|
 |
A hybrid kind of grammar | ||
Share with:
| ||
 Language Learning Forum : General discussion |  | |
 Iversen Super Polyglot Moderator Denmark berejst.dk Joined 6700 days ago 9078 posts - 16473 votes  Speaks: Danish*, French, English, German, Italian, Spanish, Portuguese, Dutch, Swedish, Esperanto, Romanian, Catalan Studies: Afrikaans, Greek, Norwegian, Russian, Serbian, Icelandic, Latin, Irish, Lowland Scots, Indonesian, Polish, Croatian Personal Language Map | Message 1 of 8 30 July 2007 at 10:13pm | IP Logged | ||||||||||||||
|
In a post in this thread I briefly mentioned that I had concocted my own kind of grammar, and Tmesis asked me what that was all about. Even though it is now 3.10 a.M. in Denmark I can't resist answering that question. Prepare for one more immensely long and boring post.
The most talked-about grammar in recent years has been the transformational grammar of Chomsky. I have never liked it: it starts out with some empty symbols and through a series of transformations it ends up with empty symbols that corresponds one-to-one with the elements of a meaningful sentence in a real language. My gut feeling tells me to start with something that has meaning (basically words, but affixes and certain combinations of words also qualify). The entities have certain construction possibilities, and on that basis I want to construct all the sentences in the universe. This is in fact how an old-fashioned grammar functions. Chomsky has coined the term constituent-structure for this kind of grammar, but some of its adherents prefer to speak about fields, and therefore you could also call it a field structure grammar. But sometimes it is evident that sentence structures have some kind of internal connection, that cannot be expressed in a more elegant way than through transformations. So the ideal grammar for me is fundamentally a field structure grammar with the addition of transformation possibilities. The transformations can be something as simple as adding an adjective to a nominal phrase, but they can also transform a whole sentence. The most complicated field structures are organized around finite verbs, and in any field structure a sentence is basically seen as the field around one or more finite verbs with all the subfieklds included therein. However there are cases where a sentence lacks a finite verb, for instance the Russian combinations of a subject and a subject predicat without any verb at all to connect them. That construction is called a nexus. Or it could be a sentence structured around an infinite verbal form like an infinitive ("Que faire?" in French). Or even just a nominal phrase, if the context is clear enough to interpret it ("Two one-way tickets to Tombouctou, please"). But let's for illustration purposes - and because the time is now 3.29 a.M.) just consider the true sentences that have a finite verb as their center. This verb is the main ingredient of the verbal field, and around this you find infintie verbal forms, subjects, several kinds of objects and something called adverbials. We'll return to those later. If a field of any kind (except the verbal field of the complete sentence) contains a finit verb as the center of its own verbal field, then we have a construction with a subordinate sentence. This sentence is inside the 'big' sentence like a Chinese box inside another box, and it can contain its own subordinate sentences. And now for something completely different: there is something called an pronoun and another called a conjunction. An pronoun is a word that points to something else. If it has an implied reference in the real world it is a demonstrative, if the reference is just something vague somewhere then it is an indefinite, if the solution is expected to come in an answer to a question then it is interrogative, and if the reference goes to something in the (main) sentence then it can in certain cases be a relative pronoun, or it can be a demonstrative. I'm not going into any philosophical hairsplitting about these traditional categories, - show me a language where they are irrelevant, and we'll work out a solution from there. Conjunctions are things that tie subordinate sentences to something in the main sentence, or they can tie two or more elements together within one sentence. For some reason the subordinate conjunctions have a tendency also to be pronouns, at least in the Indoeuropean languages I know, but there are a few seemingly meaningless conjunctions, as for instance "that" in English completive sentences, and furthermore we have to reinterprete the word pronouns. This word literally says that it is something that replaces a substantive, though most grammars in spite of the name also accept adjectival pronouns. For me there is no reason to exclude words that typically have adverbial functions from the pronouns, so I consider words like "then", "how", "so" and "sometimes" as just as much pronouns as "me" or "what" or "somebody". As for "that" and "if": the completive conjunction "that" is an easy case: it is an old demonstrative, referring to the rest of the subordinate phrase, who somehow has slipped inside it to assume the role of a subordinate conjunction. "If" is more enigmatic: it probably derives from a form of the word for "give", and the thing to give was the assumption. The same kind of arguments show that the presumed 'neutral' conjunctions normally have a past where they had a reference, and the reference generally was the content of the subordinate phrase as a whole. It is now 4.10 in the morning, so I will just sketch the main types of subordinate clauses, using the notions that I have just introduced. The first group is called the 'partial' interrogative clauses. Their conjunctions are interrogative pronouns (but not "if"), and the whole clause functions in a substantival role somewhere inside the main sentence. The typical case is "Subject + (transitive) verb + interrogative clause", but it can also be the regimen of a prepositional clause, - the prepositions of these constructions tend to 'slide' into the verbal field, leaving the subordinate clause outside, but basically it is still a combination of a preposition and an interrogative clause. It is important to note that the function and meaning of the pronoun/conjunction inside the clause cannot be predicted from the rest of the main clause, - the clause so to say forms it own little world. The same applies to the next group: the completive clauses, in English typically with "that" as the conjunction (though I include the interrogative clauses with "if" in the same group). Here the reference of the conjunction/pronoun is in principle the content of the subordinate clause, but in practice you don't care about this reference, you just see a completive or interrogative clause with a conjunction (that in some cases can be missing). If you have a 'partial' interrogative clause then the conjunction/pronoun has a role in the subordinate clause. In "I don't know where she is" it is an indication of a place, which normally is seen as an adverbial funktion. You could detach the subordinate clause and insert an indefinite or demonstrative pronoun instead: "she is SOMEWHERE" or "she is HERE" (note that when the function as a conjunctional is lost the field is no longer tied to the start of the sentence"). A subordinate sentence that has been treated like this could be called an "intrapositional sentence" (I have put something inside it!), and of course there must be a welldefined transformation that takes this figment of the imagination and transforms it back into a subordinate clause inside a main sentence. With the relative clauses the reference is somewhere in the main clause: "I don't know the house that she has bought". This "that" isn't empty, it refers to "(the) house" and the whole subordinate clause has the role of a complement inside a nominal phrase. In many cases you could use another kind of modifier instead: "I don't know HER house". The intraposition sentence is something like "She has bought A HOUSE", i.e. the reference can replace the relative conjunction/pronoun. This is the defining characteristic of relatives, and - the funny thing is that it also works with adverbial stuff: "She has found somewhere where she can hide" --> "she can hide somewhere". The really funny thing is that 'correlative constructions', typically comparisons, generally can be seen as relative constructions. Here you normally have a demonstrative pronoun as a modifier (in my extended definition) in the main clause and a relative conjunction/pronoun in the subordinate clause: "He is not SO BIG as his sister is" ---> "His sister is SO BIG" In Latin and its offspring, the Romance languages, you often have a neat pair of a demonstrative versus a relative pronoun: talis - qualis, tantus - quantus, but the mechanism is the same. However comparison constructions have a tendency to degenerate: first you skip the verb, and with time certain constructions develop into something that can hardly be seen as clauses: "as good as new". The last kind of subordinates are the "independent clauses". Many traditional grammars speak about "independent relative clauses" in a case like this: "Whoever enters through this door must be shot without warning". We don't know who it is, so we make up an intraposition sentence: "Somebody enters through this door". Whoever it is can also be smuggled into the main clause, - "That mister somebody-who-enters-through-the-door must be shot without warning". So the characteristic feature of this construction is that