To Sandhi or not to Sandhi?

[Yeleixingfeng]

By the way, when reading classical texts, how to know when or not to sandhi?

[SimL]

By the way, I haven't heard of anyone saying m-thang. I will ask around.

Oh! Sounds like a case where the shortened form becomes the only known form.

Do you know the word "thang1" by itself though? As in: "hmmm, m7-cai1 lang2 thang1-co3-bo5" (= "hmmm, not sure whether we are supposed to do it or not"). [Ah-bin: have you captured in your rules that "-bo" doesn't cause sandhi of the preceding syllable? Perhaps I shouldn't write it joined with a hyphen...]

[Ah-bin]

By the way, when reading classical texts, how to know when or not to sandhi?

What I have observed from the 三字經 is that the verb in verb-object compounds sandhis, and the first syllable in compound words. 四書 五常, and in things that we wouldn't consider compounds in themselves. 之 acts like 个, sandhing in front of things, but not affecting what comes before it, unless it is used in its meaning as the object of a verb. Negators like put sandhi in the same way.

I'll just write a bit out here of the beginning of the Sam Chu Keng, with hyphens to show where the sandhi is. I'm just doing it from memory of how the chant sounds in the Taiwanese recording I have of it. I've added a few notes.

人之初　性本善 Jîn chi-chho•, sèng pún siān,
性相近　習相遠 sèng siang kūn, síp siang-oán,
苟不教　性乃遷 Kó• put-kàu, sèng nái-chhian,
教之道　貴以專 kàu chi-tō, kùi-í-choan, 貴以專 sandhis, but I don't know why
昔孟母　擇鄰處 Sek Bēng-bú, ték-lîn-chhú. 孟母 鄰處 are compounds
子不學　斷機杼 Chú put-hák, toān-ki-thú. 子 doesn't sandhi because it is the subject
竇燕山　有義方　 Tō•-Ian-san, iú-gī-hong.
教五子　名俱揚 Kàu ngó•-chú, bêng kū iâng.
養不教　父之過　 Iáng put-kàu, hū chi kò.
教不嚴　師之惰 Kàu put-giâm, su chi-tō.
子不學　非所宜　 Chú put-hák, hui-só•-gî. 非 is a negator of 所宜
幼不學　老何為 Iù put-hák, ló hô ûi? this one is hard, since 何 is the object of 為, I can't recall if it sandhis or not
玉不琢　不成器　Giók put-tok, put-sêng-khì. 不= negator 成 = verb 器 = object
人不學　不知義 Jîn put-hák, put-ti-gī.
為人子　方少時　Ûi-jîn-chú, hong siàu-sî,
親師友　習禮儀 chhin-su-iú, síp-lé-gî.

[amhoanna]

Ah-bin, I think there's even more words that sandhi, or "run". I'm pretty sure the 本 in 性本善 runs. Also 昔 in 昔孟母. U probably knew about the 相 in 性相近.

Since "taⁿ" and "khahcá", etc., always take citation, hard to explain why 昔 would run.

I wonder if the sandhi rules for tha̍kche·h are as shot through with exceptions as the regular sandhi rules. There's probably not that many people left that would know. At least we have recordings of the 三字経. And Buddhist recordings.

[Abun]

Great that there already is a thread about this, I almost started a new one
However, even though I know I’ll repeat quite a lot of stuff that’s already been said, but in order to get my understanding straight, please let me recapitulate:

In general, the last word of a phrase appears in citation, everything before is running. So the question comes down to what is defined as a phrase. In order to determine this, there are the following rules:

• 1. Most prominently of course the last syllable of a sentence (excluding sentence closing particles such as --ah(矣), --bô(無), --lah(啦) ect.) marks the end of a phrase and therefore appears in citation.
  2. The topic (which is often but not always identical with the subject) also classifies as a phrase and as such, its last syllable usually appears in citation tone (usually means that some speakers do not apply this rule, apparently especially in TW). The exception are set phrases (such as lí-hó 你好 for example) which I can only guess are perceived as a single entity rather than a sentence, so only the last syllable of the whole phrase is in citation tone. Like the whole sentence, the topic may also be followed by certain exclamatory particles in neutral tone, e.g. --honnh(乎).

