03:09
@Argyll I can't figure out what you mean by "topic + comment structure is not a real thing". I never once used that expression.
In fact, I don't agree with analyzing grammar using those (which is why I didn't use those terms), because that is pragmatics, not grammar.
In every language, the topic can be fronted in some way, but how that is accomplished is restricted by grammar.
And I'm not saying that lexical units correspond directly to grammatical functions. There is solid scientific evidence that humans have an internal language (e.g. see this article), so when we say something we are effectively translating from that internal language into the language of communication.
The simplest explanation of this internal language that appears to fit the data is something like a multi-stack push-down automaton with a small constant number of stacks each of which has small maximum size, and these parameters are constrained by the person's working memory.
This mechanism is capable of generating something similar to CFG but a bit more complex, even though it is in fact nothing more than a regular language (generated by a DFA with many many states). If you're not familiar with those terms, you might want to look them up.
Each word in a natural language corresponds to a set of possible internal representations, each of which has a certain internal grammatical syntax paired with a semantic content. This does not mean that the surface text will necessarily manifest the same grammatical function, but it is always easily generated from the underlying internal statement.
For example in English we learn that a declarative sentence of the form "S H V [O]." where "H V" is the verb (split into helper verb and main verb) can be transformed into a question form "H S V [O]?" This is a simple transformation from the underlying internal representations, which we might perhaps express here as "S,((HV),O),statement" and "S,((HV),O),question" respectively.
03:55
For example, every adjective A can be used as the object of an equative verb (also called a predicate complement) in the internal language. How this translates to various natural languages is secondary. In chinese the internal statement ( S is A ) is covered by "S A" (because the equative verb is implied) or by "S D A" where D is a suitably weak degree adverb such as non-emphasized "很". The latter is often chosen if "S" is too short, for reasons of euphony.
For example "那朵花漂亮" sounds okay, but "你漂亮" sounds odd if not part of a larger whole, and so one usually adds a degree adverb such as "很" or "真" or "好".
Notice that it is difficult to translate "你很漂亮" back to English accurately without knowing the exact intonation, because emphasis on "很" would make it mean "you are very pretty", but lack of emphasis would make it convey just "you are pretty".
@Argyll By the way, I have no idea why you say "the version I learned is to group verb with the subject and say that's a verb phrase". That is wrong. The term "verb phrase" does not include the subject in the standard approach by computational linguists for English (in line with the internal language) The top-level structure is ( S VP ) where S is a subject and VP is a verb phrase. VP in turn is comprised of the verb and the appropriate number of objects and optional adverbial phrases.
04:15
Note that prepositional phrases modifying the verb are adverbial phrases. Prepositional phrases modifying the noun are adjectival. This is an internal grammatical structure and crosses majority of natural languages.
If you're interested in evidence for an internal structured language and not just internal representation of lexical units (as per the article that I linked above), see this article especially points (19) to (25). Also note the grammar involved in point (29); "{ the dog ( { the children } like ) } chased { the cat }", where you cannot replace "the dog" by a pronoun because pronouns cannot be modified by adjectival phrases.
7 hours later…
11:12
@user21820 interesting discussion, and great articles you linked along the way.
Since you seem to know what you're talking about, which is a rare quality, I would like your opinion on the following:
in another thread I was debating with others about whether certain linguistics concepts are general enough that it's possible to use them to analyze different languages, irrespective of whether said concepts are commonly used in such analyses. The keyword being "commonly used".
Specifically we were talking about whether the notions of "direct object" and "indirect object" make sense in Chines…
Since you seem to know what you're talking about, which is a rare quality, I would like your opinion on the following:
in another thread I was debating with others about whether certain linguistics concepts are general enough that it's possible to use them to analyze different languages, irrespective of whether said concepts are commonly used in such analyses. The keyword being "commonly used".
Specifically we were talking about whether the notions of "direct object" and "indirect object" make sense in Chines…
5 hours later…
15:48
@blackgreen Yes I do agree with you that verb valency exists in every language, and that what manifests as direct and indirect objects in English also manifest in other languages, including Chinese. For instance, "What did you call me?" = "你叫我什么？". And you're surely correct that if languages share a certain kind of grammatical structure then we can analyze them both using the same concept (sufficiently generalized).
4 hours later…
19:48
@user21820 I remembered the VP definition wrong. I browsed some literature and did find the custom being VP being the latter half. That makes sense per what we discussed.
I did not say spoken/written language represent internal language. Though I am not convinced about internal language --- whether it actually looks like a language.
The point of phrasal structure is to reduce sentences into a recursive formula. So if you make them VP + NP with verb being in the first half, at least with English, it is unnatural because whether a linking verb is used depends on the latter half and that latter half would look strange if we characterize adjective as noun phrase.
Predicate complement is not a phrasal category concept. I also object to the straight claim that predicate complement is internal.
I dont really think 你漂亮 is grammatically special. It sounds odd because you rarely hear 1st person pronoun + description in Chinese. Change it to something else. 你吃饱了吗？ (Which actually has the predicate complement structure) 我饿了。 烟花好看。 etc etc
Predicate complement or topic comment etc obviously relate in somewhere to internal processes of the brain. Listing topic followed by comment is just the least structural way to put them together. I don't see any relation to how internal "language" looks like. And I don't see the evidence that it should look like a language.
As for topic comment not being a thing, I mean that you don't need topic comment as special cases to explain Chinese sentential structure. Not that you cannot use the idea to good effect elsewhere.
Re adjective. When I say adjective, it is strictly for in a phrasal structure -- where we would need to list all the definitions of phrases, adjective should be somewhere. The assignment of adjective to words isn't inherent. It is due to recursively breaking down sentences into components. And if a word falls in the position of adjectives, it is an adjective.
So the process starts with a best attempt at proposing recursive definition of phrases, and then proposed a (probably) simple definition of sentence. Multiple possibilities exist here. And then reduce sentence into components until hitting base case --- ambiguity is probably a necessity along the way. But you should have good successes with any decent initial proposal.
1 hour later…
21:23
Btw, to both of you, verb valency is not about indirect vs direct object? Valency is the number of agents a verb can pertain to. Included in the number is both subject and object. It's very useful for a English learner to get the labeling in dictionary because it predicts the use of preposition when using an intransitive (valence = 1) verb with an "object". Object in a semantic sense. And dictionaries do differentiate intransitive and transitive (valence = 2) verbs. In theory, valency helps differentiating avalent (valence = 0) and ditransitive (valence = 2) verbs as well. (e.g. "He walked"…
« first day next day → last day (17 days later) »
Transcript for
Aug19
Aug '2020
Aug21
Discussion on answer by dan: Using 有 …
Imported from a comment discussion on chinese.stackexchange.co...
