Two thousand eight

The sentence went

(1) Two thousand eight was bad for the wallet, but perhaps good for the soul.

This from a posting to, as printed in the NYT “Op-extra” (Week in Review 1/11/09, p. 12). In it, “Two thousand eight” must be understood as referring to the year 2008. Now this is entirely comprehensible, but it might give a reader a few centiseconds of pause, as would

(2) The year two thousand eight was bad for the wallet …

It’s a style sheet thing: most editors would replace (2) by

(2′) The year 2008 was bad for the wallet …

and would absolutely not allow

(1′) 2008 was bad for the wallet …

instead of (1).

I looked at related matters in a Language Log posting a while back, but I didn’t look at year names then. To recap a bit: the expression iPod begins with a lower-case letter. But style sheets for English generally require that every sentence begin with an upper-case letter, which would bar iPod as the first word of a sentence. We then have a conflict between faithfulness (Faith), preserving the lower-case i, and well-formedness (WF) according to some set of conventions, calling for upper-case I if the word begins a sentence. In this particular case, some authorities (the Chicago Manual of Style, for one) allow Faith, but recommend avoiding the issue by re-wording so as to have some other material at the beginning of the sentence.

There are many different ways in which Faith-WF conflicts can arise with respect to the Sentence-Initial Capital-Letter Rule (SICLR). One subset of cases has to do with numerals. The SICLR requires that the first alphanumeric character of a sentence (quotation marks and some other punctuation marks are disregarded) be an upper-case letter, so that numerals are barred from this position.

There are subcases. Numeral expressions can be used to represent quantities (4,572 people), a number in the abstract (The most random number is 17), a year (2008), and other things. Insisting on well-formedness (according to style guide rules) would then require:

(3) Four thousand, five hundred, seventy two people attended the concert.

(4) Seventeen is the most random number.

(1) Two thousand eight was bad for the wallet.

My own reactions to these three examples differ. (3) strikes me as ponderous and also wasteful of space (note to those who take brevity to be next to godliness), but I’m not really happy with the 4,572 alternative; it looks odd to me (more on this below). On the other hand, (4) strikes me as just wrong: I’d normally take seventeen to denote a quantity, 17 a number, and these are very different things. Finally, there’s (1), which seems to me to be in a middle zone, though I’d prefer the briefer 2008.

Note: these are my personal reactions, and they’re not absolute. I’m not proposing to ban one form of expression in favor of another here. I do bridle, however, at those who would claim that my sentence, above,

(3) strikes me as ponderous …

is simply incorrect, illiterate, ignorant, and so on, and treat it as if it were some sort of moral lapse. According to these critics, this sentence would have to be expanded to something like

Example (3) strikes me as ponderous …

Three final points about these numerical examples (plus a bonus proscription at the very end). One: clever people sometimes suggest to me that there’s nothing wrong with (1′) (or the numeral-initial versions of (3) and (4)), because they do in fact begin with an upper-case character. It’s just that the upper- and lower-case versions of a numeral are identical. I’ve yet to find anyone who was convinced by this argument.

Two: yes, many people are taken aback by things like (1′), especially if they’ve been explicitly taught the SICLR. We’re all inclined to favor what we’re used to and to reject what we experience as novelty. But there is a considerable virtue in generously accomodating to practices other than your own (you don’t have to adopt them, merely to recognize that they are reasonable and comprehensible). And there is absolutely no matter of consequence here, no reason why WF has to trump Faith in this particular case.

Three: people have occasionally defended the SICLR to me on the grounds that it makes clear when a new sentence is beginning. True, it helps to do that, but the work of indicating sentence divisions in English writing is already done well by other means. When a piece of text begins, you know that you’re at the beginning of a sentence. And when one sentence ends, you have a period (or other sentence-final punctuation) to tell you so, so that what follows is the beginning of a new sentence. That is, the SICLR enforces a redundant mark for ‘beginning of sentence’ (in the technical, not invidious, sense of redundant). (And in so doing it opens the way for an occasional ambiguity, between words that have initial caps on their own and words that get them only by virtue of being sentence-initial.)

Not that there’s a problem with redundant (in this sense) marks — systems of agreement and government are significantly devoted to marking relationships that are calculable by other means — but, though useful, they are in principle dispensable in some cases. Still, there’s a difference between the conventions of the spoken (and written) language, which (for the most part) have “grown” implicitly, and the conventions specifically of the written language, where explicit prescription plays a major role, and could easily be replaced by a different prescription, or by free choice.

An extra quibble: any number of manuals tell you that you must not write {January 13th} (curly brackets enclose written material); only {January 13} is acceptable. The usual defense is that {January 13th} is prolix, because it has an unncessary {th}. Omit Needless Letters, or something like that.

What makes this proscription especially bizarre is that {January 13} must be read as “January thirteenth”. I cannot say “I met him on January thirteen”. That is, {January 13th} is faithful (but, to some people’s measures, not well-formed).

So there’s an orthography-to-pronunciation convention. Ok, I guess. But what riles me is all those advice sites that dump on {January 13th} and the like, as if they were signs of idiocy. Why do people care so much?

7 Responses to “Two thousand eight”

  1. The Ridger Says:

    I guess I was never taught that. I happily start sentences with punctuation or numerals. I have on occasion wondered if a name like “van der Horst” should become “Van der Horst” at the beginning of a sentence. I can see both arguments – ordinary lower case letters become capitalized, but then again a name is always capitalized so why not always not?

  2. mollymooly Says:

    Never start a sentence with a numeral! Really? I suspect this is one of the many many rules that U.S. copyeditors have collected which British and Irish publications have never heard of. At any rate, (1′) is all over The Irish Times.

  3. arnoldzwicky Says:

    To mollymooly: I suspect that practices are different in different publications, and different for different subcases. I do know that the Cambridge Grammar of the English Language (from the British, not the American, branch of the company) is meticulous about not starting a sentence with an example number: not {[5i] presents a different problem}, but {Example [5i presents a different problem}.

  4. Ian Preston Says:

    I don’t believe it has ever occurred to me that starting a sentence with a number as in (1′) might be problematic and if asked to justify the lack of concern I think I might straight-facedly have made the argument which you say noone you have ever met has found convincing.

    What does trouble me is incorporation of lower-case algebraic symbols into sentences. Since case matters to the meaning, Faith seems imperative but the pull towards WF makes placing them at the start of a sentence feel unacceptable. Hence a sentence beginning “x is the sum of y and z …” feels wrong whereas one beginning “X is the sum of y and z …” feels fine. The discomfort extends even to symbols drawn from other alphabets; “Σ denotes the variance …” causes me no distress but I can’t begin a sentence with “σ denotes the variance …” In cases such as the latter I feel impelled to reword to relocate the symbol while feeling a combination of resentment at the rule that is preventing me from writing what I want to write and foolishness at feeling bound by it.

  5. arnoldzwicky Says:

    And the beat goes on, at least in in the NYT: from the Science Times of 20 January 2009:

    Of the1,154 pharmaceutical plants mentioned in generic drug applications to the Food and Drug Administration in 2007, only 13 percent were in the United States. Forty-three percent were in China, and 39 percent were in India.

  6. September 19(th) « Arnold Zwicky's Blog Says:

    […] blast from the past: from a 2009 posting on writing dates: … any number of manuals tell you that you must not write {January 13th} (curly […]

  7. Ellen K. Says:

    Interesting. I wasn’t taught SICLR; at least not in any way that covers numerals or was otherwise memorable. I was, however, taught that numbers are written out when they begin a sentence (and along with that, the numbers 1 through 10, or maybe 9, anywhere in a sentence). But no explanation for why was given.

Leave a Reply

%d bloggers like this: