An HR bureaucrat, whom cannot write

« previous post | next post »

When I give lectures on why you should not listen to prescriptivists' dimwitted prattle about the wrongness of constructions that are fully grammatical and always were, people sometimes ask me what I would regard as bad grammar, as if such cases were going to be hard to find. So occasionally I note down striking cases of failure to get English syntax right (especially written English, naturally enough), and discuss them here.

A friend (don't make me say who) with a middle-rank managerial position in a large bureaucratic organization (don't make me say which) recently received a memo informing him about which of his recommendations for staff promotions and pay increases had been successful, and part of it said:

…it is strongly recommended that you meet with staff, whom have been unsuccessful, in order to provide support after their receiving the disappointing news.

That's a rather astonishing ungrammatical case of whom, used without a shred of justification as subject of a tensed verb to which it is immediately adjacent; but also a crashingly salient case of punctuating a restrictive relative incorrectly. And the email version of the memo, amazingly, was even worse.

There are two major ways in which a relative clause may function. One is that a relative clause may be a fully integrated modifier of the noun in a noun phrase, often providing some sort of semantic restriction on the reference of that noun. Thus person can be used to denote the entire class of human beings, while person who has been unsuccessful denotes only the smaller subset of those who have failed at something. The underlined part is what The Cambridge Grammar of the English Language calls an integrated relative clause. They are often called "restrictive" relative clauses, or "defining" relative clauses.

The other major function for relative clauses is to serve as a parenthetical interruption of the main flow of a sentence, contributing supplementary information about someone or something immediately after it is referred to in the main content. Thus You can talk to John if you like just says that if you want you can talk to John, but You can talk to John, who has more experience, if you like adds some supplementary (and definitely secondary) information about John's experience level. This kind of relative clause is the one that CGEL calls a supplementary relative clause.

There are all sorts of differences between the two, but the one that is crucial here is that supplementary relatives must be separated off with commas and integrated ones must not be.

Thus, since supplementary relatives are often appended to proper names while integrated ones almost never are, it is obligatory to have the commas in You can talk to John, who has more experience, if you like.

The really important communicational difference between the two kinds of relative is that they have utterly different semantics. If I say Politicians who I admire never get elected, with an integrated relative, my claim (possibly a true one, e.g. if I only admire third party candidates) is about politicians who I admire; but if I say Politicians, who I admire, never get elected, I'm making the false claim that no politicians ever get elected, with the (astonishing) supplementary claim that I admire all politicians.

The memo my manager friend received makes the semantically disastrous grammatical mistake of putting commas before and after the relative clause, and thus makes (unintendedly) a completely false claim. I'll just ignore the distracting extra morphological error of choosing whom (rarely used these days except after prepositions, and never used as the subject of an immediately following verb). Let's just replace whom throughout, whether it's understood as a subject or not (that's what nearly everyone does: we say Who did you tell?, not Whom did you tell?). To tell a manager whose recommendations for raises have in some cases been denied that he should meet with staff who have been unsuccessful is sensible enough; but it is very different to say meet with staff, who have been unsuccessful. Putting it that way instructs him to call a meeting with all staff, or call in each staff member individually. And it claims via the relative clause that all staff have been unsuccessful. No raises or promotions whatsoever. That is not what the memo writer meant.

I mentioned earlier that the email version of the same memo was worse still. It contained two prime syntactic blunders suitable for holding up before the young as an example of how not to write. The whom is not there in the email version, but this is what it said:

We do recommend that you meet with colleagues, who have been unsuccessful, in order to provide support after receiving disappointing news.

First, it seems to say that he must meet with all colleagues. (This is not true: he is being advised to have sessions only with those whose promotion or pay raise applications were turned down by the relevant higher authority.) Second, because there are commas flanking the relative clause who have been unsuccessful, it must (under the usual rules of written Standard English) be interpreted as a supplementary one, so the sentence asserts that all colleagues have been unsuccessful (and this is also not true).

But there is a third blunder: a fairly astonishing dangling modifier. No subject is provided for the verb receiving, so one has to guess, and the guess that is probably the most syntactically plausible one is semantically implausible and clearly not right. If I said to you something like You should pour a stiff drink after receiving disappointing news, you would think I meant pour a stiff drink for yourself after you have received news that is disappointing for you. But in the quotation above, provide support after receiving disappointing news doesn't mean provide support for yourself after you have received news that is disappointing for you. It means provide support for the relevant staff members after they have received news that is disappointing for them. But it doesn't clearly say that. It leave things dangling, tempting you to delay yourself by picking up the stupid meaning and then backtracking. It fails to provide you with visible guidance.

Both the hard copy memo and the email version are atrocious pieces of writing. Someone in the Human Resources department of that organization needs to go take an elementary grammar and writing class.

And not a course that uses The Elements of Style as a text. All that E. B. White wants to impress upon you (for it was he, not William Strunk, who wrote the relevant section) is that only supplementary relatives are allowed to begin with which; integrated relatives are forbidden to use it, and must begin with that instead. This is not true: respectable users of the English language for hundreds of years have used both which and that to begin integrated relative clauses (think of Roosevelt's famous description of the date of the Pearl Harbor attack was "a day which will live in infamy"). It's a fetish that obsesses amateur grammar buffs and ill-informed copy editors, and it's not important at all. What's important are things like those commas round supplementary relatives, and the proper construction of sentences so that understood subjects of non-finite subjectless clauses will be immediately guessable.