the whole subordinate clause has a role in the main sentence that is compatible with the role of the pronoun/conjunction alone in the subordinate clause, - and there is nothing 'relative' about that, so I prefer to use the term 'independent clause'. Again we can make the same kind of construction with adverbial elements: "I leave when that man enters through the door" ---> "that man enters ... sometime tomorrow" ---> "I leave tomorrow". The whole subordinate clause indicates a point in time, and so does the conjunction/pronoun ("when"). Normally the pronoun in an independent subordinate clause is an interrogative pronoun or a variation of one of these. There are complications (which I have studied in detail in the case of French long ago), but now it is 4.57 in the morning, and I have to stop soon. So the only thing that I have left to say is that there is a connection between the independent clauses and the relative clauses: if you take something neutral, for instance a demonstrative, and put into the main clause, you can transform almost any construction with an independent clause into a construction with a relative clause. That explains for instance the French constructions with "ce qui", "là où" and so forth. But now it is 5.01 and I have to get some rest. Goodnight. Edited by Iversen on 31 July 2007 at 7:06am 1 person has voted this message useful | |||||||||||||||
| Seth Diglot Changed to RedKing’sDream Senior Member United States Joined 7221 days ago 240 posts - 252 votes Speaks: English*, Russian Studies: Persian | Message 2 of 8 30 July 2007 at 11:11pm | IP Logged | ||||||||||||||
Im curious, so Ill throw out a few comments quickly. What exactly do you mean by creating a grammar? Do you mean a tool which helps you understand verious languages from a certain point of view, or do you mean a generative theory which is supposed to make predictions about the structures of languages (similar to what Chomskian grammar tries to do)? If you are trying to construct the latter, then I am not clearly seeing what predictions your theory makes. However, some of your suggestions such as the inability to predict the meaning of a pronoun in some circumstances seem to make very wrong predictions, as would your contention that words like "that" are really pronouns. (If it was, we would expect to behave as other pronouns do, but it does not; so emperically it seems that it is best to conclude that is of a different funtional type). You also seem to mention verbless "sentences" as instances which contradict the verb-centered approach of some forms of generative grammar. Actually, what you have brought up is usullaly not felt to be a problem for syntacticians. These are usually called speech acts. They are not usually the subject of analyses because it is debatable whether they have truth values or not. There is a certain evasiveness to linguistics theory, however, at times. Sentences with no verbs, for example, are often viewed as having a null verb or some sort of agreement function--in other words, there is not simply nothing there. I am also not quite clear about your take on Chomsky. You mention "empty symbols." Are you talking about functional heads and unpronounced elements like PRO and OP? Chomskian grammar does start out with words: the idea is that words are inserted from a mental lexicon where their meanigns are stored and place into the derivation. Transformational grammar is also more or less old hat now. Chomsky (and his followers in generative grammar) are all doing what has been dubbed "Minimalism." 1 person has voted this message useful | |||||||||||||||
| Iversen Super Polyglot Moderator Denmark berejst.dk Joined 6700 days ago 9078 posts - 16473 votes Speaks: Danish*, French, English, German, Italian, Spanish, Portuguese, Dutch, Swedish, Esperanto, Romanian, Catalan Studies: Afrikaans, Greek, Norwegian, Russian, Serbian, Icelandic, Latin, Irish, Lowland Scots, Indonesian, Polish, Croatian Personal Language Map | Message 3 of 8 31 July 2007 at 5:17am | IP Logged | ||||||||||||||
|
No, I don't want to assume the responsability of making a language-machine that can produce all correct sentences and only those. That was Chomskys lofty dream (at least before he gave up and became a minimalist), and the result was a kind of grammar that I can't use when I try to learn a new language because it is too abstract. Maybe the programmars who try to make translation software carry that dream on, but that's their project, not mine, and I won't interfere in their choice of methods. I just noticed that the words - which in my world are the main carriers of meaning - are much more central to old-fashioned grammars, and that makes them easier to use during the learning process - in spite of some deficiencies that can be remedied.
