"No telling is neither complete nor accurate"
« previous post | next post »
Emily Yahr, "Read George W. Bush’s speech at the African American Museum, 13 years after signing the bill to build it", Washington Post 9/24/2016:
Our country is better and more vibrant because of their contributions and the contributions of millions of African Americans. No telling of American history is neither complete nor accurate without acknowledging them.
Full audio is here.
Daniel Deutsch sent me the link, with the comment that "Bush 43 gave a beautiful speech at the museum opening, but this seems overly negative" — referring to the "No telling … is neither complete nor accurate" phrase.
This seems to be an example of a phenomenon known variously as "Multiple Negation" or "Negative Concord". For a discussion of negative concord in languages around the world, see Anastasia Giannakidou, "N-words and Negative Concord", The Linguistics Companion 2002, who observes that "NC is observed in many languages; e.g. Romance, Slavic, Greek, Hungarian, Nonstandard English, West Flemish, Afrikaans, Lithuanian, Japanese".
On the other hand, Johan van der Auwera & Lauren Van Alsenoy ("On the typology of negative concord") claim that "contrary to what is often stated, negative concord is not all that frequent […] based on a world-wide sample of 179 languages".
But be that as it may, the English language has been trying to make up its mind about negative concord for the past millennium or so — for a detailed study, see Amel Kallel, "The Loss of Negative Concord in Standard English", 2011. (Or see the 2007 paper by the same author in Language Variation and Change.)
More accurately, as Kallel describes it, Old English was pretty firmly in the Negative Concord camp; Middle English and Early Modern English "exhibited variable use of [+NC] and [-NC] systems, i.e. speakers belonging to those two periods used both single and double/multiple negations to express the same meaning", while "Modern Standard English exhibits a virtually uniform [-NC] system". In a footnote, Kallel explains that "This study excludes the non-standard varieties of English which exhibit an NC system".
For information and references about those "non-standard" varieties, see the Yale Grammatical Diversity Project's informative page on Negative Concord in English.
But I think that there's a stronger negative-concord residue trying to diffuse into "standard" English than these works suggest. That influence has been one of the alternative explanations that I've suggested for (some instances of) the phenomenon of misnegation — for a few selected examples see:
"I challenge anyone to refute that this negative is not unnecessary", 1/21/2004
"Not doubting that the door could not be opened wider", 6/5/2006
"It's hard not to read this and not do a double-take", 8/1/2006
"Multiplex negatio ferblondiat", 7/14/2007
"I wouldn't be surprised if they didn't", 2/23/2009
"Never no one without Cornish", 8/1/2011
"Everything cannot not be unbelievable, either", 8/10/2011
"The things neither of them don't do", 9/17/2014
David Marjanović said,
September 25, 2016 @ 10:50 am
I agree. Another example is the frequent use of "I don't think" rather than "I think" after sentences that contain a negation.
Karen said,
September 25, 2016 @ 11:14 am
I think negative concord is *extremely* common in IE languages, regardless of the "world-wide sample of 179 languages".
TR said,
September 25, 2016 @ 11:46 am
I found it surprisingly hard to work out what this sentence would compute to if taken literally. Eventually I concluded that it would mean "Any telling of American history that did not acknowledge them would fail to be both incomplete and inaccurate", but I still can't look at the sentence and get it to mean that.
Gregory Kusnick said,
September 25, 2016 @ 12:25 pm
Bush was not speaking extemporaneously; the sentence in question is part of a prepared text that likely went through several revisions. So it seems possible that the apparent negative concord in this instance is an artifact of authorship-by-committee that none of the individual contributors would have produced naturally. (But on the other hand, apparently nobody thought it wrong enough to warrant another rewrite.)
[(myl) This might not be relevant, since there are quite a few obvious ad libs, apparent re-phrasings, and other likely departures from the script in the recording. I looked for a "remarks as prepared for delivery" text, and failed to find one. So maybe the transcript is how the original text read — in which case the leakage of negative concord into formal written English is even more remarkable — or maybe W reformulated the sentence in tune with his native West Texas vernacular.]
Coby Lubliner said,
September 25, 2016 @ 2:05 pm
"Native West Texas"? GWB is a native Connecticut Yankee like you, Mark.
[(myl) It's true that he was born in New Haven, but according to Wikipedia, "Bush attended public schools in Midland, Texas, until the family moved to Houston after he had completed seventh grade. He then went to The Kinkaid School, a prep school in Houston for two years."]
Guy said,
September 25, 2016 @ 2:17 pm
On a slightly related note, I recently produced a sentence with a form like "there are no X's or any Y's", with the same meaning as "there aren't any X's or any Y's". This construction seems totally natural to me, but the semantic scope of the negation has to be interpreted as covering the second coordinate, even though it's not syntactically modifying that coordinate or any constituent higher in the phrase structure. How do other people feel about this construction?
Gregory Kusnick said,
September 25, 2016 @ 2:25 pm
MYL: My guess is that the original (i.e. first draft) text probably said something like "No telling of American history is complete or accurate…" and that it was revised into the neither/nor form in order to punch it up.
