What's wrong with this passage?

« previous post | next post »

Elizabeth Daingerfield Zwicky has forwarded to me a site with (yet another) little "grammar test" (a "Google Grammar Test", from Tyler Cowen) — this one has only two items — that makes me scratch my head.

(I guess I should remind you that in some quarters, "grammar" covers absolutely anything in language that can be regulated: discourse organization, syntax, word choice, morphological forms, stylistic choices, politeness formulas, punctuation, spelling, whatever. So, ahead of time, I had no idea which features of this very short passage might be seen as reprehensible. Was it, for instance, spelling homepage as a solid word, rather than as two separated words?)

Here's the passage:

Here’s what’s on Google’s home page on May 16, 2009:
  Over 28,000 children drew doodles for our homepage.
  Vote for the one that will appear here!
Test yourself: Can you find the two grammar errors?

and here are the answers (from Penelope Trunk):

The AP Stylebook says "over" is a way to move—a preposition. And “more than” must precede a number. Also, if you are voting for one, specific doodle, then the AP Stylebook tells you to use “which” rather than “that.”

Here I'm going to talk about the which/that issue. I'll save a return to the adverbial-over question for another time.

That vs. which in restrictive relative clauses has a history at least a century deep (the actual history is not important in the current context); we've posted about it endlessly on Language Log, arguing that both variants are available in modern standard English as restrictive relativizers. But many have denied this and propose to allow only one. The usual scheme for choosing relativizers is what I've called Fowler's Rule: that in restrictive relatives, which in non-restrictive relatives (it's more complicated than that, but this is the slogan version).

In fact, that's exactly what the AP Stylebook recommends — that in this case, which is punctuated as a restrictive — so the reference to it above is puzzling. Certainly, the AP Stylebook doesn't mention specificity or uniqueness. Here's the main part of what it says (in its current on-line version):

that, which (pronouns) Use that and which in referring to inanimate objects and to animals without a name. Use that for essential clauses, important to the meaning of a sentence, and without commas: I remember the day that we met. Use which for nonessential clauses, where the pronoun is less necessary, and use commas: The team, which finished last a year ago, is in first place.

(Note that this formulation folds in a recommendation against that used to refer to human beings.)

Uniquely referring NPs like the only solution to the problem can have either sort of relative:

The only solution to the problem that will satisfy them is to abandon the project. [restrictive]

The only solution to the problem, which will satisfy them, is to abandon the project. [non-restrictive]

But there's no denying that

(1) Vote for the one (doodle) that will win.

is weird, and which doesn't improve things:

(2) Vote for the one (doodle) which will win.

Making the relative clause non-restrictive makes things even worse (in this context, though you can imagine a context in which it might make sense — a context in which there is only one doodle in question):

(3) Vote for the one (doodle), which will win.

This is not an issue of grammar, but of content. The problem with (1) and (2) is that they appear to be saying that the reader should know which doodle will win, and telling the reader to vote for that one. That's just silly.

Whoever wrote the passage probably got into trouble by packaging two things into a single sentence using a relative clause:

Thing 1: Vote for one (doodle). That is, vote for your favorite doodle.

Thing 2: The winning doodle — the one with the most votes — will appear here.

You can put them together, but not with a relative clause; what you want is a coordination:

Vote for your favorite doodle, and the winner [quite probably not your favorite, out of 28,000] will appear here.

Both Thing 1 and Thing 2 involve uniqueness, but in different ways.

Note that

(4) Vote for one (doodle) that/which will appear here.

is neither grammatically nor pragmatically peculiar, but it's surely not what the Google writer had in mind, since Thing 2 implicates that only one doodle will appear on the site the Google posting links to. The non-restrictive counterpart,

(5) Vote for one (doodle), which will appear here.

is also neither grammatically nor pragmatically peculiar, but, again, it doesn't capture the writer's intent.

In the end, I suppose that most readers of the Google posting figured out its intent. I'm just saying that this was an inept way to try to get it across. That is, it's poor, but not ungrammatical, writing, and that vs. which is beside the point.


  1. fev said,

    May 22, 2009 @ 8:26 am

    The author manages to get AP style exactly backwards on some other points too — considering her claim elsewhere that "real grammarians … have memorized the AP Stylebook," it's one of the truly charming features of her style.

  2. A Reader said,

    May 22, 2009 @ 8:31 am

    It might be worth noting that I had no trouble grasping Google's intent- and indeed had a lot of trouble getting why you said (1) was weird until you had explained yourself. I now see how you took it, but I really found Google's way pretty natural and unconfusing. I can't tell you why, unfortunately.

  3. kuri said,

    May 22, 2009 @ 9:04 am

    The writer could have said, "Vote for the one that you want to appear here!" I think that would have captured the writer's intent. Although "and the winner will appear here" is only implied, almost everyone is familiar enough with voting to understand that.

  4. bianca steele said,

    May 22, 2009 @ 9:08 am

    I was amused by her implying that people who have memorized the AP Stylebook have better social skills than people who have gone to graduate school for English. At first it seems counterintuitive, but it made me think.

    Her style is very entertaining. It takes nerve to concede that it's not a mortal sin not to sent 100% typo-free e-mail, 100% of the time.

  5. bianca steele said,

    May 22, 2009 @ 9:10 am

    Sorry, that should have been "send," not "sent."

    I do think it would be interesting to look at why the rules are interpreted differently at times. Usually it feels like dueling stylebooks.

  6. Michael Yuri said,

    May 22, 2009 @ 9:24 am

    I think it has something to do with two different senses of the phrase "vote for X"

    In some contexts it means: of the available options, choose X. In other contexts it means: choose among the options for position X.

    If you say "In the last election, I voted for the President," you're saying you voted for Obama. If you say "In the last election, I voted for President," you're saying that you cast a vote for the office, but not indicating for which candidate.

    In that context, the "the" seems to make all the difference, but it's not clear to me that this is always the case. To me, "Vote for the one that will appear here" most naturally reads like "Vote for President."

  7. D.Sky Onosson said,

    May 22, 2009 @ 9:25 am

    I think kuri makes a good point. Our real-life knowledge of voting makes it pretty clear that the winner will not be decided by our single vote. This means it is pretty doubtful that anyone would take "Vote for the one that will appear here" absolutely literally – so I don't see what there is to be confused about. Of course, people who are determined to find such things, will.

  8. Vincent said,

    May 22, 2009 @ 9:27 am

    I'm with "A Reader" on this one. The meaning is clear to anyone who understands this kind of popular competition.

    The word "for" stands proxy for the well-understood mechanism, and any reader would make the mental substitution as follows:

    Submit yourvote to participate in a competition to determine the one that will appear here!

    It would be very pedantic to complain of the grammar errors suggested by Penelope Trunk.

  9. James said,

    May 22, 2009 @ 9:42 am

    I think Vincent and A Reader may have misunderstood. Nobody is saying that the Google blurb is hard to understand, or that it's unclear what its author meant.

  10. GD said,

    May 22, 2009 @ 9:50 am

    The first 'error' I spotted was ambiguity – is it the child or the doodle that/which will appear?

  11. Stephen Jones said,

    May 22, 2009 @ 12:09 pm

    The interesting thing is that nobody spots the only grammatical error there is.

    In the world, all kinds of different people have a favorite song because there is so many different types of music.

    'There are so many different types' is perfectly grammatical, and so, I would say is 'There's so many different types' since there appears to be a tendency to see 'there's' as a complete phrase instead of splitting it into its constituent parts. But splitting it into its constituent parts is precisely what 'there is' does, and the full form of the verb is ungrammatical. I suspect the child said 'there's' and the idiot that transcribed it for Google, transcribed it as 'there is' because he was told you should only use apostrophes in direct speech.

  12. Bloix said,

    May 22, 2009 @ 12:17 pm

    "Vote for the one that will win!" is perfectly intelligible and I would think that it's more likely to encourage people to vote than the proferred alternative. It's true that it has a literal meaning that is an impossibility except for clairvoyants; but the motivational effect of associating the reader's vote with the winner is probably more important to the contest sponsors than literal logical coherence would be – particularly in light of the fact that the target audience appears to consist of children.

  13. Language Log » What’s wrong with this passage? Writer River said,

    May 22, 2009 @ 12:28 pm

    […] Language Log » What’s wrong with this passage?. Tom Johnson | May 22, 2009 | permalink Tags: grammar   […]

  14. Rolig said,

    May 22, 2009 @ 1:00 pm

    I agree with Michael Yuri that the issue has to do with how we perceive voting. "Vote for the one that will win" is a little awkward, though clear enough in its meaning. But the underlying thought is "Vote for the winner!" i.e. "Vote for the doodle you want to be the winner." The double "for" in the question: "Who did you vote for for the person most likely to succeed?" shows the ambiguity in the phrase "to vote for". In other words, we can vote for one of the available options, but we can also vote for the position whose holder tha vote determines. It is the latter that is understood in the command, "Vote for the doodle that will appear on the homepage" (= "vote for the winner of the contest" = "vote for the one that will win").

  15. Karen said,

    May 22, 2009 @ 1:11 pm

    They're folding two things into one sentence, which then sounds a bit odd.

    (a) Vote! (b) The winner will appear…

    After all, people rarely exhort you to vote for the loser.

  16. mollymooly said,

    May 22, 2009 @ 3:16 pm

    Expanding on Rolig's point: one does see readers'/viewers' polls where one voter will win a prize, provided they voted for the poll-winner; which obviously skews the poll. Not that that's what Google meant, but in some contexts the wording could make such a difference.

    Expanding on kuri et al's point: there are cases where every reader will process the message and extract the intended meaning, even if the literal meaning is different. [I'm excluding figurative language, which always behaves thus.] Some people may not even notice the literal meaning and blithely acquire the intended meaning, but some people may do a double-take before this happens. Whether the writer ought to rewrite depends on how many people do how big a double-take.

  17. Carl said,

    May 22, 2009 @ 4:21 pm

    "Vote for one to appear here!"

  18. bianca steele said,

    May 22, 2009 @ 4:51 pm

    Oops. I misread. I took her point about social skills to be–since she says pretty explicitly that only people with social skills get hired–she had to be more like the AP stylebook wienies in order to remain employed. She's not saying that. I stand by "it makes you think" though.

  19. Stephen Jones said,

    May 22, 2009 @ 6:57 pm

    After all, people rarely exhort you to vote for the loser.

    You've clearly never heard of the Lib-Dems.

  20. A Reader said,

    May 22, 2009 @ 9:09 pm

    @James, Arnold Zwicky called the construction 'weird' and that it was 'an inept way to try to get it across' and 'poor, but not ungrammatical, writing'. This is what I was responding to.

  21. A Reader said,

    May 22, 2009 @ 9:14 pm

    *and said that it was

  22. Ran Ari-Gur said,

    May 22, 2009 @ 9:33 pm

    I wonder if this might relate to the "nerd-view" that Dr. Pullum has discussed several times on Language Log: perhaps they are telling the readers, as a group, to vote collectively for the one that will appear (just as Americans voted for Obama for President, even though many Americans voted for other candidates, and the majority did not vote at all).

  23. Theo Vosse said,

    May 23, 2009 @ 7:04 am

    Apart from "Vote for the winner" (isn't it odd that you're required to vote for something that is going to win anyway?), you should have googled the answer:
    "the one that": 58,300,000 hits
    "the one which": 2,350,000 hits
    I think we have a clear winner and I know which one I want to gladly vote for.

  24. [links] Link salad hangs out at BayCon | jlake.com said,

    May 23, 2009 @ 10:04 am

    […] What's wrong with this passage? — Language Log with more grammar nuttiness. […]

  25. The other Mark P said,

    May 23, 2009 @ 6:04 pm

    Test yourself: Can you find the two grammar errors?

    I would suggest that if people need to actively search to find grammar errors that they are trivial at best.

    If they cannot be noticed on the first reading, what harm has been done?

  26. Noetica said,

    May 24, 2009 @ 7:58 pm

    I'm not sure about this, in the original post:

    Note that this formulation folds in a recommendation against that used to refer to human beings.

    This concerned an excerpt from the AP Stylebook, including the following sentence:

    Use that and which in referring to inanimate objects and to animals without a name.

    The sentence does not, by itself, unequivocally rule that that and which are to be used only for such references. Indeed, it does not rule that these are the only words – nor even the only pronouns, nor (to be ruthlessly strict) even the only relative pronouns – to be used in referring to inanimate objects and to animals without a name. The entire excerpt does not mention persons, or animals that do have a name.

    Here is what we find at the entry "essential clauses, nonessential clauses" (with my emphasis in bold):

    That is the preferred pronoun to introduce essential clauses that refer to an inanimate object or an animal without a name.

    Preferred marks a strong qualification. (This sentence itself, by the way, employs that to refer to "essential clauses". Are essential clauses inanimate objects? I suppose we can allow that they are!)

    And we find this at the entry "who, whom":

    Who is the pronoun used for references to human beings and animals with a name.

    Arguably this gives AP Stylebook's definitive ruling. But I'm still not sure: Clearly we are not to take it literally, yet nothing in the context limits the domain even to relative pronouns; and we must assume that the book allows he, she, they, whom, and whoever for references to human beings! This entry illustrates something about how to interpret AP Stylebook entries generally: loosely, along with context that goes well beyond nearby text.

    Interpreting that way, I see nothing in the book unequivocally and "knowingly" against that for references to human beings. Have I missed it somewhere? I think those responsible for the AP Stylebook probably did not turn their mind to cases like this:

    DOCTOR to RECEPTIONIST: Who is there that I need to see immediately?

    As a critic of many styleguides, I find such insouciance common. Doesn't Gore Vidal say that the trouble with our intellectuals is that they are not very intelligent? One trouble with our guides to style is that their style is not very clear.

RSS feed for comments on this post