My father, a psychiatrist, puts a prudent amount of faith in how meaningful a substitution error is in speech. He doesn't believe a speech error tells us what you're thinking. But it does tell us what you're thinking about. If you slip up in your wedding vows and say "for better or for hearse" or "in sickness and in hell" there's something on your mind. Who knows why it's on your mind but it is.
I agree to an extent. But I also trust that there's a reasonable amount of linguistic gravity, even if it's largely phonological. And frozen or reinforced phrases easily add mass to these bits and pieces of linguistic matter. Such a phrase is one that everyone has heard and everyone uses as a single unit. Often idiomatic, sometimes perfectly grammatical. Think of the Match Game. I say 'Keep off' and you say 'the grass.' I say 'Buy one' you say 'get one free.'
It's also possible that the bare structure of a phrase is planned before it's spoken and some flourishes are added on the fly. So when a bare phrase is similar to a common phrase it can take the familiar form. That's probably what happened to Pat Buchanan while talking Rachel Maddow on her new MSNBC show. The topic was Sarah Palin. Isn't always these days?
Buchanan: The press threw an apocalyptic fit on this thing--an apoplectic fit if you will--when she was announced. And I think she was treated horribly.
His bare phrase is The press threw a fit. Before he says it he decides to add an adjective. What adjective would indicate a loss of composure and control. How would you indicate that the press was confused and slightly crazed? Well he can't go there of course. But it's fair to say that epileptic likely came to his mind pretty quickly. He avoids it and instead called it
an apocalyptic fit.Well that one doesn't really capture it. He's going for an adjective that captures the behaviour of the press, not the historical or global implications. So he reaches into the bag again. This time he pulls out apoplectic, landing safely on a phrase that is out there. But his pronunciation of apoplectic provides a smidgen of evidence that epileptic is still rattling around his mind. The first vowel is pretty close to ɛ making apoplectic sound very close to it.
Now as I said, apoplectic fit is out there and it's attested. It even gets an early citation from 1611 in the OED. It's not as common as epileptic fit according to a quick Google™ search. 32K vs 124k. But as an object of the verb threw the numbers are much closer. Threw an epileptic fit: 90 hits. Threw an apoplectic fit 64 hits:
So what might have happened with Buchanan's statement is almost like a magnetic repellant from a dangerous phrase. He was thinking apoplectic, which is an appropriate phrase for what he's saying. But it's so close to a phrase he wants to avoid that he steps away from it and goes to apocalyptic which is fine but has an odd meaning. So he regathers and again approaches his original intention and goes juuuust a little past it to the first vowel of the word he was trying to avoid in the first place.
Now this is mapping out quite a visual path. And it would be nice if language worked so tangibly as to give our processes a reasonable space model for analysis. But I've probably already gone too far. And for too little.*
Mark Liberman has put up a few posts about such things recently.
*How about the rest of Buchanan's points? That's where the meat is.