What's the plural of syllabus?

« previous post | next post »

Reader A.T. writes:

When I can't sleep, I go onto TED.com. I'm watching a talk by Pinker and he says syllabuses at one point (about 15:36). Not sure if you've blogged about syllabuses versus syllabi in the Language Log, but I think it'd be a pretty cool topic to discuss.

I'd say "syllabuses" as well, though I can't recall the last time I had occasion to use the plural form.

The thing is, the word is a fake to start with, a misinterpretation due to scribal error. Here's what the OED sez:

So if we were going to be etymologically exact, the singular should be "sittyba" and the plural should be "sittybes", or something like that. Why should we invent a fake Latin plural to go with the fake Latin singular? My advice is to stick with plain English syllabuses.


  1. J. Goard said,

    October 4, 2010 @ 9:08 pm

    Various levels of regular/irregular inflections get to be the way they are through frequency-influenced analogy, anyhow. Why would one expect borrowing or historical error in a word's origin to permanently block any analogy but the single most regular one?

  2. Stephen C. Carlson said,

    October 4, 2010 @ 9:12 pm

    Looks like σιττύβα is first declension, so the plural should be σιττύβαι = sittybae.

  3. John Cowan said,

    October 4, 2010 @ 9:17 pm

    It's no more fake than ginkgo, and at least that has a regular plural.

  4. Niels G. said,

    October 4, 2010 @ 9:17 pm

    But it's not quite fake Latin. – Alcuin has syllabus, and Bede an acc. syllabum, so the o-stem is attested in mediaeval Latin, although in both cases it seems to be synonymous with syllaba.

  5. J. Goard said,

    October 4, 2010 @ 9:20 pm

    To follow up a little on that:

    -uses, at the end of a longish, non-Germanic-sounding word, feels rare and therefore strange. Meanwhile, the kind of person who says syllabus will probably have in their active vocabulary sufficiently many similar singular-plural pairs to establish the template -us/-i. The historical information is kewl, but I figure to keep on saying syllabi.

  6. Niels G. said,

    October 4, 2010 @ 9:26 pm

    Although, looking at the examples again (i.e. actually looking rather than just glancing), I suppose it's much more likely to be a u-stem:


    Nisi forte notaria manus verba syllabus vel literas inmutasset.

    Beda Venerabilis:

    N quoque littera pari ratione, ni fallor, cum in medio uerbo consonanti alteri fuerit subiecta, praecedentem syllabum siue natura seu positione semper longam habet, ut 'regna,' 'calumnia'.

    So you could make an argument for pl. syllabūs, just to confuse matters further.

  7. Twitter Trackbacks for Language Log » What’s the plural of syllabus? [upenn.edu] on Topsy.com said,

    October 4, 2010 @ 9:29 pm

    […] Language Log » What’s the plural of syllabus? languagelog.ldc.upenn.edu/nll/?p=2684 – view page – cached October 4, 2010 @ 8:58 pm · Filed by Mark Liberman under Morphology, Words words words Tweets about this link […]

  8. David Rivers said,

    October 4, 2010 @ 9:34 pm

    I have a running gag with a couple of my colleagues that you add the suffix -cum to make practically any Latin-sounding noun a plural (e.g., "abacus" becomes "abacum"). Not really sure where that came from to be honest! (Please refresh me, if you think you know which word(s) this might have actually derived from!)

    Anyway, I've been reading your Log for you a few months now, and have been enjoying it :)

  9. Henning Makholm said,

    October 4, 2010 @ 9:39 pm

    Sticking to "syllabuses" is certainly good advice, which I have no problem with. But what can we say descriptively about the OP's (apparent) preference for "syllabi"?

    It's easy to spin a nice condemning story about how certain groups are using Latinate endings in English as a shibboleth for having the proper education to fit into the cultural and economic elite, and so forth, blah blah blah. In this story it could add to the ridicule that the shibboleth the OP proposed is not just wrongly declined, it is not even Latin in the first place.

