SC readers know that Andrew Sullivan is often a source of material (here, here, and most recently, here). Your host tries to avoid discussing Mr. Sullivan's (he's entitled to Dr., but SC hasn't seen him insist on it) views in this forum; our purposes here are largely linguistic. This one's a beaut.
In his latest article for the Sunday Times of London, AS attributes a coinage to Mickey Kaus:
Here's a word that deserves to be entered into the political lexicon. The blogger Mickey Kaus coined it. It's "pandescender."
As Mr. Sullivan explains it, the word is a combination of "pander" and "condescend", and is a verb meaning to simultaneously pander and condescend. In the usages of both Sullivan and Kaus, it accepts the suffix "-er" to derive the noun "pandescenderer", meaning "someone who both panders and condescends". Since the plain verb form contains the string "-er" itself, we must assume that it derives from one of the words that was used to make the compound, and thus that Mickey Kaus' grammar (and Andrew Sullivan's, since this doesn't bother him) features a productive infixation rule, like so:
pander + condescend -> pan- + descend + -der -> pandescendder
Since English does not generally permit geminate (for nonlinguists, "long") consonants, we may also assume that some phonological process reduces the output of the rule to "pandescender".
Now, it is true that most English speakers actually allow at least some infixation, even though it's not taught in schools. SC still remembers his very first linguistics class, where the professor demonstrated this fact by uttering "in-f***ing-credible". Let's see if we can take a stab at characterizing the morphological process that allows this (note: the following analysis is less serious than SC's usual standards of rigor, such as they are).
Note that in both cases, the infixed word is inserted after the first syllable of the infixed word. So there's a positional effect. It's hard to identify a clear phonological rule; while place of articulation remains stable at the infixation point for "pandescend", there is no common phonological feature at the junction of "in" and "f***". So maybe it's just a positional issue. Let's try a few new coinages, trying to keep things in the same domain of lexical semantics to eliminate effects there:
Based on the obliviousness of the average cell-phone user to the outside world, how about "chatter" and "ignore"? "chatignorter". Maybe just "chatignorer". Or maybe "igchatternore". The first two sound too much like someone in the process of ignoring a chat, which is the one thing cell-phone users don't do; meanwhile "igchatternore" has a certain onomatopoeic quality to it that your host rather likes. It sounds like something that would come out of the mouth of someone not paying attention.
Or perhaps from the combination of "mobilize" (which Merriam-Webster gives as a synonym for "drive") and "imprecate" (a nice way of saying "curse"), we can fashion a useful description of what people do while stuck on freeways during rush hour. "moimprecatebilize" probably violates most people's intuitions about the sounds of English, never mind the grammar, but "immobilizeprecate" felicitously adds the sense of being immobilized.
It seems abundantly clear to your host that infixation is actually capable of being a quite productive phenomenon in English. Unlike other languages, which make use of morphemes with well-defined syntactic and semantic properties to perform infixation, it only really serves to create compound words, perhaps in a more emphatic manner than merely joining them end-to-end. SC rather likes this idea, and will continue to try his hand at inblatherfixating.
(Edited to correct spelling of "pandescender", as per reader Russell's comment, at 5:46 a.m. on 3/8/04)