When you see my remarks about constructions without a finite verb as an attack on "the verb-centered approach of some forms of generative grammar" I'm slightly puzzled. After all the rest of my post describes at system where the finite verb is always seen as the center of a sentence if it is there, - and if it is not then the center might be an infinite verbal form with almost the same kinds of fields around it. A few constructions (such as the Russian combination of a subject and a subject predicat without any copula verb) are so close to 'complete' sentences that you might be tempted to operate with a null-verb. The same applies to certain Swedish subordinate clauses without a finite verb. I don't see this as evasiveness, but rather as an attempt to include some aberrant structures in the general framework in the most economical way. But what about the rest? I don't see any reason to introduce the notion of speech acts just for verbless constructions (and as far as I know it is not how it is defined by Searle and co.). In fact "speech acts" refer to the use of language, not to the internal structure of the utterances, and it is in every respect a relevant task for a syntactician to dissect such constructions. My main problems with the description of subordinate clauses in traditional grammars are 1) they tend to put clauses with a substantival or adjectival role in one corner and those with an adverbial role in the opposite corner, 2) the very fact that they divide subordinate clauses according to their perceived 'word class' instead of looking at the way their conjunctions function. I'm aware of certain cases where there aren't any conjunctions: direct quotes, some conditional constructions and cases where you could have used a completive conjunction or relative conjunction/pronoun, but omitted it. I use this formulation in deference to your observation that a completive "that" doesn't behave like a pronoun, - not even in the extended sense where adverbial roles are accepted. In fact it doesn't refer to anything particular, - historically it did (most completive conjunctions that I have checked are old demonstratives), but with the change that happened when this old demonstrative was reanalysed as a member of the subordinate instead of something outside it, for all practical purposes it lost its reference and with that is role as a pronoun. I should not have referred to it as a combined conjunction/pronoun. Another formulation that needs a comment is this one (about the interrogative clauses ): "... the function and meaning of the pronoun/conjunction inside the clause cannot be predicted from the rest of the main clause, - the clause so to say forms it own little world." The meaning of this is that whatever the choice of verb in the main sentence you are free to choose any interrogative pronoun you fancy in the subordinate clause, and that is still a valid statement for the languages that I know (though I would not be surprised to see restriction rules in more exotic languages). However it is a fact that for instance the mood of the subordinate verb (indicative or subjunctive) can be influenced from the outside. Postscriptum: I knew that I took a risk when I formulated a general synopsis of my views about syntax in general and subordinate clauses in particular. But is there anybody else who could do the same thing without leaving some loose ends and some formulations that don't fit into the framework of other grammatical traditions? In that case please try to see whether you could do a better job - the authors of the grammars on my shelves couldn't, and that is why I originally started out to make my own system. Besides I did in all modesty use the word "concoct", not "create", because I get all the concrete details from the old grammars. Edited by Iversen on 31 July 2007 at 7:16am 1 person has voted this message useful | |||||||||||||||
| Seth Diglot Changed to RedKing’sDream Senior Member United States Joined 7221 days ago 240 posts - 252 votes Speaks: English*, Russian Studies: Persian | Message 4 of 8 31 July 2007 at 12:25pm | IP Logged | ||||||||||||||
Ok. If you are not attempting to create a generative grammar, then I suppose I have fewer comments to make. However, it sounded like you were proposing something in place of Chomsky's grammar, which sounds like it would have to take on the same responsibilities. I wouldn't say that Chomsky (or most syntacticians) have given up that dream. Inventing minimalism only meant introducing a new, simplified way of looking at how the grammar should (ideally!) work. You also mention that his grammar is something that doesn't work for you when you try to learn a new language. Once again, that's fine as long as you understand that that was never the point, and as long as your version is something that you use as a way of thinking about grammatical relations. You also mention that words paly a different role in more traditional grammars. Keep in mind that more traditional grammars are not grammars in the sense of what is studied in modern linguistics. It's kind of link comparing apples and oranges. If thinking about language in those terms helps you, then that's great. The point is simply that traditional grammars and generative grammars view languages in significantly different ways.
I think we're talking about two different things here. I had something much more technical in mind about the verb's role. Also, I was not accusing you of being evasive; rather, I was mentioning that there are instances in which some traditional liguists can be evasive.