    But the very fact that fake Latinate endings are being applied to non-Latin words seems to suggest a more interesting inquiry. It means that those who expect "syllabi" cannot be relying on rote learning of Latin vocabulary (which is how a working shibboleth for having an actual classical education would work); they must actually be using rules to decide which words to apply which fake Latin endings to. What are those rules? Who tends to use them, in which situations?

    For example, could we hypothesize an emerging morphological rule in certain registers of certain English sociolects that, say, words that end in unstressed -us preceded by another unstressed syllable make plurals with -i? Probably that particular rule (which I just invented without thinking too deeply about it) would not withstand empirical testing. But in principle, can one imagine morphological rules like that? Or is that just silly?

  10. D Nakassis said,

    October 4, 2010 @ 9:41 pm

    Ah, but LSJ has the following entry: "σίλλυβος, o, parchment-label, appended to the outside of a book, Cic. Att. 4.4a.1, 4.8.2, dub. in 4.5.3: pl. σίλλυβα, τά … (In Cic. Att. 4.5.3 the vv. ll. sit tybis and sic tu iubes are perhaps traces of sittybis; but σιττύβαι (q.v.) has an inappropriate meaning, unless it can mean leather case of a book)." Indeed, σιττύβαι are defined by the lexicographer Hesychius as leather garments. So it seems that the OED is probably wrong in this instance. Not that it really matters. I say syllabi.

  11. Army1987 said,

    October 4, 2010 @ 9:45 pm

    On the COCA there are 290 occurrences of "syllabi" and 20 of "syllabuses"; on the BNC there are 238 "syllabuses" and 11 "syllabi". So it's an AmE/BrE thing, it seems

  12. George Amis said,

    October 4, 2010 @ 9:57 pm

    OED actually gives syllabi as the first plural and syllabuses as the second. Of course, since the word is suspect anyway, we could act as though it's fourth declension masculine and say that the plural is syllabus.

  13. Jason Eisner said,

    October 4, 2010 @ 10:45 pm

    @Henning Makholm: When I was 8, I felt that the plural of "toothbrush" really ought to be "toothbri," although I wasn't entirely sure where that feeling came from — it bothered me that I didn't feel the same way about "brush" or "bus." You're presumably right that it's about stress.

    I don't think I could be accused of putting on classical airs in this instance, since I didn't go around actually saying "toothbri." (Although I did find it an amusing affectation to ask for some "Kleenices" when I needed to blow my nose …)

    I say "syllabi," by the way.

  14. john riemann soong said,

    October 4, 2010 @ 10:51 pm

    "uses, at the end of a longish, non-Germanic-sounding word, feels rare and therefore strange"

    no no it doesn't. /syllabi/ feels strange. so would "feti".

  15. J Lee said,

    October 4, 2010 @ 10:57 pm

    Can someone run the numbers on 'octopus,' a Greek word with an ending mistakenly thought to be Latinate? Maybe it's my youth, but 'octopi' and 'foci' sound preposterous.

    [(myl) The numbers as of 2004 are here.]

  16. J. Goard said,

    October 4, 2010 @ 10:58 pm

    @Henning Makholm:

    could we hypothesize an emerging morphological rule in certain registers of certain English sociolects that, say, words that end in unstressed -us preceded by another unstressed syllable make plurals with -i?

    I think it's certainly more complex that that, as you suggest. Alumnus and bacillus are penultimately stressed, and then there are locus, focus (in the mathematical sense) and cactus. However, it is apparent that the -us/-i pattern has extended beyond the borrowed forms from Latin second declension, based upon some kind of analogical relationships that are significant to English speakers. My argument is that this is not different in kind from applying the -s plural to borrowed nouns — it's just a question of whether the analogy that wins the day is a strong local one or a more distributed one.