Thus, both theme (topic) and rheme (comment, i.e. the rest) classify as a phrase, going conform with the nature of Chinese languages as topic-prominent languages.

• 3. Attributes which are attached with the particle ê(的) are a phrase whose last syllable appears in citation tone and ê is neutral tone. However, if the attribute is a single personal pronoun, a demonstrative pronoun or a numeral (in the last two cases the particle is spelt 个 by the MoE), this rule does not apply.
  4. If the verb is followed by what I was taught to call a complement of direction (i.e. khì(去), lâi(來 (lâi), khí(起), lo̍h(落), ji̍p(入), chhut(出) or a combination of one of the latter four with one of the former two), it is perceived as the end of a phrase and appears in citation tone while the compliment is neutral tone. This rule also applies if the compliment appears in a figurative meaning rather than a literal one (for example khòaⁿ--khí-lâi(看起來))
  5. If the verb is followed by what I call a complement of result (i.e. an adjective or verb denoting the result of the verbal action, e.g. chia̍h--liáu(食了)), the verb usually appears in citation tone. However, I’m pretty sure I occasionally heard it as a running tone, too. I have a feeling this occurs mainly with very common compliments such as tio̍h(著) and kìⁿ(見) for example.
  6. I am not sure about the thing I call a complement of possibility (cf. question 4 below)
  7. If the verb is followed by anything else, it is not perceived as the end of the phrase and therefore the verb appears in running tone. These things may include:
  • a) an object.
    b) the sort of compliment that I call a complement of degree (i.e. a phrase introduced by the particle kah(甲) which expresses the degree/extent to which the action is performed). I’m not sure if the same is true if liáu(了) is used instead of kah, but my feeling says yes.
    c) a prepositional phrase (for example using the prepositions tī(佇), tòa(蹛), kàu(到), hō͘(予)).
    d) a phrase indicating the duration of the action.
    e) a verbal sò͘-liōng-sû(數量詞, numeral-measureword phrase) such as chi̍t-ē(一下), saⁿ pái(三擺), nn̄g chōa(兩逝) ect.
  8. Surnames take citation tones when followed by the given name or sian-seⁿ(先生), the latter of which also takes neutral tone then. I am not sure about the situation with other forms of address such as sió-chiá(小姐), lāu-su(老師), su-hū(師傅) ect.
  9. There are a few words whose last (or only) syllable apparently always appears in citation tone, no matter where in the sentence they appear. These include for example the 3rd person pronouns i(伊) and in(亻因) and the aforementioned adverbs taⁿ(今) and khah-chá(較早).
  10. Furthermore, there also are a few words in which a syllable other than the last appears in citation tone for some reason. Some of these words can be understood as little sentences in itself and thus consist of a sort of micro-theme and micro-rheme, such as thiⁿ--kng(天光), but some cannot. This latter group includes for example āu--ji̍t(後日, in the meaning of “the day after tomorrow, as opposed to āu-ji̍t(後日 as well) „some day in the future“, where the first syllable is running tone), chûn--nî(前年) and kéng--chhat(警察, I know some people analyse this as kèng-chhat, but according to my information, the first syllable should be 上聲). I have conflicting information about whether or not the following syllable should be read in neutral tone, in fact, this might be different from word to word.

Additionally, a phrase may be followed by appendixes, which in most cases comprise of a particle (chō͘-sû 助詞, this includes for example the exclamatory particles, ê(的) and the compliments of direction) and which usually appears in one of the two variants of neutral tone (khin-siaⁿ 輕聲).

So my questions would be:

