Zompist Bboard Again

Clearly, the one Good and Correct approach to Mandarin's word classes is that espoused by 郭锐 Guō Ruì in his book 现代汉语词类研究 xiàndài hànyǔ cílèi yánjiū ["A Study of Word Classes in Modern Chinese"] (2002), published by 商务印书馆 Shāngwù Printing House, where he analyzes the language as having twenty (20) different top-level word classes.

(No, I haven't read the book, but there's a paper by Lukas Zadrapa I recently read where he mentions in passing that Guo proposes about twenty word classes in it.)

akam chinjir wrote: ↑Sun Oct 20, 2019 10:53 pm

Ser wrote: ↑Sun Oct 20, 2019 10:14 pmThere remains an important distinction between stative verbs and action verbs nevertheless, and there are not many stative verbs that aren't adjective-ish.

I was reading a bit about hěn 很 yesterday. There are contexts in which a bare gradable adjective gets a comparative sense if it's not accompanied by a degree word, most often hěn (so Zhāngsān gāo 張三高 is ill-formed or, given the right context, means Zhangsan is taller; zhāngsān hěn gāo 張三 很高 is Zhangsan is tall). You don't get the same effect with stative verbs like xǐhuān 喜歡 like, love. Meanwhile nongradable adjectives mostly can't be used as predicates, as far as I know, so obviously they're not verbs.

(There are adjectives like bīnglěng 冰冷 ice cold that are ungradable but can be used as predicates; many such adjectives are compounds in which the first element---like bīng 冰 ice here---could be thought of as a degree word. Are there any predicable nongradable adjectives that aren't like this?)

There's also an argument that adjectives and verbs behave differently in prenominal modification (Paul, Adjectival modification in Mandarin and related issues).

I like the way you're thinking about word classes, but I don't know what you want to eventually get at. I mean, what if there is good reason to think of all four of Mandarin's regular/action verbs, stative verbs, non-gradable adjectives, and gradable adjectives (or "adjectival verbs" if you prefer) as different in important ways? Personally, so far, you've convinced me well enough that there is? Nevertheless we're left with the philosophical problem of what is worth classifying at the top level and what is worth lumping as subcategories of the same supercategory.

I mean, I'm all in favour of calling those four categories distinct at the top level, but maybe there is some value in putting stative verbs and adjectival verbs (gradable adjectives) together.

akam chinjir wrote:Are there any intransitive stative verbs that aren't adjectives?

This is a great question I hadn't noticed. Are there any? It is a little surprising that all the usually-mentioned Mandarin stative verbs are transitive. I think that's actually cooked into why they're considered "verbs" at all, at least when assuming a priori that the problem in question is the Eurocentric "what counts as verbs and what counts as adjectives?" (as opposed to "what categories are worth distinguishing at the top level to explain Mandarin in its own terms?").

vegfarandi wrote: ↑Sat Nov 02, 2019 12:28 pm

Ser wrote: ↑Sun Oct 20, 2019 10:14 pm

dhok wrote: ↑Sat Oct 19, 2019 6:46 amThere isn't really a whole lot to report so far, other than a reminder that Mandarin (all of Sinitic? can anybody who knows Canto chime in?) adjectives are really just stative verbs. There are, afaik, virtually no real tests to distinguish the two that aren't strongly contrived.

There remains an important distinction between stative verbs and action verbs nevertheless, and there are not many stative verbs that aren't adjective-ish.

I recommend this Haspelmath paper on "Escaping ethnocentricism in the study of word-class universals" https://www.academia.edu/2244681/Escapi ... universals

I don't know if you're telling me about that article in support of what I said or against what I said, but I liked it. It is fun to think that although people often blame the study of Latin for the use of the usual word categories, it's true that the ancient and medieval grammarians lumped nouns and adjectives into the same category of the nomen.

By the way, I think some of the discussion on word classes across languages partly comes from a misunderstanding of the terminology. This is highlighted in Haspelmath's paper with the use of such terms as "transitival" and "intransiverb". Apparently, people hear people use "stative verbs" and "action verbs" and for some reason they immediately think they're being classified together as "verbs", even though they may be different categories at the top level. Two-part terms like these are often used simply out of both convenience and a certain reluctance to make up new single-word terms (such as "intransiverb"). When I say that there's an important distinction between stative verbs and action verbs, I do mean that it's best to consider them different at the top level.

