Talk:Relative clause/Archive 1

From Wikipedia, the free encyclopedia

Archive This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Contents


Hey Marksweep, thanks for the useful copyedit. I accept most of it. But not all. First of all, the examples are variations of a famous nursery rhmye "This is the house that Jack built". The rhyme says "that", not "which", so I won't let you off with changing that. I will accept you making the point that SOME syntacticians don't accept "that" as a relative pronoun, but they are a minority, so you can't on that basis go removing "that" throughout the whole article.

I would disagree that they are a minority. At any rate, there is agreement about the wh- series in English, but there is no agreement about the status of "that". So we shouldn't say in the first example that "that" is a relative pronoun when this is the one contentious item. Rather start with something uncontentious that everyone can agree on, then get to the issue of "that" later. (See more below.)

Next point. I think you are wrong about "who" after a preposition. "Whom" is not much used in colloquial English, but colloquial English seldom puts the perposition before the pronoun. Hand on heart: would you really say "Jack is the boy with who Jenny fell in love"? I think you've been misled by what you would do when the preposition is at the end. I've expanded this into a separate paragraph, because it IS interesting.

I don't know what the basis for that judgement is. Searching for "in who" on Google gives about as many hits as a search for "in whom" does, though this is of course unreliable because the first search also returns hits for "WHO". (This has to be done properly; see below.)

In particular, I've just done a corpus search: "of who" is not frequently enough used to rate a mention in the table. I'm not trying to be prescriptive in this table - the point is not that nothing else can be heard or should be allowed, but that these are what are usually observed. (I take it that is what you meant by POV - you thought I was trying to be prescriptive; but everything I wrote was descriptive of normal English use, and based on descriptive grammar books, not just my own observations.)

I take it it was a slip that you added "none" to the table in the subject row. I challenge you to produce a contact clause with the subject pronoun suppressed.

No slip. The issue is not "subject" vs. "object". Here's a simple example: "This is the man (who) Jack believes will be king." The relative pronoun must be in subjective/nominative case, because it refers to the subject of "will". Omitting "who" is fine, but omitting "Jack" is not. So there you have an example where a relative pronoun referring to a missing subject was omitted; it's just that you can't do this when "who" is the subject of the main verb inside the relative clause.

If you disagree with anything I write, by the way, please say why here and seek a consensus; simply changing substantive content without comment is not kosher. --Doric Loon 10:29, 19 Mar 2005 (UTC)

Well, I'm operating on the general principle of being bold. I don't think that any and all changes have to be discussed first before they can be implemented, and I usually pay no attention to who wrote what. This is not about you personally; it's about this article, which could have easily been misread as being prescriptivist. I realize that it was not your intention to be prescriptive, so I think we're already in agreement regarding what the article should be about. Since there remain a few things to be discussed, let's talk about them here. I've tried to organize the following discussion around your specific objections. Please add or reorganize if you think I've left out some of your points. --MarkSweep 22:28, 19 Mar 2005 (UTC)
Never occurred to me that this was personal. I'm cool! --Doric Loon 22:43, 20 Mar 2005 (UTC)

Status of that

The status of "that" as a relative pronoun is contentious. (Agreed?) Traditional grammars treat "that" as a relative pronoun, but not all contemporary grammars do: e.g. the Cambridge Grammar of the English Language pp. 1056–7 makes a good case for treating "that" as a subordinator instead of a relative pronoun; and the British National Corpus treats "that" as a subordinating conjunction even when it introduces relative clauses. This is not an empty terminological debate: there are important differences between "that" and "which" (e.g. you can say "in which" but not "in that", etc.). It is precisely those differences that have led some to say that "that" behaves unlike the other relative pronouns and should therefore not be grouped with them. --MarkSweep 22:28, 19 Mar 2005 (UTC)

"That" certainly behaves slightly differently from the other relative pronouns, but it behaves far more differently from anything else it can be compared with. To me, the principle that language abhors a vacuum is enough to explain why, given two near synonyms we have evolved patterns of preferring one or other in particular circumstances. But whether I agree with the latest theory is neither here nor there. Traditional grammar certainly makes it a relative, and so do most of the books being published today. To me, this IS a fairly empty terminological debate. I am interested in theory and don't mind you writing it up, but I doubt if this distinction is actually helpful to anyone trying to use the language. My proposal: put a heading "Controversial status of that" at the bottom just above the overview table and discuss this there, but it would be important to name the linguist who proposed the theory and note that it is new and not universally accepted. Then, having dealt with that there, write the rest of the article on the assumption that the traditional terminology is good enough.--Doric Loon 22:43, 20 Mar 2005 (UTC)
Of course I meant "language abhors a synonym" - a rather corny catch-phrase, but it has a grain of truth. --Doric Loon 10:29, 21 Mar 2005 (UTC)

Subjects vs. nominative case and bare relative clauses

The conditions under which a relative pronoun can be omitted need to be discussed. I'm not sure if this article is the right place or if relative clause would be more appropriate. Anyway, we all agree that "who" can appear as the subject or object of verbs (leaving aside arguments of prepositions for now), whereas "whom" can only appear as the object. However, a complex relative clause can have several subject (of several nested clauses), in which case "who" can be omitted if it refers to an internal subject (just not the main subject):

This is the man who Jack believes will be king.
This is the man Jack believes will be king.

I would suggest that "(none)" simply be omitted from the summary table. This is because the conditions on bare relative clauses are more complex and are independent of case. For example, the Cambridge Grammar also points out the following kind of contrast (my example, based on a Google search):

There are traps that/which the player can be released from.
There are traps the player can be released from.
There are traps that/which only under special circumstances the player can be released from.
But arguably not: There are traps only under special circumstances the player can be released from.

In other words, "that" can only be omitted when it's immediately adjacent to the main subject of the relative clause.

Given that these conditions are fairly complex, the case of an omitted relative pronoun is best not mentioned in the summary table. --MarkSweep 22:28, 19 Mar 2005 (UTC)

You are right that this is complex. Is that a reason to omit it from the table? BTW, the table is based on one in a Van Dale dictionary - these are not my ideas, and yes, the dictionary distributs "none" the same way I did. I think you are wrong about your example, though it is a very complex and interesting sentence. Think about this again:
This is the man who Jack believes will be king.
Try this test. Can you substitute "whom"? Yes you can:
This is the man whom Jack believes will be king.
Therefore "who(m)" in this sentence is an object pronoun. It is in fact the object of "believes". That it also has a subjective relationship to "will be" is a very interesting problem, but that doesn't make it an exception to the rule that a contact clause is only possible when the relative pronoun being suppressed is objective.--Doric Loon 22:43, 20 Mar 2005 (UTC)
So do you accept this then?
Jack believes him will be king.
How about the following?
This is the man whom Jack thinks will be king.
This is the man whom Jack said will be king.
Just curious. Do you have a source for the rule that you state? I've cited the Cambridge Grammar, which says that what we've called contact clauses here are possible even though the missing constitutent inside the relative clause is in the nominative case. The main criterion is that the top-most subject cannot be omitted. Note that we don't necessarily need to resolve this debate here: if you add the Van Dale dictionary to the list of references and put a note near the overview table stating that it is based on a certain page or passage from your source, that would be almost all we need. --MarkSweep 03:07, 21 Mar 2005 (UTC)


As I am away from my books at the moment I can't cite you an authoritative source, but popular one is easily found with a google search. For example [[1]] ("who ist Subjekt und darf nicht weggelassen werden") or [[2]] ("the relative whom or that may be omitted").
I certainly wouldn't say *"Jack believes him will be king." But yes, I personally would say "the man whom Jack thinks/says ..." It feels to me a bit like an a.c.i., though of course it is not one. That's why said it was such an interesting and complicated example, which took me by surprise. Thanks for bringing it up. If you want to put this into the article then go ahead. It is certainly interesting enough. But you will have to do quite a bit of (non-original) research to see who is saying what. But I wouldn't put special cases into the overview table, which is complicated enough already.
BTW, that Google search turned up another fact which surprised me. The term "contact clause" seems mainly to be used in German EFL contextst, though not exclusively - one of those links I just gave you was an American university. Possibly you would prefer to give prominence to a different term if you have one? --Doric Loon 10:57, 21 Mar 2005 (UTC)
OK, I TAKE IT ALL BACK. Having discussed this with several colleagues I now see that saying This is the man whom Jack said will be king is just an illogical quirk of mine - it is not standard, and in prescriptive grammar it is not correct. So you are right, who in those sentences is subject. --Doric Loon 14:49, 22 Mar 2005 (UTC)
While it is illogical, I suppose, and violates the usual prescriptive grammar of English, it's rather natural-sounding, and I'm sure I remember seeing once a Web page that gave examples from important writers (such as Shakespeare) of sentences like it. (The point being, as I recall, that the distinction between who and whom has never quite been so simple as subjective vs. objective.) Ruakh 16:48, 22 Mar 2005 (UTC)

