Language Log

Funny how people talk ways they claim they don't; I've even had them deny stuff I've recorded (or even shown them on spectrograms).

Well, not too funny when you consider the reasons.

Probably everybody who's collected data systematically or taught a course in which variation played a role has come across this phenomenon.  Does anyone have a name for it?

(A related phenomenon -- Do As I Say, Not As I Do -- is familiar to anyone who's looked at the advice literature.  An adviser will sternly proscribe some variant -- restrictive relative which, for example, or logical rather than temporal since, or a pronoun with a possessive as antecedent -- and then use it, often within sentences of the proscription's formulation.)

If we want a suitably scientific-sounding label, we could base it on anosognosia (coined by Babinski in 1914, as French anosognosie), which the OED defines as 'unawareness of or failure to acknowledge one's hemiplegia or other disability.'  (It's usually the result of right brain injury of some kind.  In my partner Jacques's case, the cause was radiation.)  The word has the parts:


We can then replace the Greek stem noso- with some more appropriate one, like praxi- 'performance, practice'.  Ta-da!

I've found over and over again that people who have been surreptitiously tape recorded in sting operations often refuse to admit that they said the inculpatory stuff on the tape in spite of what the recording clearly shows that they said. They usually claim that the government tampered with the tape. In rare occasions, the tape has been tampered with, but 99% of the time it hasn't.

This looks like something a bit different from Dennis Preston's observation, where people don't seem to believe that they sound the way they were tape recorded (usually their phonetics, I assume, although perhaps grammar as well). In the sting tape cases, the person denies the CONTENT of the statements, claiming that they never could have said such things.

Your suggested label, apraxignosia, seems to fit both categories.