Linguistic Miscellany Thread

[Qwynegold]

Yeah, that's how I've learned it. But I even found some source claiming that the whole box is done first. *sigh*

In Chinese? Who is that? I tried finding any website that said that, and every one I came across showed the same order as that of the Chinese government's standardization body I mentioned above.

Hmm, now I can't find that either. Strange, I'm sure I saw that, because I wrote down a note about it.

[Kuchigakatai]

What ‘prosodial rules’ are those? I’ve never heard of such a thing, but they sound interesting!

Mandarin has an "end weight" restriction which applies to simple VPs. A monosyllabic action verb does not normally end a sentence as a bare VP, unless it is done as some kind of rhetorical turn partly evoking poetry, or as part of parallelism. In such a case, you either need to use a "prosodial variant" or "elastic variant" of it (typically a synonymic verb-verb disyllabic compound, or a verb-object compound with a generic object like 念書 niàn-shū "study-book" for 'study'), or you need to provide it with extra prosodial weight using an adverbial, or a resultative/potential/directional complement, or the negators 不 bù and 沒 méi, or an aspect particle. This is how 寫字 xiě-zì "write-character" becomes simply 寫 xiě "write" in all the examples I provided in my previous post.

Mandarin also has certain prosodial restrictions over the length of words that can combine directly one after the other in a phrase:

1. In noun + noun compound nouns, the first (modifier) noun is not normally monosyllabic when followed by a (modified) disyllabic noun
2. In verb + object VPs, a disyllabic verb is not normally followed by a monosyllabic noun
3. In modifier adjectival verb + noun NPs, a monosyllabic verb does not normally directly modify a trisyllabic noun

To get around restrictions 1 and 2, many monosyllabic nouns similarly have a "prosodial" (or "elastic") variant, which when used changes the construction to disyllabic noun + disyllabic noun, and disyllabic verb + disyllabic noun. This does mean that all or nearly all of Mandarin's normally monosyllabic nouns have lexical disyllabic prosodial variants. For example, to express 'book knowledge', you shouldn't say *書知識 shū-zhīshi, book-knowledge, with monosyllabic shū 'book', because that violates restriction 1. You need a prosodial variant of shū 'book', such as 書本 shūběn. So, 書本知識 shūběn zhīshi "book knowledge".

Alternatively, another option in case #1 is to use a monosyllabic form or abbreviation of the modified noun, so that you have monosyllabic noun + monosyllabic noun. Given enough context or a previous mention of 書本知識 shūběn zhīshi, perhaps you could get away with 書知 shūzhī or 書識 shūshí (Taiwan shūshì), although it might sometimes strike people as a bit literary/rhetorical/joking/artificial, although it can also be picked up as conveniently concise. Similarly, in case #2, you can sometimes abbreviate the disyllabic verb into a monosyllabic form. And in case #2, you can also use a demonstrative or quantifier before the monosyllabic noun, to give it extra length.

Seeing a disyllabic verb + a monosyllabic noun is in fact very common, but it is understood as an NP, not a VP. 說明 shuōmíng 'to explain' + 書 shū 'book' forms 說明書 shuōmíng-shū 'instruction manual', not *to explain (the/a) book(s). For the latter, you'd need to use an elastic variant of 'book': 說明書本 shuōmíng shūběn.

For restriction #3, all you need is to add the linker 的 de (which marks relative clauses and possessor NPs) between the monosyllabic adjectival verb and the trisyllabic noun.

Finally, it is normal to drop the number 一 yī 'one' in the construction "verb + 'one' + classifier + noun", so that 完個遊戲 wán ge yóuxì ("play CL game") stands for 'play a game' with the classifier (個 ge "CL") bearing the singular meaning of the dropped 一 yī 'one' (the full construction would be 完一個遊戲 wán yíge yóuxì). However, this can't be done if the verb must be disyllabic, except under some further conditions namely, both A) the disyllabic verb being internally a verb + resultative/directional complement, and B) the whole clause being the first in a sequence express a condition or time or concessive contrast...

