonstrated, so it is not recommended for learners to fret about whether a vowel is shifting position slightly between for example [a] and [ɑ], when native speakers themselves do not even reliably perceive or produce such minute distinctions.

As an orthographic convention, the medial i and u become written as <y> and <w> when paired with the null initial. Standalone <i>, <u>, and <ü> are written as <yi>, <wu>, and <yü> respectively. This is as in Hanyu Pinyin.

From a precise phonetic point of view, low tones trigger lowering of the vowel position compared to high tones. Among these rimes, the <o> rime has perhaps the most perceptible allophonic vowel change, with many speakers pronouncing it as a slightly higher [ɔ] in high tones compared to low tones [ɔ̞].

In the sequence /iɔ/, there is a degree of natural rounding influence from the /ɔ/ vowel onto the preceding medial, so though it has historically been transcribed as <io>, the medial may be pronounced as /y/ [ɥ] by speakers as well.

In contemporary Fuzhounese, there is no difference in the pronunciation of checked coda -h and checked coda -t; the two letters both represent the glottal stop /ʔ/. Previously, the -t class of rimes was pronounced with a checked coda -k, which was the result of a coalescence of the checked codas -p, -t, -k, found in words from both the literary and vernacular registers of the language. Meanwhile, the -h class of rimes had a glottal stop coda, and was found exclusively in words from the vernacular layer of the language. As a result, there are actually more words with -t than with -h in Fuzhounese. While the two classes of rimes are now both pronounced as a glottal stop, they still behave differently in both consonant assimilation and tone sandhi, and therefore the LexxRom romanization retains a distinction between them in spelling.

Alternating Tense-Lax Rimes

This class of rimes exists in tense-lax pairs; the first rime form given is the tense rime form, used with the higher four tones Yinping, Yangping, Yinshang, and Yangru, and the second rime form given is the lax rime form used with the lower three tones Yinqu, Yangqu, and Yinru. However, in non-final position, syllables with the lax rime form not only change tone (see below for a description of the tonal system), but also change their rime to the corresponding tense rime form. Syllables that by default already have the tense rime form do not ever alternate to become the lax rime form, so learners should focus on making sure to remember to alternate the lax rime forms into their corresponding tense rime forms.

From a diachronic and comparative point of view, the original form of these rimes was the tense rime form; the Gutian lect and other Mindong varieties north of the urban Fuzhou area do not exhibit this alternating rime pair phenomenon, and in these lects the phonetic values of the rimes align with Fuzhou’s tense rimes. What happened is that the low pitch of the low tones naturally triggered physiologic conditions that favored overall lowering of the vowels from an acoustic perspective. Technically, one can posit for the stable rimes to also have lax equivalents in the three lower tones, and this is occasionally seen in phonetic interpretations of Fuzhounese, but the change in vowel timbre is phonetically very slight and unnoticeable by native speakers, and therefore does not need to be brought to the attention of learners in the same way that the alternating rime pairs above are pointed out.

The lax rimes /ɑʊŋ/, /ɑʊʔ/ can also be seen transcribed as /ɔʊŋ/, /ɔʊʔ/.

Tonal System

Citation Tones

Citation Tones

Tone 1 is a high flat tone, similar to the Tone 1 in Standard Mandarin.

Tone 2 is a high falling tone, similar to the Tone 4 in Standard Mandarin.

Tone 3 is a mid-level tone that is generally flat, but may have a slight bit of down-trailing near the end for some speakers when spoken in isolation, so it has sometimes been described as ˧˨. In terms of tone category, Yangshang tone words with sonorant initials have merged into the Yinshang Tone 3 to form a larger Shang tone category, whereas other Yangshang tone words have merged into Yangqu tone.

Tone 4 is a dipping tone, but the rising component is generally quite short and subtle, so it has also been described as ˨˩. This is a low tone category in Fuzhounese, which triggers lax rime forms in rimes with alternating rime pairs.

Tone 5 is a circumflex tone with a clear rising component that trails down afterwards. Yangshang tone words with voiced non-sonorant initials merged into this tone category. This is another low tone category in Fuzhounese, triggering lax rime forms in rimes with alternating rime pairs.

Tone 6 is a short rising tone paired with checked coda rimes (-h, -t). It is another low tone category in Fuzhounese, triggering lax rime forms in rimes with alternating rime pairs.

Tone 7 is a short high tone paired with checked coda rimes (-h, -t).

Tone Sandhi

Fuzhounese has an extremely robust tone sandhi system. While the tone is stable for the syllable in the final position of a multisyllabic tonal sequence, non-final syllables can change their tone in an intricate manner based on the tone(s) of subsequent syllables. Descriptions of exact tone sandhi values vary slightly across different linguistic descriptions of Fuzhounese, which may reflect some idiolectal variation between speakers.

Tone 1 (ā) changes to a high falling tone <à> like Tone 2 before the low tone category tones (ǎ, a̭, áh/át) and the middle tone category (a̍). Before the high tones (ā, à, āh/āt), it is a high tone <ā>, like its original value.

The low tone categories 4, 5, and checked -h 6 (ǎ, a̭, áh but NOT át) follow the same tone sandhi behavior as Tone 1, changing to either <à> or <ā>. They also change the rime of the syllable to the corresponding tense form.

Tones 2 (à) and 7 (āh) change to a middle tone <a̍> like Tone 3 before the high and middle tone categories (ā, à, a̍, āh/āt), but before Tone 1 (ā), it can assimilate its tone to be the same as Tone 1 <ā>. Before the low tone categories (ǎ, a̭, áh/át), they change to a low falling tone <ǎ>, like Tone 4 but without the subsequent rising tail at the end.

Tones 3 and checked -t Tone 6 pattern together in tone sandhi. Before the middle tone category Tone 3, they become rising tone <á>, before the high tone categories (ā, à, āh/āt) they are a low falling tone <ǎ> like Tone 4 but without the subsequent rising tail at the end, and before the low tone categories (ǎ, a̭, áh/át) they are a high tone <ā>.

In reduplicated syllables, the first syllable can change into a low falling tone written <ǎ>, like Tone 4 but without the rising tail.