Metaphysics intruding on morphology

« previous post | next post »

I received this email message this morning:

Dear Student Systems User

There are currently problems with the main database server, affecting NESI, EUCLID, WISARD, STUDMI, etc.

IS are investigating, but we have no timescale for a resolution. Sorry for any inconvenience

Student, Admissions & Curricula Systems

You might like to reflect awhile on the linguistic lessons you can learn from this. Then read on…

Of course, one thing I learned is that this is going to be a terrible day for anyone relying on databases of information for or about students. Admissions, courses, registration, timetabling, everything. I don't know why university database systems are so flaky, ill-designed, and prone to disaster, but they are. But this is Language Log, not Database Systems Log. I said linguistic lessons. And I think the major linguistic points are these.

  1. The claim that some have made about regular plurals not being permitted in syntactic nominal compounds (about how allegedly mice eater is grammatical but rats eater is not) are just not tenable. Student Systems user is a counterexample, and so is Student, Admissions & Curricula Systems (or more simply admissions system, which is what EUCLID is supposed to be when it's up and running).

  2. Computer support guys these days are saving bytes on punctuation: nothing after the salutation line (Dear Student Systems User); nothing after the pre-signoff sentence; nothing after the signoff (Regards).

  3. We have no timescale for a resolution is a nice example of the jargony official talk that people switch to for institutional announcements (an individual was apprehended rather than someone was picked up, and so on). Why could they not say we don't know when it will be fixed ? Because they are an office writing to the university at large and they need to feel that they have behaved officially. Not, in my view, the appalling ethical catastrophe that George Orwell made of it in his vastly overrated, flailing essay "Politics and the English language", but definitely a known feature of English produced under the constraint of being in a bureaucratic role.

  4. And somewhat more interestingly, the British English plural verb agreement with subjects denoting collectivities is clearly alive and well. IS in the second sentence of the message text is the university's Information Services division, but Information Services is not exactly a plural NP here. It's a singular NP with plural morphological form, like cornflakes as in Cornflakes is my favorite breakfast. It should be clear semantically that the verb are does not agree syntactically with services: after all, it is not the services that are investigating (they are abstract objects). It is the collectivity consisting of the people working in IS that can be said to be investigating. The noun phrase IS denotes the collectivity as a whole. So the sentence has the same structure as IBM are investigating. Most Americans would write that as IBM is investigating. This contrast in preferences is a minor syntactic dialect difference between UK and US English: British speakers tend to prefer plural agreement with subjects denoting companies, sports teams, departments, committees, governments, and so on, at least when they are being referred to in a way that highlights their collective nature. Metaphysics intruding on morphology.


  1. Rachael said,

    November 4, 2009 @ 12:43 pm

    Doesn't point 4 undermine point 1? I don't have any more context than is in the post, but I read "Student Systems" as a "singular NP with plural morphological form, like cornflakes" (before I even read as far as point 4). So I don't think "Student Systems user" (or indeed "cornflakes eater") works as a counterexample to the "mice eater"/"rats eater" observation.

  2. Mr Fnortner said,

    November 4, 2009 @ 12:52 pm

    I imagine the first rough cut at "IS is investigating, but…" left a techie aghast with nowhere to turn but to change the repellent second is to are.

  3. Rich said,

    November 4, 2009 @ 12:57 pm

    Geoffrey –

    Do you think the issue of point #2 is more than just saving effort typing? That it is an actual move toward the use of periods as sentence separators, instead of sentence terminators? In the Sorry sentence, there is no need to indicate the end of the sentence, because the end of the paragraph — the blank line — implies it. Have you seen other examples like this?

    I've seen this sort of thing in computer code comments many times. Single line comments that are whole sentences have no terminal periods. Periods are only used when there are two sentences on one line. (But then the second sentence sometimes gets its period, too. Nobody's consistent.)

    On the other hand, the missing period after inconvenience may be nothing more than a typo…

  4. Otter said,

    November 4, 2009 @ 1:14 pm

    As a member of the IS collective, this is the equivalent of a telegraph from a correspondent in a muddy trench under heavy artillery fire. I would hate to have any linguistic lessons drawn from the (thankfully few) similar messages I have been forced to send.

    The hope that the rest of my career will pass without the need for another keeps me from drafting a pleasing and grammatically unremarkable form letter for future use.

