In the course of my editing this afternoon I hit the word, or rather non-word, incalcitrant. I thought perhaps it was some arcane medical term, but a little googling convinced me it was simply an error for recalcitrant; furthermore, there was an entire webpage about it (created by Jeffrey K. Parrott, Ph.D. candidate in linguistics, member of Punks in Science and, if that’s actually his picture, a fellow hat-wearer):
Summarizing, naïve Modern English speakers encounter the opaque recalcitrant, with the meaning ‘stubborn, etc.’ They semantically reanalyze -calcitrant as meaning something like ‘cooperative, etc.’ and replace the semantically unrelated prefix re- with the semantically relevant negating prefix in-. This yields the fully transparent new word incalcitrant, meaning ‘stubborn, etc.,’ literally ‘not cooperative, etc.’ The reanalyzed derivation of incalcitrant seems clear enough, and it should dispel any notion that English speakers are stupid or ignorant. The successful reanalysis of recalcitrant requires sophisticated (unconscious!) knowledge of morpho-lexical semantics, morphological constituency, morphotactics, and morphophonology (this latter because English speakers never mistakenly use the wrong variant of in-, e.g. *ilcalcitrant, *ircalcitrant, *imcalcitrant).
One question remains: why is the less productive prefix in- used instead of the fully productive prefix un-? That is, why don’t we ever see *uncalcitrant (a Google search on this non-word brought back one single result, compared with 193 results for incalcitrant)? The answer is simple, but has a fascinating implication. As noted above, the prefix in- attaches to Latinate vocabulary. Because -calcitrant is a Latin root, it will be negated with in- and not un-. But that means that naïve Modern English speakers have unconscious knowledge about the Latinate/non-Latinate distinction in their vocabulary items! They retain this knowledge in spite of the fact that the Latin meaning of -calcitrant is not only lost, but changed in the reanalysis. So speakers of Modern English are much smarter than they are portrayed by prescriptivists and their ilk. The English language is in no danger of decay, whatever that would mean.
Elsewhere on the morphological-analysis front, Erin O’Connor has a post about the English reanalysis of the plural tamales as tamale + -s, creating a new singular tamale to replace Spanish tamal; she compares “those crazy Latin and Greek words like stadium/stadia and octopus/octopodes, which may have English pluralization rules.” As I said in her comment thread:
I used to have the same reaction to “tamale,” but then I relaxed and accepted that it’s the English singular, while tamal is the Spanish singular, and there’s no more point trying to get English speakers to use it than there is trying to get them to say Ciudad de México instead of Mexico City. (Also, note that even the Greeks and Romans often got the declined forms of octopus and other -pous words wrong according to Justin the Mad Latinist.)