  17. Disfraz said,

    October 4, 2010 @ 11:01 pm

    +1 for syllabi and, oddly, Kleenices. It also bugged me for a long time that the plural of paradox isn't paradoces. And on the basis of those, I'd support @Henning Makholm's view that it's a more or less regular overgeneralisation in my dialect. And no matter how hard you try to make me ashamed of it with phrases like 'fake Latin' and 'plain English', I won't be.

  18. GEW said,

    October 4, 2010 @ 11:15 pm

    Anecdotally, I tend to use an '-i' plural, whether or not it's appropriate etymologically, when a word is three syllables or more and ends in '-us'. (With the exception of 'prospectuses', for reasons I cannot currently fathom – maybe the rule Makholm suggested is correct?) With one- or two-syllable words, I'll use an irregular plural if the final syllable is unstressed.

  19. Niels G. said,

    October 4, 2010 @ 11:15 pm

    My gut (which, as attached to an ESL-speaker, should count for very little) also favours syllabi, don't think I've ever come across the -es plural before. Kleenices is new to me, but brilliant.

  20. Garrett Wollman said,

    October 4, 2010 @ 11:33 pm

    Certainly the application of "irregular" plurals to modern coinages is a long-standing feature of hackish, although only four are mentioned in my copy of AIWORD.RF[UP,DOC]: "FROBNITZEM", "FROBBOTZIM", "QUUCES", and "VAXEN". There's a line between playful overgeneralization and snobbery, but I'm not sure where it is, or even if we'd put it all in the same place. (I'll admit that "txen" is a part of my productive vocabulary, and I do tend to use the "Latinate" plurals to words that have both traditional and regular English -s plurals.)

    There's probably a connection here as well to the not uncommon preference for zero plurals for fictional nationalities, particularly when formed in -i or -in.

  21. J. Goard said,

    October 4, 2010 @ 11:36 pm

    MYL links above to an old Bill Poser post, which states that:

    octopi is neither the English plural nor the classical plural.

    This is precisely the kind of thinking that I find problematic: the idea that -us/-i is a "classical plural" in present-day English, or, to put it another way, that present-day English speakers (at least on a significant scale) have a rule that says in essence, "don't change borrowed forms X, Y, Z from how they work in the source language." My opposing view is that -us/-i is a construction of English, whatever this might entail for different individuals, sociolects, etc.

    Of course, Latin origin and a certain declension are the causal explanation for the words that originally seeded the construction, but the construction subsequently became a pattern of English. (In contrast with the vast majority of Latin morphological patterns, which have left no such effect on ordinary English.)

  22. David L. said,

    October 4, 2010 @ 11:49 pm

    Since some of the speculation here is about which form feels rare or unnatural, the frequencies seem worth documenting at first pass. Google searching yields about 628,000 results for +syllabuses, but about 17 million for +syllabi. COCA has 20 "syllabuses", but 290 "syllabi." While there are well-documented devils in the details of Google counts, it seems quite likely that, descriptively, syllabi is the entirely dominant plural form.

    You prescriptivists here at Language Log are, of course, still welcome to hand out your advice on what we all ought to say! :-)

  23. J. Goard said,

    October 4, 2010 @ 11:49 pm

    By the way, how does anybody think classical Latin declensions got to be the way they were? Carved in alabaster from the beginning of time?

    "Declension" is just a way of classifying analogical patterns, at varying degrees of generality, which exerted pressure upon similar word forms to assimilate, through the same means as has produced the allegedly incorrect English octopi. Moreover, it is the very same forces which produce octopuses, differing only in degree, i.e. the breadth of the analogy.

  24. Qov said,

    October 5, 2010 @ 12:17 am

    I'm going with the mongoose trick. "This year's syllabus and those of other years, too."