• 1. Have I gotten anything wrong? Or have I forgotten something?
  2. What happens if a sentence has both a theme and a distinct subject? Examples for this would be “I ba̍k-chiu chin súi (伊目睭真媠)” or “chit-ê-lâng sin-khu chin lò (這个人身軀真躼)". Does the subject (in the examples ba̍k-chiu or sin-khu respectively) take citation tone or running tone? My feeling would tell me to use running tone (which is the reason for my theory that it’s not the subject which is a separate phrase, but the topic), but I don’t know for sure.
  3. What is your feeling about the tones when using a compliment of result (cf. point 5 above)? I seem to recall the verb should be in citation tone, but at the same time I’m absolutely certain I heard some of these constructions in running tone (I clearly remember thiaⁿ-kìⁿ(聽見), khòaⁿ-kìⁿ(看見), thiaⁿ-tio̍h(聽著) and khòaⁿ-tio̍h(看著))…
  3. What happens if the part in neutral tone consists of more than one syllable (for example in khòaⁿ--khí-lâi(看起來)? Will the intonation just drop or do you maybe still here the relation in tone register vaguely indicated or something the like.
  4. I don’t quite know about the effects of what I know as compliment of possibility (the one where you insert ē(會) or bē(袂) between the verb and either a compliment of result or direction, e.g. khòaⁿ-ē-tio̍h(看會著), kiâⁿ-bē-ji̍p-khì(行袂入去)) on tones. Is the verb in citation or running tone then?
  5. What happens if liáu(了) after a verb introduces a complement (in fact I know very little about this complement, also as to the way in which it is different from the one using kah(甲))?
  6. What happens if a surname is followed by a form of address other than sian-seⁿ(先生)? Does it still take citation tone then, and if yes, does the form of address still take its normal tone or does it change to neutral?

[amhoanna]

A valiant effort by aBun. I salute that.

Now, I've been able to find time to come here, but your post is a little too involved for me right now. Rather than ignore your post, though, how about this: if there are structures or phrases U're not sure how to work, tone-wise, U can post them here and we'll label them for U.

Be aware that tone sandhi rules are very different btw Penang and Taiwan. In my estimation, this is probably THE biggest stumbling block for communication btw the two dialects. I also touched on this in the Bisayas thread.

[Abun]

Hey amhoanna,

yes I admit, it's pretty long and theoretical, I should have given practical examples. I will try to formulate my questions as example sentences:

1. Chit-ê-lâng sin-khu chin lò (這个人身軀真躼).
Would you use running or standing tone on the "khu"?

2. Will need a couple of examples here...
Góa thiaⁿ--bô lí kóng--ê ōe (我聽無你講的話).
Lí kám ū thiaⁿ--tio̍h (你敢有聽著)?
I-ê chá-tǹg bô chia̍h--liáu (伊的早頓無食了).
Pháiⁿ-sè, góa bô khòaⁿ--kìⁿ (歹勢，我無看見).
Would you use running or standing tone on the verb?

3. khòaⁿ--khí-lâi (看起來)
Here the last two syllables are neutral, right? Does that mean they have the same tone quality or is there a difference between them?

4. Chit tiâu hāng-á sioⁿ oe̍h--ah--lah. Sái khì-chhia tō sái bē ji̍p-khì (這條巷仔傷狹矣啦。駛汽車就駛袂入去).
How are the tones on sái bē ji̍p-khì (駛袂入去)?

5. I have difficulties making an example sentence with liáu (了) before a complement because I don't know about the grammatical restrictions on that one, only that in some cases it can be used in a similar fashion kah (甲) in "I kóng-ōe kóng kah hō͘ lâng khùn--khì--o͘h (伊講話講甲予人睏去喔)".

6. How would you pronounce the surname in expressions like "Lîm sió-chiá (林小姐)", "Koeh lāu-su (郭老師)", "Tioⁿ thài-thài (張太太)"?

[amhoanna]

1. Chit-ê-lâng sin-khu chin lò (這个人身軀真躼).
Would you use running or standing tone on the "khu"?

Standing!

The sentence is technically correct, but "sinkhu" is redundant here.

Góa thiaⁿ--bô lí kóng--ê ōe (我聽無你講的話).

running; standing
There should not be a double dash after the thiann.

I would say Lí kóng saⁿh góa thiâⁿ bô. (Sannh and bô in standing, everything else running. There is a question as to whether the sannh is really a neutral form of sáⁿ, but U might wanna leave that for Advanced Hoklo.)

Lí kám ū thiaⁿ--tio̍h (你敢有聽著)?

standing

I-ê chá-tǹg bô chia̍h--liáu (伊的早頓無食了).
Running; no double dash. In my idiolect, at least, there would be a hǒ͘ between cia̍h and liáu.

Keep in mind that in Malaysia, Singapore and Sumatra, liáu can function as an aspect marker equivalent to Mandarin le0 (which may well be a cognate, in some way). In that case, both cia̍h and liáu would be standing. (Their Hoklo stands a lot; Twese/Amoy Hoklo runs a lot. This is a small part of what gives Twese that smooth, snakey sound when spoken right.)

Pháiⁿ-sè, góa bô khòaⁿ--kìⁿ (歹勢，我無看見).

standing

3. khòaⁿ--khí-lâi (看起來)
Here the last two syllables are neutral, right? Does that mean they have the same tone quality or is there a difference between them?

There are two different kinds of "neutral" in Hoklo.

"Echo neutral" is level, unstressed, and takes on the tone level where the previous syllable left off.

"Fallaway neutral" is low, unstressed, and falls toward the bottom of the tone scale. When two or more fallaway neutral syllables sequence together, there may be a waterfall effect.

Here, both syllables are fallaway neutral. They are basically the same. The previous syllable ended at the bottom of the tone scale; there was no room for a waterfall.

4. Chit tiâu hāng-á sioⁿ oe̍h--ah--lah. Sái khì-chhia tō sái bē ji̍p-khì (這條巷仔傷狹矣啦。駛汽車就駛袂入去).
How are the tones on sái bē ji̍p-khì (駛袂入去)?

running, running, standing, fallaway

The end of the first sentence is interesting too. "Ah" is echo neutral in Mainstream Taiwanese, although it might be fallaway neutral in other dialects (even on Taiwan). "Lah" is fallaway. "Ah" is level, and takes the tone level of the end of "oe̍h" -- this can be high, mid, or poss. even lowish depending on the dialect. "Lah" then falls away from there.

U mentioned building all this into the writing system. That would be nice indeed.

5. I have difficulties making an example sentence with liáu (了) before a complement because I don't know about the grammatical restrictions on that one, only that in some cases it can be used in a similar fashion kah (甲) in "I kóng-ōe kóng kah hō͘ lâng khùn--khì--o͘h (伊講話講甲予人睏去喔)".

Just listen or observe.

As a guideline,

liáu deals with actions that could be regarded as a "performance" (not a piáu'ián, but a piáuhiān)

kàu >> kà >> kah deals with extents, or distances; 地步s, to use a Mandarin word.

6. How would you pronounce the surname in expressions like "Lîm sió-chiá (林小姐)", "Koeh lāu-su (郭老師)", "Tioⁿ thài-thài (張太太)"?

All running.

Surnames stand in front of --siansiⁿ and --ka. These are the only ones I can think of off the top of my head right now.

Great questions.

[Abun]

1. Chit-ê-lâng sin-khu chin lò (這个人身軀真躼).
Would you use running or standing tone on the "khu"?

Standing!

Okay there goes my theory that it's the topic, not the subject which is resposible for standing tone

It's also interesting that in some of the cases of "complement of result" (tsia̍h-liáu, thiann-bô), running tone is used, while on others (thiann--tio̍h, khuànn--kìnn) it's standing one. I have to think about the reason for that... maybe the former group has been so much lexicalized that it's a single verb instead of verb+complement already? But then again, it's not like they are used that much more often than thiann--tio̍h or khuànn--kìnn...

5. I have difficulties making an example sentence with liáu (了) before a complement because I don't know about the grammatical restrictions on that one, only that in some cases it can be used in a similar fashion kah (甲) in "I kóng-ōe kóng kah hō͘ lâng khùn--khì--o͘h (伊講話講甲予人睏去喔)".

Just listen or observe.

As a guideline,

liáu deals with actions that could be regarded as a "performance" (not a piáu'ián, but a piáuhiān)

Hm, then I guess I should have used liáu in my example, right?

[amhoanna]

In your sentence, I would've used "kà", same as U. He spoke on TO A POINT (地步) WHERE the others fell off in a slumber. I guess my guideline is not very clear. The natives will be a better resource on this anyway.

Semantics will not get U all the way with tone sandhi. Notice that "the̍h tiāu" and "the̍h cáu" are running, standing. But "the̍h--khì " and "the̍h--khílâi" are standing, fallaway, (fallaway).