Treading down the thorny path

Two evergreen topics in grammar and usage: so-called “split infinitives”, where some usage critics have insisted that they must always be avoided, however unnatural the results of this avoidance are; and modifier attachment, where jokes are often made about one of the potential attachments, however preposterous the interpretation associated with this attachment is.

The two topics are connected through their unthinking devotion to dogmas of grammatical correctness: avoid split infinitives, avoid potential ambiguity. A devotion that leads adherents down the thorny path of usage rectitude to using unnatural syntax and entertaining preposterous interpretations.

But first, the thorny path. The (tough) counterpart to the (easy) primrose path.

The primrose path. From the Phrase Finder site on primrose path (especially in being led down the primrose path, or going down the primrose path, with dire consequences):

meaning: the pleasant route through life, of pleasure and dissipation

origin: This phrase was coined by Shakespeare, in Hamlet, 1602. It is evidently a simple allusion to a path strewn with flowers.

An allusion to a path strewn with flowers, yes, but heavily figurative already in Ophelia’s injunction:

Do not, as some ungracious pastors do,
Show me the steep and thorny way to heaven;
Whiles, like a puff’d and reckless libertine,
Himself the primrose path of dalliance treads

On the primrose path tread the puffed (arrogant, boastful) and reckless libertines, the gay party boys, and of course the scarlet women, as in the 1940 film Primrose Path:

(#1) Primrose Path is a 1940 film about a young woman determined not to follow the profession of her mother and grandmother, prostitution. It stars Ginger Rogers and Joel McCrea. The film was based on the play of the same name by Robert L. Buckner and Walter Hart and the novel February Hill by Victoria Lincoln (Wikipedia link)

The image of the primrose path serves as a warning about the dangers of pursuing pleasure without responsibility, especially when this leads to disastrous consequences. But the thorny path, pursuing rectitude without empathy, can equally lead to disastrous consequences. A caution that applies to the pursuit of grammatical correctness without regard for actual usage and actual interpretations of language in context.

Avoiding split infinitives. The dreadful example. In The Economist‘s 2/13/21 issue, p. 81, in a review of Fall, by John Preston, a biography of bigger-than-life media mogul Robert Maxwell:

([Maxwell] was briefly a Labour MP who drove the House of Commons mad with his refusal ever to shut up)

This was so awkward, so inept that I had to pause for a moment to consider what it was supposed to mean, before I saw that it was a labored avoidance of the very natural split infinitive in

… with his refusal to ever shut up

(while I was typing up the paragraph above, a Medicare Helpline commercial came by on tv with this nicely split infinitive:

Call now to instantly find out …

where the split-avoiding

Call now instantly to find out …

is not only immensely awkward, to my ear, but is also likely to be misunderstood as having instantly modifying call now rather than find out.)

The Maxwell sentence in 2021 was also somewhat of a surprise, since several years before, the publication had officially softened its policy on split infinitives, and I would have thought that his refusal to ever shut up (vs. his refusal ever ever shut up) would have been a star candidate for the Economist‘s kinder, gentler treatment. But the publication clearly backslides now and again, permitting eruptions of irrational anti-splitting prejudice.

My posting on the new rationality: from 5/20/18, “A zombie lurches in the NYT”, on the Times‘s (a) it sits inert, ready mostly to be ignored — rather than (c) ready to mostly be ignored (split infinitive) or (d) ready to be mostly ignored (“split verb”):

A great many writers and editors would avoid (c) because it’s a (so-called) “split infinitive” (SI), a construction with material intervening between the infinitive marker to and its VP complement. Irrational aversion of SIs has a long sad history, but even peevish and sticklerish usage advice has been shifting in their favor in many circumstances: the Economist and the AP Stylebook, among others, have newly gotten on board.

From an earlier time, in my 10/13/12 posting “Avoiding a split infinitive”,

on “split infinitives” and their avoidance. These are perennial topics on Language Log, this blog, and other linguablogs; at this point, I see no useful purpose in running through the routine one more time, when MWDEU laid things out almost 25 years ago [in 2012, so the MWDEU treatment is now approaching 35], and you can find the details in many well-informed sources. But here’s a sharply worded assessment of the baleful effects of ill-informed advice, from “How awkwardly to avoid split infinitives” on Stan Carey’s blog:

Grammatical misinformation can be countered, but it leaves a residue. It feeds what the Language Log linguists call “nervous cluelessness” about language, and it leads to avoidance of a legitimate usage because a subset of vocal pedants have fomented a petty prejudice against it.

And here’s Geoff Pullum confronting the prejudice on Language Log:

All I’m saying that, squirm though you may, it is fairly common for placing an adverb between infinitival to and the following plain-form verb to be not just grammatical (it is always grammatical), but also the best stylistic choice.

(For an inventory of important Language Log postings on split infinitives , through mid-2009, see my 6/13/09 posting “Postings on split infinitives”.)

Attachment humor. There’s a Page on this blog inventorying postings on attachment of modifiers. Many of the postings are about attachment humor, in which an intended, contextually entirely straightforward attachment is disregarded by some chucklehead, who fixes instead on a preposterous alternative.

So it is with the Mother Goose and Grimm comic strip from 3/7/21:

(#2) How to interpret a bad guy with one arm named Fred? In particular, what does named Fred modify?

Grimm’s intention is high attachment (HA): named Fred modifies a bad guy with one arm. But the notoriously dim-witted Ralph gets the low attachment (LA) interpretation, in which named Fred preposterously modifies one arm.

The set-up is decidedly artificial: it looks like named Fred is in the quote only to set up the possibility of an absurd LA: why would anyone mention the name of a Fugitive bad guy with one arm in retelling the story of an episode? How would it be relevant?

An earlier, and more complicated, MGG strip on attachment of modifiers: from my 5/27/17 posting “Memorial Saturday 4”, on cartoon #1 there:


The relevant expression is the NP anyone he sees on a bike. The PP on a bike is serving as an adverbial here, but the question is where the adverbial is attached: high, at the level of anyone; or low, at the level of the immediately preceding V sees.

Mother Goose’s intention is that the adverbial is understood as a postnominal modifier of anyone, “extraposed” from anyone on a bike: anyone he sees on a bike is understood as ‘anyone on a bike who he sees’.

But the doctor takes the adverbial to be an adverbial denoting the place where the seeing takes place, in which case the adverbial locates the referent of the subject (he), that is, the dog. So if the problem with the dog comes when the dog is on a bike, Mother Goose should take the bike away.

The problem here is all potential ambiguities are being treated as equally available, when in fact actual speakers in real-life contexts disregard most of the potential interpretations as being irrelevant to the context or too preposterous for consideration; people either don’t notice these potential interpretations at all, or they abandon them after the briefest consideration — at the level of centiseconds — and entirely out of consciousness. Nevertheless, the dogma that ambiguity, even merely potential ambiguity, is an offense against grammar (because it introduces unclarity) endures.

I wrote a series of Language Log postings on potential ambiguity back in 2008; the first two are especially relevant here:

— from AZ on LLog on 6/1/08, on “Avoiding ambiguity: a pattern”:

potential ambiguity is EVERYWHERE. There are lexical ambiguities, constructional ambiguities, multiple potential referents for anaphors, and more complex cases. Almost any sentence potentially has more than one interpretation. Given that, ambiguity should be seen as a feature of language, not a bug

— from AZ on LLog on 6/2/08, on “Avoiding potential ambiguity: does it improve clarity?”: you will not be surprised to hear that my answer to that query is a resounding NO, argued at some length.

Leave a Reply

%d bloggers like this: