The Astonishment Effect in negation
« previous post | next post »
I wrote in my posting on "forbidden OSR":
Every so often we post about some comprehensible examples that strike us and our correspondents as unacceptable — examples like ["the well is forbidden to play near"] — and then our task is to try to decide whether these examples are all inadvertent errors, or whether at least some of the instances represent a non-standard system different from our own. (Not infrequently, the latter turns out to be the case, to our astonishment.)
Call this the Astonishment Effect. You think that something is just flat-out ungrammatical, and then you find piles of examples.
My posting elicited e-mail from Paul Postal, who reported on a couple of Astonishment experiences of his own, having to do with negation.
[What follows is a guest posting from Paul Postal.]
… here are two cases of the above which actually arose in my research on negation, one in 2005, one a couple of weeks ago.
One topic I was studying was negative inversion in complements, discovered by Larry Horn in 1978, as in (1):
(1) I didn't believe that ever had he understood the Kravowitz Theorem.
I wrote that subject/auxiliary inversion was as obligatory here as in main clauses like (2):
(2) Never did he/*he did understand the Kravowitz Theorem.
But for some reason I googled this 'obvious' fact and found enormous numbers of counterexamples, all totally inconceivable for me, e.g.:
a. I don't believe that at any time in history there has been a more important necessity for the voices of peace and reason to sing their song at the top of their … (link)
b. I don't believe that at any time Truman ever had on board the full six Administrative Assistants authorized by law. (link)
c. I don't believe that at any time we’ve entered into discussion it was driven by somebody who wanted to buy down the annual harvest. (link)
More recently, I was thinking about the negative reading of:
(3) Sheila has yet to finish her thesis.
I wanted to argue there is a deleted negative, and cited among other evidence:
(4) *Sheila has not yet to finish her thesis.
I would give this actually ten stars and my wife fifty. But google supplies e.g.:
a. … as the Shipping Board has not yet to use the lakes in training and recruiting mariners, but has left that field to the United States Navy. (link)
b. … review has not yet to begin. This procrastination has denied the public the opportunities to search for an appropriate design to facilitate conservation of … (link)
c. … ideal, who has not yet to be put to use. (link)
I don't know what is wrong with these people but am pretty sure even modern medical science is helpless to fix it.
M. Logan said,
July 20, 2008 @ 2:38 pm
Just to clarify, the starting question is (1) as stated versus (1') with "… ever he had…"?
For me, both (1) and (1') sound iffy, and I would have to phrase it as "… had he ever…"
(For what it's worth, (2a) and (2b) sound OK to me, but there's something a bit off about (2c).)
Ed said,
July 20, 2008 @ 2:48 pm
(1) sounds stilted, if not ungrammatical, to me. (a-c) under (2) are all perfectly normal. be astonished if you like =)
(a-c) under (4) also sound unnatural. i could go with either a phrasing like (3), which is, again, a bit stilted, or drop the "to" and use the past participle:
a) …has not yet used…
b) …has not yet begun…
c) …has not yet been put to use…
S Onosson said,
July 20, 2008 @ 3:05 pm
I agree with Ed on all counts.
Q. Pheevr said,
July 20, 2008 @ 3:24 pm
I guess everyone triggers the astonishment effect in someone sometimes. I find (1) totally ungrammatical, and the "I don't believe" sentences in the first a–c set completely fine. (4) and the second set labelled a–c are ungrammatical for me, although I don't think I need more than one asterisk for each of them. (If I encountered (4) in the wild, I would assume that it was a performance error in which the writer/speaker had started to produce "Sheila has not yet finished her thesis," and then switched over to "Sheila has yet to finish her thesis" in midstream.)
mollymooly said,
July 20, 2008 @ 5:45 pm
I would certainly give (1) more stars than (2*). I think fronting "ever" just isn't possible in that subordinate clause. BTW I think that, to parallel (1), (2) should read
– Never had he/*he had understood the Kravowitz Theorem.
Most of 2,3,4 look like overnegation errors from using hifalutin language. This happens a lot with "can('t) but", "could(n't) but", "(don't) doubt (but) that"
I could imagine saying but not writing (2b), with its extra "ever".
Nick Z said,
July 20, 2008 @ 6:31 pm
Of course, the concomitant of the "astonishment effect" is the inability to understand why supposedly correct examples are right…
Am I misunderstanding the point of negative inversion if I say that I would have to say 1) I didn't believe that he had ever understood the Kravowitz Theorem? And 2) He never did understand the KT.
(Standard British speaker)
blahedo said,
July 20, 2008 @ 10:24 pm
I gather from your phrasing that (1) is an example used by Horn—could you give it a little more context? Like the other commenters, I'm finding it pretty dubious, but I can get it with a lot of work: to me, it's unedited spoken only, where if the utterer got as far as "I don't believe that *ever*…" and then had to choose between "he had" or "had he", the inverted choice is at least less bad (probably from analogy to (2)). But it requires stress on the "ever" and is still sort of questionable.
With (3) and (4) there is a simple explanation of how it arose: (4) is a conflation of the impeccable (3) and the similarly impeccable
(5) Sheila has not yet finished her thesis.
and the frequency of examples online could easily be attributed to that if you wanted to maintain (4) as ungrammatical. (I certainly find (4) much less astonishing than (1)!) Or, of course, (4) could just be the vanguard of a new grammatical construct. Vive la différence!
The Ridger said,
July 21, 2008 @ 8:45 am
I think that 1 is "supposed" to have inversion because the "ever" was fronted. But that sounds impossibly stilted to me – "I don't believe that ever had he …" isn't emphatic, just strange. I'd have to say "I don't believe that he had ever", myself – I can't say "that ever had he…" though "never had he" is okay. It's a negation thing, I think.
Examples 2 all sound fine to me, though 2c seems a bit odd – possibly because I don't have the context. It seems to want a "but" , but then again maybe he does mean that such people never drive the discussions, in which case it's fine.
The 4s all seem off to me, but they seem like a simple error – maybe in revision.
Graham Campbell said,
July 21, 2008 @ 12:21 pm
1 is less astonishing to me, and is in fact something I would say (although never write). The emphasis with 1 all goes on the "ever" and then the admonishment of his inability to grasp the Kravowitz Theorem becomes clear.
For me, 2c is absolutely referring to such people as would never drive the discussions as The Ridger suggests. It's certainly a more complicated way of making that point but, as with 1, is much more emphatic (or, depending on the speaker, dramatic).
The rest of 2 and 3 are sensible enough. 4 is most certainly astonishing.
S Onosson said,
July 21, 2008 @ 5:04 pm
"I would give this actually ten stars and my wife fifty."
At first glance, I took this to mean you would give fifty stars to your wife, and wondered what she had done to displease you so!
Janice Huth Byer said,
July 22, 2008 @ 6:53 pm
I'm with Ed and S Onosson who sense "to" in #3 contraindicates the need for a negative. The "to" turns "has" into a modal auxiliary of future obligation that "yet" serves to make negative. Which is to say, the phrase "has yet to" entails not having done something.
On the other hand, remove "to" and "has" becomes a totally different word that aids a regular past tense verb"finished" to become a past participle, which, even with the help of "yet" needs a negative, being a good ol' regular verb.
"She has not yet finished her thesis."
"She has not yet to finish her thesis" is the kind of hypercorrection or rewriting error, that native speakers, including children, are reported never to make in extemporaneous talk, whereas writers do, usually after too many rewrites.