  25. Betsy McCall said,

    October 5, 2010 @ 12:26 am

    Thanks for the post on this. Syllabus is a word that has interested me and it certainly seems to have an origin that is just as interesting. I confess, I used to say "syllabi", but after looking it up in my Latin dictionary several years ago and discovering it's given as a -u- stem, mostly for the sake of being perverse, I guess–I was in a Classics program at the time–decided to switch to "syllabuses". Classicists were typically appalled that they had been doing it wrong all this time! :) As I recall, the plural of "virus" originally prompted our discussion. In any case, it's just another example of educated speech that gets internalized into a language and loses any connection to the original.

  26. Aaron Toivo said,

    October 5, 2010 @ 1:00 am

    I am with J. Goard on this. Most average people in the present day do not know enough about Latin morphology to make any conscious attempt to mimic it. They wouldn't even know where to begin.

    You do not need to know anything about any other language to use English well (a principle I believe I first encountered from Languagehat, ages ago), so Latin can have no bearing on the question of how to pluralize the English words octopus and cactus and syllabus. If we observe -i in substantial use by average people who are not trying to be playful when doing so, this must be treated as a fact about English morphology.

    As for the rule that sanctions it, the singular must be spelled -us and this syllable must be unstressed, but there is no requirement that the previous syllable also be unstressed, or else "cacti" would never work — instead, it's one of the few words that strongly prefer -i.

  27. Gordon Campbell said,

    October 5, 2010 @ 1:35 am

    One 'bus, two 'bi? Any Latin scholars out there with advice on this one?

  28. Gordon Campbell said,

    October 5, 2010 @ 1:38 am

    Or maybe I could just search the archives: http://itre.cis.upenn.edu/~myl/languagelog/archives/000814.html

  29. Adrian Bailey said,

    October 5, 2010 @ 2:29 am

    "-uses, at the end of a longish, non-Germanic-sounding word, feels rare and therefore strange"

    Not to me, but as has already been pointed out, the preference for "syllabi" appears to be leftpondian. I'd say syllabuses, omnibuses, prospectuses, radiuses…

    Once upon a time, when I was working for AOL, I had a hard time persuading my boss that the plural of virus wasn't virii or viri.

  30. groki said,

    October 5, 2010 @ 2:34 am

    @Garrett Wollman, @Aaron Toivo:

    intentional playfulness certainly is a legitimate source of new language–embiggen, for example.

    I enjoyed "taxen" (if that's what it is): nice chime. two favorites of mine: plural computers = "boxen"; plural a**holes = "ani."

  31. Jayarava said,

    October 5, 2010 @ 2:51 am

    It's an odd legacy left to us by the literati of bygone days, this desire to use Latin and Greek inflections in English. We have a perfectly good language, with a straight forward enough way of making plurals – one simply adds a z, doesn't one?

  32. Richard M Buck said,

    October 5, 2010 @ 2:59 am

    @Aaron Toivo:
    …how to pluralize the English words octopus and cactus and syllabus. If we observe -i in substantial use by average people who are not trying to be playful when doing so, this must be treated as a fact about English morphology.

    I mostly agree: certainly I'm happy to accept the historically-wrong octopi as 'correct' in modern English from a descriptive point of view: it is what people say, and I understand it, so it seems pointless to argue about it.

    However, it is also the case that some people — presumably those who, as you say, "do not know enough about Latin morphology" to get it right — feel a need to inflict unhistorical (and in some cases non-existent, from the point of view of Latin) pseudo-Latin-ish endings on innocent words in what appears to be a non-playful way. You may have come across **penii for penises, and **virii for viruses; and I have now come across **rectii for the first time. Every time I put my students through an online exam with a particular piece of software, I have to click through to a screen entitled "Assessment Statii". It's possible, I suppose, that this was intended as a joke; but I've always read it just as pretentiousness on the part of someone who felt "statuses" to be insufficiently classy.

  33. Jac said,

    October 5, 2010 @ 4:08 am

    As one who has endured many education policy conferences, I would say that the plural of syllabus is tedium.

  34. Pflaumbaum said,

    October 5, 2010 @ 4:44 am

    Re octopus, the 'correct' etymological plural would be 'octopodes', but anyone who says that deserves to be strangled eight times.

    I was recently discussing number in latinate words and foreign loan-words with a bunch of friends (yes it was a terrific party), and several of them claimed to use the terms 'fora', 'amaretti' and even 'pizze' in normal conversation. One of them claimed 'graffito' was natural to her. I was sceptical, but maybe they really do say these things. I would use 'fungi' and maybe 'cacti' and a few others, but to do it generally as a matter of principle seems really pretentious. Or am I just being a reverse peevist?

  35. eriphania said,

    October 5, 2010 @ 5:06 am

    @Jayarava indeed. I haz many cheezeburgerz…

    English of course has many ways of pluralising words, of which -us to -i is merely one. The English rules for selecting which is correct are, um, not trivial (no-one ever codified them for me as a child, I merely had to learn by rote which was proper).

    Many people play with using the 'wrong' pluralisation for comic effect, or simply because they like the sound of the resulting word better. You might dislike but would you misunderstand virii, mongeese, boxen, sheeps, spice … some of these words are even in common use in some sub-cultures.

    [For those wishing to assert that "us to i" is "only for Latin words" I recommend "octopodes" as a very easy (fake) Greek plural or "geckogecko" (we get gecko from Indonesian)]

  36. army1987 said,

    October 5, 2010 @ 5:14 am

    @David L.: but in the BNC "syllabuses" predominates. See my post above.

  37. Matthew Kehrt said,

    October 5, 2010 @ 5:37 am

    In the subsubsubfield of computer science I used to study, the term "reducible expression" is abbreviated to "redex". I do not think I have ever seen it pluralized as anything but "redices". I once had a professor, in lecture, tell the class that this was how it was pluralized.

    Similarly, with varying degrees of seriousness, I have heard this pattern applied to other CS abbrieviations: "regices" (for regular expressions) and "mutices" (for mutual exclusion locks).

  38. Pflaumbaum said,

    October 5, 2010 @ 6:11 am

    I might be wrong, but I think that's a slightly different thing Matthew. In your case you have people coming from a mathematical background dealing with neologisms, rather than existing words with commonly used, non-latinate plurals. I reckon your professors are working (albeit playfully) by analogy with familiar words like 'matrix' and 'index'. That kind of analogical formation could even end up becoming productive in English.

  39. Aaron Toivo said,

    October 5, 2010 @ 6:32 am


    Language play is absolutely a legitimate source of new language, yes. But there's a difference between play, which is always deliberate, and language as you would use it without being conscious of it. Only when something enters the latter category can we truly say it part of the language.

    @ Matthew Kehrt:

    Don't forget unices and linices!
    My personal favorite language play with this one is "defenestratrices". That unpacks to 'defenestrate' (to kill by pushing out of a high window) + feminine agentive -rix + plural modification to -rices. "Women who kill others by pushing them out of high windows". I've waited for ten years now for an excuse to use it.

  40. Acilius said,

    October 5, 2010 @ 6:45 am

    @ D Nakassis: When I was in grad school one of my fellow students spent a few weeks looking into the origin of "syllabus," reaching the conclusion that the likeliest etymon was sillybos, sillybotos. He didn't publish his findings, unfortunately, or I would link to them. Though to this day when I hear "syllabi" I suppress an impulse to mock-correct it with "sillybotes."

    It may not be as silly for us to use Latinate plurals in English today as it was for Alcuin, Bede, and other medieval types to break with classical practice and decline Greek words in Latin texts. But I still object to it. Latinists may chuckle the first time they see a form like "syllabi" or "Kleenices," but will soon weary of the humor; and in other ears they are simply oddities.

