Saturday, July 31, 2010

One millennium, two ...

I have this notion, which is admittedly somewhat idiosyncratic, that when foreign words join the ranks of English, they don't get to bring along all their foreign-word baggage -- their plurals, their conjugations, or anything else that would require an English speaker to somehow have to intuit the rules for that word in its original tongue. No, because the word has joined English, which already has rules for these sorts of things, and in the fine print of the contract that the word signed upon joining the language, it explicitly says that it must abide by house rules.

Nonetheless, people get theyselves worked up these things. Not terribly long ago, I essentially laid all this out when someone asked what the plural should be of the car known as the Prius. Easy, sez me: in English, the default way to make plurals is to add -(e)s, so, hey, Priuses. All this fancy talk about what the plural would be if this were a Latin word is just piffle, because we're not talking about Latin, we're talking about English.

Ask folks what the plural is of octopus, and yer more edumacated types will tell you it's octopi, because something-something-Greek-something-something. Or cactus, or focus, or schema, or -- heh-heh -- opus.

Now, I would not say that it's wrong to say that the plurals are octopi or cacti or foci or schemata or opera. But I will also staunchly defend octopuses and cactuses and focuses and schemas and opuses. Because these latter words all follow normal rules, and because if I were a native speaker of English, I would find those to be perfectly reasonable ways to make the plural forms. (Hey, wait, I am, and I do!)

Ok. Surely among the more conservative publications in terms of these things is the New York Times, yes? Yet behold this thing that they have printed:

(I make you a picture in case some editor at the Times get a gander at this, haha.)

Surely this is unusual for a publication, is it not? Google reports a mere 425,000 hits for this term, versus around 6 million for the term millennia (where the plural ends in -a because something-something-Latin-something-something).[1]

Having read this far, tho, you will not be surprised to hear me say that I cheer the NYT for this. I'm sure they'll get letters from folks who want us all to know Latin declensions, but fie on those people. And I say to the word millennium, welcome to English, after all these years!

[1] This is not to be confused with the Mazda Millenia (sic), a singular car with a plural name, which we forgive because marketing has yet another set of rules, among them being "let's not let grammar get in the way of a good brand."

Update 1 Aug: Apparently I had a thought about millenniums in an earlier post. I guess this obsesses me, I just keep forgettting that it does.

Update 1 Aug: Per Ben, fixed all the spellings, oops. :-) This actually upped the count for the non-native plural (millennia) to more like 10 million.

Update 5 Oct: Mark Liberman posts on the Language Log about the plural of syllabus. Interesting comment in the thread from Henning Makholm speculating a bit about the use of -i as a productive plural marker in English, not just a fossil on foreign borrowings.


Malti said...

Ask folks what the plural is of octopus, and yer more edumacated types will tell you it's octopi, because something-something-Greek-something-something.

"octopodes", surely?

And I find the attitude that irregular plurals are somehow un-English interesting. I presume you also stand by the use of "childs", "foots", "mouses", "sheeps", "womans", etc?

I suspect even if English-speakers were to collectively shun "-um" > "-a", "-us" > "-i", etc. plural rules and make all borrowings perfectly regular we'd still end up making new irregular nouns from words we already have, as we do with verbs. No sooner as we'd regularised "melt"s past participle to "melted" from "molten", and suddenly "sneak" has decided it wants to be "snuck" in the past-tense rather than "sneaked". It'd probably be the same for other word classes. So you can have "Priuses", "cactuses", "octopuses", but it just brings us a step closer to the plural of "house" becoming "hice".

Ben Zimmer said...

Apparently I had a thought about milleniums in an earlier post. I guess this obsesses me, I just keep forgettting that it does.

Do you also keep forgetting about the second n? :->

WordzGuy said...

@Malti, you'll note that I didn't say that I think irregular plurals are wrong. (In fact, I explicitly said that they're not.) What I'm saying is that when a word is borrowed into English, it's perfectly fine for it to use English morphology.

All the terms you note are in fact native Anglo-Saxon terms that were not borrowed into English; they came by their plurals honestly, via plural makers that happen no longer to be productive.

As for "octopi," RHD seems ok with that:

And in fact, this illustrates the point -- I have no idea how to decline octpous in Greek, and I don't actually care, because we're not talking about Greek.

@Ben -- I dunno, there are an awful lot of m's and n's in that word. A body gets tired of typing them all, ya know? :-)

Malti said...

I know all the words I mentioned are Anglo-Saxon ones, that was my point. So why insist on the "s" in new words? Why can't the plural of "cactus" be "cectus", or "Prius" be "Prisen"?

("octopi" is accepted, but I'd have thought not by "edumacated speakers" who'd mention "something-Greek-something", because referring to them as "octopi" requires the misapprehension that the word would form it's plural like in Latin.)

Alon said...

@Malti: "Why can't the plural of "cactus" be "cectus", or "Prius" be "Prisen"?". Because English users have not gravitated towards any of these forms, and this is unlikely to change in the near future.

While English speakers occasionally devise pseudo-strong forms for formerly regular verbs (sneak, which you mention, being a notable case, as is dive), this is far less common for nouns. Other than the humorous boxen that geeks like to use for (the specialised meaning of) box, the -en plural of OE weak nouns has long been improductive, and the same applies to ablaut. Some pedantic confusions have stemmed from unetymologically applying the rules for the Latin first, second and third conjugations to nouns not derived from that language, but that's it.

arnie said...

Your difficulty in spelling millennium, let alone knowing its Latin plural, isn't surprising to me. Around 1999 I was writing a Web site which included information about the Millennium Dome, being built at the time in Greenwich, London as part of the Year 2000 celebrations.

I started getting feedback that it was difficult to find my site through the search engines and discovered that many people couldn't spell millennium to save their lives. Search engines then were not as sophisticated as today's, and spelling a word incorrectly meant that the search would almost certainly fail. I added keywords mis-spelling "millennium" and traffic improved by around 20%.

Syz said...

Amen. Can we add "corpuses" to the list? It would replace the infidel plural "corp%#@" so adored by writers of linguistics thesises (ok, that's a bit ugly).