  5. Tim Silverman said,

    November 4, 2009 @ 1:34 pm

    @Rachel: that was my reaction, too. I also interpreted Admissions as a singular in plural form, though I wasn't very clear over it was a process or a department (or indeed something else).

    @Mr Fnorter: no, that really is the British collective plural in action. Although IS is collectively investigating, it is doing so by virtue of its individual members investigations, so the plural is appropriate (in British English). If the verb can only apply to IS as a unit, then the singular would be necessary ("IS is one of the largest departments"—obviously its members are not departments at all, a fortiori not the largest one.).

    @Rich: I've seen this a lot too (and done it myself!) and also suspected that something like your explanation was the correct one: the full stop is treated as unnecessary because of the paragraph break.

  6. BlueBottle said,

    November 4, 2009 @ 1:42 pm

    Other than the missing full stop at the end of the pre-signoff sentence, Point 2 just looks like the standard style for business correspondence that's been current in the UK for some decades, now.

  7. John Cowan said,

    November 4, 2009 @ 2:07 pm

    It seems like our Geoffrey has passed too much time in the States, where official writing continues to be in a commatose condition.

  8. E. Pyatt said,

    November 4, 2009 @ 2:21 pm

    The only thing that stood out was Lesson 4 (IS with plural noun) which seems very British to me. I'm pretty sure if our U.S. tech support group were in the midst of a similar server crisis it would be "IT is investigating…"

    I agree with Otter that the text shows the hallmarks of an IT organization working on a major meltdown. On the one hand they are using official jargon to make it look like things are "under control" (irony quotes) peppered with references to arcane acronyms the IT group is familiar with. However the passage "Sorry for the inconvenience" seems a little too brisk/informal (I would expect "We apologize for any inconvenience" if a copywriter had been involved.) Sounds like a quick one-off from a programmer.

    In any case, I hope IS is/are able to get the servers back online soon.

  9. Adrian said,

    November 4, 2009 @ 2:44 pm

    The fact that nouns used as modifiers are sometimes plural doesn't nix the general rule that they should be singular. And there is usually a good reason for "breaking" the rule, e.g. "admissions" is a recognised term which it would be odd to change, and "systems user" may emphasise that it isn't just one system that's being referred to.

  10. Nathan Myers said,

    November 4, 2009 @ 2:53 pm

    I notice particularly the "currently". It's rather a long word for what amounts to a stutter, but that's how I see it almost always used.

    While I'm here, I'd like to express appreciation for the honest use of "problems". Most of IS in the States would have written "issues" in that spot. I wonder what even more vague term they will migrate to when "issue" becomes too freighted with negativity.

  11. Alan Palmer said,

    November 4, 2009 @ 2:55 pm

    I'm Britsh, so use and agree with the use of IS are investigating. I very much doubt that the entire personnel of the division are actually investigating the problem; some must be left to handle other things.

    I must say that I'm a little pleased at the reference to problems, though. I wrote a similar notice for our corporate website a while back, and my boss substituted issues "because that way it doesn't sound so bad".

  12. jfruh said,

    November 4, 2009 @ 2:55 pm

    Frankly what I find more interesting about the last paragraph is the kind of abrupt shift from the third to second person plural: "IS are investigating, but we have no timescale for a resolution." (I'm assuming that the we in question are members of the IS team.)

    I am reminded of the dilemma faced with Facebook status updates, which by default begin with the user's name. "Josh is eating breakfast" is easy enough, but somehow typing "Josh is eating his breakfast" brings home the fact that you're referring to yourself in the third person, which feels faintly ridiculous, whereas "Josh is eating my breakfast" just looks odd.

  13. D.O. said,

    November 4, 2009 @ 2:56 pm

    I am surprised by the use of the word timescale. Is it because of the "muddy trench under heavy artillery fire" situation or the word does have the meaning of timetable as one of the possibilities?

  14. Spell Me Jeff said,

    November 4, 2009 @ 3:30 pm

    I actively participate on a board set up for web developers, which includes programmers as well as HTML designers.