Furthermore, it is also the case that in some languages it may be convenient to have supercategories like "verbs" even though, say, transitive and intransitive verbs may be very different. The ancient Romans surely had a good reason to have their nomen category, namely that nouns and adjectives share the same morphological endings and can be the only word of a noun phrase (and therefore be the subject of a verb, etc.), even though the latter also had gender agreement.

In other words, who says that some very particular morphosyntactic behaviour makes different word classes? Haspelmath's criticism of Chung hinges on one different behaviour (#7 "Specific External Argument") between "go"-type intransiverbs and "big"-type adjectivals. But intransiverbs and adjectivals are otherwise the same for behaviours #1-6 and #8 in Table 1, and as he says, intransiverbs and adjectivals don't have any behaviour unique to themselves. It might be useful to lump intransiverbs and adjectivals as subtypes of the same category, whatever you want to call it.

Ser wrote: ↑Sun Nov 03, 2019 5:09 pm

vegfarandi wrote: ↑Sat Nov 02, 2019 12:28 pm

Ser wrote: ↑Sun Oct 20, 2019 10:14 pm
There remains an important distinction between stative verbs and action verbs nevertheless, and there are not many stative verbs that aren't adjective-ish.

I recommend this Haspelmath paper on "Escaping ethnocentricism in the study of word-class universals" https://www.academia.edu/2244681/Escapi ... universals

I don't know if you're telling me about that article in support of what I said or against what I said, but I liked it. It is fun to think that although people often blame the study of Latin for the use of the usual word categories, it's true that the ancient and medieval grammarians lumped nouns and adjectives into the same category of the nomen.

By the way, I think some of the discussion on word classes across languages partly comes from a misunderstanding of the terminology. This is highlighted in Haspelmath's paper with the use of such terms as "transitival" and "intransiverb". Apparently, people hear people use "stative verbs" and "action verbs" and for some reason they immediately think they're being classified together as "verbs", even though they may be different categories at the top level. Two-part terms like these are often used simply out of both convenience and a certain reluctance to make up new single-word terms (such as "intransiverb"). When I say that there's an important distinction between stative verbs and action verbs, I do mean that it's best to consider them different at the top level.

Furthermore, it is also the case that in some languages it may be convenient to have supercategories like "verbs" even though, say, transitive and intransitive verbs may be very different. The ancient Romans surely had a good reason to have their nomen category, namely that nouns and adjectives share the same morphological endings and can be the only word of a noun phrase (and therefore be the subject of a verb, etc.), even though the latter also had gender agreement.

In other words, who says that some very particular morphosyntactic behaviour makes different word classes? Haspelmath's criticism of Chung hinges on one different behaviour (#7 "Specific External Argument") between "go"-type intransiverbs and "big"-type adjectivals. But intransiverbs and adjectivals are otherwise the same for behaviours #1-6 and #8 in Table 1, and as he says, intransiverbs and adjectivals don't have any behaviour unique to themselves. It might be useful to lump intransiverbs and adjectivals as subtypes of the same category, whatever you want to call it.

Yeah, I just happened to have finished reading it when I came across this thread, it wasn't meant to support or go against any arguments being made here specifically, just a worthy additional input to the discussion. I've very often hit discussions of grammar that have a very clear L1 or at least Western bias and it confounds me that trained, degreed linguists are not more attuned to these issues in 2k19. I like his note about the Whorfian effect on linguists, about how the words shape our thinking. As you note about action verbs and stative verbs, it's unfortunate that both contain the root "verb" because it's easy for it to suck a casual reader into focusing on that part of it, when really the action/stative distinction is truly key.

Ser wrote: ↑Sun Nov 03, 2019 5:09 pm

akam chinjir wrote:Are there any intransitive stative verbs that aren't adjectives?

This is a great question I hadn't noticed. Are there any?

Now I'm wondering about intransitive uses of zhù 住 live, be fixed in place: 我住在台北 I live in Taipei, and so on.