So don't ever imagine, on the basis of all my preaching against Strunk and White, that I think all honest attempts at using English are just as good as any others. The two passages quoted above are terrible writing. It's inexcusable to write as badly as that when conveying serious material to managers in the pursuance of your administrative duties. It's a disgrace to the profession of bureaucrat. Such writing needs to be fixed. But let's make sure we fix the right things.



49 Comments

  1. David Cantor said,

    April 21, 2010 @ 10:12 am

    Of course, there *could* come a day when most careful writers adopt the habit of using commas every which way they choose, forcing readers to use context to distinguish between integrated and supplementary relative clauses. In such a hypothetical case, insisting that everyone return to the old rules of punctuation would, in fact, be prescriptivist.

    [In principle, I suppose, things could go that way. There have been changes in Standard English punctuation rules between the 18th century and the 21st. But remember, there's nothing wrong with prescriptive teachings about grammar. A good language teacher owes it to the students to lay down some rules about what to do. The crucial thing, though, is not to prescribe the following of rules that are not, and never were, descriptively accurate of error-free use of the language. —GKP]

  2. Charles Gaulke said,

    April 21, 2010 @ 10:26 am

    Only if there first came a day where we radically redefined "careful" so that "careful writers" didn't connote the class of writers who are careful to write clearly. Nice try, though.

    The diference between prescriptivism and descriptivism isn't whether or not there are rules, it's whether we think our rules determine what works or we think our rules ought to be determined by what works.

  3. Ben said,

    April 21, 2010 @ 10:36 am

    Replacing nonrestrictive whiches with thats is not just the fetish of ill-informed copy editors, but of us well-informed ones too. The Chicago Manual of Style's usage section (5.202) calls for doing so. If I'm told to apply Chicago Style to a book, I change a lot of whiches to thats (unless the writing is really good). It's harmless, mostly. And I don't confuse what I'm doing for "correcting a mistake"; I'm applying a style–a set of guidelines for choosing between alternatives. I don't think there's anything wrong with nonrestrictive which, just as there's nothing wrong with not using a serial comma, or hyphenating non-restrictive, putting a period outside quotation marks. But I change these things, because it's my job. Please understand–us copy editors are just following orders.

    [Yes, it is perfectly reasonable for a publisher to have a house style: en dashes or em dashes, parentheses around bibliographical reference dates or not, capital letter after a colon or not, inflected negatives or not… But that doesn't legitimate publishers insisting on bossily changing grammar when there's no need to. The Chicago Manual of Style is a masterful work on publishing and typesetting. But it does indeed stipulate which-hunting, and it shouldn't. I wish a few more copy editors would go to their bosses and exert some gentle pressure against such bossiness. I don 't ding people for having to work for a misinformed prescriptivist boss; but an experienced copy editor could make some recommendations once in a while. Couldn't you, Ben? —GKP]

  4. James Davis said,

    April 21, 2010 @ 10:39 am

    “if I say Politicians, who I admire, never get elected, I'm making the false claim that no politicians ever get elected, with the (astonishing) supplementary claim that I admire all politicians.”

    No, you're not making either of those claims, semantics makes it clear which you mean. To make the ‘weird’ claim, you'd need to clarify that you meant the normally unthinkable.

    [Semantics certainly does not make clear what anybody means. Only people who have utterly confused semantics with pragmatics could think that. —GKP]

    Ultimately, all a comma is is a breath or short pause. I find that even though it results in ‘comma splices’ and other ‘bad things’, it is generally /easier/ to read something as being natural if commas are used for any short pause, periods for longer pauses, semicolons for medium pauses, etcetera.

    If we stop thinking of punctuation as logical connectives of some kind, and instead started thinking of punctuation as having phonetic/prosodic value, we could both do away with all these silly rules, make learning English easier, and make written English more representative of the spoken word it represents.

    [Oh, dear. The foregoing two paragraphs would be best read as a remarkably full and clear exposition of a set of claims that are widely believed but entirely untrue. Punctuation is not prosodic markup. The relation between these obligatory grammatical elements of the written language and the highly variable intonational phrasing and pausing of the spoken language is loose, complex, and indirect. —GKP]

    Oh, but that “whom” is a fsking unjustifiable hyper-correction. You'll get no argument from me on that one.

  5. Mel Nicholson said,

    April 21, 2010 @ 11:07 am

    Charles said:

    The diference between prescriptivism and descriptivism isn't whether or not there are rules, it's whether we think our rules determine what works or we think our rules ought to be determined by what works.

    I agree, but I'd feel more comfortable agreeing if you had used a period (or better yet a long dash) instead of a comma.

    On dangling modifiers (I've taken the liberty of replacing the disastrous clause with an adjective to focus just on that issue):

    We do recommend that you meet with colleagues who were unsuccessful in order to provide support after receiving disappointing news.

    I find that inserting the direct object "them" after "provide" makes the missing subject of "receiving" much less alarming to my grammar radar.

    We recommend that you meet with colleagues who were unsuccessful in order to provide them support after receiving disappointing news.

    Having edited that, a fourth problem becomes clear. The final noun phrase is referring to specific news (namely the news that the application was unsuccessful) and as such needs a definite article "the" before it. Otherwise this memo seems to refer to any disappointing news rather than the specific disappointing news.

    We recommend that you meet with colleagues who were unsuccessful in order to provide them support after receiving the disappointing news.

    The use of "[person] was unsuccessful" in place of "[application] was denied" still annoys me, but that is more about taking responsibility for a decision than anything to do with grammar.

  6. CLS said,

    April 21, 2010 @ 11:07 am

    Somewhat off topic, but another example of a missing subject leading to a potentially unintended interpretation. This comes from a television commercial for a gubernatorial candidate in New Mexico.

    "Criminals take advantage of weak laws like giving driver’s licenses to illegal immigrants as governor, that will change."

    (I put a period at the end of this extract because that is the only place in it where I perceive sentence-final prosody.) The first time I heard this, I thought the criminals were going to be governor … To me (a native English speaker), this utterance is totally ungrammatical.

    [Not off topic at all. As governor, that will change is a classic dangling modifier case — one of the shortest and clearest I have ever seen. Nice one, CLS. —GKP]

  7. Ray Girvan said,

    April 21, 2010 @ 11:13 am

    James Davis: If we stop thinking of punctuation as logical connectives of some kind, and instead started thinking of punctuation as having phonetic/prosodic value

    That's a moot point, because the recent trend – at least among among some US style pundits – has been in the other direction. In Bad Comma, Louis Menand's New Yorker diatribe about Lynne Truss, Menand claimed that punctuation just exists as structural markup “to add precision and complexity to meaning” and “increase the information potential of strings of words”. He says: “As [Truss] points out, in earlier times punctuation did a lot more work than it does today, and some of the work involved adjusting the timing in sentences. But this is no longer the norm, and trying to punctuate in that spirit now only makes for ambiguity and annoyance”.

    Personally I think Menand's view on this point is bilge; its role in timing isn't remotely defunct.

  8. Randy Alexander said,

    April 21, 2010 @ 11:13 am

    @Geoff: Now that's what I'm talkin' about!

    More and more of this kind of stuff is what the public needs. Not everyone's writing can go through a copy editor, and anyway, a copy editor's job ought to be editing for consistency, not fixing the horrible grammar blunders of people who got too confused by spending too much time trying to follow "rules" in S&W, and not enough time learning syntax (or even just reading good writing).

    I do sympathize with these confused people though. English grammar education is a mess the world over, and it's only through efforts like the ones you are making (especially through your "campaign" and your posts) that the mess will have any chance of being cleaned up in our lifetime.

    There needs to be a lot more communication between grammarians and educators.

    Posts like this clearly demonstrate what kinds of things need to be targeted in grammar education and what kinds of things can be left for copy editors. I hope a lot of grammar teachers are reading this post, and I hope to see more and more great posts like this.

    [Thank you, Randy, for the vote of confidence. And if I may just utterly pervert the function of the Language Log comment columns for a moment: I took up the recommendation you sent to me by email and had dinner at Casa Romero my last night in Boston, and it was indeed very good. Thanks for that as well! —GKP]

  9. Mel Nicholson said,

    April 21, 2010 @ 11:21 am

    I wrote that I had replaced the clause in question with an adjective, but changed it back when I realized I had changed the meaning. Here is a reduced contrast:

    Please console unsuccessful employees.

    vesus

    Please console employees who were unsuccessful.

    I find that the scope of unsuccessful is restricted to the application in question when using the latter form, but takes on a much wider scope and seems to describe a characteristic of the employee in the former.

  10. Adrian Bailey (UK) said,

    April 21, 2010 @ 11:25 am

    I tend to side with James Davis because in English (unlike German, for example) a comma is not used because of logic or grammar but to signify a physical stop. In other words, the "rules" of punctuation, such as they are, spring from the way we speak, rather than being imposed artifically on the way we write.

    When spoken, "Politicians, who I admire, never get elected" has two such stops, thereby conveying a different meaning from "Politicians who I admire never get elected."

    Perhaps GKP can provide some examples of how this is a myth.

    [(myl) There are many examples where there is a comma and no pause, and even more examples where there is a pause but no comma. One standard case of comma-but-no-pause is a sequence like "Portugal, Spain(,) and France" (with or without the "Oxford comma"). And to find an example of pause-but-no-comma, even in fluent and formal speech, listen to any expert newsreader for five or ten seconds, and you'll almost certainly find one.

    To demonstrate this, I just picked a random NPR story, where the first example occurs 1.6 seconds from the start of the story (pause indicated by //):

    File this in the category of // work-related nightmares.

    Another one comes up in the very next sentence:

    An Apple engineer goes to a bar to celebrate his 27th birthday and leaves behind // a working prototype of the super-secret, next-generation Apple iPhone.

    These are both places where a comma is, by orthographic convention, impossible — but they're both places where it makes good communicative sense to put in a pause, in order to set off and emphasize what comes next.

    Another random NPR story includes a short phrase where there's a comma with no pause, and a pause with no comma:

    Scotland, France and probably more will follow // tomorrow.

    For more evidence, please simply compare a correctly-punctuated script (or transcript) with the audio recording of any similar segment.]

  11. Mel Nicholson said,

    April 21, 2010 @ 11:38 am

    @Ray: Exactly. If the role of punctuation to establish timing were defunct, a certain novel way to use the period for emphasis would. not. work.

  12. Aaron Toivo said,

    April 21, 2010 @ 11:38 am

    Your points about whom and the commas are well stated and obviously correct, but I'm having serious trouble with condemnation of this dangling modifier. While not a syntactician I think I'm a reasonably competent with English, and I find this dangling-modifier example idiomatic and wholly lacking in any kind of problem. I'm curious what the "most syntactically plausible one" is that's semantically implausible, because I can't see how any other subject for it than "colleagues" could even be considered a potential interpretation. It certainly can't be "we" or "you", though I'd have trouble explaining why this is… but I suspect a pragmatics issue. The "colleagues" are front and center in a way that the pronominal participants just aren't.

  13. Greg B said,

    April 21, 2010 @ 11:46 am

    If I read Adrian Bailey and James Davis correctly, people who are short on breath when they write will use more commas, for commas 'spring from the way we speak'. Or maybe they mean if we want the reader to feel winded we will provide extra commas (to indicate a 'breath or short pause')? They seem to say that commas and semicolons regulate the rhythms or cadence of reading. Does any reader really ask 'if this string of words were spoken out loud with these pauses in place, it would sound as follows, and thus I understand it, for that is the function of punctuation'? I can't say I read things as though they had to be uttered to be understood; at least I don't think I do. If I misunderstand, please let me know.

    I find GKP's explanation more convincing, but perhaps we should all pause and reflect on it ; ; ; ; ; .

  14. James Davis said,

    April 21, 2010 @ 11:51 am

    “Semantics certainly does not make clear what anybody means. Only people who have utterly confused semantics with pragmatics could think that.”

    Eh, that's not the first time I've written too quickly and used the wrong word, but you're right on that one. It probably won't be the last time either, what with internet posts being generally unedited and all.

    “Oh, dear. The foregoing two paragraphs would be best read as a remarkably full and clear exposition of a set of claims that are widely believed but entirely untrue. Punctuation is not prosodic markup. The relation between these obligatory grammatical elements of the written language and the highly variable intonational phrasing and pausing of the spoken language is loose, complex, and indirect.”

    No, punctuation isn't prosodic mark-up, I agree with you on that. It was at one time, and it could be again, if people wanted it that way. However, that doesn't change that it has become “obligatory grammatical elements” (as you put it).

    Really though, they're not truly grammatical elements, they have no clearly defined phonetic, morphologic, or syntactic value, serving instead as some kind of logical connective that must be used in a particular way. The rules for how they are used are generally prescriptive and not particularly related to syntactic structures. After all, “Politicians who I admire” and “Politicians, who I admire” are both noun phrases, all that comma does is mark how we should interpret the relative ‘who I admire’. I prefer to think of it as marking the pause that comes between a noun head and a non-restrictive relative clause in spoken English (this gives the comma, in my opinion, a clear value); however most people (including you, it seems) think of it as simply visually representing a logical connection showing the type of relative clause.

    And this brings me back to my point:

    *If we stop thinking of punctuation as logical connectives of some kind*, we could do it in an easier way. I didn't claim that punctuation does represent prosody, I claimed that it could, and that if it did, it would be clearer. This is an arguable position, and there are reasons it could be wrong. Nevertheless, it is not the same position you just attributed to me and argued against. There is a nuanced difference between wanting a change in language and thinking that the change is already there.

  15. Alan Gunn said,

    April 21, 2010 @ 11:55 am

    I see that use of "whom" a lot in police reports. It's also standard police-department English to use "myself" in place of both "I" and "me," so you get sentences like "Myself and Officer Smith approached the suspect, whom identified himself as …."

    No native speaker of English would write that way without having been taught to. I know that some police academies teach the "myself" substitution; I know nothing official about the "whom." Perhaps it is an attempt by people who took Thurber seriously to express respect for the subject, like the standard police practice of calling vicious criminals "gentlemen."

  16. James Davis said,

    April 21, 2010 @ 12:06 pm

    @Greg B

    “I can't say I read things as though they had to be uttered to be understood; at least I don't think I do. If I misunderstand, please let me know.”

    Please note my longer response, where I clarified my position.

    Now, I can't speak for everyone, but whenever I read something, I repeat what I'm reading in my head as if it were being spoken. So, I don't know, for sure, but, I imagine, that the, frequent, use, of commas, might, make, the reader, read, the sentence, like it, is spoken, by, William, Shatner, with all, those, silly, pauses.

  17. chris said,

    April 21, 2010 @ 12:07 pm

    There's no question that the original use of "whom" was wrong, but am I the only one not completely comfortable with "Politicians who I admire…", in which "who" is the direct object of the verb "admire"?

    I would probably finesse the issue with a contact clause: "Politicians I admire never get elected." But that solution isn't available if you want a supplementary rather than an integrated clause.

  18. Sam said,

    April 21, 2010 @ 12:08 pm

    This question has been torturing me since I started reading this blog, and now seems like an appropriate time to ask.

    What do you propose as the ideal basis for that "elementary grammar and writing class," and does such a class exist? This is personally relevant because my significant other is about to begin a career as an English teacher, and neither of us can differentiate between prescriptivist dross and "real" guidelines for quality writing. What should be teaching his students? (Enjoy this opportunity to indoctrinate the next generation.)

  19. Troy S. said,

    April 21, 2010 @ 12:16 pm

    Wouldn't "Politicians whom I admire" be correct? The reasoning being that "Politicians" is a direct object of "admire." Even if it's not correct, I can see how easily the mistake could be made here. In the main clause, "meet with the staff" has "the staff" as the object of the verb phrase "meet with" and he has mistakenly made the pronoun in the subordinate clause agree in case with its antecedent in the main clause.