Yip and Rimmington's grammar (2nd ed.) covers this in chapter 26, pp. 429-452. The publications mentioned in that LanguageLog link in the 3rd paragraph can also be useful. As I often say though, this is best experienced while trying to learn the language, when you see native speakers correcting your Mandarin by adding longer word variants. Natives who, when asked why, respond saying your Mandarin just doesn't have good rhythm...

[bradrn]

What ‘prosodial rules’ are those? I’ve never heard of such a thing, but they sound interesting!

In such a case, you either need to use a "prosodial variant" or "elastic variant" of it (typically a synonymic verb-verb disyllabic compound, or a verb-object compound with a generic object like 念書 niàn-shū "study-book" for 'study') … many monosyllabic nouns similarly have a "prosodial" (or "elastic") variant, which when used changes the construction to disyllabic noun + disyllabic noun, and disyllabic verb + disyllabic noun. This does mean that all or nearly all of Mandarin's normally monosyllabic nouns have lexical disyllabic prosodial variants.

In that case, is there any reason why those verbs and nouns are analysed as being monosyllabic with a disyllabic variant, rather than disyllabic with the second syllable being elided in many situations?

[Kuchigakatai]

In that case, is there any reason why those verbs and nouns are analysed as being monosyllabic with a disyllabic variant, rather than disyllabic with the second syllable being elided in many situations?

I've never seen an argument for why not, but I imagine it's because the monosyllabic-disyllabic counterparts are pretty unpredictable. It isn't obvious that a disyllabic can be reduced at all, sometimes they can't. The main effect of the prosodial restrictions above is to force monosyllabic words into longer forms. So, that, and also the overall weirdness of this phenomenon, as it's not something you find in European languages. It's not something that non-specialist linguists, Chinese lexicographers, Mandarin textbook writers or Mandarin teachers are normally aware of, even though there's probably some 35 or 40 years of research on the topic now... Or if they're aware of it, they don't include it in works (much as you see in English w.r.t. separable and inseparable phrasal verbs).

Another thing is that many monosyllabic stative verbs that express what English uses adjectives for have disyllabic variants, and since prosodial restrictions don't affect such words much, they're really just allomorphs in free variation, with the same meaning. All the basic monosyllabic colours have them with -色 sè as a suffix for example, e.g. 藍 lán ~ 藍色 lánsè 'blue', so you can say 藍的繪畫 lán de huìhuà or 藍色的繪畫 lánsè de huìhuà 'blue painting(s)'. Also, disyllabic verbs that are not verb-object compounds can often be used transitively without being reduced, so they effectively become clunkier allomorphs in free variation of their transitive monosyllabic form if they have one: 學英文 xué yīngwén ~ 學習英文 xuéxí yīngwén 'to study English'. The clunkiness tends to give them a cold, formal, officialese air. Maybe this leads researchers to think of them as allomorphs that exist at the same level, rather than deriving the monosyllabic from the disyllabic. The counterparts are not 1:1 either, and a monosyllabic word may have more than one disyllabic expansion.

It's not always the second syllable that is elided by the way. In disyllabic verbs that are not verb-object compounds, it's often the first syllable that is elided, e.g. 戰勝 zhànshèng ~ 勝 shèng 'to defeat [an enemy team]'.

[zompist]

This stuff is really interesting— thanks, Ser!

I wonder if a lot of this developed due to burgeoning homophony— where Middle Chinese had about 5700 possible syllables, Mandarin has just 1300. So a lot of morphemes get thrown in just to ensure the meaning is communicated. Yet the more concise wényán has always been available as a model.

學習 xuéxí 'study' is etymologically 'learning-practice', so it's not surprising that it can be expressed 學. From a modern perspective it's an abbreviation, but it's also a perfectly valid OC verb.

戰勝 zhànshèng 'defeat' is etymologcially 'war-victory', which may explain why the abbreviation is 勝 which is what actually contains the idea of defeat.

I wonder if there's an influence from poetry as well. OC poetry was built from 4-character units; from Hàn times 5-character lines were used. That makes 4-syllable constructions very useful, and (say) 6-syllable ones very awkward.

[Kuchigakatai]

學習 xuéxí 'study' is etymologically 'learning-practice', so it's not surprising that it can be expressed 學. From a modern perspective it's an abbreviation, but it's also a perfectly valid OC verb.

