Using 'where' in relative clauses

[Nic]

In all my grammar books, it states simply that 'where' can be used in relative clauses as a relative pronoun to describe a place. But in fact it's not that simple, and I find myself unable to explain why sentences like this are wrong:

We often go to visit our friends in Bristol, where is 30 miles away.

Beijing, where will hold the 2008 Olympics, is my hometown.

...and why sentences like this are right:

Glasgow, where my brother lives, is the largest city in Scotland.

Why sometimes do I have to use 'which' instead of 'where'? I can't find any references to this in any of the grammar books that I have!

Thanks so much for your help!

[sonya]

which is the subject in this.. I forget what this type of clause is called. (relative clause?).

where is the object..

[Stephen Jones]

Use where in place of "preposition plus which" and you have a useful rule of thumb.

Glasgow, which my brother lives in, is the largest city in Scotland.
Lebanon, which I am going to next week, is a fascinating country.

[abufletcher]

This is a tough nut to crack from the "no-grammar/chunking" perspective but I'd like to at least give it a shot. The key is to see if it's possible to come up with a linear solution that doesn't involve inversion. The thing would be to see if there is any way that these two alternative "turn beginnings" condition what is to follow in lexical and or logical ways.

London, where...

London, which...

To say this in another way, talk in the real world is both produced and heard incrementally moment-by-moment and bit-by-bit. In conversation possible “next speakers” will need to be carefully monitoring the unfolding trajectory of “turns-in-progress” for clues as to when in the future this speaking turn might be “projectably complete” at which time unmarked speaker transition would be relevant.

l-
lo-
lon-
lond-
Londo-
London-
London wh-

Note that at this point this “turn-in progress” could still turn out to be “London weather…” up to the moment that the /r/ in “where” gets produced at which point a different sort of trajectory presents itself. So the question becomes what sort of projections about possible upcoming completions are occasioned by “London, where…” vs. “London, which.”?

Traditional approaches to linguistic analysis are based on analysis of “what came to be produced.” Yet, this mode of analysis is simply not available to next speakers who can, of course, only monitor that part of an utterance which has actually been produced up to sme given moment.

One issue that complicates the projection here is that, at least in naturally occurring talk, these aren't so much “relative clauses” as they are “parentheticals” i.e. talk that is designed as discontinuous with the prior talk and is typically marked prosodically. But this also gets us an analytic toe-hold, namely, if there is an identifiable “left bracket” that marks the detour into a parenthetical then there must also be a “right bracket” to mark the termination of the parenthetical talk. And this is where we might consider what German grammarians refer to as the “satzklammer” (“sentence bracket”). In German once a aux. verb is used the “remaining verbs” got post-positioned to the end of the sentence:

Ich HABE…………..GEMACHT.

So Germans monitoring for that point of possible completion, can upon hearing one of the modals or aux verbs (though I doubt they think of it in these terms) know that things will not be over until “the fat lady provides a verb in the proper shape, so to speak.

I think that's the situation for English speakers with “London where…” Once a speaker starts like this talk-recipients monitoring this turn can just hang back and wait for the verb that will marked the “right bracket” of the parenthetical (“London, WHERE my father LIVES,…”) at which point the original trajectory projected by “London” become relevant again.

This is still far from being completely clear in my mind and there's still a lot to work out , including the differential projectability of “London, which…” but I'll save that for a later installment. Maybe someone here can run with the ball.

[abufletcher]

It just occurred to me that "London where-" would also be hearable as "London wear-/ware- which present other trajectories. So it would be the next part of this utterance which might clearly establish that a paranthetical was underway.

"London where they..."

[fluffyhamster]

'London, where the weather is...'.

Just thought I'd give you a bit more to play with, Abu.

[mesmark]

I like the linear idea as it fits in well with what I'm trying to do over here.

When I hear the speech above

London, where ... or London, which ...
I'm expecting a relevant tangent description of London, either way.

Londo, where ... tells me a full thought on London is coming. Once that thought is finished then we're back on track
London, which ... signals to me a partial description of London is coming. Once that's finshed we're back on track.

I think the problem with the example Nic has is they are forcing it. It goes back to working with grammar for grammar's sake and distorts usage and meaning.

Beijing, where will hold the 2008 Olympics, is my hometown.

I probably wouldn't put these 2 sentences together in the first place. It doesn't seem like a relevant tangent. Maybe like this:

They are going to hold the 2008 Olympics in Beijing. That's my hometown, you know.

Sorry, Nic! We've stopped trying to answer your question. Did you get the answer you needed?

[lolwhites]

When I started my Linguistics degree there was a demonstration of how we process speech. Basically they played us a piece of natural speech, one word at a time. They started by playing the first word, then the first two words, then the first three and so on. The phrase was "I wasn't good enough for them", so we heard "I", then "I was", then "I wasn't" etc.

What happened was that what we heard was totally incomprehensible until we heard at least four words. The "I" on its own sounded more like a gasp, "I was" sounded like someone trying not to be sick. It wasn't until we heard the chunk "I wasn't good enough" that suddenly everything seemed to "click" and it all made sense.

My point is this: doesn't this have implications for the "trajectory" assertion? Until listeners have heard enough language to work with they can't make any predictions.

Abu - what do you say to the student who produces the phrase *London, which my father lives? Do you simply tell him/her that "native speakers don't say that" or do you give some sort of general priciple to avoid making that kind of utterance in the future. Convention or "rule", the student who wants to make a good impression on a job application letter needs something more substantial, surely?

[abufletcher]

My point is this: doesn't this have implications for the "trajectory" assertion? Until listeners have heard enough language to work with they can't make any predictions.

There are two kinds of incontrovertible evidence that participants in talk-in-interaction are perfectly capable of projecting upcoming "possible completion." First, it is simply an observable fact of any and all naturally occurring talk that participants are massively able to produce perfectly coordinated and precision timed next speaker start-ups. In the following A has produced a turn-at-talk. Participant B has been able to project the trajector sufficiently accurately to begin speaking at the very instance that A stops speaking.

A: So how's it going?=
B: =OK

For a host of neurolinguistic reasons this would be physically impossible had B being waiting until he heard A come to actual completion before beginning to talk. The result would have been a interturn gap:

A So how's it going?
(0.2
B: OK

This is observably NOT the case in talk. Believe me when I say that this turn transition thing is VERY finely coordinated in talk.

Second, there are ample cases where a next speaker is actually able to co-complete the turn-in-progress in ways that demonstrate their accurate projections of the trajectory of the talk. Participants in talk-in-interaction most observably do NOT wait until the end of speaking turns to process "meaning."

In fact, the ways that turns are begun is so significant that turn-beginnings which happen to get produced in overlap with another speaker are regularly recycled.

B: [yeah]
A I was...

The problem of course with the demonstration in your linguistics class was that the talk was presented out of context. In the real world talk is ALWAYS situated both within a specific talk-sequential location and a real world context. This by the way was one of the points that Brazil made in his Grammar of Speech book. One of of the primary tenets of Chomskyan linguistics is that linear grammars can't work (thus the need for deep structure arguments). But that argument only works if presented in the abstract. In a contextualized world were not all possible hearings are possible-in-THIS-context, linear grammars turn out not to be a problem.

Clearly though there is still much I have to think out here.

[abufletcher]

Beijing, where will hold the 2008 Olympics, is my hometown.

Actually, what occurred to me as the probably higher frequency construction would be:

"Beijing, where the 2008 Olympics (will be held)/(are going to be held), is my hometown."

But I'm sure the following is entirely possible/likely as well.

"Beijing, (0.2) where they'd going to hold the 2008 Olympics,(0.2) is my hometown."

As I said this is a tough nuts to crack but that mundanely observable fact that next speakers can and do project the trajectory of such things in talk-in-interaction is just too well-established to deny.

[abufletcher]

'London, where the weather is...'.

Just thought I'd give you a bit more to play with, Abu.

Hmm..that does rather mess up my initial argument about the "satzklammer" idea. It would seem that the word immediately following "where" is of utmost importance here

"London, where the..."

"London, where Joe..."

But again we are trying to argue this stuff out of the blue while participants in actual talk would have access to the sequencing to limit possibilities.

[lolwhites]

I wasn't asserting that projection doesn't happen; it's something any good teacher encourages their students to do. What I was saying is that in order to project/predict, one needs a minimum amount of language to go on.

I am certainly not convinced by the argument that Beijing, where will hold the 2008 Olympics, is my hometown is wrong simply because it is a low-frequency item. I think there is something about the semantic (note I'm avoiding the word "grammatical" here) qualities of where which means it can't be used in this way.

[abufletcher]

I wasn't asserting that projection doesn't happen; it's something any good teacher encourages their students to do. What I was saying is that in order to project/predict, one needs a minimum amount of language to go on.

I think we agree here.

I am certainly not convinced by the argument that Beijing, where will hold the 2008 Olympics, is my hometown is wrong simply because it is a low-frequency item. I think there is something about the semantic (note I'm avoiding the word "grammatical" here) qualities of where which means it can't be used in this way.

I think we agree here too. There is definitely something "wrong" with "Beijing, where will hold..." that is not just frequency related. I think the argument here was that even the "corrected" version ("Beijing, where they will hold the 2008 Olympics...") might be a bit "forced." I think you're also right that this has something to do with the "semantic" projections of "where."

Let me play around with the concs and see where that leads me.

[abufletcher]

one needs a minimum amount of language to go on.

...and a considerable amount of context.

[mesmark]

Actually, what occurred to me as the probably higher frequency construction would be:

"Beijing, where the 2008 Olympics (will be held)/(are going to be held), is my hometown."

But I'm sure the following is entirely possible/likely as well.

"Beijing, (0.2) where they'd going to hold the 2008 Olympics,(0.2) is my hometown."

I was actually thinking...

'My hometown is Beijing, where they will hold the 2008 Olympics.'

That would seem to have the relevant tangent qualities I was think of. Th other way, it seems we're talking out of the blue. But, like it was stated, I guess they're all out of context.