  20. Henning Makholm said,

    April 21, 2010 @ 12:17 pm

    I find it interesting that both of the "staff(,) who have been unsuccessful" and "politicians(,) who I admire" examples are ones that pragmatically need no disambiguation between supplementary and integrated relative clauses, simply because one of the readings is too insane to have been meant.

    In order for the argument to differ in form from a random prescriptive peeve, I think you need to extend it with examples where disambiguation is desperately needed because both readings are plausible. I think plenty such examples could be found with a little searching.

    This is not to say that getting the punctuation wrong in these cases is excusable if only the meaning gets through. The disambiguation device will only work in the cases that need it if readers are used to how it works, that is, if they are used to seeing pragmatically unambiguous examples punctuated correctly. Otherwise it would be hard to develop an intuitive awareness of the meaning of whether there are commas. (I suppose that is a kind of Kantian principle of orthographic correctness).

  21. Sili said,

    April 21, 2010 @ 12:18 pm

    By George! I think I got it!

    Thank you.

    Of course, there *could* come a day when most careful writers adopt the habit of using commas every which way they choose, forcing readers to use context to distinguish between integrated and supplementary relative clauses.

    That happens to be the case in Danish. At least under the old system – we have two, just be the difficult. (At least it isn't Nynorsk.)

  22. Sravana said,

    April 21, 2010 @ 12:35 pm

    @Henning At least in the case of the staff example, I think both readings are plausible — it can easily be construed that all of that staff were unsuccessful, unless further information was given in the letter.

  23. a soulless automaton said,

    April 21, 2010 @ 12:41 pm

    Hmm. I can actually imagine two ways of speaking the phrase "Politicians, who I admire, never get elected", with pauses at each comma, such that the intended meaning comes out as either the correct or absurd interpretation.

    The absurd interpretation gives the clause in question the same pacing as the rest of the sentence, while the correct interpretation gives the clause a faster pacing, change in pitch, and emphasis on the word "admire".

    The latter I would never render in writing as a comma-delimited phrase, of course–it sounds more like a parenthetical aside, implying a shortened version of something like "[at least those] who I admire". But technically, it does fit the bill of "short pause between spoken words".

    I tend to doubt that there's any consistently meaningful relationship between punctuation placement and the pacing of spoken language; the mapping between them is neither injective nor surjective.

  24. ClockwerkMao said,

    April 21, 2010 @ 12:51 pm

    I think the prosody vs. logical connectives misses what's going on by mistaking prosody as supplemental to the meaning rather than critical to it. For example, while Spoken English uses a pitch contour to indicate a question, (and I profess unfamiliarity apart from little book-learning) Spoken Japanese requires the question word 'ka'. It seems to me that Written English, not using pitch contours, does the same thing as Spoken Japanese. It adds the question word '?', which has the phonetic value of '?', which tricks us because it's only got a graphic phonetic value rather than a sonic phonetic value, and we were all taught that writing is simple transcription. This. Thing. Works. Exactly because writing is not simple transcription, at least not anymore.

    Now of course the closely related nature of Written English and Spoken English, plus the notion that they're the same language, means that we'll associate similar emphatic technique. We'll say, The period means a spoken stop. So when we're fluent in both, we'll easily see that period-stopped sentences are emphasizing each word rather than the sentence as a whole, just as putting audible space between each word does.

    It's too bad I don't have anything but anecdote and intuition to back this up, because I sure think it's worth studying.

  25. Nathan Myers said,

    April 21, 2010 @ 2:13 pm

    Usually when I encounter a comma in such a position, the writer is German. Some such writers have reported that in German, intended meaning doesn't determine comma placement, so they're not used to placing or interpreting them that way.

  26. Henning Makholm said,

    April 21, 2010 @ 2:29 pm

    @Sili: "Every which way" does not apply to Danish, but "forcing readers to use context to distinguish between integrated and supplementary relative clauses" does.

    The traditional Danish comma rules, borrowed from German, are purely syntactic and prescribe commas around all clauses. The official language committee has for several decades to championed various alternative systems which would, among other things, distinguish between integrated and supplementary relative clauses through the use of commas. In the most recent alternative ruleset, all subordinate clauses end with a comma whereas only supplementary relative clauses take a comma at the beginning.

    Unfortunately, teachers have freedom to choose between one system and the other, and the older commas-everywhere system is certainly easier to teach. So not many readers are aware of the semantic distinctions, and what you get when you try to exploit the greater expressivity of the new system is not better communication but "there's a comma missing" complaints.

    The new system is a better one than the old one, but there appears to be no way to get from here to there.

  27. Jarek Weckwerth said,

    April 21, 2010 @ 3:10 pm

    @Nathan, Henning: Polish also has a punctuation rule that requires a comma before all occurrences of który 'that/which'. As a result, you tend to find quite a few "unnecessary" commas in Pole-produced English :)

    @Henning & others: A standard example in Roach's English phonetics and phonology is this:

    The conservatives(,) who like the proposal(,) are pleased.

    Does this need disambiguation? I think it does. Maybe not desperately, but both versions are plausible. Coincidentally, the example comes from a section on "tone unit boundaries"…

  28. Bloix said,

    April 21, 2010 @ 3:13 pm

    The use of commas in ways that would be incorrect today is common in the US Constitution. E.g.,

    The judicial Power of the United States, shall be vested in one supreme Court…
    Treason against the United States, shall consist only in levying War against them…
    The Ratification of the Conventions of nine States, shall be sufficient for the Establishment of this Constitution …
    The enumeration in the Constitution, of certain rights, shall not be construed to deny or disparage others retained by the people.

    The obsolete punctuation usually just seems quaint, but in one celebrated case (involving an integrated relative) it causes genuine confusion:

    A well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed.

  29. Spell Me Jeff said,

    April 21, 2010 @ 3:51 pm

    I'm noticed that the thing we call a comma splice was quite common in learned American writing in the early 20th c., and seems to remain common in learned British writing of today. I also notice that when it was thus used, author and/or editor had sense enough to make sure it did not create any confusion. This makes the successful splice more a matter of finesse than rule. Very difficult to teach, which is perhaps why we do not.

  30. Rubrick said,

    April 21, 2010 @ 3:59 pm

    Your analysis is right on the mark, but your use of adjectives like "atrocious", "horrible", and especially "astonishing" suggest you don't read a lot of corporate communications. The examples given strike me as fairly typical, perhaps even a little better-written than average — it was at least easy enough to figure out what the writer must have meant.

  31. Army1987 said,

    April 21, 2010 @ 4:59 pm

    The commas-around-supplementary-but-not-around-integrated-clauses thing also applies to Italian, even though Italian has typically more commas than English. For example, you would use a comma even in places such as after but in "The committee dealing with the question of commas agreed on a final text, but despite the importance of the matter, the relationship with semicolons was not considered" (example borrowed from the EU style guide), and "comma splices" are completely normal.

  32. Simon Cauchi said,

    April 21, 2010 @ 5:14 pm

    GKP says "whom" is "rarely used these days except after prepositions". Really?

    I don't know how to use the various corpora that could be consulted to determine the point, but I for one use bare "whom" quite readily in relative (but not interrogative) clauses. E.g.

    That woman with long red hair whom we saw at the supermarket this morning, I saw her again this afternoon at the beach.

    Who did you ask to see, Mr Jones or Miss Smith?

  33. Stephen Jones said,

    April 21, 2010 @ 5:35 pm

    GregB is being deliberately obtuse. The point is that the spoken language uses pauses and intonation to mark off phrases and distinguish relationships between them; this function is taken over by various punctuation marks including the comma.

    File this in the category of // work-related nightmares.

    Another one comes up in the very next sentence:

    An Apple engineer goes to a bar to celebrate his 27th birthday and leaves behind // a working prototype of the super-secret, next-generation Apple iPhone.

    These are both places where a comma is, by orthographic convention, impossible — but they're both places where it makes good communicative sense to put in a pause, in order to set off and emphasize what comes next.

    If you want to indicate an abnormal pause in these cases then you can use … or — or even :

    I suspect we're talking about a particular oratorical technique peculiar to the newsroom.

    [(myl) There are certainly characteristic techniques of newsroom oratory, but when it comes to the relationship between the norms of English punctuation and the location of pauses in speech, I'm highly confident that you're completely wrong. This opinion is based on examining of phrasing patterns in many styles of speech — informal conversation, story-telling, political speeches, broadcast talk of many kinds, audio books, and so on. Can you find a single extended example of English speaking in which all commas correspond to pauses, and all pauses correspond to commas or other marks of punctuation? There are thousands of audiobooks at http://librivox.org/newcatalog/, and I invite you to find even one that supports your argument.

    This is not to say that there's never any connection between phrasing and punctuation. To quote the chapter on punctuation from CGEL (written by Geoff Nunberg, Ted Briscoe, and Rodney Huddleston), "We have emphasised that punctuation cannot be regarded as a means of representing the prosodic properties of utterances, but there is no doubt that there is some significant degree of correlation between the use of delimiting commas and the likelihood that the constituent concerned would be set apart prosodically in speech."]

  34. Jean-Sébastien Girard said,

    April 21, 2010 @ 5:39 pm

    I think one of the citation has an extraneous word. In the first version, the sentence ends "after their receiving the disappointing news.", while the second quote (following "but this is what it said: ") lacks this "their".

    While including that word might not be the best choice, it makes it almost impossible to me to argue for a dangling modifier ("their" MUST refer back to "staff"), so I can only suppose the word has no business there.

    [I agree that after their receiving sounds very strange; but that is apparently what was in the hard copy message that my friend received (or at least, it is what was in the email telling me about it" — I might recheck). —GKP]

  35. Robert T McQuaid said,

    April 21, 2010 @ 5:46 pm

    My attention fell on the "m". In non-professional writing on the internet the choice between who and whom is made by rolling dice. The distinction is lost except among language aficionados

  36. Bloix said,

    April 21, 2010 @ 5:49 pm

    But Simon Cauchi, you know and work with Latin and other Romance languages that are more highly inflected than English. Your sense of grammar must be more highly and consciously developed than that of most people.

    Most monolingual English speakers have a very superficial understanding of subject-object declension. Anything other than the simplest pronoun usages leads to error. And the who-whom distinction has disappeared from the speech of most people almost as thoroughly as you-thou and will-shall – that is, when people do use it, they do so in order to appear more formal, and they are frequently in error.

    PS- I assume "who" in your second example was meant to be "whom" – I won't go so far as to call the error Freudian, but I'll note that the way you wrote it is the way most people would say it.

  37. Dougal Stanton said,

    April 21, 2010 @ 6:06 pm

    @Alan Gunn,

    No native speaker of English would write that way without having been taught to. I know that some police academies teach the "myself" substitution; I know nothing official about the "whom."

    Either there are a lot of ex-police officers in my area or this is more common than you suggest. It seems to appear wherever someone who is not used to public speaking has to perform in some formal or official capacity — police in the witness box, staff announcing things on trains, and so on. It sounds like an attempt to emulate some ultra-formal, learned register and ends up sounding downright odd. I have given up even wondering why the conductor announces that the train will be "arriving into Edinburgh Waverley".

  38. Bloix said,

    April 21, 2010 @ 6:17 pm

    I don't know where "myself" comes from, but I do suspect that it solves a problem that many people have. I think that people simply cannot learn when to say "John and I" and when to say "John and me." It's just impossible for them to distinguish between the subject pronoun and the object pronoun in the compound setting. Their natural inclination is to say "John and me." The ones who try to get it right hyper-correct and say "John and I" all the time. If you learn to say "myself" whenever you have the "John and I/me" problem you don't have to make the subject-object distinction and so you're never wrong.

  39. Simon Cauchi said,

    April 21, 2010 @ 6:44 pm

    @Bloix: No, the "Who" in my second example ("Who did you ask to see?") wasn't a slip. I was reporting my own speech habits: I find I quite often use the relative "whom" but hardly ever the interrogative "whom".
    The point of my comment was merely to express some modest disagreement with GKP's "whom (rarely used these days except after prepositions . . .)".

  40. Scriptor Ignotior said,

    April 21, 2010 @ 9:00 pm

    Mel Nicholson:

    … to provide them support …

    Eeeuw! Surely to provide them with support. Or is your version now to be judged "error-free use of the language", to adopt GKP's meta-prescriptive formulation?

  41. Stephen Jones said,

    April 21, 2010 @ 10:26 pm

    Can you find a single extended example of English speaking in which all commas correspond to pauses, and all pauses correspond to commas or other marks of punctuation?

    You're setting up a straw man. For a start there are places where there are pauses and commas can't be put, and as far as the other way round goes it depends on the comma-placing habits of the transcriber.

    What I am saying is that with normal intonation if you don't put in a comma without a pause you will do fine. The standard American rules of punctuation, which treat placement of commas as an exercise in check box filling, tell us to put a comma after initial 'However, Nevertheless, Afterwards' and a host more words. Often it simply makes the page seems busy and suggests a pause where none would be, so I will argue it's fine to miss them out, and often do. (Note the purposeful placement of the comma there, in defiance of another bizarre rule: the one that states you put a comma between coordinate clauses with different subjects but not the same subjects).

    I doubt you will ever go seriously wrong if you never put in a comma where there wouldn't be a pause or some equivalent differentiation marker in speech.

    eeuw! Surely to provide them with support.

    The Shorter Oxford English Dictionary gives both forms. The COCA has 55 hits with the double object and 192 with 'with'.

  42. Garrett Wollman said,

    April 21, 2010 @ 11:13 pm

    @Robert McQuaid: Fred Vultee (headsuptheblog.com) comments regularly on how his local paper seems to have a knack for getting the "who/whom" distinction wrong at a worse-than-uniform-random rate.

    @several commenters: You can add Finnish to the list of languages that require a comma before all relative pronouns. (Assuming they haven't change the rules in the last twenty-one years.)

    @Bloix and others: My boss is much given to using "myself" as the generic first-person pronoun. No police background AFAIK; he's an electrical engineer by training.

  43. Scriptor Ignotior said,

    April 22, 2010 @ 1:31 am

    Stephen Jones, you write:

    The Shorter Oxford English Dictionary gives both forms. The COCA has 55 hits with the double object and 192 with 'with'.

    Hmm. My SOED (from 1997) includes this for "provide":

    5 v.t. Equip or fit out with what is necessary for a certain purpose; furnish or supply with something. (Foll. by with, †of, double obj.). LME.

    But it gives no citation. OED (same vintage) has this:

    5. trans. To supply or furnish for use; to yield, afford. Const. †to (obs.), for, or with dative.

    It gives citations "with dative" from 1581 and 1898. It gives no such citations under this later section:

    8. To furnish or supply (a person, etc.) with something. Often in indirect passive.

    Just three constructions are dealt with under this section. The first uses with:

    1860 Tyndall Glac. i. xxii. 151 The waiter then provided me with a ham sandwich.

    The other two constructions use of and with in, and both are marked obsolete.

    So according to OED, presumably, such forms as this are not to be found:

    *They were not provided support.

    Nevertheless, I find 12 instances of provide them in the full text of OED; the 2 of these that have a double object are from 1747 and 1774; the other 10 are followed by with, and 7 of these are from the 20th century.

    Especially by failing to give "indirect passive" instances lacking with, OED's section 8 weighs against any robust acceptability for to provide them support in British English. What's more, the very thorough and permissive (and American) Merriam-Webster's Dictionary of English Usage does not give any pertinent example lacking with, saying only this (at "Provide"):

    But when provide is used transitively to mean "to make something available to," the thing being provided is usually named in a propositional phrase introduced by with.

       The route is well provided with signs.

    Distinguished (by that initial but) not from similar cases lacking with, but rather from cases like these:

    Provide schools and teaching to all children and illiteracy goes down dramatically.

    Now, Stephen, I cannot see what search you did exactly at COCA. Care to provide the details? When I did a search on "to provide them" restricted to "Academic" (which perhaps respects traditional norms more than other registers do), I got 103 hits. Of these, 7 were not relevant (they terminated in them); only 4 of the remaining 96 had a double object, and were not followed by with. That's American English, and it's academic. From my own experience, the double object construction is less common outside of American usage.

  44. Stephen Jones said,

    April 22, 2010 @ 1:54 am

    provide + personal pronoun + noun
    provide + personal pronoun + with + noun

    With regard to your comments on the OED examples if you allow a double object then you are allowing the passive construction. There'll be no need to list them separately, just as when it states a noun is transitive we presume the passive is allowed even if there are no examples.

    The COCA produces 132 examples of be + provided + noun to 138 for the use with 'with' so the construction is more common in the passive than in the active.

  45. Scriptor Ignotior said,

    April 22, 2010 @ 5:06 am

    Stephen Jones:

    I see. Well, restricting your search to Academic we get these results ("with present" first; "with absent" second):

    68 (85.00%); 12 (15.00%)

    Restricting to Fiction:

    24 (92.31%); 2 (7.69%)

    Restricting to Spoken, by way of contrast:

    27 (52.94%); 24 (47.06%)

    And restricting to various ranges of years:

    1990-1994
    61 (82.43%); 13 (17.57%)

    1995-1999
    57 (91.94%); 5 (8.06%)

    2000-2004
    45 (72.58%); 17 (27.42%)

    2005-2009
    29 (58.00%); 21 (42.00%)

    [I suppose that the composition of the corpus changes since the first of these periods; this would account for the anomalous dip between it and the second period. Or perhaps the numbers are just skittishly low.]

    Two tentative conclusions:

    1. In American English, the construction lacking with is more common in speech than in writing, and least common in the written registers that pay most respect to literary, traditional "propriety".

    2. In American English, the construction lacking with has become markedly more frequent over the last two decades; and it looks set to overtake the construction using with.

    As I have said, in my experience that with-less construction is less common in non-American usage. It would be interesting to replicate my survey with a British corpus.

    Now Stephen, an answer to this:

    With regard to your comments on the OED examples if you allow a double object then you are allowing the passive construction. There'll be no need to list them separately, just as when it states a noun is transitive we presume the passive is allowed even if there are no examples.

    But my comment was explicitly about OED's section 8:

    Especially by failing to give "indirect passive" instances lacking with, OED's section 8 weighs against any robust acceptability for to provide them support in British English.

    Your objection is immaterial, since section 8 does not allow at all for constructions lacking with (and of and in). OED cites at section 8 "His valet [was] provided with phosphoric matches …". Section 8 would clearly be the place for it to give coverage to parallel forms like "*His valet [was] provided phosphoric matches …", in a separate subsection. But it does not give coverage to such with-less constructions, because it does not detect any.

    Last, I ask this: How are we to interpret the expression "error-free use of the language" in the light of such divergence by year, register, and (I would add) national variety?

  46. Mark P said,

    April 22, 2010 @ 7:03 am

    but I for one use bare "whom" quite readily

    And no doubt you will continue to do so while the word drops out of general usage. I know how to use it but generally avoid it so as to not sound excessively posh.

    "Whom" won't be missed. And for a large part of the population is already not missed.

  47. Army1987 said,

    April 22, 2010 @ 11:59 am

    The use of nominative "he" as the complement of "was" in the otherwise not-too-formal post threw me off for at least 500 milliseconds.

    [Yes; I wrote "it was he, not William Strunk, who wrote the relevant section". One of the rare cases in which he seems appropriate to me after the copula is when it is the foregrounded element in a cleft. I guess I would accept either it was him who wrote the relevant section or it was he who wrote the relevant section. Both are Standard English, but the former seems substantially more informal than the latter, and on this occasion I seem to have chosen to be more formal. I don't know why. Perhaps it was in unconscious mockery of those old weasels Strunk and White (and they deserve mockery; White suggests you should write The culprit, it turned out, was he, which strikes me as lunacy). But I honestly can't remember whether I felt like mocking them at that point; perhaps I did, but I simply don't know. —GKP]

  48. Mark F. said,

    April 23, 2010 @ 11:06 pm

    I think people hear in their heads comma-associated pauses as they read, and when they write they think they are recording the pauses they would say, when fact they are recording the pauses they want their readers to hear as they read.

    Except that it's not just pauses. The prosodic difference between "politicians who I admire" and "politicians, who I admire" seems to involve changes in pitch contour, syllabic rhythm, and emphasis.

    I recently claimed that I punctuated by sound, but I'm rapidly coming around to the conclusion that what I have really done is internalized some version of the syntactic rules for comma use just by seeing it in print. I certainly don't consciously apply a list of rules.

  49. Rodger C said,

    April 28, 2010 @ 11:27 am

    Re "which" v. "that" in restrictive clauses: In junior high school ca. 1960 I was taught that either was acceptable. That this wasn't my teacher's idiosyncrasy I've confirmed with my collection of old style manuals. In subsequent decades there was a growing movement, now triumphant apparently, to segregate the functions of the two pronouns. I don't understand why we're inventing MORE arbitrary rules; I suspect this one is driven by the manufacturers of word-processing software.

RSS feed for comments on this post