In one of Grammar Girl’s podcasts, I came across a term that was unfamiliar to me: the squinting modifier, a modifier placed so that it’s unclear whether it’s modifying something just before it or something just after it. Her example:
Children who laugh rarely are shy.
Are we talking about laughing rarely, or rarely being shy? A genuine ambiguity, although the name didn’t make sense to me, and Grammar Girl didn’t explain its origin. To me, this is just a variety of attachment ambiguity. In the attachment ambiguities I’ve talked about before, it was a question of whether a modifier phrase attached “low” or “high”, but both possibilities had it looking backward for what it modified. (Refer to the tree diagrams in the other posts under this category to see how “low” and “high” make sense.) For example, in I resolve to call her up a thousand times a day, we could be talking about a thousand calls (low) or a thousand resolutions (high), but either way, the modified phrase is before the modifier a thousand times a day. In a squinting ambiguity, it’s a question of whether the phrase attaches forwards or backwards. So to my mind, a more transparant name would be forward/backward attachment ambiguity. But what does that have to do with squinting?
It turns out squint has a meaning I never knew. When I think of squinting, I think of this:



I checked to see what the Merriam-Webster Dictionary of English Usage had to say on the subject. It was skeptical that such a thing was found very much in the wild:
The squinting modifier resides chiefly in college-level handbooks. … [In our opinion], the squinting modifier is more of a theoretical possibility — with, it must be admitted, a catchy title — than a real problem. It would seem most likely to occur when a split infinitive is being carefully avoided by putting the would-be splitting adverb ahead of the infinitive….
Their reasoning was that a native English speaker would not write, for example, rarely are shy; they’d write are rarely shy, unless they’re trying to follow some half-baked rule about not letting adverbs split up whatever they think they’re not supposed to split. Well, I’ve seen squinting modifiers. They’re real. They’re out there, and in sentences that weren’t written with an eye toward avoiding splitting infinitives or other things. A couple that I’ve had sitting in the draft posts:
Quitting smoking now greatly reduces serious risks to your health.
Place the appliance on a hard, flat level surface only to avoid interruption of airflow underneath it.
In the first sentence, does the now modify quitting smoking, or greatly reduces serious risks to your health? For real-world reasons, we know it modifies quitting smoking: Otherwise it would be implying that previously, quitting smoking didn’t greatly reduce serious threats to your health. In the second example, does the only modify hard, flat level surface, or to avoid interruption of airflow underneath it? Well, if it modified to avoid interruption…, it would be saying in essence, “There’s only one reason you should be placing this appliance on a hard, flat level surface, and that’s to avoid interruption of the airflow underneath it. Otherwise, place it on a surface that isn’t hard, flat, and level.”
Hmm. Well, I guess MWDEU‘s point still stands after all. Even though both meanings were phrased in a natural way, the ambiguity was still only a theoretical possibility in each, this time because of real-world knowledge.
But that still doesn’t answer my question: Why are these things called squinting modifiers?
Through Google Books, I tracked the term back to George Campbell’s The Philosophy of Rhetoric, published in 1776. Here’s what Campbell said:


Apparently, Campbell thought he was explaining the term by telling what I already knew: The modifier could attach forwards or backwards. But he did give a clue: squinting construction is a calque (i.e. a term borrowed from another language, but translated first) of the French term construction louche. In my French-to-English dictionary, louche is translated as “cross-eyed,” as well as “ambiguous, suspicious.” It’s related to the verb loucher, which means to slant, or wander, like an eye that can’t focus correctly. Now that made some sense: A modifier that seemed to be looking in two directions. My question now: Why was louche translated into English as squinting?
The answer turns out to be simple: Because squint can mean to be cross-eyed, and also to be wall-eyed. The earliest definition in the OED is “To have the axes of the eyes not coincident, so that one or both habitually look obliquely; to be affected with strabismus,” followed by “To look with the eyes differently directed; to glance obliquely or in other than the direct line of vision; also, to glance hastily or casually, to peep,” and a couple of definitions later, “To move or branch off in an oblique direction.” My definition isn’t in there at all, but I know I’m not the only one with that meaning for squint: I got my narrowed-eyes illustrations by doing a search for squint in Google Images.



So squinting modifier (or squinting construction) is an apt name after all, if you have the right definition for squint. I never did until today. Did you?