It might make sense to distinguish words that are taking arguments (verbs) from words that are heading arguments (nouns) from words that aren't doing either thing (adjectives would be in this group). (Impressionistically this is close to the distinctions Baker ends up with in his Lexical Categories, though it's less precise, and also I guess less theory-internal.) Distinguishing things that way, zhù 住 would end up as a verb, I guess?

Ser wrote: ↑Sun Nov 03, 2019 5:09 pm

vegfarandi wrote: ↑Sat Nov 02, 2019 12:28 pm

Ser wrote: ↑Sun Oct 20, 2019 10:14 pm
There remains an important distinction between stative verbs and action verbs nevertheless, and there are not many stative verbs that aren't adjective-ish.

I recommend this Haspelmath paper on "Escaping ethnocentricism in the study of word-class universals" https://www.academia.edu/2244681/Escapi ... universals

I don't know if you're telling me about that article in support of what I said or against what I said, but I liked it.

Did you (or anyone else) figure out what if anything he means by "aprioristic"?

akam chinjir wrote: ↑Sun Nov 03, 2019 9:23 pm

Ser wrote: ↑Sun Nov 03, 2019 5:09 pm

vegfarandi wrote: ↑Sat Nov 02, 2019 12:28 pm
I recommend this Haspelmath paper on "Escaping ethnocentricism in the study of word-class universals" https://www.academia.edu/2244681/Escapi ... universals

I don't know if you're telling me about that article in support of what I said or against what I said, but I liked it.

Did you (or anyone else) figure out what if anything he means by "aprioristic"?

He’s referring to the tendency to fit understudied languages to prior theories rather than adapting the theory based on the new discovery, assuming what you held before remains correct, even in the face of empirical evidence to the contrary. I think.

Maybe? Though at the end of the day, at best he's got a reasonably good question, something you might ask after a conference presentation on a language you don't really know anything about. That's a pretty slim basis for the accusation that someone, along with the research tradition she supposedly represents, is impervious to counterevidence.

In Chung's response, she actually supposes that by his lights a theory is aprioristic if it has deductive structure---I think all that means in context is that it has consequences that can be checked---which fits pretty well with what he says but doesn't really make his argument seem especially sensible.

Anyway, I find it hard to believe that a neutral judge could read both their pieces and conclude that she's the one who's being more dogmatic, or less engaged with the details of a particular language.

dhok wrote: ↑Thu Oct 31, 2019 4:32 am (an example of a cognitive-quirk constraint: the way humans operate in the world seems to mandate or at least very strongly encourage physical objects in the world acting as arguments and physical actions as predicates. Sensory attributes like "green" or "tasty" seem to be up for grabs. There is no language where the equivalent of "there's a green bottle on the table" looks like "a bottle-ish green tables upon-ly". You could maybe do something like that as a literary experiment, but it doesn't line up with the way humans experience the world so no language operates that way.)

This reminds me of English, as well as many Romance languages. It always strikes me as peculiar how attributes end up as arguments and objects end up as modifiers or predicates. e.g. "lots of books", "a group of students", "three counts of manslaughter", or "blobs of paint". I would have expected the modification relationship to go the other way around, based on Mandarin. To me, this is sort-of-an equivalent of your "bottle-ish green" ("green of bottle"?) situation.

Also, I couldn't quite decipher what you mean by "tables upon-ly", but if one goes by the model of "bottle-ish green", is it meant to mean the "table" acts as a predicate and "upon-ly" acts as an argument/nominal? If so, then such a construction may be exceedingly common: https://en.wikipedia.org/wiki/Relational_noun.

Seirios wrote: ↑Wed Nov 06, 2019 7:41 am Also, I couldn't quite decipher what you mean by "tables upon-ly", but if one goes by the model of "bottle-ish green", is it meant to mean the "table" acts as a predicate and "upon-ly" acts as an argument/nominal? If so, then such a construction may be exceedingly common: https://en.wikipedia.org/wiki/Relational_noun.

