Complaint(s) Department

« previous post | next post »

Today's Non Sequitur:

This illustrates the protean nature of English compound-noun semantics. But it made me wonder whether the commoner form was "Complaint Department" (as in this strip) or "Complaints Department", since I think I've seen both.

A bit of poking around turned up something that I didn't expect: The COCA corpus (425 million words of American English) has 8 instances of "complaint department" and none of "complaints department". In contrast, the British National Corpus (100 million words of British English) has 1 instance of "complaint department" and 9 instances of "complaints department".

This might be taken by crypto-Whorfians to suggest that the Brits are (10/100)/(9/425) = 4.7 times more concerned about complaining than Americans are, but I won't go there, among other things because these counts are rather small, and may simply reflect a difference in the distribution of source-text types. But these results do suggest that there's a trans-Atlantic difference in the choice of "complaint" vs. "complaints" as the first element of this compound noun.

This in turn suggests several questions that I don't have time to look into this morning: Is this apparent difference (between British "complaints department" and American "complaint department") a real one? How strong are the geographically-linked preferences? (Not strong enough for me to have noticed, anyhow.) Is this an instance of a more general trans-Atlantic difference in noun-compound morphology? Whatever the difference is, when and how did it arise (if, of course, it exists at all)?

Perhaps someone has already explored these questions, somewhere in the extensive linguistic literature on the form and meaning of English noun compounds and complex nominals more generally; if so, no doubt a helpful reader will tell us about it in the comments.

Meanwhile, for some insight into one of the reasons that this extensive literature has come into being, see this exchange between Geoff Pullum and me: "Postcard from Vegas, 3: Regularly-inflected plurals exclusion? I don't think so", 12/1/2003; "Activities Centers in Paradise and Santa Cruz", 12/1/2003; "The rigors of fieldwork trips", 12/1/2003.

Update — A bit more data.

Guardian page counts via Google "site:" search, actual pages returned rather than estimates; NYT page counts from, 1981-present.      "activity centre" 190 "activities centre" 264    (58% plural)
New York Times:    "activity center" 149 "activities center"   95    (39% plural)

guardian "complaint department"  13 "complaints department" 179  (93% plural)
NYT        "complaint department"   43  "complaints department"  10  (19% plural)

So the early returns suggest that there is a real difference, but one that's probabilistic rather than categorical, and one that may differ for different compounds.


  1. Colin John said,

    December 7, 2011 @ 8:37 am

    'Complaints' would be normal for me (BrE – in my 50s). I wonder if singular 'complaint' in the UK carries more of a suggestion of 'illness' than it does in the US. The exchange:
    'What's your complaint?'
    seems normal enough to me.
    Or maybe we always have lots of complaints to make, not just the one!

  2. Robin Lovell said,

    December 7, 2011 @ 8:47 am

    Two questions occur to me.

    1. Is this UK/US usage difference connected, however tenuously, with the 's' that gets dropped somewhere in Mid-Atlantic from 'towards', 'backwards', etc.?

    [(myl) I don't think so.]

    2. Haven't all complaint(s) departments been replaced by customer service (sic) departments?

    [(myl) Apparently not.]

  3. Ian Tindale said,

    December 7, 2011 @ 8:50 am

    I’m glad you’ve done the maths.

  4. Jon H said,

    December 7, 2011 @ 9:22 am

    My thought was the same as Robin's second point. However, in the UK I would expect this to be "Customer Services". I do hope that in the US it's usually "Customer Service".

  5. Martin J Ball said,

    December 7, 2011 @ 9:31 am

    UK: Road Works
    US: Road Work

  6. David Denison said,

    December 7, 2011 @ 9:39 am

    Is this relevant? From the Cambridge History vol. 4 (1998: 130):

    The modifying noun is usually singular [example omitted, where e.g. _a bill worth ten dollars_ corresponds to _a ten-dollar bill_].

    Quirk et al., having pointed that this holds even for nouns which otherwise have no singular, as in trouser press, go on to suggest that nevertheless ‘the plural attributive construction is on the increase, particularly in BrE where it is more common than in AmE’ [American English], citing examples like a grants committee (1985: 17.108). We might also compare the aurally identical variation in BrE doll’s house vs. AmerE doll house.