I hope I didn't misspeak. Yes, speech acts of course are not defined by Searle or Grice or whomever as verbless sentences. Rather, I was trying to point out that some of your verbless examples are examples of speech acts, not normal truth-bearing sentences. A speech act does, more or less, refer to the way language is used. However, that is a major point. Linguists study knowledge of language, not language use. When you go through the derivation of a sentence, you are not supposing that these are steps that one goes through in real time to produce a sentence; rather, it is meant to eplain how sentences which he have knowledge of are derived. Regardless of whether speech acts propose a problem for linguistic theory, I primarily wanted to point out that linguists are more than aware of them and how they fit into their system. Moreover, it is quite easy for syntacticians to deal with speech acts (that contain a verb at least). I think it may be another story for semanticists.
This seems to be mostly about traditional grammatical analysis and historical linguistics. I don't know too much about, so I don't really have much to say here.
I thought you were trying to say that "that" is a pronoun in the sense that "he" and "she" are. Once again, I think I just thought you were trying to argue something else. Anyway, I appreceate that you seem to have done your homework. I think, though, the issue here is that the grammar you are working is one which allows you to anylyse constructions you come into in various languages, and not the purely generative kind I had in mind. 1 person has voted this message useful | |||||||||||||||
| Iversen Super Polyglot Moderator Denmark berejst.dk Joined 6700 days ago 9078 posts - 16473 votes Speaks: Danish*, French, English, German, Italian, Spanish, Portuguese, Dutch, Swedish, Esperanto, Romanian, Catalan Studies: Afrikaans, Greek, Norwegian, Russian, Serbian, Icelandic, Latin, Irish, Lowland Scots, Indonesian, Polish, Croatian Personal Language Map | Message 5 of 8 31 July 2007 at 5:59pm | IP Logged | ||||||||||||||
|
I think there are some cultural issues here. In the USA linguistics went straight from the ultra-concrete field-work structuralism of the Bloomfield school to the ultra-theoretical transformationalism of Chomsky and his followers. Practical language learning didn't really benefit from either of these extremes.
In Europe neither Bloomsfield nor Chomsky have had nearly the same impact as in America. Instead European linguistics has been dominated by practical grammarians who continued in the footsteps of their illustrious predecessors. For me as a practical language learner who dropped any thought of doing linguistics on a professional level as early as 1981, the European tradition has clearly more relevance. If I had to construct digital translation systems the situation might be different. Edited by Iversen on 31 July 2007 at 6:02pm 1 person has voted this message useful | |||||||||||||||
| Seth Diglot Changed to RedKing’sDream Senior Member United States Joined 7221 days ago 240 posts - 252 votes Speaks: English*, Russian Studies: Persian | Message 6 of 8 31 July 2007 at 6:21pm | IP Logged | ||||||||||||||
|
Yeah, probably.
If by relevance you mean learning languages, then I entirely agree. There are very good (theoretical) linguists who have almost no interest in learning foreign languages. I should also mention that there are aspects of the generative linguistics field that I am skeptical of. I just wanted to know for sure what we were discussing here. 1 person has voted this message useful | |||||||||||||||
| tmesis Senior Member Mayotte Joined 6645 days ago 154 posts - 146 votes Speaks: English* | Message 7 of 8 31 July 2007 at 10:25pm | IP Logged | ||||||||||||||
|
- Edited by tmesis on 17 February 2008 at 3:11pm 1 person has voted this message useful | |||||||||||||||
| Iversen Super Polyglot Moderator Denmark berejst.dk Joined 6700 days ago 9078 posts - 16473 votes Speaks: Danish*, French, English, German, Italian, Spanish, Portuguese, Dutch, Swedish, Esperanto, Romanian, Catalan Studies: Afrikaans, Greek, Norwegian, Russian, Serbian, Icelandic, Latin, Irish, Lowland Scots, Indonesian, Polish, Croatian Personal Language Map | Message 8 of 8 01 August 2007 at 4:41am | IP Logged | ||||||||||||||
|
I'm afraid that this is in fact the case. You can write down which prepositions are used with which cases, and you can write down which verbs have become associated with which prepositions, and you can try to make rules about when the prepositions can be separated from the nominal clauses they govern, - but ultimately the whole thing is a mess. Edited by Iversen on 01 August 2007 at 6:12am 1 person has voted this message useful |
If you wish to post a reply to this topic you must first login. If you are not already registered you must first register
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
