Saturday, December 22, 2012

A comma puzzle: The false-interjection error

If you enjoy puzzles about the comma—and who doesn’t?—here’s an elegant but very difficult one, courtesy of Daily Writing Tips where Mark Nichol proved it’s difficult indeed by getting it wrong, as did my wife, a short-story author with a postgraduate degree in English. But it’s not impossibly hard, since the first commenter on Mark’s blog got it exactly right. (I’ll delay the link so you can try it.)

In the form I’ll use, the puzzle requires you to choose the two correct versions:

Version 1. Residents decide driving, and shorter trips to places like Canada are safer options.
Version 2. Residents decide driving and shorter trips, to places like Canada, are safer options.
Version 3. Residents decide driving, and shorter trips to places like Canada, are safer options.
Version 4. Residents decide driving and shorter trips to places like Canada are safer options.

My wife chose Version 4 alone.

Mark Nichol chose Version 3 and Version 4.

The correct answer is Version 2 and Version 4 (with Version 2 the more likely intended meaning).

Everyone agrees that Version 4 is correct; the questions are why does Mark erroneously think Version 3 is also correct and why does my wife fail to recognize that Version 2 is correct? Mark’s explanation supports my previous claim that among grammatically literate writers the most important comma errors derive from or at least implicate grammar errors. My grander claim is that comma errors create useless cognitive disfluencies, since they affect the reader’s grammatical parsing. In fields like brief writing, where the highest levels of clarity are advantageous, repeated comma errors—even if they’re subtle or controversial—summate to undermine Clarity.

Mark’s reasoning expresses a more straightforward error in grammatical analysis than the error I analyzed in The fundamental error of comma usage, as Mark claims that the string, and shorter trips to places like Canada, in Version 3, is an interjection, but an interjection (like Oh!) is grammatically isolated from the rest of the sentence. If the string were an interjection, the clause’s verb, are safer options, should be singular rather than plural. Since there’s no way to punctuate the sentence to make the verb singular, the italicized string, which must form part of its clause’s subject, can’t be an interjection.

But Version 2 is the correct answer if you’re allowed only one choice. The difference from Version 4  is that to places like Canada is a restrictive modifier in Version 4 and a descriptive modifier in Version 2, and the descriptive meaning is more probable. Read closely, Version 4 advises shorter Canadian trips, whereas the writer almost surely intended to advise limiting the length not just of trips to Canada (and similar places) but trips in general.  

Wednesday, December 5, 2012

Emphasis by brevity of sentences, paragraphs, and sections

To emphasize an idea, put it in a short sentence. To emphasize a sentence, put it in a short paragraph. To emphasize a paragraph, put it in a short section. In general,

Readers will give information relative emphasis in inverse proportion to its density.

I haven’t seen this principle articulated, and it only became apparent to me through the lens of construal-level theory; although the tip to use very short sentences for occasional emphasis is a commonplace, to use long sentences for de-emphasis isn’t. That the principle hasn’t been generalized might be because the effect is often subtle: it is only one of at least five means of emphasis, but a more interesting reason that the effect has gone unnoticed will emerge, in that the construal processes explaining emphasis by brevity also explain why writers aren’t apt to notice.

I’ll begin with an example at the sentence level. Compare this very long sentence to the constituent propositions:

With capitalism’s evolution, a decreasing proportion of the value produced is constituted of labor directly employed, an increasing proportion from labor already concretized in capital goods, since mechanization of production is the fundamental means to increasing economic efficiency, where capital goods contribute to the value of a product to the extent they are consumed in its production. (Context: Juridical Coherence.)

The simple ideas the sentence contains, such as that mechanization of production is the fundamental means to increasing economic efficiency, are commonplace ideas others have expounded at length. To subordinate their importance to the ideas I deemed novel, I demoted them by including them in one complex sentence.

Construal-level theory explains why emphasis by brevity works, by the low granularity of far-mode. The theory predicts and experiments find that reading occurs in far-mode, whereas writing occurs in near-mode (I conclude that the latter is lamentable), where far-mode apprehends in global units as we see from afar. In far-mode, each sentence has equal value; the more thoughts occurring in a sentence, the less the relative value of each. The theory also explains why emphasis by relative brevity isn’t common knowledge. Even while writing in far-mode, the writer is nearer his work than the reader because the self-other axis is a major dimension of construal level, and in near-mode, the longer sentence is more important than the shorter, rather than the reverse—near-mode adds when far-mode averages.

The phenomenon of emphasis by brevity confirms some standard writing advice and rebuts other standard advice. Commentators have expressed surprise at the degree to which variation in sentence length improves comprehension, suggesting more is at work than maintaining interest by variety. Varying sentence length makes writing clear by informing the reader how important the writer regards each component idea.

The misguided advice includes limiting sentences to one idea, implying writers should avoid compound sentences (and semicolons). Compound sentences serve to de-emphasize the ideas they contain, so their avoidance sacrifices emphatic contrast. Other misguided advice concerns paragraphs. Consistently short paragraphs have the same leveling effect on importance as consistently short sentences. And routine use of separate paragraphs for transitions between paragraphs is bad practice because the merely transitional usually doesn’t merit emphasis.

For a document’s sections, one all-too-common practice gravely offends against construal-level theory. A conclusion is almost mandatory in legal briefs and is necessarily short, but nonetheless, the emphasis it receives is often bestowed on a platitude with an initial “whereas,” in all caps no less. The better practice is to reserve a memorable idea for the short concluding section.