Guy: I'd say "no Xs nor any Ys".
Rebecca said,
September 25, 2016 @ 2:36 pm
GWB may not be a native west Texan, but he moved there as a young toddler and went to school there through high school, so there's a good chance that his speech was as influenced by the local dialect as it was by his parents'.
TR said,
September 25, 2016 @ 2:50 pm
Guy's construction sounds totally natural to me too — I wouldn't even have noticed the syntactic anomaly.
Guy said,
September 25, 2016 @ 2:55 pm
@Cory Lubliner
Given that Bush's family apparently moved to Texas when he was 2, according to my quick Googling, I think that West Texan could fairly be described as his native dialect. One could make an etymological objection to that usage, if etymology is to be regarded as important, but then we would need to acknowledge that no one is born already knowing how to speak a language.
Guy said,
September 25, 2016 @ 3:41 pm
That should be @Coby Lubliner, my apologies.
David Marjanović said,
September 25, 2016 @ 4:13 pm
Influence from German? German would have "there are no Xs or Ys" (lacking an equivalent of "any"), while in English, I thought, "there aren't any" is normal and "there are no" carries extra emphasis.
Guy said,
September 25, 2016 @ 4:34 pm
@David Marjanavić
"There are no Xs or Y's" would be normal in English too. I think the inclusion of "any" was motivated in part by the fact that the X phrase was a little heavy and in part because the X and Y were two very different things, not things whose absences were relevant for the same reason.
Also, I could be wrong, but my impression is that negation on the noun phrase is relatively neutral but negation on the copula expresses contrastive emphasis on the polarity (as if you are contradicting a suggestion that that there are Xs and Ys). But I agree if you wanted to emphasize the completeness of the absences, rather than polarity, you would negate the noun phrase(s) and put stress on the "no" (instead of the nominals, where stress would lie in a more neutral case). But in that case it would probably be "no Xs and no Ys", not "no Xs or any Ys".
FM said,
September 25, 2016 @ 4:58 pm
@TR: I think it would mean "Every telling without acknowledging them is either complete or accurate (but not necessarily both)."
AntC said,
September 25, 2016 @ 5:02 pm
I never thought I'd say this, but after so much linguistic Trumpery, I'm getting all nostalgic about plain down-home Bushisms.
Coby Lubliner said,
September 25, 2016 @ 5:44 pm
@David Marjanović: German es gibt keine X oder Y is a very different construction from the English "there are no…" since keine is clearly an adjective modifying both X and Y and is the equivalent of English "not any"; "no" is not so clear.
Coby Lubliner said,
September 25, 2016 @ 5:47 pm
Beating about (the) Bush: I was overly literal about "native"; a personal thing with me because I am a primary but not native speaker of English.
AntC said,
September 25, 2016 @ 7:24 pm
@myl But I think that there's a stronger negative-concord residue trying to diffuse into "standard" English than these works suggest.
I agree. Is it the influence of non-standard varieties: Hip-hop/gangsta rap/etc? But if so, how come the same AAVE influences didn't 'stick' from the jazz era?
Thank you for the Kallel references. That seems dubious about the claim that it was the influence of Latin-inspired peevers (with their complaints about the 'illogicality' of double-negatives) that squished out [NC] and [MN] from early/middle English.
Am I right in thinking that French is [+NC]? ne … pas ; ne … jamais ? How come Latin didn't have that same influence?
Jerry Friedman said,
September 25, 2016 @ 8:14 pm
AntC: Many nonstandard varieties of English have NC. John McWhorter has written, "Double negation is par for the course in nonstandard dialects all over Great Britain." I believe the same is true elsewhere.
Guy said,
September 25, 2016 @ 8:26 pm
I suspect that the transition from "mixed" to "no negative concord" in standard English is not really accompanied or driven by a substantial change in the prevalence of negative concord among speakers. Imagine if it became universally agreed in the English-speaking world that rhoticity should be the only acceptable standard for pronunciation in formal communications addressed at the general public but the non-rhotic dialects do not disappear, but are merely stigmatized.
AntC said,
September 25, 2016 @ 10:51 pm
@Jerry Friedman, aha! thank you.
Yes, ain't … nowt … rings a bell from living in Yorkshire — Leeds/West Yorkshire (just over the border from Farnworth/Manchester that McWhorter quotes), not so much York/posh Yorkshire.
Do I also remember an ain't … [NC] construct in Peter Wimsey's idiolect? (Fiction, so not reliable.)
AntC said,
September 25, 2016 @ 11:24 pm
The Giannakidou paper (thanks Mark) answers my q about French:
We can conclude then, that patterns with French n-words and ne are patterns of negative spread, which squares with the observation that ne can also be dropped. French appears to be unique in that it employs only negative spread and strictly excludes NC proper.
The paper reports quite a variation in the forms of [+-NC] across Romance languages. Which doesn't square with the blame being laid on Latin for [NC] disappearing from Standard English. Furthermore Germanic languages are strongly [-NC], Which doesn't square with Old English being strongly [+MN]. I wonder how it was in Norman French?