戰勝 zhànshèng 'defeat' is etymologcially 'war-victory', which may explain why the abbreviation is 勝 which is what actually contains the idea of defeat.

I'm not quite sure if it's really appropriate to say the monosyllabic form is an abbreviation though... Another thing is that many of these transitive verbs and some of these nouns more commonly appear in their monosyllabic form than the disyllabic one, using the disyllabic form mostly when they need that end weight or greater length. 'Table' is usually 桌子 zhuōzi rather than 桌 zhuō (which does often get used when it has weight, as in the NP 這張桌 zhè zhāng zhuō "this CL table", it's not a mere abbreviation), but 'book' is usually 書 shū rather than 書本 shūběn. The verb for 'to eat' is typically just 吃 chī, and gets its allomorph 吃飯 chī-fàn (eat-rice) only when it needs that end weight. Impressionistically/informally, I think 學(習) xué(xí) is usually monosyllabic but the disyllabic form gets some use, and then (戰)勝 (zhàn)shèng is often used in both the monosyllabic and disyllabic forms, maybe 60% - 40% or so (I don't really know).

This prosodial sensitivity and allomorphy is a pretty different thing from abbreviations in the usual Western sense, like 美 Měi, 家 Jiā and 薩 Sà for 美國 Měiguó 'the US', 加拿大 Jiānádà 'Canada' and 薩爾瓦多 Sàěrwáduō 'El Salvador'.

I wonder if there's an influence from poetry as well. OC poetry was built from 4-character units; from Hàn times 5-character lines were used. That makes 4-syllable constructions very useful, and (say) 6-syllable ones very awkward.

The use of 5-char lines from Han onward and also 7-char lines from Tang onward likely says something about the shortening of words, but I don't see how that's relevant... Funnily, you can find some individual 5-, 6- and 7-char lines in the Shijing already (like the 7-char lines 2/4/6 of the first poem here), but full 7-line poems didn't become a thing until a millenium and a half later.

[Raphael]

The new thread on Alien Places of Articulation over in Conlangery leads me to a question about human places of articulation that I didn't want to ask in that thread: why is humming so comparatively little used in human communication?

[bradrn]

The new thread on Alien Places of Articulation over in Conlangery leads me to a question about human places of articulation that I didn't want to ask in that thread: why is humming so comparatively little used in human communication?

Humming is just a nasal consonant prolonged to make a tune — and I think that we can all agree that nasals are pretty common in human languages. So I’d that human languages do very commonly use humming.

[Moose-tache]

To be proper "humming," it would need to appear without anything of higher or lower sonority within the same syllable. So /km:a/ wouldn't count as humming, even though it has a long nasal in the middle. Phrased that way, it's easy to see why humming isn't used more: it's incompatible with an information-rich syllable structure.

[bradrn]

How common is it for a language to have articles which are not placed at the edge of an NP? I’ve managed to find exactly one possible example so far (Koromfe, in which articles may be placed before a long relative clause), but I’d quite like to know if there are any more languages like this.

[Moose-tache]

How common is it for a language to have articles which are not placed at the edge of an NP? I’ve managed to find exactly one possible example so far (Koromfe, in which articles may be placed before a long relative clause), but I’d quite like to know if there are any more languages like this.

Well, technically Romanian articles appear between the noun and the adjective. Does it have to be a free-standing article?

Also, can anyone point me to a chronology of the fortition of w across Europe? I know it happened in Romance before Germanic, and was probably complete in Slavic by the advent of written records. But beyond that, when/where/how did this areal feature spread? For example, did it happen in West Germanic sooner than North Germanic? Was it an early or late feature in Vulgar Latin? Did it happen in Balto-Slavic before or after it broke into subfamilies?

[bradrn]

How common is it for a language to have articles which are not placed at the edge of an NP? I’ve managed to find exactly one possible example so far (Koromfe, in which articles may be placed before a long relative clause), but I’d quite like to know if there are any more languages like this.

Well, technically Romanian articles appear between the noun and the adjective. Does it have to be a free-standing article?

I did actually mean free-standing articles, sorry for the ambiguity! (Lots of languages have affixal articles — Hebrew’s another one — but of course those are almost never at the edge of the NP, since they’re usually just attached to the noun.)