    [(myl) There's no question that the first-element nouns in noun compounds are often singular, and the "trouser press" case is a nice illustration of the fact that this can happen even when the singular form is otherwise unexpected.

    It's nice to see that Quirk noted the (apparent) trans-Atlantic difference. I wonder whether his claim that "the plural … is on the increase" is really true, given e.g. this.]

  7. Duff L said,

    December 7, 2011 @ 11:06 am

    1. I wonder whether there might be more consensus between BrE and AmE when using the construction 'Department of Complaints'.

    [(myl) There's no reason to suppose that English speakers from either group would be inclined to use the singular in such cases — this issue, I think, is not about singular/plural usage in general, but rather about the specific case of the first element in a noun compound.]

    2. YOU want to complain?! Look at these shoes. . . .

  8. Coby Lubliner said,

    December 7, 2011 @ 11:31 am

    BrE: drugs dealer, sport section
    AmE: drug dealer, sports section

  9. Mr Punch said,

    December 7, 2011 @ 11:45 am

    "Complaints department" immediately struck me as a Briticism. But, as Coby L. points out, the pattern isn't consistent. I reckon it's just a random variation, like drink/drunk driving.

  10. Boris said,

    December 7, 2011 @ 12:28 pm

    I think sport(s) is a special case, and therefore cannot be used as a counterexample. For example, the sport(s) section of a newspaper would be entitled sport in the UK and sports in the US, but a complaint(s) section (if there were such a thing) would never be called simply "complaint" (unless there is only one complaint in the section).

    [(myl) Exactly. In more detail: BrE has a common abstract noun "sport", used in phrases like "interested in sport", for which the AmE equivalent is the plural "sports". So "sport section" is like "fashion section" — it's a section about sport, and there's no reason to add an 's' in such cases.

    It's interesting, though, that in AmE the 's' is pretty much obligatory in "sports section".]

  11. Andrew (not the same one) said,

    December 7, 2011 @ 1:57 pm

    I'm British and I would say 'drug dealer'. I don't think I'm alone in this either – there was an episode in former British soap opera Brookside where a character had 'drug dealer' painted on the front of his house (by his enemies, obviously), and another character misread it as 'rug dealer'.

    [(myl) The British National Corpus has 37 instances of "drug dealer" and only 8 instances of "drugs dealer" — only 18% "drugs".]

  12. Robin Lovell said,

    December 7, 2011 @ 1:58 pm

    Although I've lived in USA almost 30 years, there are some American usages that I resist, because they still sound wrong to me, no matter how many times I hear them. I just can't bring myself to say "math" and evade the issue by using "mathematics" in full.

  13. Chandra said,

    December 7, 2011 @ 2:22 pm

    As a British-parented Canadian resident who lived in the U.S. for several of my formative years, "complaints department" sounds more natural to me. I'd like to be able to verify this against general usage in Canada; however, every company I can think of offers only a much more benign-sounding "customer service / customer care / client care" department instead.

  14. J.W. Brewer said,

    December 7, 2011 @ 4:01 pm

    The google n-gram viewer has "suggestion box" dominating "suggestions box" for both AmEng and BrEng, although perhaps more overwhelmingly so for AmEng.

    [(myl) COCA is 18/1 for the singular, BNC is 8/2 for the singular. All the data so far seems to be consistent with a general BrE effect towards a higher proportion of plurals, interacting with strong individual differences for particular compounds. (If we ignore the cases where for one reason or another, only one possibility exists…)]

  15. Ralph Hickok said,

    December 7, 2011 @ 4:39 pm

    I consider myself a sports historian, but I belong to the North American Society for Sport History, which publishes the Journal of Sport History.

  16. mollymooly said,

    December 7, 2011 @ 4:59 pm
    The top plural-noun + noun collocations in COCA:
    arts education
    rights movement/group/violation/activist
    sales tax/manager
    earnings growth
    securities fraud
    communications director
    pants pocket
    operations manager
    gains tax
    graphics card
    appeals court
    arms control/race/embargo
    sports car/team
    weapons inspector/system
    skills training
    savings account
    awards ceremony
    services company

  17. rkillings said,

    December 7, 2011 @ 8:39 pm

    No question that pluralized attributives in noun compounds are commoner in BrE than in AmE. But is this not just one manifestation of a more general tendency in BrE to inflect attributives?

