Recoiling from complexity

Nathan Heller has critically reviewed Steven Pinker’s The Sense of Style. His main complaint is that Pinker’s advice includes too much leeway based on context and subjective judgement and too little respect for the “basic principles of English grammar”. For example:

English is complex. To help reduce ambiguity, modern usage attaches specific words to specific functions. The restrictive-nonrestrictive division between “that” and “which”—two particularly common and shifty words—is one attempt at clarity.

Heller cites “usage”. To remind you: usage is what people do with language. So, do people use ‘that’ to open restrictive clauses and ‘which’ to open nonrestrictive ones? Some do, but many don’t.

A great little study by Jonathon Owen, for his Master’s thesis at Brigham Young University, looked at the kinds of changes copy editors make. He reviewed edits made to manuscripts written by academics and found one of the commonest to be imposing the “the that/which rule, proscribing the use of which as a restrictive relative pronoun”. This tells us that this distinction may be widely observed by copy editors but less so by even educated writers of formal prose – and presumably less still by the public as a whole.

The American Heritage Dictionary (AHD) goes further, noting that “this use of which with restrictive clauses is very common, even in edited prose”.

So yes, people have attempted to create this distinction: EB White recommended it, so did HW Fowler, so does Garner’s Modern American Usage. And dedicated copy editors have done their best to make usage follow this distinction (including me, until I wised up). But usage persists in going its own way (Garner concedes that restrictive ‘which’ is ubiquitous). And no lack of clarity has resulted, because we use commas to highlight nonrestrictive clauses.

Heller has further concerns:

Another is the rule that “like” joins noun phrases, while “as” or “as if” is for verb phrases. (“It looks as if my date is here!” “You look like Mom in that dress.”) Pinker doesn’t see the point of that one, either.

His grammar terms aren’t quite right here: “my date is here” is a clause, not a verb phrase; “is here” is the verb phrase.

Again, this is a rule insofar as people like Heller insist on it, but by the standards of usage there’s far more leeway, depending on the situation. The AHD again:

They don’t make them like they used to. I remember it like it was yesterday. As these familiar examples show, like is often used as a conjunction meaning “as” or “as if,” particularly in speech. While writers since Chaucer’s time have used like as a conjunction, the usage today has a somewhat informal or conversational flavor. Language critics and writing handbooks have condemned the conjunctive use of like for more than a century, and in accordance with this tradition, like is usually edited out of more formal prose.

But it adds:

Like is acceptable at all levels as a conjunction when used with verbs such as feel, look, seem, sound, and taste: It looks like we are in for a rough winter.

And Garner accepts that conjunctive ‘like’ is “acceptable casual English”, although not yet “unimpeachable”. Again, genuine confusions caused by using ‘like’ in this way are vanishingly rare. Heller raises another issue:

It’s for grammatical consistency, not beauty or gentilesse, for example, that correct English has us say “It was he” instead of “It was him.” Pinker calls this offense “a schoolteacher rule” that is “a product of the usual three confusions: English with Latin, informal style with incorrect grammar, and syntax with semantics.” … [But] the nominative and accusative cases are the reason that we don’t say gibberish like “Her gave it to he and then sat by we here!” No idea is more basic to English syntax and grammar. In the phrase “It was he,” “it” and “he” are the same thing: they’re both the subject, and thus nominative.

I agree: it definitely isn’t for beauty that anyone says “It was he”. It sounds so prissy I can hardly bear to type it.

And Heller is right about the difference between nominative and accusative, but then he misapplies it because of another grammatical misunderstanding. In “It was he”, “It” and “he” are not both the subject; “he” is the object.

You can see this more clearly in the grammatically equivalent “It was they”. Subjects govern verbs, and here, the singular subject “it” gives us the singular form of the verb “was”. Then “they”, being plural, is obviously not the subject. Heller’s mistake is precisely the confusion of syntax with semantics that he quotes Pinker warning him against. The two pronouns have the same referent but not the same grammatical role.

All this nit-picking is cracking good fun, but it’s beside the point. According to usage, “It was he” lives happily enough at the formal end of the spectrum but struggles elsewhere. More from the AHD:

Traditional grammar requires the nominative form of the pronoun in the predicate of the verb be: It is I (not me); That must be they (not them), and so forth. But in less formal contexts the nominative pronoun can sound pretentious and even ridiculous, especially when the verb is contracted, as in It’s we. The traditional rule creates additional problems when the pronoun following be also functions as the object of a verb or preposition in a relative clause, as in It is not them/they that we have in mind, where the plural pronoun serves as both the predicate of is and the object of have.

Garner also agrees that “it is me and it’s me are fully acceptable, especially in informal contexts”. The resultant alleged lack of “grammatical consistency” causes no harm. One final complaint from Heller:

The same is true of “who” and “whom,” another nominative-accusative pair to which Pinker objects, sort of. He writes, “The best advice to writers is to calibrate their use of ‘whom’ to the complexity of the construction and the degree of formality they desire.” Yet who wants to undertake that calibration all the time? The glorious thing about the “who” and “whom” distinction is that it’s simple.

Whom the hell is this guy trying to kid? It’s easy to calibrate formality to context. We all do it all the time.

(And, as Ben Zimmer reports, even writers as good as Nathan Heller trip themselves up on the ‘who’/‘whom’ distinction.)

Garner also likes the nominative/accusative distinction here, but finds that “in certain contexts, whom is stilted”. The AHD agrees: “in practice whom is uncommon in speech and everyday writing because it has a formal tone”.

So, overall, what lies behind Heller’s objections?

This tendency to add complexity, ambiguity, and doubt is a troubling feature of Pinker’s rules.

This is telling. Heller wants language (and guidance on it) to be simple. That’s admirable, but his insistence on unbending rules can lead to a conception of language that is not just simple but simplistic. While he knows in general that language is complex, he seems to recoil from glimpses of this complexity.

Language is alive and its conventions are diverse: as such a vast category of human behaviour, how could it be otherwise? But to accept that is to abandon the certainties of the rulebook, and some people find that troubling. I know I did, at first. Then I found it exhilarating.

While I’m here: I mostly liked Pinker’s book. In the tradition of Joseph Williams’s Style, Pinker doesn’t just set out a list of dos and don’ts – he takes pains to explain what promotes clarity in writing. He covers how to structure a sentence so it’s easier to understand, how to show logical links between sentences and create a sense of flow through a paragraph, how to fight the “curse of knowledge” and focus on the reader’s perspective rather than your own, and a lot more.

These things matter so much. In my copy editing, I daily come across writing that needs these sorts of principles applied. They are far more important than the piddling little worries – like split infinitives and ‘literally’ and ‘very unique’ and double negatives and Heller’s points above – that dominate public debate about language.

On the negative side, I could have done without all Pinker’s sentence diagrams, which I wasted a lot of time trying to understand before realising the details were barely relevant. And the section near the end, where he lists a load of words that he thinks people should stop misusing, goes against the spirit of the rest of the book by not giving reasons. I would also have liked more detail on some of the academic studies he cites to back up his arguments – I don’t have the time or resources to chase up references.

But on the whole, it’s a damn good book.

Both comments and trackbacks are currently closed.


  • Luke  On November 4, 2014 at 2:53 pm

    I am 49, consider myself reasonably educated and never came across the “that/which” rule until Word started having grammar check.

    I clearly remember Latin and French classes in which we were taught how to distinguish the two uses of which so as to translate into French and Latin. Far from seeing either use as incorrect, our teachers devised lots of sentences using “which” in one or other sense for us to translate. If we were going to translate into Latin/French, we had to realise that the word had two uses. And we found it quite difficult to learn.

    I think the fact that we all had to be taught it (and struggled) shows how ingrained the loose use is. (Of course, you know that.)

    The whole prescriptive nonsense falls away when you study another language. The French just have to accept the way we use which, and that we say “it’s me.”

    • Tom Freeman  On November 4, 2014 at 4:25 pm

      I didn’t come across the ‘that’/’which’ rule until I started doing editorial work in my 20s. But I was keen to know all the rules, so I took it up, and it became second nature.
      Then, a few years ago, I was persuaded that the rule is pointless, so I stopped applying it in my editing. But I still prefer restrictive ‘that’ over ‘which’ – it sounds better to me. Following this convention changed my aesthetic judgement.
      So, even though I don’t see it as a matter of right and wrong any more, the rule still influences me. No big deal, of course, but it shows how hard it can be to shake these things – ‘but it feels wrong!’

      • Luke  On November 4, 2014 at 9:49 pm

        I don’t disagree with you, but to me “that”, particularly when applied to humans, sounds odd. FWIW, I think translating is significant. We don’t say Homer was wrong to use rather ahistoric Ionian Greek. We just accept he did so. But thanks for response (and original post).

  • Rui  On November 6, 2014 at 12:21 am

    There’s a typo in the first sentence.🙂
    Thanks for the article. I’m planning to read the book since you think it’s worth it.

    • Tom Freeman  On November 6, 2014 at 7:58 am

      Dammit! Well spotted, thank you! Fixed now.

  • TW  On November 12, 2014 at 3:45 pm

    ‘In “It was he”, “It” and “he” are not both the subject; “he” is the object.’

    Object? Predicative complement, surely?

    • Tom Freeman  On November 12, 2014 at 7:33 pm

      Yes, I think you’re right. I’ll add a note.

      • Tom Freeman  On November 12, 2014 at 7:39 pm

        Actually, I’m not going to add a note. If I start to edit the post, WordPress eats all my paragraph breaks, and I don’t have the will to put them back in.
        But you are right. With other verbs it would be an object, but not with ‘be’.
        Heller’s still wrong, though, so I’m going to give myself half a point for that sentence!

%d bloggers like this: