Talk:Verb argument

This needs expansion (as do many ling articles).

a few points:

• syntactic argument vs semantic argument
• nouns & other things can have arguments too

- Ish ishwar 05:34, 2005 Jan 6 (UTC)

Compulsory arguments

I've just reverted Jorge Stolfi (talk · contribs) on the issue of compulsory arguments. This topic can be viewed from several points of view but, in general, I think it safe to say that some arguments cannot be ommitted in English. Changing the intro text to mention English is also not a good idea for generality. There are some idiomatic uses of English that allow you to drop the subject. In other languages you can drop all the arguments, but precisely in those cases there has been discussion about the convenience of using a syntactic argument approach to describe them (I remember a paper on Mandarin that argues that subject and object are nonsensical categories to describe Mandarin grammar). Please let's discuss this here before making sweeping changes. --Pablo D. Flores (Talk) 12:51, 9 February 2006 (UTC)

Well, aren't the following sentences grammatical:

it is not safe to put a cat in the dishwasher

to put or to take, that is the question

In what sense is an argument "compulsory" if it may be omitted?

As for making generalizations: the concept of subject makes sense in Indo-European languages only because (1) they classify all words into four major classes — verbs, nouns, adjectives, adverb; (2) the typical clause consists of one verb, preceded or followed by several nouns or noun-like phrases adverbs or adverb-like phrases; (3) the verb is inflected so as to match certain attributes of one, and only one, of the noun-like components.

In IE languages, therefore, you can identify the subject as being the noun-like phrase in a clause that constrains the inflection of the verb. Languages that do not satisfy (1)--(3) may not have any concept equivalent to the IE subject.

The object is even more problematic. In Latin, besides the verb and subject, aclause also contains (a) up to three nouns inflected in distinctive ways - accusative, dative, or ablative; and (b) an indefinite number of adverbs and adverb-like phrases, the latter including preposition+noun constructs. Clearly the three noun arguments are special, but none is so much different from the other two. So speaking of "the object" is not meaningful; it is more logical to say that a verb has four noun arguments - subject(nominative), accusative, dative, and ablative (and each can be omitted).

In English and Romance languages, however, most verbs accept only one additional noun argument besides the subject, which is not inflected in any special way, and must follow the verb (whereas the subject must precede it). Hence, those languages have the a well-defined concept of "object".

That was my reason for adding "in English, for example". Without a general definition of subject and object for arbitrary language, the current paragraph does not make sense. All the best, Jorge Stolfi 14:17, 9 February 2006 (UTC)

There are some syntactic tests that have been proposed to establish subjecthood. Some are weaker than others cross-linguistically. I dont think that morphological agreement is always a very reliable indication of subjecthood. These tests are not mentioned in Wikipedia (as you might guess). Many would agree that subjects are generally definable. There are some who claim that the notion of syntactic categories are unnecessary in describing certain languages (e.g. D.N.S. Bhatt's ideas about Kannada). Your Shakespeare quote is cool: I need to think about that one. – ishwar  (speak) 07:29, 22 February 2006 (UTC)

I don't think the "Shakespeare" quote is representative of standard usage; it's more like a play on words mimicking "to be or not to be". I simply don't agree with the statement that in English all arguments can be ommitted. Of course you can say "Got milk?" or "Just remembered something" but that's more like an ellipsis in rapid speech, not the wholesale pronoun dropping you have in Spanish for subjects or in Japanese for practically all arguments. And there's something definitely different between "the cat" and "the table" in the sentence "The cat is sleeping on the table". There's a good reason why linguists distinguish between core compulsory arguments vs. complements or optional arguments

In any case, this can only be solved by good sources, not by refuting each other's opinions. Ish, I know you're a true fountain of sources. Can you dig something up? --Pablo D. Flores (Talk) 17:48, 22 February 2006 (UTC)

Examples are still standing...

I don't see how a source could negate the existence of the examples given above. More generally, the subject can be omitted in English in the following cases, at least:

1. In a nominal subordinate clause with the verb in the infinitive — as in the example it is not safe {to put a cat in the dishwasher}.
2. In an adverbial subordinate clause with verb in the gerund, as in he sings while {carrying the piano}
3. In coordinate sentences, as in He bought the car and {sold the house}.