    Compare open-end vs. closed-end fund in AmE, which becomes open-ended vs. closed-ended [sic!] fund in BrE.

    Or adjectival forms. As an American reader of The Economist, I always find it jarring to read phrases like "a Californian judge" or "the Texan press". But these very usages are commanded by the newspaper's own style guide ("PROPER NOUNS: if they have adjectives, use them.") Odd, given that the USA accounts for well over half of its paid circulation, and AmE typically doesn't inflect the names of states in this kind of expression.

    The mystery is why "appeals court" in AmE is "appeal court" in BrE. (However, I see that The Guardian preserves the AmE usage when reporting about US court decisions.)

    [(myl) The question of whether place names as modifiers are used in nominal or adjectival form is not so simple — see e.g. "All your base are belong to which lexical category?", 5/15/2004; "W's conundrum", 2/23/2006; "More political morphology: Democrats, Great British, and geese", 2/19/2007.

    Note also that (according to stress and sense alike) these constructions are modifier-noun phrases, not noun-noun (or adjective-noun) compounds; so this (though interesting) is a different question from the one that we started with.]

  18. Brett said,

    December 7, 2011 @ 9:36 pm

    @rkillings: I believe that the difference between "appeals court" and "appeal court" is inherited from the more formal terms: "court of appeals" versus "court of appeal." I interpret these formal terms as being slightly different in structure, analogous in meaning to "vale of tears" and "insurer of last resort," respectively.

  19. o. said,

    December 7, 2011 @ 9:57 pm

    I have in the past noticed that in BBC and other British sources, noun-noun NPs sometimes have the first noun plural where the AmE equivalent would be singular; often enough that I mentally tagged it as a generically British.

    [(myl) A Google site search of turns up 101 "drug czar" to 57 "drugs czar". A similar search on the turns up 147 "drug czar" to 200 "drugs czar". A search of the NYT archive since 1981 yields 523 "drug czar" to 0 "drugs czar". So there's certainly a difference, but at least on the British side, it's not a categorical one.]

  20. njun hung said,

    December 7, 2011 @ 10:45 pm

    I am a non-native English speaker. Things like these confuse me so much! Anybody can help with a proper choice?
    What about 'languages centre, language centre, languages center, and language center'?. Which one is more commonly used?

  21. Will said,

    December 7, 2011 @ 11:33 pm

    I decided to compare n-gram counts for "complaint department", "complaints department", and "customer service department" in the google n-gram viewer.

    For American English:

    For British English:

    This confirms that (in this corpus of books anyway) that around 1980, "customer service department" began to rise in both AmE and BrE, and really started to skyrocket in the 90's (maybe related to the rise of the internet?).

    And interestingly, for both AmE and BrE, "complaints department" has been on the rise.

    In BrE, "complaint department" is barely there, and in AmE, "complaint department" has been fairly consistent since the 30's following a rise from nothing to a peak in 1920.

  22. Ben Hemmens said,

    December 8, 2011 @ 7:01 am

    German often inserts an -s- when creating compounds. E.g. Verwendung-s-zweck, Bestimmung-s-ort, Empfang-s-bereich, Elektrizität-s-gesellschaft, Gravitations-s-gesetz, Konzession-s-bereitschaft, Mord-s-brocken, Personenstand-s-register, Scheidung-s-anwalt, Universität-s-platz, Verkehr-s-planung. If I owned a Duden I would probably be able to look up the rule, which however should be fairly easy to see even in these few examples: the S is just a phonetic spacer for certain collisions of last and first syllables.

    -n- or -en- are often inserted for a similar reason. Both the s and n versions could also mark plurals or genetive (Verkehrsplanung as Planung des Verkehrs makes sense as a genetive; Universitätsplatz as Platz der Universität doesn't) but there are plenty of examples where they clearly don't.

    The mice eater/rat eater in German to my ear could be called either a Mausfresser or Mäusefresser (with a semantic preference for the plural, because surely the monster eats more than one mouse – on the other hand, the fact that a Fahrzeughersteller produces thousands of vehicles doesn't tempt me to call it a Fahrzeugehersteller; I'm not sure why not) whereas *Rattefresser doesn't work and the naturally inserted n of Rattenfresser is indistinguishable from a plural marking of Ratte.