The "tables upon-ly" example seems to be supposed to have a predicate-type word that specifically refers to being {in some position relative to}? a table. (I also find it hard to tell exactly what the logical structure is supposed to be.) I haven't learned or studied any languages with relational nouns, but my impression was that relational nouns tend to have fairly broad meanings and constitute a fairly closed class (similar to preposition inventories in languages with prepositions), so I would find it surprising if a natural language had a relational noun that would be best glossed with the term "table". I thought relational nouns were things used like "top" in "The hat is on top of the table" (where "table" is used as a non-relational noun) or "foot" in "The house is at the foot of the mountain" (where "mountain" is used as a non-relational noun).

Clearly in "it tables uponly", "table" is the verb, and "uponly" is the adverb, describing how it tables (i.e. in an upon-ish fashion).

Estav wrote: ↑Wed Nov 06, 2019 10:37 am

Seirios wrote: ↑Wed Nov 06, 2019 7:41 am

dhok wrote: ↑Thu Oct 31, 2019 4:32 am (an example of a cognitive-quirk constraint: the way humans operate in the world seems to mandate or at least very strongly encourage physical objects in the world acting as arguments and physical actions as predicates. Sensory attributes like "green" or "tasty" seem to be up for grabs. There is no language where the equivalent of "there's a green bottle on the table" looks like "a bottle-ish green tables upon-ly". You could maybe do something like that as a literary experiment, but it doesn't line up with the way humans experience the world so no language operates that way.)

Also, I couldn't quite decipher what you mean by "tables upon-ly", but if one goes by the model of "bottle-ish green", is it meant to mean the "table" acts as a predicate and "upon-ly" acts as an argument/nominal? If so, then such a construction may be exceedingly common: https://en.wikipedia.org/wiki/Relational_noun.

The "tables upon-ly" example seems to be supposed to have a predicate-type word that specifically refers to being {in some position relative to}? a table. (I also find it hard to tell exactly what the logical structure is supposed to be.) I haven't learned or studied any languages with relational nouns, but my impression was that relational nouns tend to have fairly broad meanings and constitute a fairly closed class (similar to preposition inventories in languages with prepositions), so I would find it surprising if a natural language had a relational noun that would be best glossed with the term "table". I thought relational nouns were things used like "top" in "The hat is on top of the table" (where "table" is used as a non-relational noun) or "foot" in "The house is at the foot of the mountain" (where "mountain" is used as a non-relational noun).

Thanks, I misread "tables" as a plural noun.

I wonder if it would be difficult to test the existence of this construction. What if your language does have a whole set of verbs or verbalisers you attach to nouns to mean "to be in some position relative to N", except most kinds of spatial relationships require an additional argument?

The two cases I have specifically in mind are Mandarin and most languages with a locative case and without an audible copula.

A): Mandarin effectively has a wide-purpose locative morpheme 在 zai4 which broadly correspond to in/at/on, etc. You can say it is just "to be {in some position relative to}". In many cases you need an additional morpheme to specify the relation. E.g. 在学校 and 在学校里 both work as "at school, be located within the bounds of the school". This leads to three questions:

1) I feel you could reasonably tweak this into a logical/syntactic structure where 在N is the predicate and words like 里 "in, inside" are just such an argument. (Which may not be Mandarin, but the amount of changes seem to be small, so some Chinese topolect might do it, or some other language might.)

2) Suppose "在 + N" is one single structure which requires an obligatory argument of LOC if N is not a place. This would in effect treat 在N like a transitive verb: the reason why *在桌 is ungrammatical but 在桌上 "on the table" is grammatical would be same as why *"I devoured" is bad but "I devoured it" is good. I'm having some difficulty seeing how you can tell this structure apart (在N is an inflected form of N which takes an obligatory argument of LOC) from, say, a structure where N + LOC is one single structure, where LOC is obligatory if N is not a place, and this N + LOC structure takes 在 as an argument (N LOC is an inflected form of N which takes an argument of 在).

3) How do we treat words like 学校? Do we just say there's a class of nominals which can co-occur with 在 directly? Would this imply that these nominals, in this position, belong to the same syntactic category as the N + LOC structure?


B): Many languages have nominal inflections aka cases, and many of them have locative cases. Many languages also possess no audible copula. This leads to the following logical problem. First, the locative case would literally mean "being in some position relative to N". It is functionally the same as a hypothetical participle of such a verb. Second, without an audible copula, most languages simply express predication as if the copula is somewhere but silent, e.g. "X Y" would mean "X is Y" and "X of-Y" would mean either "X of Y" or "X belongs to Y"/"X is Y's". So, a nominal in a locative case would literally, in its formal logic, be of the structure of the proposed "tables". Now, what if more complex/specific relations are expressed with an additional modifying word, like Russian? Sure, Russian copula is audible in non-present tenses, so it might not work as such a language, but given Turkish has no audible copula in any tenses (it doesn't qualify as such a language because it happens to use relational nouns),I don't feel such a language would be that unlikely to exist.

There's a language in Taiwan that has a verbal infix that means "spearing". I'll look it up later for more details. It might have been an infix on something else.

dhok wrote: ↑Thu Oct 31, 2019 4:32 am(an example of a cognitive-quirk constraint: the way humans operate in the world seems to mandate or at least very strongly encourage physical objects in the world acting as arguments and physical actions as predicates. Sensory attributes like "green" or "tasty" seem to be up for grabs. There is no language where the equivalent of "there's a green bottle on the table" looks like "a bottle-ish green tables upon-ly". You could maybe do something like that as a literary experiment, but it doesn't line up with the way humans experience the world so no language operates that way.)

This sounds reasonable, but it seems much less so after browsing my Nishnaabemwin grammar, which has intransitive verbs such as

be empty land
be a very heavy snowfall
be a slippery road
be a boatload
be three things
be a high ridge
be damp grass
be a hump in the floor
be many butterflies about
be a dog

Tzeltal is also an interesting study, because often what we'd consider nominal information lives in the verb, while nouns often just give the material. E.g. the verb pach means "place a bowl shaped vessel upright on a surface." You would then use an 'object' whose only purpose is to say what the bowl is made of. It can be shown what when Tzeltal children learn a new noun, they assume it refers to a new type of material.

zompist wrote: ↑Thu Nov 07, 2019 10:58 am

dhok wrote: ↑Thu Oct 31, 2019 4:32 am(an example of a cognitive-quirk constraint: the way humans operate in the world seems to mandate or at least very strongly encourage physical objects in the world acting as arguments and physical actions as predicates. Sensory attributes like "green" or "tasty" seem to be up for grabs. There is no language where the equivalent of "there's a green bottle on the table" looks like "a bottle-ish green tables upon-ly". You could maybe do something like that as a literary experiment, but it doesn't line up with the way humans experience the world so no language operates that way.)

This sounds reasonable, but it seems much less so after browsing my Nishnaabemwin grammar, which has intransitive verbs such as

be empty land
be a very heavy snowfall
be a slippery road
be a boatload
be three things
be a high ridge
be damp grass
be a hump in the floor
be many butterflies about
be a dog

Tzeltal is also an interesting study, because often what we'd consider nominal information lives in the verb, while nouns often just give the material. E.g. the verb pach means "place a bowl shaped vessel upright on a surface." You would then use an 'object' whose only purpose is to say what the bowl is made of. It can be shown what when Tzeltal children learn a new noun, they assume it refers to a new type of material.

With Nishnaabemwin, at least some of those are derived, because what we would think of as a copula is a verbal final added to a noun-stem. I'm guessing "be a dog" is nimee or something like that? The PA stem for 'dog' was *aθemw-, and its most basic realization is the noun *aθemwa, not a verb. Caddoan and Iroquoian are more notorious for this, though. At least in (some of?) Caddoan, kinship terms are verbs "be the mother/father/sister/etc. of".

I hadn't heard of the Tzeltal example...that's fascinating.

Thanks to this thread, I'm going to verb numbers in my Cat Conlang. Also, tangentially appropriate since it's either related or borrows heavily from Middle Chinese

Just posting to say I couldn't find the language in Taiwan. Its possible the Wikipedia article has changed and thus doesn't come up on a search anymore when I type the words i remember.