[Kuchigakatai]

I did actually mean free-standing articles, sorry for the ambiguity! (Lots of languages have affixal articles — Hebrew’s another one — but of course those are almost never at the edge of the NP, since they’re usually just attached to the noun.)

Romanian could still count, since the definite article is not necessarily attached to the noun. The Romanian definite article is some sort of phonologically merged clitic that attaches to the first adjective or noun in an NP, basically. If your NP has an adjective + a noun, it attaches to the adjective.

[Richard W]

I did actually mean free-standing articles, sorry for the ambiguity! (Lots of languages have affixal articles — Hebrew’s another one — but of course those are almost never at the edge of the NP, since they’re usually just attached to the noun.)

The Hebrew article is almost always at the start of the noun phrase.

Isn't the Lancastrian definite article proclitic? It's most prominent on a preceding preposition, as in in t'kitchen.

[bradrn]

I did actually mean free-standing articles, sorry for the ambiguity! (Lots of languages have affixal articles — Hebrew’s another one — but of course those are almost never at the edge of the NP, since they’re usually just attached to the noun.)

Romanian could still count, since the definite article is not necessarily attached to the noun. The Romanian definite article is some sort of phonologically merged clitic that attaches to the first adjective or noun in an NP, basically. If your NP has an adjective + a noun, it attaches to the adjective.

It sounds like that makes it a first-position clitic — which is sort of an example of what I was looking for, but not really.

I did actually mean free-standing articles, sorry for the ambiguity! (Lots of languages have affixal articles — Hebrew’s another one — but of course those are almost never at the edge of the NP, since they’re usually just attached to the noun.)

The Hebrew article is almost always at the start of the noun phrase.

From my limited Hebrew knowledge, it goes after prepositions like lə-, hə-. (I tried to write those in Hebrew, but the combination of LTR and RTL text wasn’t working properly, sorry!)

[Richard W]

From my limited Hebrew knowledge, it goes after prepositions like lə-, hə-. (I tried to write those in Hebrew, but the combination of LTR and RTL text wasn’t working properly, sorry!)

I wouldn't count those prepositions as part of the noun phrase. I'm getting very confused as to what you're after. Preposition-article-noun is the order you see in English, French and Hebrew, with various degrees of fusion between and within languages.

[bradrn]

From my limited Hebrew knowledge, it goes after prepositions like lə-, hə-. (I tried to write those in Hebrew, but the combination of LTR and RTL text wasn’t working properly, sorry!)

I wouldn't count those prepositions as part of the noun phrase. … Preposition-article-noun is the order you see in English, French and Hebrew, with various degrees of fusion between and within languages.

Oh, that is true — sorry! (You may be able to tell that I’m really bad at syntax, though I hope to read a bit more about it soon.)

I'm getting very confused as to what you're after.

I’ll try to explain in more detail. In languages like English — and most languages I’ve seen so far — the article is placed at the edge of the noun phrase, outside any modifiers which may accompany the noun. Are there any languages where this does not happen? I’m looking specifically for languages which have a free-standing article with a defined, consistent place in the noun phrase, but which occurs inside nominal modifiers like adjectives and relative clauses.

[akam chinjir]

Would something like "all the young dudes" count?

[bradrn]

Would something like "all the young dudes" count?

Interesting example! But it only works for the single modifier all: you can’t say *some the young dudes, for instance. So I’d say it isn’t what I’m looking for.

[zompist]

Would something like "all the young dudes" count?

Interesting example! But it only works for the single modifier all: you can’t say *some the young dudes, for instance. So I’d say it isn’t what I’m looking for.

You can say "both the young dudes".

In syntax, we say that (English) articles and quantifiers both live in the Det node, which always precedes the N' that makes up the rest of the NP.

Biblical Hebrew has some interesting rules regarding the construct state, which is used for expressions like "the word of the king", "the god of Israel", "the house of peace", etc. The first word is in a distinct morphological form, the contruct state. The interesting bit is that the article can apply only to the other word, and makes both words definite. Thus dəbar ha-mmelek "the word of the king" where the article ha- attaches to the second (non-construct) word, though it is semantically only a modifier to the first.