Language Log

Back in 2008, Arnold Zwicky described a category of typos that he called  "completion errors":

…a "completion error", a typo that results you start writing or typing a word and then drift part-way in to another word.  I do this all too often with -ation and -ating words — starting the verb COOPERATING but ending up with COOPERATION, for instance.  And several people have reported on the American Dialect Society mailing list that their intention to type LINGUISTS frequently leads them into LINGUISTICS, which then has to be truncated.  (This discussion on ADS-L followed my typing "original Broadway case", with CASE instead of CAST, and commenting on it.) 

26 years earlier, David Rumelhart and Donald Norman used the term "capture errors" for this phenomenon ("Simulating a skilled typist: A study of skilled cognitive-motor performance", Cognitive Science 1982:

This category of error occurs when one intends to type one sequence, but gets "captured" by another that has a similar beginning (Norman, 1981). Examples include:

efficiency – > efficient
incredibly – > incredible
normal – > norman

President Joe Biden is known for having overcome a serious stuttering problem as a child — see e.g. "Biden’s Stutter: How a Childhood Battle Shaped His Approach to Life & Politics", or "Joe Biden's history of stuttering sheds light on the condition". It also seems clear that the techniques that he developed to overcome the problem are still present in his speech today, as I discussed in "Calling all linguists", 10/21/2023. My conclusion in that article, agreeing with others more knowledgeable than I am, was that the main effect is selective lenition, probably related to what are called "gentle onset" techniques.

But what's less clear is whether this effect is different in kind from things that happen in (almost?) everyone's speech.

In a YouTube video yesterday, Michael Popok explained the differences (in New York State law) among a "verdict", a "decision and order", and a "judgment", in the context of the latest stage of Donald Trump's civil fraud case. Those intricacies are an interesting aspect of the sociolinguistics of the law, but the topic of this post is Popok's word-exchange speech error at about 4:45:

uh it crosses the i's and dots the t's
sorry
dots the i's and crosses the t's

Joe Biden recently said "the president of Mexico" when he meant "the president of Egypt". A couple of days earlier, he said "Mitterand" when he meant "Macron". Of course this fed into the flurry about his age, which was both-siderized by references to Donald Trump's calling Victor Orbán "the great leader of Turkey" when he should have said "Hungary", saying "Obama" when he should have said "Biden", saying "Nikki Haley" when he meant "Nancy Pelosi", and so on. And there've been lots of references to similar substitutions by other public figures like Sean Hannity.

However, my focus in this post is not political or journalistic, though there's plenty to be said about both of those topics. Rather, it's a question of psycholinguistic terminology. Similar proper-noun substitutions are common — but what should we call them?

From Jon Stewart's 1997 interview with George Carlin (starting at about 1:17.6):

well- well uh to- to go backward with the question,
don't forget, what we do is oratory.
It's rhetoric.
It's not just comedy, it's a form of rhetoric
and- and with rhetoric, you- you look and you listen for rhythms,
you- you look for ways
to sing at the same time you're talking, and to go
[skat-like phrases, based on rhythmic patterns of /d/-initial syllables…]

…in New Zealand. Phil Pennington, "Analysis: National opposed bilingual road signs, so what does the evidence say?", RNZ 62/2023:

Analysis – Bilingual road signs send a signal – that the country values te reo Māori. But going bilingual was confusing and National would not support it, National's Simeon Brown told voters in blue-ribbon Tauranga recently.

Accusations of racism and a walkback by the party leaders followed. But what evidence is the choice to go bilingual based on?

Helpfully, finding the answer to that is easy. The answer Waka Kotahi is relying on is in a 39-page "research note" into international experiences and outcomes.

However, a quick scan reveals the answer itself is not as straightforward as some of the commentary on the debate has suggested – that it is a straw man.

In Maria Bartiromo's recent interview with James Comer (R-KY), there's an interesting speech error — "quid pro crow" for "quid pro quo":

Rep. Steube (R-Fla.) announces he’s supporting Trump pic.twitter.com/3bL2cBl9z2

— Acyn (@Acyn) April 17, 2023

By now I must have listened to Taylor Swift's "Blank Space" a hundred times.  The first fifty times I heard a crucial line in it as "Got only Starbucks lovers" or "Not only Starbucks lovers", and it was driving me crazy because I couldn't make sense of it.  Sometimes I forced myself to believe that she was saying "Got only starcrossed lovers", but that didn't make sense either.  Then, on December 4, 2014, I read Mark Liberman's "All the lonely Starbucks lovers" on Language Log, and I learned — much to my astonishment — that, according to the lyrics, she was supposedly saying — repeatedly in the song — "Got a long list of ex-lovers".  Still today, after listening to the song and watching the video countless more times, plus reading the printed lyrics, I hear her sing "Got / Not only Starbucks lovers", never "Got a long list of ex-lovers".

Thus I am simultaneously assailed by multiple Taylor Swift mondegreens and polyphonic earworms ("trouble, trouble, trouble; shake, shake, shake it off").

One conventional view of "disfluencies" in speech is that they're the result of confusions and errors, such as difficulties in deciding what to say or how to say it, or changing ideas about what to say or how to say it, or slips of the tongue that need to be corrected. Another idea is that such interpolations can serve to "hold the floor" across a phrase boundary, or to warn listeners that a pause is coming.

These views are supported by the fact that fluent reading lacks filled pauses, restarts, repeated words, and non-speech vocalizations. And as a result, (human) transcripts of interviews, conversations, narratives, and speeches generally edit out all such interpolations, yielding a text that's more like writing, and is easier to read than an accurate transcript would be. Automated speech-to-text systems also generally omit (or falsely transcribe) such things.

The result is a good choice if the goal is readability, but not if the goal is to analyze the dynamics of speech production, speech perception, and conversational interaction. And in fact, even a brief examination of such interpolations in spontaneous speech is enough to tell us that the conventional views are incomplete at best.

I've noticed recently that automated transcripts from rev.ai do a good job of transcribing ums and uhs in English, though repeated words are still omitted. And in the other direction, I've noticed that the transcripts on the site of the U.S. Department of Defense include (some of the) repeated words, but not the filled pauses.  It's interesting to compare those transcripts to the audio (where available) — I offer a sample below.

Ruth Blau sent a link to a law firm's page on the "Difference Between Judges and Magistrates", which was probably created in response to the role of a magistrate in the recent FBI search of Donald Trump's Mar-a-Lago estate.

The linguistically relevant bit is the substitution of "illicit" for "elicit":

I was puzzled for a while by a interesting error in yesterday's The Hill. A story by Jared Gans, under the headline "What Weisselberg’s guilty plea means for Trump", ended like this:

Weissmann said defense counsels requesting coverage in a plea agreement for other crimes that may have been committed is “standard,” so someone knows “there’s nothing waiting in the wings.”

He said its exclusion from the agreement is “striking” and makes him believe Bragg more when he said the investigation is ongoing.

“That made me think that we all need to sort of take a deep breath and wait to see what happens after the Trump Organization trial, and so whether other churches get brought,” Weissmann said.

It's been a while since my last Breakfast Experiment™, but a conversation yesterday spurred me to run a simple data-analysis script with interesting results, presented below. The script and the results are simple, but the issues are complicated — consider yourself warned.