    Because most punctuation symbols serve as programming operators, their indiscriminate use in a programming context can lead to catastrophic errors. Here is a typical bit of advice:

    Try writing $var = isset($_POST['var']) ? $_POST['var'] : null;

    The writer in me longs to finish off the remark with a period; but what if this period were misconstrued as a continuation of the programming statement? The statement would fail. I wonder if some hardcore IT people have simply abandoned punctuation qua punctuation in the interest of unambiguous code.

  15. Andrew (not the same one) said,

    November 4, 2009 @ 3:47 pm

    Rachael: I think in this case 'Student Systems user' does indeed mean 'user of student systems', just as 'rats eater' might mean 'eater of rats'. 'Student Systems' isn't parallel to 'Information Services', which is a grammatical plural with singular meaning; it refers to the actual systems, not to the department which runs them.

    E. Pyatt: in this case the arcane acronyms would actually be understandable by users (who indeed might well know what systems they referred to without knowing what they stood for).

  16. Alex Dodge said,

    November 4, 2009 @ 3:53 pm

    In regards to periods being used as sentence separators:

    This could be reading too much into it, but since this is a university IT department, I can infer that most of their work consists of hacking Perl together. Perl uses semicolons as line separators, not terminators as in some other languages.

    Alex Dodge

  17. rpsms said,

    November 4, 2009 @ 5:22 pm

    Rather than being a jargon term, "timescale" more likely refers to the inevitable question from 100s of users about time in the broadest sense (e.g. "Are we talking hours, days, or weeks?")

  18. Nathan Myers said,

    November 4, 2009 @ 5:50 pm

    Would a "ratseater" be the usher at an informants' symposium?

  19. mollymooly said,

    November 4, 2009 @ 5:53 pm

    point #1 is, to a lesser extent than point #4, a facet of British-American dialect difference. Plural nouns as attributives are somewhat commoner here than there.

  20. OhioBob said,

    November 4, 2009 @ 6:07 pm

    The author would receive a warning from Word when writing "IS is investigating" to prompt about the double use of "is". As Bill Clinton said, it all depends on what the meaning of IS is.

  21. Scriptor Ignotior said,

    November 4, 2009 @ 6:20 pm

    This is the part that intrigued me:

    The claim that some have made about regular plurals not being permitted in syntactic nominal compounds (about how allegedly mice eater is grammatical but rats eater is not) are just not tenable.

    "The claim … are just not tenable." Mes félicitations! Delightfully oblique. The attraction of the plurals some, plurals, and compounds certainly are strong.

  22. Matthew Kehrt said,

    November 4, 2009 @ 6:57 pm

    @D.O. I think "timescale" is close to verging on stadard geek slang these days. If I wanted a vague idea of when something would be done, I would almost certainly ask "what's the timescale on that?"

    @Spell Me Jeff I have the exact opposite problem, I think due to the influence of programming. If I'm not careful, I tend to write single paragraph sentences of a sequence of clauses delimited by semicolons.

  23. Karen said,

    November 4, 2009 @ 9:10 pm

    Might it be that all rats are more or less alike (at least to the eater) while each system is different?

  24. Nathan Myers said,

    November 4, 2009 @ 9:52 pm

    "Timescale" is just slightly geekier than the management-speak "timeframe".

  25. Aaron Davies said,

    November 5, 2009 @ 1:46 am

    "timescale" is jargon for something like "estimated time of delivery". it probably comes from staring at gantt charts.

    re: "problems" vs. "issues", the current candidate for next runner on the euphemism treadmill seems to be "challenges".

    re: redundant paragraph-terminating periods, i find myself dropping them fairly frequently, but my typing has recently been influenced by a lot of time spent chatting while playing networked games, where milliseconds really can make the difference between victory and defeat.

  26. Picky said,

    November 5, 2009 @ 5:38 am

    A parallel to this use and non-use of periods can be found in (at least British) newspapers where the convention is that picture captions do not finish with a full point. If the caption comprises two sentences the first will end with a full point and the second without. (Although one can sometimes see that the sub-editor has been sufficiently disturbed by this to insert the second full point, in defiance of house style).

  27. Steve F said,

    November 5, 2009 @ 8:41 am

    I would agree that the British tendency to use plural verbs for collective nouns is more a matter of metaphysics than morphology, because we do it in a very specific and logical way, and differentiate between singular and plural collectives in quite subtle ways. For example, 'England' is obviously a singular noun, but on the – admittedly infrequent – occasions that the English national team (football, cricket, rugby etc.) is beating its opponents, it is perfectly natural, almost obligatory, to say 'England are winning'. On the other hand, it would sound odd, to my ears anyway, to say 'England are winning the war' and totally ungrammatical to say 'England are my native country'. 'The police' is generally plural – 'The police are coming', 'The police have arrested the suspect' – but sometimes 'the police' denotes a single body, so that it would be weird to write 'The police are an arm of the state'. Inconsistent, but logical.

  28. Paul Power said,

    November 5, 2009 @ 10:08 am

    There's an old joke, forgive me if you've heard it before..

    A tailor wants order more than one goose (a sort of iron). He starts to write:

    "Please could you send me two geese.."

    This does not sound right, so he tries again..

    "Please could you send me two gooses.."

    Again he's not happy. Finally he solves his problem:

    "Please could you send me one goose…………..

    PS While you're at it could you send me another goose?"

  29. language hat said,

    November 5, 2009 @ 10:57 am

    There's a very similar short story by Zoshchenko in which a department wants to order more pokers but can't decide on the genitive plural of the Russian word kochergá 'poker.'

  30. army1987 said,

    November 5, 2009 @ 12:50 pm

    @Paul Power and lanugage hat: there are similar jokes in Italian ("oranghi" or "orangi"?), too.
    As for "timescale", as a geek myself I think it's what rpsms says. One can say "on geological timescales" or "on atomic timescales" to mean very long/very short times. When using this meaning speaking seriously, both an hour and three months would be "human timescales", but …

  31. Ellen said,

    November 5, 2009 @ 1:07 pm

    The period after "etc", being as it's after an abbreviation, doesn't really truly break the pattern of no punctuation at the end of lines/paragraphs.

  32. Boris said,

    November 5, 2009 @ 1:17 pm

    Re: goose, kocherga, etc: apparently, at least at one point, Microsoft insisted that the plural of "mouse" (as in pointing device) is "mouse devices".

  33. Eddy said,

    November 5, 2009 @ 2:41 pm

    @Geoffrey "Most Americans would write that as IBM is investigating." If relaying this to a colleague would they then say "It doesn't know when it will be fixed." or "They don't know when it will be fixed."?

    I think that a lot of the use of plurals, here in the UK, is to avoid the she/he PC maelstrom.

  34. Jan Schreuder said,

    November 5, 2009 @ 3:19 pm

    I would like to repeat Rachael's point (because I had exactly the same reaction and I would like to know what Geoffry thinks ):

    "Doesn't point 4 undermine point 1? I don't have any more context than is in the post, but I read "Student Systems" as a "singular NP with plural morphological form, like cornflakes" (before I even read as far as point 4). So I don't think "Student Systems user" (or indeed "cornflakes eater") works as a counterexample to the "mice eater"/"rats eater" observation"

  35. Joe said,

    November 5, 2009 @ 4:01 pm

    Not sure what all the whinging is about regarding jargon. So Orwell is wrong when he says that there is a ethically catastrophic component to breaking the rule, "never use jargon if there is an everyday equivalent" but GP is right if the same rule is broken used for "official talk". What's wrong with breaking the rule if everyone gets the sense, " we don't know when it will be fixed"? Is it because there is an implicit sense of authority in such language that some people find distasteful?

  36. Paul said,

    November 5, 2009 @ 4:42 pm

    I think the "timescale" sentence is indeed jargon, and that rpsms is being too charitable here.

    Speaking of jargon, why do people in the military always refer to people as "individuals"? And what's up with "improvised explosive device"? Nothing makes an IED anything other than a "bomb," and I have a feeling things are called IEDs without any real thinking as to whether they're actually improvised.

  37. Tom said,

    November 5, 2009 @ 5:17 pm

    When I read "IS are investigating…," I assumed that the sentence originally read, "We are investigating…" but the author thought that perhaps the reader wouldn't know who "we" was so they changed that without checking the rest of the sentence.

    As to "problems" vs. "issues," we used to use a product called BugTracker to keep track of errors found in our code but now we use Jira Issue Tracker because we don't have any bugs in our code anymore.