Granted, the last example above may be analyzed as a single clause with a coordinated predicate He {bought the car and sold the house} (although this does not seem to be a universal approach). In the first example the subject is not only syntactically omitted but also semantically indeterminate. Jorge Stolfi 06:15, 23 February 2006 (UTC)

(Non)universality of "subject" concept

As for the (non)universality of the concept of "subject", I stand by my comments above. To define a "syntactic subject", the language's syntax must have the concept of "verb" and must give special treatment to one of the verb's arguments, that is somehow comparable to the treatment given by IE languages to their subjects. I suspect that there are languages where this is quite impossible. On the other hand, if we try to define a "semantic subject", we cannot do much better than "the subject of a clause in language X is the entity which, in a 'straightforward' translation of that clause into English, gets mapped to the subject of the English cause." Besides being useless, this definition often fails. E.g., should one translate Spanish me gusta el vino into "I like wine" or "wine pleases me"? (Note that el vino is the subject of the Spanish sentence, while me is a dative pronoun.) Jorge Stolfi 06:15, 23 February 2006 (UTC)

Required semantic arguments

Later on it says that "put" requires 3 semantic arguments - subject, object, and locative. However, in this sentence

he quickly put the cat in the box through the opening with his hands

why is "in the box" different from "through the opening" or from "with his hands"? Or from "quickly"?

All the best, Jorge Stolfi 14:17, 9 February 2006 (UTC)

cool. these are fun questions.

I think that the English verb put requires 3 semantic arguments: (1) actor, (2) patient, (3) and what I would call components of PATH. The PATH argument seems to be inherent in the meaning of put which distinguishes it from something like move or throw where the PATH is not specified. The PATH could be defined as a path of motion which could be divided into 3 parts: (1) SOURCE, (2) TRAJECTORY, (3) GOAL. I think that put requires either TRAJ or GOAL: in the box would be GOAL, through the opening would be TRAJ. I say this because both I put the cat in the box and I put the cat through the opening are completely grammatical (for me).

I put the cat from within the cage or I put from within the cage the cat are rather ungrammatical - I feel that the GOAL/TRAJ is missing. Which is weird... (why is just TRAJ ok?) Anyway, this contrasts with throw as in I threw the baseball from (within) the building, where this sentence is grammatical.

The instrumental with his hands seems irrelevant and not specified in the meaning of put just as time in the afternoon is irrelevant as in I put the cat in the box in the afternoon. Of course, all events occur within time frames (at least in the realis), but these are not required by the verbs themselves. (I wonder are there any verbs that require the time to be mentioned??).

So, because these components of PATH are kind built into the meaning of put I would say that they are semantic arguments. Even in Japanese where you could omit all of the arguments in the syntax, the parts of PATH built into put distinguish it from other actor-movement verbs.

This doesnt feel so certain. Semantics is not one of my developed areas of knowledge. peace – ishwar  (speak) 07:18, 22 February 2006 (UTC)

That makes some sense. However, the GOAL and TRAJECTORY are distinct arguments, since both can be specified in the same clause (he put it outside through the window); so saying that "put requires three arguments" is too simplistic. Rather, it has several adverbial arguments, but at least one of those two must (?) be specified. Jorge Stolfi 06:30, 23 February 2006 (UTC)

Well...?

If there are no conter-arguments to the arguments above, I plan to restore those edits of mine that were reverted recently. All the best, Jorge Stolfi 05:11, 21 February 2006 (UTC)

Usage of the term "complement"

The term "complement" as described in this article (complement as adjunct):

Non-core arguments are also called "oblique arguments" or "complements". They are usually adpositional phrases showing time ("in the morning"), location ("at home"), beneficiaries ("for her"), etc.

Is in direct opposition with the usage in the primary article for complements in linguistics: Complement_(linguistics)

We therefore need references of schools using the term in different manners.

--Spinksam (talk) 10:54, 15 December 2009 (UTC)