Yonatan Zunger (zunger) wrote,
Yonatan Zunger
zunger

Random English question

Question for all you native (and fluent) speakers out there, especially language geeks:

I generally don't split infinitives in English. There's one case that I'm stuck on, though, because I'm not sure if there's another way to indicate the difference I have in mind: "not to do X" versus "to not do X." The former implies that X is not done, but possibly through inattention or accident; the latter, a usage borrowed mostly from the speech habits of computer scientists, implies that the not doing of X is a primary objective of one's actions.

Is there a more correct way to say this? It feels clunky every time I say it.

(What brought this to mind was a news article about the Clintons' married life, where they say that Mr. Clinton "has told friends that his No. 1 priority is not to cause her any trouble." When I read that, it seemed that "not" was modifying "is" rather than "cause," which would suggest that his next line ought to be "It's to make sure other people do! Wahahahaha!")
  • Post a new comment

    Error

    switch
    default userpic

    Your reply will be screened

    Your IP address will be recorded 

    Help
  • 24 comments

susancalvin

May 23 2006, 19:24:03 UTC 9 years ago

If I was thinking really hard about it, I might say, "to do no X". But it probably depends on X, to see if that makes sense.

en_ki

May 23 2006, 19:24:12 UTC 9 years ago

The rule against splitting infinitives is dumb and totally ignorable in the cause of proper semantics. Rise up and write!

zunger

May 23 2006, 19:24:52 UTC 9 years ago

I completely agree... it's a stupid rule. But this particular case still clunks against my ears.

johnnybrainwash

May 23 2006, 19:29:39 UTC 9 years ago

I agree with your usage. Once you use it a thousand times, it might start sounding normal.

Your alternative, however, would be something like: "his number 1 priority is to avoid causing her any trouble".

mrissa

May 23 2006, 19:34:19 UTC 9 years ago

Yes. "Avoid" or "refrain from" vs. "neglect to" will clear that right up.

troglodyteking

May 23 2006, 20:18:35 UTC 9 years ago

I think this is probably the best way to get around the issue.

anthologie

May 23 2006, 19:46:23 UTC 9 years ago

Agreed. The sentence is awkwardly constructed no matter which way the infinitive is used, and awkward sentences need to be rewritten, period.

The split-infinitive rule is a holdover (along with not ending sentences with prepositions) from efforts to Latinize English and "rescue" it from its Anglo-Saxon/German roots. It has left us with a legacy of awkward construction problems and a bunch of More Grammatical Than Thou smugness (I admit it, I'm guilty of this as well, but for other reasons).

I do agree that the original example raises the question of intent. This can be resolved contextually, or again, by rephrasing the sentence so that the meaning and intent is clear.

There are plenty of times when splitting infinitives *is* in its own way awkward, and therefore not recommended.

en_no_gyoja

May 24 2006, 00:06:55 UTC 9 years ago

"Boldly to go where no man has gone before!"

See?

The "Never split an infinitive" thing is dopey.

anthologie

May 24 2006, 00:12:07 UTC 9 years ago

"To go boldly where..." would be just fine. And sounds quite regal.

zunger

May 24 2006, 00:18:54 UTC 9 years ago

But I'd say that "to boldly go" doesn't sound bad per se. It's one of those cases where the rule feels too strict.

faerieloch

May 23 2006, 21:07:49 UTC 9 years ago

I think a few years ago the OED declared it no longer incorrect to split your infinitives. In your above example, however, "not" negates the action, whatever it is. So "not to cause trouble" and "to not cause trouble" do actually mean the same thing, I think, we just parse it differently. "not" after a verb puts the action in the negative whereas putting it after the "to" makes it a positive action in our brains. However, they both mean he's going to avoid causing trouble.

zunger

May 23 2006, 21:24:51 UTC 9 years ago

I think there really is a parse error in "is not to cause trouble," since (to cause trouble) is a verbal phrase, so the "not" is left to modify "is." The best advantage of "is to not cause trouble" is that it mangles the verbal phrase badly enough that nobody mistakenly parses it as an atomic object. :)

faerieloch

May 23 2006, 23:28:41 UTC 9 years ago

Oh dear, now I'm think of how you would diagram/graph this sentence.... Can verbal phrases not be modified?

zunger

May 24 2006, 00:17:29 UTC 9 years ago

Hmm. They can; the question is whether the more natural parenthesization is "(my purpose) is (not to cause trouble)" or "(my purpose) (is not) (to cause trouble)."

In general, English admits a number of parenthesization ambiguities which have to be resolved by context; one of my favorites is the Society of Black Physics Students. That one can't be resolved (AFAICT) without non-grammar inputs. The "cause trouble" example seems simpler; it looks like a case of assigning probabilistic weights to two different bindings, and finding that one binding (the "is not" one) is more likely in English as a whole than the other. So a case for natural misunderstanding.

finagler

May 23 2006, 22:13:29 UTC 9 years ago

As quasi-quote Winston Churchill, "This is the kind of pedantic nonsense up with which I will not put!"

rubicantoto

May 23 2006, 22:50:40 UTC 9 years ago

I do agree with you that the quote about Clinton is a little off-putting.

In normal, casual speech, if I were to say "[Clinton] has told friends that his no. 1 priority is not to cause her any trouble" I would have phrased it that way so I could finish with with "but rather to go on a speaking tour" (or some other clause).

English is chock-full of lazy scoping like that. Take this sentence: "I don't think English is a silly language.". In usage, this tends to mean 'the speaker has a specific belief that English is not a silly language'.

Whereas if we parse the word sequence 'not+verb' as 'verb [negated]' as most other 'not+verb' sequences are parsed, the meaning that comes out would be more like 'the speaker lacks conviction regarding English being a silly language'.

:(
</soapbox>

To address the question you actually asked, I agree with johnnybrainwash and anthologie that there's no way to move the 'not' to fix that sentence.

zunger

May 23 2006, 22:57:01 UTC 9 years ago

Is that a parse error? Either way, I would parse it into

think:negated
subject: I
object:
is[verbal phrase]
subject: English
predicate nominative:
language[noun]
adjective: silly


The difference would be that in the first case, emphasis is implicitly on silly (the most weakly-bound object is the most mutable, and thus an implicit comparative), whereas if the speaker explicitly emphasizes 'think,' it implies that this is a subjective opinion and others may disagree.

anthologie

May 24 2006, 00:13:54 UTC 9 years ago

This reminds me of a terrible little joke I made to myself a week or so ago, when I dug out a copy of a childhood book called something like, "A hole is to put things in." Basically it's a list of items and what they're for. I made up my own:

"A preposition is to end a sentence with."

There's an infinitive there, but it doesn't work with the rest of the storyline to split it. Ah well.

zunger

May 24 2006, 00:43:19 UTC 9 years ago

And a conjunction is to begin one. :)

anthologie

May 24 2006, 00:44:10 UTC 9 years ago

Word.

This has been on my mind lately

xaosenkosmos

May 24 2006, 03:04:19 UTC 9 years ago

I'm not touching your question, because i don't know the answer with certainty. But i will happily explain why that is =)

The semantics of negation are very subtle. To give some sense of how obnoxious it is: the text for my intro to semantics class has exactly one paragraph on negation. Somehow they turn propositional logic's ¬ (or !,/,...), into a whole paragraph. Because that's all they cover. Assuming that natural language is as clear-cut as propositional logic is hah-hah funny.

More specifically, what is the difference between a theist, not a theist, an atheist, and an anti-theist? The latter are both "negations" of the former, but one is a binary opposition and the other a polar opposition. We can explain them, but try to squeeze them into a truth table or two. It just doesn't feel right. There's something missing there.

I'm fairly convinced that boolean logic is impotent once you step out of the happy land of contrived boolean problems (which includes a lot of really damned useful stuff). The lay of the land here evokes program verification, and specifically, the standard quip about program verification:
The problem with engineers is that they tend to cheat in order to get results.
The problem with mathematicians is that they tend to work on toy problems in order to get results.
The problem with program verifiers is that they tend to cheat at toy problems in order to get results.
-- Anonymous

Re: This has been on my mind lately

zunger

May 24 2006, 04:19:04 UTC 9 years ago

I think there's still some value in Boolean logic, but you have to very carefully parse the sentences, and occasionally annotate them with other data, before it's useful. Turning any of those (!)theists into a Boolean negation of "theist" would definitely be nonsense, but parsing them into

theist -> "believer in the existence of a god"
not a theist -> "not a (theist)"
atheist -> "believer in the nonexistence of a god"
anti-theist -> "opponent of (theist)s"

leaves us in much better shape. Parse enough and you're left with syntactic particles of definite Boolean meaning plus nouns and verbs with (hopefully) either irreducible definitions or agreed-upon ones. ("God" being one of the infamously bad cases; so many of the arguments for the existence thereof fall apart completely when you expand the definition of that noun)

this lesson is hammered in linguistics classes like "social construction" is in anthropology

evan

May 24 2006, 07:12:14 UTC 9 years ago

Language is whatever we all agree it to be. Whichever construction sounds right to you, as a native speaker, is "correct" for any meaningful definition of the word: it's what will make others understand you.

Re: this lesson is hammered in linguistics classes like "social construction" is in anthropology

zunger

September 22 2006, 23:48:07 UTC 9 years ago

Hmm... I just came back to this post after a while.

The problem is, I'm not actually a native speaker. :)