"Articles currently living in the Hamilton area"

« previous post | next post »

From Jeffrey Shalit, "Funny Word Order in a Poster Advertising a Study on Word Order", 12/11/2011:

The thing that caught Peter's attention is an apparent attempt to move the post-nominal modifier "currently living in the Hamilton area" past the prepositional phrase "for a linguistic study on the relation between word order and articles".

We can clarify the situation by simplifying the example:

[Original] We are seeking German language speakers from Austria, Switzerland, Liechtenstein or Germany for a linguistic study on the relation between word order and articles currently living in the Hamilton area.

[Simplified] We are seeking German speakers for a word-order study living in Hamilton.

The logic of the sentence is "We are seeking X for Y", where X = "German speakers who are living in Hamilton" and Y = "a word-order study".

In the last form, with the business about "living in Hamilton" expressed as a relative clause, the sentence can work with the relative clause shifted ("extraposed") to the end:

We are seeking German speakers for a word-order study who are living in Hamilton.

And it works even better if the extraposed modifier is heavier, e.g.

We are seeking German speakers for a word-order study who have been living in Hamilton for at least three of the past five years.

It's easy to find real-world examples of this general type:

We are seeking a man for our Group Actuarial staff who has completed four or more examinations and is presently engaged in Group Actuarial work.

Employers, however, are increasingly seeking people for these technician positions who have four or more years of postsecondary school training in this field …

As Peter Fries observes ("Post nominal modifiers in the English noun phrase", in Peter Collins and David Lee, Eds., The clause in English, 1999):

Extraposition of post-nominal modifiers of noun phrases is in general rare […] In general an element can be extraposed to the end of its sentence only when the extraposed element is sufficiently important to the purposes of the text and the text segment in which it occurs to warrant the prominence it receives as a result of being postponed to the end.

So one problem here is that "currently living in the Hamilton area" is not heavy enough — whether that's because it's not "sufficiently important to the purposes of the text" or because it's just too short and simple — to warrant extraposition from "German language speakers from Austria, Switzerland, Liechtenstein or Germany".

Another problem may be that there's a lot of other stuff in the verb phrase: the basic object noun phrase is already a complex one ("German language speakers from Austria, Switzerland, Liechtenstein or Germany"), and then there's the prepositional phrase "for a linguistic study on the relation between word order and articles".

But there's an even more important problem: it's somewhere between hard and impossible to extrapose a post-nominal modifier headed by a gerund-participle (like "currently living in the Hamilton area"), no matter how heavy or important it is, and no matter how light-weight the intervening stuff is. In fact, Joe Emonds (A unified theory of syntactic categories, 1985) used examples like the following to support an argument that "bare reduced relative gerunds should not extrapose" at all:

That man (who is) cleaning the table has a nice shirt.
That man has a nice shirt who is cleaning the table.
*That man has a nice shirt cleaning the table.

Somebody (who is) visiting from the capital wants to give a talk.
Somebody wants to give a talk who is visiting from the capital.
*Somebody want to give a talk visiting from the capital.

Somewhat similar things do occur:

At this moment a servant came into the room, and said that a man was in the porch beyond the hall, waiting to speak to Sir Peter.

But it was like when someone is in the toilet with you waiting for you to pee, and so you can't, even though you have to.

However, these are probably not really what Emonds calls "bare reduced relative gerunds", since they also work in cases where the un-extraposed version would be questionable at best, e.g.

Someone waiting to speak with Sir Peter was in the porch beyond the hall.
Someone was in the porch beyond the hall waiting to speak with Sir Peter.
?He waiting to speak with Sir Peter was in the porch beyond the hall.
He was in the porch beyond the hall waiting to speak with Sir Peter.


  1. Nick Lamb said,

    December 11, 2011 @ 3:04 pm

    "That man has a nice shirt who is cleaning the table" is problematic for me. I'm pretty sure I would never say that and I would be momentarily confused if someone else uttered it. I think the problem is that "who is cleaning the table" seems to be there to tell me who is meant by "that man". But if it's necessary I think I expect this information up front, or else in its own sentence as an answer to a (perhaps unspoken) question.

    "That man has a nice shirt. (Which man?) The one with the bowler hat and dark hair, on the corner by the taxi"

    Whereas "Somebody wants to give a talk who is visiting from the capital" is less troubling, because maybe the fact that they're visiting from the capital isn't there to tell us who "Somebody" is, but for some other purpose, perhaps to emphasise their importance or the need to book a specific date.

  2. Ben Hemmens said,

    December 11, 2011 @ 3:46 pm

    Kudos to the author of the ad for knowing that Liechtenstein exists. But since it is home to perhaps 30,000 German speakers, completeness might suggest mentioning also the Autonome Provinz Bozen – Südtirol / Provincia Autonoma di Bolzano – Alto Adige, home to about 300,000 German speakers; Luxembourg (around 300,000 German speakers) and the German-speaking community in Belgium (74,000). Harder to estimate is the number of native German speakers in Alsace, though it might be more than in Liechtenstein. I doubt whether the Liechtensteiners or any of these others living in the Hamilton area would feel excluded by saying Germany, Austria and Switzerland; on the other hand, I feel confident that putting in the entire list would have headed off the extraposition problem.

  3. Derek McIver said,

    December 11, 2011 @ 5:12 pm

    Related to Ben's comment above, I'm a bit shocked that in an article analyzing word order, nothing is mentioned about the order of the countries listed. Why is Germany last? Wouldn't that be the most logical first item in the series? Are the researchers suggesting that speakers of Hochdeutsch are less desirable for their study?

  4. Ben Hemmens said,

    December 11, 2011 @ 6:09 pm

    Germany itself is home to a variegated patchwork of dialects. The division between someone reliably being able to communicate in Hochdeutsch is much more a matter of education and what practice the person gets in the course of their everyday communication than whether they live in Germany, Austria or Switzerland. Anyone who needs to carry on written communication for work can write and speak Hochdeutsch. People who speak Hochdeutsch natively (ie don't speak a recognisable dialect among family and neighbours) are to be found only in certain regions of Germany (roughly speaking, the middle) and in mobile elites.

  5. a George said,

    December 11, 2011 @ 6:34 pm

    using a flash on a camera to photograph a b/w text very frequently does give these green hues when you try to correct the contrast by software. So, it is authentic, allright.

  6. Nick Lamb said,

    December 12, 2011 @ 5:50 am

    Derek, I think maybe "German speakers from Germany" (even when you then go on to list several other countries) sounds and looks strange, as if it's redundant, even though strictly it is not. So that explains not putting Germany first in the list, at least.

    [(myl) FWIW, the heading of the original flyer uses the German-language names of the countries involved, and orders them with Deutschland first.]

  7. Ben Hemmens said,

    December 12, 2011 @ 8:18 am

    In German, we often use the abbreviation / acronym DACH.
    D, A and CH are used as prefixes for the zip codes and Dach is the word for "roof".

  8. Warsaw Will said,

    December 12, 2011 @ 3:15 pm

    I can quite see why you refer to gerund-participles, just like in EFL we often refer to both gerunds an present participles as '-ing' forms. And I can see there is a good argument for fusing the two terms. But I cannot see how participles in reduced relative clauses can be referred to as 'reduced relative gerunds', as they are not functioning in the slightest as nouns.

    I also agree with Nick Lambe that that these two examples are very unlikely, and can't imagine a native speaker ever saying them.

  9. J. Goard said,

    December 12, 2011 @ 11:21 pm

    @Nick Lamb:

    I think you may be suffering a bit of written language tunnel vision. (We all get it from time to time.) In spoken discourse, this seems like a perfectly common structure: the speaker makes a concise personal assessment, which omits some information potentially relevant to the addressee, and then subsequently fills in the addressee's gaps with a relative clause:

    (1) Dude, that chick was hot who we met in the library.
    (2) That video was stupid that the professor told us about.
    (3) A: What should we bring to the party? B: Those sandwiches were great that we made last time.

    For the original sentence in question, just imagine yourself suddenly noticing the man with the nice shirt and remarking on it, then (because there are other men in the vicinity) tacking on a relative clause in case your listener doesn't know who you meant. Is it still strange?

  10. Warsaw Will said,

    December 13, 2011 @ 4:46 pm

    @J.Gourd – I can't agree. Yes, of course we hesitate and leave things out in spoken language, but we don't usually come out with such unnatural sounding sentences as a result.

    These are all defining relative clauses where in spoken English we don't even pause for breath. What's more the relative pronoun represents the object, and so in spoken English is usually left out. Presumably we can only do this because what follows is absolutely automatic, we don't even need the thinking time of the pronoun.

    (1) Dude, that chick we met in the library was hot.
    (2) That video the professor told us about was stupid.

    And even if we did miss out this information, a much more likely solution would be something like:

    (1) Dude, that chick was hot. You know, the one we met in the library.
    (2) That video was stupid. That one the professor told us about.

  11. Chad Nilep said,

    December 15, 2011 @ 12:25 am

    @Warsaw Will
    "we don't usually come out with such unnatural sounding sentences"

    Engagement with spoken corpus data reveals that we (language users) are very bad at knowing what we usually say.

  12. Jake said,

    December 15, 2011 @ 6:04 pm

    there's not much of an issue here. when you see the construction "between x and…" you can be 99% confident it's going to be followed by "y," thus completing the "between x and y" clause

RSS feed for comments on this post