Most Latin words in -us have plural in -i, but not all, & so zeal not according to knowledge issues in such oddities as hiati, octopi, omnibi & ignorami …
H.W. Fowler, A Dictionary Of Modern English Usage (1926)
Writing about the noun form of bogus recently made me think about nouns ending in -us, and how some of them have irregular plurals derived from their Latin origins: alumnus and alumni; cactus and cacti; stimulus and stimuli—and so on. What these words have in common is that they derive from Latin second declension masculine nouns.
My three examples above are still in fairly standard use, but not all English -us words that are derived from the Latin second declension customarily take -i in the plural. Here’s the Google Ngram for crocus and its two plural forms, for instance:
Pretty much everyone is saying crocuses, while croci has always languished. Other second declension nouns in the same category are callus/calluses, campus/campuses, chorus/choruses, circus/circuses, genius/geniuses and lotus/lotuses. Virus/viruses is something of a special case, since virus (“poison”) was a non-countable noun in Classical Latin (like “music” in English) so it had no plural form. And in the singular it seems to have behaved as a second declension neuter noun, for which plural forms were rare and irregular. So viri has never been an appropriate plural.*
For -us words that do take -i, there’s been a creeping trend towards regularization. Here’s the Ngram for hippopotamus and the two versions of its plural, for instance:
There has been declining discussion of the hippopotamus in the Google corpus over the last century, but we can see that the regular plural hippopotamuses started to edge out the traditional hippopotami at some time during the 1980s.
And the two plurals of nautilus fought a brief tussle in the 1960s, but nautiluses is now the clear winner:
You can still say nautili if you want to, of course. But there are some -us nouns in English that never take -i in the plural, except by mistake.
One reason for this is that not all Latin nouns ending in -us were second declension. Some were masculine fourth declension nouns, for instance, which signalled their plural form simply by lengthening the “u” in -us—the spelling remained the same. In English, nouns derived from the fourth declension take the conventional plural form. Among the commoner examples, we have: apparatuses, censuses, consensuses, foetuses, hiatuses, impetuses, linctuses, nexuses, plexuses, prospectuses, sinuses and statuses.
Then there are the occasional neuter third declension nouns ending in -us that have carried their own elaborate plurals into English: corpus/corpora, genus/genera, onus/onera, opus/opera. And should you ever need to talk about more than one Venus, (at a classical sculpture exhibition, perhaps) you should know that as a feminine third declension noun its plural is Veneres.
Octopus was adopted into Latin from Greek oktopous (“eight-footed”), and retained its Greek plural octopodes, which is also used in English. If you’re nervous of that one, octopuses is equally acceptable. Platypus, also from the Greek, might perhaps take platypodes, but standardizes instead on platypuses. Platypus was originally the formal genus name assigned to this Australian mammal, and it’s customary for all common names derived from “modern Latin” biological nomenclature to take a standard English plural—tyrannosauruses, colobuses, ficuses, acanthuses.
Another reason for a Latin-derived English noun ending in -us to avoid the -i form in the plural is if it’s not a noun at all in Latin. So we have bonus/bonuses (an adjective in Latin); ignoramus/ignoramuses (“I do not know” in Latin); omnibus/omnibuses (“for all”) and rebus/rebuses (“by things”). I’ll spare you a few other more recondite examples, but I can’t resist mentioning the marvellous pair mumpsimus and sumpsimus. Mumpsimuses are people who cling to erroneous ideas to which they’ve become accustomed; sumpsimuses are new ideas brought in to replace old errors. Both words derive from a story told by Richard Pace, the Tudor diplomat, in his book De Fructu Qui Ex Doctrina Precipitur (1517). Pace tells how a priest, illiterate in Latin, intoned “quod in ore mumpsimus” during the Mass instead of the correct “quod in ore sumpsimus” (“what we have received in the mouth”). When challenged on the matter, he declared, “I will not change my old mumpsimus for your new sumpsimus.”
While sumpsimus was good Latin (“we have received”), mumpsimus was of course not Latin at all. Which brings me to my final group of English nouns ending in -us—the ones that aren’t Latin. Many are simply nonsense sounds: wampus/wampuses, rumpus/rumpuses, ruckus/ruckuses, doofus/doofuses, goofus/goofuses. The old noun bogus/boguses, designating a counterfeit coin or an illegal coin press is (as I’ve previously described in detail) probably Germanic in origin, and related to the word bogy. The source of caucus/caucuses is obscure, but it may come from an Algonquian word cau-cau-asu. Grampus/grampuses is an old name for a porpoise (or for someone who puffs and blows like a surfacing porpoise) and comes from Old French graundepose, “big fish”. Surplus/surpluses is likewise Old French (“more over”). And hocus-pocus/hocus-pocuses is probably also nonsense, though it may have originated as a parody of hoc est corpus from the Latin Mass. But the correct version of that phrase is hoc est enim corpus meum (“for this is my body”), which rather strains the proposed etymology.
And then (sigh) there’s Prius. The car made by Toyota. For reasons best known to themselves, Toyota invited the general public to vote on a plural for this name—and they got what they deserved, ending up with the frankly daft Prii. Now, Toyota used to be a little incoherent about what Prius means—according to a now-removed page on their website it was “[d]erived from the Latin prefix meaning ‘to go before’”. They’ve since updated that to a more reality-based statement: “Prius is Latin for ‘prior’ or ‘previous’.” Actually, prius is either a Latin adverb (“before”, “sooner”), or the neuter form of the adjective prior (“first”, “previous”). The only way to wring a plural out of that is to use the plural neuter adjective, priora. Which, I think you’ll agree, is no better than Prii.
So. Repeat after me. These cars are Priuses. Priuses.
* And virii, concocted by the hacker community as a plural specific to computer viruses, makes no sense at all, since (if anything) it would have to be the plural of the nonexistent word virius.
Let me just write that again: IT MAKES NO SENSE AT ALL.
7 thoughts on “Latin Plurals: Nouns Ending in -us”
Fascinating and really interesting. Never thought about this much, I mean at all
Glad you liked it. I have a couple more on this theme in the pipeline, if I ever get around to writing them.
Thanks for this, very interesting and, almost completely, new to me post. Australian State Schools did not teach much, if any, Latin in the 1960’s/70’s.
The only thing I can add is that if hippopotami was good enough for Flanders & Swann it is good enough for me.
Latin was still pretty big in the UK in the ’60s and ’70s, although I opted for German instead, on the grounds that I was unlikely to find a pen-pal in Ancient Rome.
My later interest in Latin is a combination of recreational and self-defence – life as a medical student was much easier if you had some grasp of Latin and Greek etymology.
I had assumed that your medical life had given you an interest in Latin.
In my working life in the Public (Civil) Service the only contact I had with Latin was with a few Legal & Bureaucratic ones – :
Mutatis mutandis, Mens Rea, Actus Rea, Mandamus (there we go!) Ad Valorem etc.
Mandamus, “we command”, was one of those more obscure “not a noun in Latin” examples I omitted above. So mandamuses.
Another legal term, habeas corpus, “thou shalt have the body”, sports two plurals for different situations. Latin habeas corpora means that a single writ requires several people to be brought to court. But if several writs are being spoken of, the whole phrase falls into the “not a noun in Latin” category, and we have habeas corpuses.
Unfortunately for me I did have experience with mandamus in the real word a couple of times. . If a company considered that we were not enforcing the law, to their advantage of course, they would serve a “Writ of Mandamus” on us.