Seems there just ain't no easy answers in Linguistics.
Joe said,
September 26, 2016 @ 7:24 am
Is NC something that Southern dialects (or its Texas variants) share with AAVE? Could this perhaps be a deliberate show of solidarity via language?
Coby Lubliner said,
September 26, 2016 @ 9:52 am
@AntC: I thought that "negative spread" was a banking term. Does it refer, here, to the use of ne with intensifiers such as pas or point, or such words as aucun or personne, which eventually acquire a negative meaning of their own?
J.W. Brewer said,
September 26, 2016 @ 2:34 pm
I think fwiw that the sentence in question is otherwise written in a sufficiently formal register to make negative concord seem jarring rather than folksy. You'd need to make various other changes (or just start afresh) if you wanted to convey the whole concept in a sentence sufficiently folksy-and-vernacular that it would be churlish to point out the double negative. So that suggests it's merely a mistake. Now, obviously there are a lot of formal-register examples in the LL misnegation archives, and whether subconscious interference from the different syntactic rules of a less formal register is one causal factor in these situations (rather than just more or less random confusion by our overloaded poor monkey brains, where once the structure reaches a certain level of complexity there are basically 50/50 odds of being lucky enough to get it right) I suppose remains an open question.
Guy said,
September 26, 2016 @ 3:28 pm
@Joe
Cladistically speaking, AAVE essentially is a Southern dialect, which is why it shares many features with white southern dialects in terms of pronunciation, grammar, and lexicon. (Non-rhoticism, negative concord, "ain't", "britches", etc.)
David Marjanović said,
September 26, 2016 @ 3:45 pm
NC, up to and including es gibt keine … nicht "there ain't no" and wenn es keine … nicht gibt "if there ain't no", is commonplace in southeastern German dialects.
AntC said,
September 26, 2016 @ 3:55 pm
@Joe, think about it. At what time period would this sense of solidarity apply: between slaves and their oppressors; or ex-slaves and their ex-oppressors? NC was always firmly established in AAVE. There are only weak analogues to NC in African languages, so it don't come from nuthin' there. And Mark is saying it has in fact always been present in historical English; has never thoroughly died out.
[What's happening in English is in line with something called the Negation Cycle. Which is an hypothesis from Jespersen 1922. I am (again) gobsmacked by Jespersen's breadth of reach and intellectual audacity.]
The explanations that Mark links to are fascinating. (See also the McWhorter link from Jerry F.) In brief: [+NC] has probably never died out in non-standard English. The English that the slaves would have observed came as much from regional/dialectal variants as Standard English, and would have shown [+NC].
So AAVE, and Appalachian speech, and some US Southern varieties represent English frozen in time. Not nobody expressing solidarity with no-one.
AntC said,
September 26, 2016 @ 4:12 pm
@Colin Lubliner: I thought that "negative spread" was a banking term. lol
Does it refer, here, … It's messy (and my knowledge of French isn't up to understanding the nuances), so read the Giannakidou for yourself. Also French seems to be in flux with ne probably dying out — it certainly gets swallowed in rapid/colloquial speech. [ne disappearing is also in line with Jesperen's Negation Cycle.]
"negative spread" means that ne … pas ; ne … rien ; ne … personne ; ne … jamais are fixed frames. Ne cannot appear independently; and must appear in a very specific position just before the verb. Contrast English's freedom of positioning not/ n't … any.
such words as aucun or personne, which eventually acquire a negative meaning of their own? Exactly [the negation cycle] words acquire a purely negative meaning; so then the n-word [**] becomes superfluous and disappears.
[**] Yes I did a double-take the first time I saw that; but it seems to be the mot juste.
AntC said,
September 26, 2016 @ 6:18 pm
@David M …[example of NC] commonplace in southeastern German dialects.
Hmm interesting — and messy, no surprise.
By 'Germanic', Giannakidou is surveying chiefly German, Dutch, (Standard) English. He also looks a little at West Flemish, Afrikaans — where the situation is less clear. Then maybe German and Dutch regional dialects are preserving older forms, as with English? So perhaps the whole Germanic family is going through a Negation Cycle? Need more data! What have Linguists been wasting their time on for the past 50+ years since Jespersen got on to it?
G says: Although there is a clear divide between languages that employ NC as a standard structure, and languages that do not, we should note that even languages that don’t have NC may allow it occasionally, e.g. Dutch , and German ; e.g. Je hebt NOOIT GEEN tijd voor mij ‘You never have time for me’, from Dutch. These cases are admittedly quite marginal, and have a clear emphatic intonation. With this precaution, statements like “a language has NC” should be taken to mean “a language employs NC as a standard structure”.
Robert Furber said,
September 27, 2016 @ 3:31 pm
@Guy
I fail to see anything specifically "southern" about "britches". It is a phonetic rendering of the standard English pronunciation of breeches, /ˈbɹɪtʃɪz/. The pronunciation /ˈbɹiːtʃɪz/ is a spelling pronunciation of recent origin.