An alert reader, distressed that I used the word “octopi” in my New Republic article, has clipped out the proper term and mailed it to me. Here is the answer:
Octopus. Because this word is actually of Greek origin—not Latin—the classical plural is octopodes. But the standard plural in American English and British English alike is octopuses. Still, some writers [like me!] mistakenly use the supposed Latin plural.
You are both wrong. It’s “Octopussy”. Are you not familiar with Dr. Moore’s work?
Whoa! Don’t let PZ find out about this!
It depends how many there are. 2 x octopus = decahexapus ; 3 x octopus = isocatetrapus. 😉
Next you’re gonna tell me that it’s hippopotamuses, not hippopotami.
Curses, learned something new today. 😛
Ah, the perennial Debate. It seems clear that “octopi” is just wrong. I still have my doubts about “octopodes” as well. I am no classical scholar (!) but it seems to me that we don’t want the plural of pus/foot; that would mean “8 feet” which is true of a single octopus. Therefore we want to say something more like “8-footed animals.” I always go with “octopuses” for that reason.
Main Entry: oc·to·pus
Pronunciation: \ˈäk-tə-pəs, -ˌpu̇s\
Function: noun
Inflected Form(s): plural oc·to·pus·es or oc·to·pi \-ˌpī\
Etymology: New Latin Octopod-, Octopus, from Greek oktōpous
Date: 1758
Merriam Webster likes octopi?
The Greek plural is “octopodes” (or, more accurately in transliteration, “oktopodes”) but the Greek singular is “octopous” (or “oktopous”), which we don’t say. It doesn’t really make sense to use a Greek plural when we don’t use the Greek singular. We’ve borrowed the word, in altered form, into English, and so we may as well treat it as we do most words we’ve borrowed (not that we’re going to give them back, anyway): by using English inflectional morphology. So “octopuses” is the most defensible pick. The OED lists all three (puses, pi, podes) as options, and “octopuses” first. (“Octopi” is certainly a reanalysis, since the “us” isn’t actually the Latin masculine singular affix, but it’s hardly the first reanalysis to gain a toehold.)