"Our government has wrong known"?
« previous post | next post »
Thursday evening, President Trump gave a speech about election security, and since it was only 25 minutes long, I listened to the whole thing rather than relying on summaries. I'll leave the political analysis to others — what struck me was a speech error that happened a little more than half way through, at about 14:49:
Tonight we're publishing a series of
previously classified U.S. intelligence community assessments
and other reports proving that our
government has wrong known these machines are
extremely exposed to attack.
Zeroing in on the relevant phrase:
Since I'm more of a phonetician than a psychologist, my topic is the difference in formant trajectories between /l/, the onset consonant of "long", and /r/, the onset consonant of "wrong", rather than the (contested) Freudian interpretation of speech errors. In the spectrogram above, you can see that the third formant of the onset consonant of the word "wrong/long" is clearly rising into the following vowel (click on the image to see a larger version).
In a pronunciation of "long", F3 would be expected remain well above F2 in the syllable onset, as in Merriam-Webster's audio for "long"
Or Wiktionary's ("general American") version of the same word:
Compare the pronunciation of "wrong" from Merriam-Webster:
Or from Wiktionary:
Because the previous word in the president's phrase was "it's", we would expect the coarticulatory effect of /s/ to reinforce the higher F3 — the F3 lowering that actually occurred is a reliable indication of a rhotic articulation. As with speech errors in general, it's possible that the President's onset articulation was influenced by /l/ as well as /r/, but (for me at least) the perception is clearly /r/.
There's an interesting morpho-lexical question about why long has an adverbial version while wrong doesn't, but that's a topic for another day.
Peter Grubtal said,
July 18, 2026 @ 10:57 am
We used to refer to phenomena like this as "lapsus linguae", but is there a subset of these which can be explained phonologically?
In the past many of these lapses often wouldn't have been recorded for posterity, but nowadays they can be analysed and pondered over indefinitely.
Trump is of course fair game for this sort of thing. The UK still just prime minister referring to the Gaza attacks famously called for the "release of the sausages" (meant hostages of course). It used to be considered gentlemanly to pretend not to notice such lapses.
Mark Liberman said,
July 18, 2026 @ 11:27 am
@Peter Grubtal:
(Psycho-)Linguists generally call these speech-error word substitutions "Fay-Cutler malapropisms".
Y said,
July 18, 2026 @ 3:06 pm
There's an interesting morpho-lexical question about why long has an adverbial version while wrong doesn't, but that's a topic for another day.
Don't get me wrong, I'm not doubting you, but…
Y said,
July 18, 2026 @ 3:37 pm
Is there a matter of comparing velar apples and coronal oranges? When I look at spectra of my own [l] and [lˠ], the latter has a distinctly lower F₂ than the former, much closer to F₁. Dark [lˠ] is what I hear in the dictionary samples too. I see the same thing when pronouncing a retroflex r (high F₂) vs. a bunched one (low F₂). The dictionary pronunciation of "wrong" appears to be of a front variety (retroflex?), and that makes it looks extra distinct from the velarized l.
Mark Liberman said,
July 18, 2026 @ 7:01 pm
@Y "Don't get me wrong, I'm not doubting you, but…"
This comment, I found it strange…
I don't think you made it clear what you meant, and you left me puzzled.
Mark Liberman said,
July 18, 2026 @ 7:02 pm
@Y "Is there a matter of comparing velar apples and coronal oranges? "
Please point us to recordings of your apples and oranges…
Peter Grubtal said,
July 19, 2026 @ 4:55 am
Coincidentally (or perhaps not?), there's an article today on the Scientific American site which sees these slips (under the heading "spoonerisms") as indicating how the brain processes speech.
Yves Rehbein said,
July 19, 2026 @ 7:24 am
Similar, though not the same: techo'gie (Katherina Reiche, German minister for economic affairs and energy)
Michael Watts said,
July 19, 2026 @ 8:36 am
The comment is in an annoying internet style where it's supposed to demonstrate your error by example. The implication is that wrong in Don't get me wrong is an adverb. That seems questionable to me, but the far more common expression do [something] wrong does make me think that wrong is an adverb. The use has broader application; consider the sentence He read it wrong.
There's still room for debate. The Saturday Night Live sketch "Hormuz Jeff" includes the line "Never do it normal". (Meaning 'never do it the normal way'; this is a good match for "do [something] wrong" meaning 'do [whatever] the wrong way' and "do [something] right" meaning 'do [whatever] the right way'.)
David Marjanović said,
July 19, 2026 @ 1:59 pm
I'm wondering if it could be a predicative adjective instead.
(Coming from German I'm in the worst possible place to tell – we don't distinguish predicative adjectives from adverbs. They're both distinguished from attributive adjectives in lacking endings, so they look like English adjectives…)
Y said,
July 19, 2026 @ 3:06 pm
@Michael Watts: Yes that was a snark. (Annoying it may be, but it has nothing to do with the internet. It was and is used in writing and in speech.)
I read don't get me wrong as 'don't interpret me incorrectly', an unambiguously adverbial meaning. Likewise you're saying it wrong or I was doing it wrong (as you mention). You can say I had him quite wrong or you're saying it a little wrong, which argue for the interpretation of the word as adverbial and not a nominal complement.
The OED has a number of examples of wrong considered adverbial (sense B). They range from Middle English to the mid-19th century, though the entry was "not fully revised" since 1928.
Mark Liberman said,
July 19, 2026 @ 7:34 pm
@David Marjanović: "I'm wondering if it could be a predicative adjective instead."
That's surely a plausible analysis, as the other examples I cited were meant to suggest (e.g. "…left me puzzled").
It's true that wrong can be adverbial in some other constructions, but that's like noting that cut can be a verb, and so it must be a verb wherever it's used.
Michael Watts said,
July 20, 2026 @ 4:22 am
Why? Are you suggesting that adjectives can't be modified by quite?
Michael Watts said,
July 20, 2026 @ 4:35 am
For get [someone] wrong, I am not comfortable stating an analysis of it; to me it is a syntactically opaque fixed expression. In particular it's noteworthy that get [someone] wrong and get [something] wrong mean very different things.
It's not obvious that the wrong in get [someone] wrong must be an adverb simply because there is no "getting" taking place and so it isn't straightforward to say that the word being modified is get.
We can reject a parallel to "leave me puzzled"; if I get you wrong, the wrong is describing me or an action of mine, not you or an action of yours.
languagehat said,
July 20, 2026 @ 6:42 am
It's true that wrong can be adverbial in some other constructions
You said "long has an adverbial version while wrong doesn't," which seems pretty unambiguous; nobody's claiming that wrong is always an adverb, but you very clearly claimed it couldn't ever be.
Mark Liberman said,
July 20, 2026 @ 12:44 pm
@languagehat:
A good point.
But wrong has adverb-like resonances only in a few cases — there are many contexts where it should fit, if it had a fully adverbial sense, but where it definitely doesn't:
He falsely claimed the prize / *He wrong claimed the prize
It genuinely arrived / *It wrong arrived
…etc…
And "*Our government has wrong known…" is such a context.
Michael Watts said,
July 22, 2026 @ 11:27 am
Is it that wrong looks like an adverb in those 'few cases', or is it that several contexts exist in which, despite the fact that you might expect an adverb, an adjective is used instead?
Here's a sentence:
He rented a room in the back of her nail salon for his legal business, which wasn't doing too hot.
I think it's legal in modern AmE. But hot clearly "should" be an adverb applied to doing, right?
How are you doing?
Not too good.
Here, it's easy to interpret "Not too good" as an ellipsed form of "I am not too good", which applies an adjective to a pronoun. But…
I'm not doing so good.
Here, it's more difficult.
Jonathan Smith said,
July 22, 2026 @ 8:33 pm
New trend is form -ly adverbs on the fly from -ly adjectives, so
Flora is a friendly friend / friendlily she fries the scones
Milo is a manly man / manlily he mans the phones
etc