Who vs. whom as objects of prepositions

I personally see nothing wrong with using "who" after a preposition, but my personal opinions don't count. We should either find and cite a reliable source that pertains to this issue, or we should investigate this phenomenon ourselves (though this is bordering on Original Research). --MarkSweep 22:28, 19 Mar 2005 (UTC)

There is certainly nothing WRONG with it. I just don't think I have seen or heard it often, and my own explanation for that is that anyone informal enough to drop the old accusative -m will certainly not be so pedantic as to want the preposition before the relative pronoun. But now that you challenge me, I can think of an exception, namely when there is a numeral: Jack has four friends, three of whom are builders. Here the "of" MUST go before the relative, so people who never use "whom" are forced to say "of who". OK, in this case you are right, though I think it is a special case. --Doric Loon 22:43, 20 Mar 2005 (UTC)
I never hear "three of who" in a relative clause. Where I am, people almost never use the word "whom," but would never say "three of who"; it just sounds silly. (Now, you might hear that in a question:
"Three of them are coming."
"Three of who?"
but never in a relative clause.) That could be a regional thing, though; maybe you come from a grammatically deficient region. :-P Ruakh 19:16, 21 Mar 2005 (UTC)
Fit's that ye'r sayin aboot the Doric? Grammatically deficient! Awa an bile yer hied! Coming back to prep + whom, Thomson and Martinet (§74C) give the following options:
The man to whom I spoke
The man whom I spoke to
The man who I spoke to
The man that I spoke to
The man I spoke to
So as far as they are concerned, everything is OK if the preposition is at the end, but only whom follows a preposition. --Doric Loon 15:01, 22 Mar 2005 (UTC)
I agree with their list exactly. Ruakh 16:50, 22 Mar 2005 (UTC)
Thanks for adding the Thomson and Martinet reference. I'll have to check later what, if anything, the Cambridge Grammar has to say about "to who". As Doric pointed out before, "to who" may just be stylistically odd. --MarkSweep 19:04, 22 Mar 2005 (UTC)

Preposition in canonical position?

I've heard of the missionary position, but what is the canonical one? Sorry to be flippant, but I really didn't understand that phase. More to the point, is the preposition MOVED to the end of the sentence or LEFT there? Since this is not worth arguing over, I've rephrased it to avoid the issue. But Old English placed the preposition where German does, before the relative. So putting it to the end is a move. You will notice I think diachronically. (If we were to discuss the indefinate article I would say "an" is the basic form, shortened to "a" under certain circumstances. It may be you would put that the other way around. Both valid!) --Doric Loon 22:55, 20 Mar 2005 (UTC)

Good point regarding "moved" vs. "stranded/left". I'm all in favor of being agnostic about whether movement takes place and in which direction it goes. The claim from transformational grammar is that the relative pronoun moves and that the preposition is stranded or left behind in the position in which it would occur in a canonical clause (hence the term "preposition stranding"). I'm only objecting to the statement that the preposition is "at the end" of the relative clause, which is an oversimplification, because there may be other things after it. Compare:
Jill fell in love with this boy during a weekend of heavy drinking. (canonical clause)
This is the boy who(m) Jill fell in love with during a weekend of heavy drinking. (relative clause)
In the second case the preposition "with" occurs after the phrase "fell in love", just as it does in the first clause. The Cambridge Grammar calls "regular" sentences "canonical clauses" (i.e. they are neither indicative subordinate clauses, nor relative clauses, nor interrogative main clauses, nor interrogative subordinate clauses, etc.). "Canonical position" just means the preposition is where you'd expect it to be in a canonical clause, even though the argument of the preposition (in this case the relative pronoun) is not. --MarkSweep 03:38, 21 Mar 2005 (UTC)
Aha, I didn't know that. As I said, I'm rooted in historical/comparative linguistics and come to all this with a different terminology from you. I had deleted "canonical" because I understood something rather different by it, but in that case I would be happy for you to put it back. In that case it needs to be explained, though. Perhaps cross-reference it to a different Wiki-article? --Doric Loon 11:05, 21 Mar 2005 (UTC)

Commas in German

In the German section, what did you mean by "sentence internal"? Can you give me an example of a German relative clause not marked off by a comma? --Doric Loon 11:29, 21 Mar 2005 (UTC)

The previous version said something like "all relative clauses are preceded by commas", which does not account for sentence-initial free relatives. For example:
Was Jack macht, ist erfolgreich.
I'm not sure if the comma after the free relative clause is required, but certainly there is no comma before the relative clause, because it occurs at the beginning of the sentence. --MarkSweep 12:03, 21 Mar 2005 (UTC)
Aha! Fair enough. Yes, the comma is still required. But point taken. --Doric Loon 18:21, 21 Mar 2005 (UTC)

Anti-traditional-grammar POV?

When it comes to that, this article seems to have an anti-traditional-grammar POV. Statements like "That as a relative pronoun may of course be either personal or impersonal" (emphasis mine) seem inappropriate to an NPOV article.

Also, this paragraph:

A further complexity is that it is often claimed that that may only be used to introduce a defining clause. By analogy, it is sometimes claimed that who or which should only be used in non-defining clauses. However, this does not reflect common usage and is problematic after a preposition.

seems wrong on several counts. I have never heard it claimed that "who" should only be used in non-defining clauses, and a quick Google search didn't pull up any pages making that claim.

According to the traditional rules, the pronouns are used as follows:

  • Personal pronouns:
    • Nominative/subjective: who
    • Accusative/objective: whom
    • Genitive/possessive: whose
  • Non-personal pronouns:
    • As a subject or direct object:
      • Restrictive: that
      • Non-restrictive: which
    • As the object of a preposition: which
    • As a possessive pronoun: whose, or recast as of which

I'm not saying that this article should be prescriptive in nature, but seeing as many people do follow the traditional rules (whether consciously or not), and seeing as many style guides still hold them up as the norm, I don't think the article should be as blatantly dismissive of them as it is. Ruakh 19:39, 21 Mar 2005 (UTC)

The words "of course" in that sentence were mine - I tend to use this phrase meaning "on the other hand" - no POV intended, but feel free to rephrase. The paragraph you refer to is a hybrid of different people's contributions, and possibly needs to be tidied up - go ahead and do it.
I think this article has to describe the traditional rules, since they in themselves are a part of our culture which should be noted, and after all they have a long history and are very well attested. It also has to allow for change and variety in English. For example, having whom as the only possibility for "personal, objective" doesn't cover it. However, I think the overview table mostly co-incides with what you have just written, so I would have thought you would not be too unhappy with it.
As an EFL teacher I would insist that prescriptive rules have their place in certain language operations, such as learning a foreign language. But modern linguistics is and should be overwhelmingly descriptive.
Something else - Ruakh as in "ruakh elohim al pene ha-mayim"? Why don't you add a sentence or so about relative clauses in Hebrew? So far this article is very Indo-european dominated!--Doric Loon 14:31, 22 Mar 2005 (UTC)
I've tidied, and tried to make the page NPOV. (I didn't change the table at all, BTW; my reason for giving the prescriptive rule above is that various paragraphs of the article's text gave misleading information about them, with statements that some people claim things I've never heard of anyone claiming.)
I've also added sections on French and Hebrew. (I'm not familiar with your quote, BTW, but yes, I assume it's the same word "ruakh.") I would add Spanish, but my Spanish is not very good, and relative pronouns are very complex in Spanish; for example, I believe that after some prepositions, any of que, quien, el que, or el cual would be correct - depending on dialect and register - if the antecedent is a person. So I don't feel comfortable attempting an overview of Spanish relative pronouns; it would be interesting to include, though, if someone is up for it. Ruakh 16:57, 22 Mar 2005 (UTC)
Using the terms personal and non-personal is not good, because personal means first-, second-, and third-person pronouns, as distinguished from other pronouns such as demonstrative pronouns and interrogative pronouns. We should use human and nonhuman instead. - TAKASUGI Shinji
Good point. - Ruakh
In addition, I think relative words should be classified like interrogative words. For instance:
  • relative determiner
    • which (when placed before a noun)
    • whose
  • relative pro-form
    • relative pronoun
      • who, whom (human)
      • which (nonhuman)
    • relative pro-adverb
      • where (location)
      • when (time)
      • why (as in reason why)
  • relativizer (not found in English)
- TAKASUGI Shinji 08:42, 2005 Mar 24 (UTC)
That makes sense, though you should note that relative determiners would still be pro-forms (they would be pro-determiners, as it were, replacing this and his and the like).
Incidentally, I suppose this is unrelated, but the article should also note the standard use of as as a relative pronoun (as can be seen in this parenthetical note, where as is the subject of can be seen) as well as the non-standard uses of both as and what: "There's a woman as/what wants to see you." I'm not sure where that could be added, though; any thoughts? Ruakh 10:03, 24 Mar 2005 (UTC)

Hold on now. This is not the traditional or usual terminology. Is it a recognised innovative terminology from some theorist you can name? If so, you can certainly add a section discussing it. But it is not going to be helpful to readers if you go throwing out the familiar terminology throughout the article and replace it with something trendy. I thought, Ruach, that was the whole point of your "anti-traditional-grammar POV" complaint. The thing to remember is that everyone who is working with language, not just experts in linguistics, have a say in the terminology they prefer, and the majority of people are conservative in such things. This is not like an obscure point of scholarship where a few experts can move the field forward by themselves. YES discuss new ideas, but use traditional categories for the bulk of the article unless there is something very wrong with them. --Doric Loon 12:43, 24 Mar 2005 (UTC)

Oh, certainly. I didn't mean that we should exclusively, or even principally, use these classifications; I simply meant that this was a useful classification scheme that bears mention. I guess I can't speak to what TAKASUGI Shinji meant by his comment, but I assumed he didn't mean something so radical as to replace the current explantions with these. (I suppose I shouldn't have assumed without clarifying, though.) Ruakh 20:51, 24 Mar 2005 (UTC)
Traditionally, relative determiners (such as whose) have been called relative adjectives. Since modern linguists distinguish determiners and adjectives, it is increasingly common to refer to them as relative determiners.
The following site is useful:
Glossary of grammatical terms used in English Grammar: Theory and Use
Read the "relative clause" and "determiner" sections.
SIL's Glossary of linguistic terms lists only relative pronouns, relative adverbs (relative pro-adverbs), and relativizer, not including relative determiners in any of the three.
They are explained in the following book, but I don't have a copy.
QUIRK, Randolph, GREENBAUM, Sidney, LEECH, Geoffrey, & SVARTNIK, Jan. (1985) A Comprehensive Grammar of the English Language. London: Longman.
I know this terminology is not shared by everyone. - TAKASUGI Shinji 01:02, 2005 Mar 25 (UTC)

Move to relative clause?

How do people feel about moving this page to relative clause? It's very difficult to discuss relative pronouns in isolation and the current version already touches on many of the larger phenomena related to relative clauses. Then "relative pronoun" could simply be left as a redirect to "relative clause", or could consist of a brief overview and a boldface link to "relative clause". Issues like preposition stranding and defining vs. non-defining relative clauses in English do interact with relative pronouns, but are perhaps more appropriate in an article about relative clauses in general. Any thoughts? --MarkSweep 19:04, 22 Mar 2005 (UTC)

I'd support that 100%. I was actually just thinking earlier today that relative clause should redirect here, but your way makes more sense. Ruakh 19:25, 22 Mar 2005 (UTC)
Go for it!--Doric Loon 21:57, 22 Mar 2005 (UTC)

No language with a distinct relative pronoun?

So thinking about it, I don't think that this:

There appears to be no language in the world with a distinct word used only as a relative pronoun. This suggests that relative pronouns are a fairly late development in the history of language.

(added in this revision) is correct. Hebrew is a very old language, and I can't think of any uses of asher other than as a relative pronoun. Further, while nowadays it is mostly replaced by she-, which admittedly does have other meanings (much like English that), asher and she- were completely distinct in Biblical Hebrew: as far as I can think, the former was exclusively a relative pronoun, the latter never one.

Unless some source can be given to support it, I think it should be removed, or perhaps made more narrow; maybe relative clauses were a late development in Indo-European languages, even if other languages had them as an essential part of the language? Ruakh 04:24, 23 Mar 2005 (UTC)

My source for that was the article Relativpronomen in the German wiki - not really authoritative enough. I did wonder myself if it would be safer not to make a claim as absolute as "there is no language in the world". (Which is why I added the cop-out "appears to be" - but that's not good enough, is it?) But I think the point is still worth making that these are seldom words with unique useages. The French que goes back to a conjunction in Latin, for example. --Doric Loon 11:12, 23 Mar 2005 (UTC)
que is also a conjunction in French, used in much the same way as the English conjunction that. But the mere fact of English, French, and German doing something one way is not justification for the claim that languages "seldom" do it differently; after all, these languages are closely related, and the fact that they do something similarly could reflect this shared ancestry, rather than a nearly-universal truth of human language. Ruakh 15:29, 23 Mar 2005 (UTC)