    My vestigial little point: some of the inconsistency in the use of -s in English compounds might be an atavistic echo of these not-necessarily-grammatical infixes from the Germanic side of the language's ancestry.

    [(myl) This is an interesting idea, which I also considered. The fact that some compounds allow irregular plurals as first members is a negative indication, it seems to me, but there might be something to it all the same.]

  23. xyzzyva said,

    December 8, 2011 @ 7:52 am

    Ben Hemmens,
    I like the idea of considering this -s an interfix, since it dodges the issue of deciding whether it's the genitive or plural or genitive-plural in each case. It's much easier to just say it's derived from those suffixes/inflections, but itself has no meaning besides marking a stem as compounded.

    On this topic, is there anything fundamentally different between compound nouns in English and those in German, etc., other than the orthographic convention of using a space in the former and none in the latter?

  24. Ben Hemmens said,

    December 8, 2011 @ 9:28 am

    I don't think so.

    I just had a quick leaf through my Understanding English-German Contrasts (Ekkehard König & Volker Gast) and found no section on compounds.

    I think there's a bit more than an orthographic phenomenon:

    a) A compound may offer a shortcut which would not be available in English. E.g. Dein Kellerabteilfenster can be used in place of das Fenster deines Kellerabteils, whereas in English we don't really have an alternative to "the window of your basement storage room".

    b) With somewhat less of the English phenomenon of using words from different origins to express related meanings (e.g. the series cow-beef, sheep-mutton, pig-pork), German turns to compounds in this case Rindfleisch, Lammfleisch, Schweinefleisch. I do remember a German teacher once saying that German was an easy language because it had fewer words …

    – that's all that occurs to me just now.

  25. Janelle B. said,

    December 8, 2011 @ 9:45 am

    @Martin J Ball Being American, I think AmE actually uses both of those NPs, but with a difference in meaning between them.
    "Road Works" would be more like the job title – "I work in the road works office." My instinct is that this usage is influenced by the fact that such an office would oversee more than one project of "road work"…
    "Road Work" would be the name of the actual undertaking – "The men are doing road work." or "There is road work on route 2."

  26. briggslaw said,

    December 8, 2011 @ 11:11 am

    Sport section (BrE) and sports section (AmE), yes: but surely sport coat (AmE) and (sports jacket) (BrE). Go figure (figger or figyer, your choice).

  27. briggslaw said,

    December 8, 2011 @ 11:13 am

    Sorry, that should have been "Go figure (figgah or figyer, your choice).

  28. Faldone said,

    December 8, 2011 @ 4:38 pm

    @ Ben Hemmens

    As a native AmE speaker I don't see any problem with basement storage room window.

  29. chh said,

    December 9, 2011 @ 5:04 pm

    It would be nice to look at a corpus sufficiently annotated to see how often some of the plurals in these compounds take singular agreement when they show up on their own.

    As in:

    "Sales is not simple" (Sales representative)

    *"Towels is not expensive" (*Towels rack)

    It doesn't look like this can be done in COCA- I guess you would need a syntactic tree annotated corpus to do it right.

    Of course you can come up with a context where 'towels' means "the issue of towels" or "the category of things that are towels", but I'm sure that context comes up much less frequently than the one supporting "sales".

    I bet a lot of the plurals in these compounds end up taking singular agreement quite often relative to the majority of nouns that appear as singular in the first element of compounds. (But probably not all of them- 'awards' and 'pants' probably don't.)

    If that's true, then it would be interesting to see whether the geographical differences are due to these plurals getting singular agreement more often in the UK than in the US.

  30. Ben Hemmens said,

    December 9, 2011 @ 5:16 pm

    I don't have time to dream up a whole list of examples, but my point is that German on the whole tolerates these things better than English.

    Originally I wrote (and then truncated) "your basement storage room window is open" and when you look at it in a clause like that, I do think "the window of" is significantly better in English, possibly because of the strong English preference for getting the subject sorted out at the beginning of the clause. This is much less important in German.

  31. Dakota said,

    December 11, 2011 @ 10:57 am

    Mathematics class may have become a math class in the U.S., but we still have civics, statistics, social studies, computer studies, economics, physics, and gymnastics classes. Calculus has not become calculi.

RSS feed for comments on this post