I have a question about the mutual intelligibility of the dialects. I recently finished Ezra Vogel's (excellent) biography of Deng Xiaoping, who was a native speaker of Sichuanese Mandarin and spoke Mandarin with a strong Sichuan accent all his life. During the Cultural Revolution, Deng was purged and sent to Xinjian County in Jiangxi, where he worked as a regular laborer at a tractor factory for several years. Deng apparently had a hobby of repairing radios, and was distraught when he learned that after twenty years of socialism none of his coworkers at the factory were able to afford a single radio.

But how was he communicating with them? Putonghua? Again: no radios, except in the big work communes, and those were mostly blasting propaganda. In 1970, before real widespread education or migration, I assume the other workers would have been speaking Hakka or Gan. Are they mutually intelligible? (Presumably even after being purged, Deng was sent to one of the better factories, so maybe the workers *did* speak Mandarin well enough to get by.)

dhok wrote: ↑Fri Nov 08, 2019 3:11 amBut how was he communicating with them? Putonghua?

I would think so. That's generally how Chinese people communicate with each other if they speak different dialects IME.

Vijay wrote: ↑Fri Nov 08, 2019 9:13 am

dhok wrote: ↑Fri Nov 08, 2019 3:11 amBut how was he communicating with them? Putonghua?

I would think so. That's generally how Chinese people communicate with each other if they speak different dialects IME.

These days, yes! But in 1970, in a not-particularly-developed corner of Jiangxi?

dhok wrote: ↑Fri Nov 08, 2019 3:11 amBut how was he communicating with them? Putonghua? Again: no radios, except in the big work communes, and those were mostly blasting propaganda. In 1970, before real widespread education or migration, I assume the other workers would have been speaking Hakka or Gan. Are they mutually intelligible? (Presumably even after being purged, Deng was sent to one of the better factories, so maybe the workers *did* speak Mandarin well enough to get by.)

I imagine Deng might have quickly started to level his speech towards the Hakka or Gan spoken around him. Not outright speaking those dialects, but just levelling his own speech enough to be understood.

This reminds me of an old recording of Dr. Sun Yat-sen from the early 20th century (he died in 1925), which we had a brief discussion of in another forum regarding whether it was in Mandarin or Cantonese.

https://www.youtube.com/watch?v=DSdfGIBfc38

The video has some sort of transcription where grammatical words have been added in but not everything has been written down faithfully.

At the time I posted:

Ser wrote:I think this is an attempt at speaking Mandarin, but he does pronounce some words as straight Cantonese. I think this because too many of his pronunciations are straight modern Mandarin and also unlike Cantonese.

For example, his 文明 sounds just like Mandarin by beginning with w- (instead of m-) and with mid-rising tones for both morphemes, his (來)朝 also ends in -ao with a mid-rising tone (instead of Cantonese -iu or Hakka -eu in a lower-register tone), and he pronounces 這一個時候 with a perfect Mandarin zhe4 yige shi2hou4, with clear high-falling tones for 這 and 候.

However, his (世)界 is straight Cantonese gaai3 and unlike Mandarin jie4, and so is his 到 dou3 [tou]. I also find it surprising he doesn't use a mid-rising tone for 國 (using a level tone instead as in Cantonese), considering 中國 is obviously a key word to know the sound of.

It is also curious that at 1:31 he pronounces 現在 as something like "sanzaa" [sɐntsaː], but a few seconds later at 1:36 he pronounces it more like Mandarin xian4zai4 [ɕjɛn tsai].

By the way, while trying to find this post, I came across another one which I think you'll find interesting:

Ser wrote:For bonus points in some homework at school, I made short analyses of some texts in various languages, including the whole of 三体 san1ti3, LIU Cixin's The Three-Body Problem, by which I mean all three books (the first book was read in LLorg's 2018 book club). Altogether, they're about 2.5 times the length of Herman Melville's Moby Dick. The books contain ~770900 Chinese characters, but there are only 3600 distinct Chinese characters. 的 de alone accounts for almost 5% of the text. Some characters that occur only once: 馋 烷 踌 祠 嘤 谵 竺 芳 嘉.

Pabappa wrote: ↑Thu Nov 07, 2019 2:14 am There's a language in Taiwan that has a verbal infix that means "spearing". I'll look it up later for more details. It might have been an infix on something else.

The language you're thinking of wouldn't happen to be Puyuma, would it?