  38. cjackson said,

    November 5, 2009 @ 6:30 pm

    @Spell Me Jeff

    I find that I most frequently do something similar to block quotes when I'm writing an e-mail with literal code. Though when I don't separate it like that, I am usually hesitant to add the period without also adding some explicit marker of the code segment (parentheses, quotation marks, etc.). I also pluralize java objects with 's, because the name is meaningful to the computer on its own, but meaningless to the computer with an s attached. So I think it's very reasonable to assume that writing about code could affect your writing in other domains.

    (I'm tempted to replace "objects" with "Object's", but I won't.)

  39. Matthew Kehrt said,

    November 5, 2009 @ 8:44 pm

    @Paul I think you are somewhat missing the point on what is meant by "jargon" here.

    I strongly suspect that the word "timescale" was used, not because the writer was switching into a mode of speech which they considered more appropriate to communicating with non-techies , but because he or she *failed* to. The word "timescale" to mean "expected timeframe" gets used consistently (in my experience) among IT and CS people in casual conversation with each other. I suspect that it did not even occur to the author of that email that others might consider that use bizarre.

    In other words, I don't think this is a case of switching to the sort of euphemistic "manager-speak" that people love to complain about, but just the normal speech for the author.

    I will freely admit that this is based solely on my limited experience, but I and people I know certainly use the word "timescale" in that way.

  40. Paul said,

    November 6, 2009 @ 9:35 am

    @Matthew Kehrt: Fair enough. I see your distinction.

    But I think that when IT people say things like "we have no expected timeframe for resolution," they're still speaking in jargon. In fact, that's the essence of jargon: "the language, esp. the vocabulary, peculiar to a particular trade, profession, or group."

    I don't think such jargon is the "normal speech" of the author. If a student called to ask when the problem would be fixed, I'd wager that the author would respond in plain English: "We have no idea when it will be fixed." When a memo is called for, on the other hand, the same author probably switches to a sort of parallel "normal mode": jargon.

  41. Andrew (not the same one) said,

    November 6, 2009 @ 4:05 pm

    Paul Power: I have always heard this story told about mongeese. Er, mongooses. Er…

  42. Martin Ellison said,

    November 9, 2009 @ 1:49 am

    " we have no timescale for a resolution" means: don't keep ringing us up and asking us when it will be fixed, because we have no idea.

  43. Fiona Hanington said,

    November 10, 2009 @ 11:45 pm

    I am so lucky to work for a company whose IT communications person is terribly funny. I eagerly ready all IT announcements, forward them to my friends and family, and actually Reply to His Messages.

    But I know I am (nearly) alone.

  44. Breffni said,

    November 13, 2009 @ 11:52 am

    I think "we have no timescale for a resolution" is just fine. For a start, Geoff's alternative, "We don't know when it will be fixed", implies total cluelessness and maybe even abdication of responsibility (partly due to the passive, in this case, though "we don't know when we'll fix it" is worse and "we don't know when we'll get it fixed" is not much better). It would be appropriate for an end-user of the system like Geoff to tell a colleague "The system's down, I don't know when it will be fixed". But, jargon aside, he'd be unlikely to say "I don't have a timescale for its resolution", because it implies that he ought to have one. The frame "We don't have X" is a good fit for service-providers – they aim to have certain things to offer their clients. "We have the skills, we have the personnel, we have the tools, we have a plan, we have good intentions, but we don't have a timescale for a resolution". And if you concede that much, then you can see that no other simple noun can comfortably substitute for either "timescale" or "resolution". At least, "time" and "fix" certainly won't.

  45. David Walker said,

    December 1, 2009 @ 4:56 pm

    @Breffni: How about "We will fix it as soon as we can" or "we don't know when we'll be able to fix it" or "We are working on it".

  46. Breffni said,

    December 2, 2009 @ 8:39 am

    David, those don't quite do the communicative job that the technicians want done. Your first and third options almost invite the question they want to forestall (namely "When will the systems be working again?"), and "we don't know when we'll be able to fix it" suggests either incompetence or that they're not making it a priority. Their version tells users not to ask when it'll be back online, but without suggesting they're clueless or don't care, and it probably sprang easily to mind: one thing the "We don't have an X" implies is "so there's no point asking for one." You could come up with other ways to do that, and I doubt they spent a long time drafting it, but their wording does the job fine.

RSS feed for comments on this post