  41. zoetrope said,

    October 5, 2010 @ 7:06 am

    @Pflaumbaum: Your parties sound a lot like the ones I go to, but in any case, 'amaretti' sounds perfectly natural to me, at least if we are talking about the little cookies. Would you say 'amarettos' instead? One usually encounters amaretti in a group, in fact, usually in a package with the word 'amaretti' on the front, so maybe this is why 'amaretti' sounds better to me than 'amaretto', unless of course, we are speaking of the liqueur, in which case 'amarettos' seems acceptable.

    The thing that really bothers me is when people say 'panini' to refer to a single sandwich. I'm not sure how this became so common in America…nobody says 'gelati' to refer to a single ice cream. Maybe it has to do with store signs – a sandwich shop would advertise that it sells panini, but a gelateria would advertise gelato. Doesn't explain 'zucchini' for 'zucchina' though…

  42. Faldone said,

    October 5, 2010 @ 7:19 am

    Omnibus is already plural. It's the dative plural of omnis.

  43. Faldone said,

    October 5, 2010 @ 7:20 am

    So, by extension, bus is also already plural.

  44. Amy Stoller said,

    October 5, 2010 @ 7:28 am

    Years ago I worked on a revue which included a sketch in which a pretentious character talked about going to musea. I sometimes use that one just to make myself laugh.

  45. James said,

    October 5, 2010 @ 7:29 am

    Pflaumbaum (and zoetrope),
    'amaretti' sounds completely natural to me, maybe because of the ubiquity of 'biscotti'? And I love the sound of 'octopodes'; also 'rhinocerotes'. They do sound a little pretentious, but it would be great if they'd catch on (and lose their pretentiousness).

  46. richard howland-bolton said,

    October 5, 2010 @ 7:45 am

    @Aaron Toivo: 'defenestrate' (to kill by pushing out of a high window)

    I believe that neither the height nor the death is obligatory. I mean the Prague guys got up and walked away (or at least limped), so you may not have to wait for a too demanding event to use defenestratrices.

    Then, in other news, there is the possibly universal use of the regular plural 'antipodes'—presumably because most of us have two feet. I wonder if Long John Silver uses antipous?

  47. Pflaumbaum said,

    October 5, 2010 @ 8:12 am

    @Zoetrope, James

    Sorry yes I meant the liqueur, not the biscuits. To me going to the bar and saying, "Can I have two Amaretti" would be similar to saying "I'd like to order some pizze".

    Yeah 'panini' also came up that night. Again, to me 'panino' is like 'graffito' – it sounds like the user is ostentatiously trying to show their command of Italian… a bit like pronouncing 'Paris' paʁi.

    But I'm sounding a bit like the people who objected to Sotomayor's use of 'Latina'… I should live and let live. You say panino, I say panini, James says octopodes… but where does he put the stress?

  48. Rodger C said,

    October 5, 2010 @ 8:15 am

    Latin u-stems in English usually take -uses: fetuses, viruses (I think). Though nurses seem to refer to decubiti.

    I say syllabi, though I had a department head once who used to write, "Please turn in your syllabii."

    There are now two of us who used to say "Kleenices."

    Those Prague guys landed (no doubt by Our Lady's grace) in a huge pile of horseshit. (Insert obvious remark here.)

  49. Alen Mathewson said,

    October 5, 2010 @ 8:32 am

    My experience as an anatomist has put me firmly in the -s or -es plurals camp; reviewed anatomical papers usually require all latin anaotmical terms to be properly declined i.e. if you use a term in the genitive, you must use the genitive declension of the term. So in strict terms, it might be argued that there is no such thing as a latin or greek 'plural', but a plural declension. It doesn't seem to me to make any logical sense then to use the 'original' nominative plural when the noun is used in, say, an accusative sense. Okay, so who says English has to be logical; but it's how I make my personal choices.

  50. Pflaumbaum said,

    October 5, 2010 @ 8:38 am

    Is virus a u-stem, Roger? Lewis and Short gives it as neuter 2nd declension forming its genitive in -i…

  51. zoetrope said,

    October 5, 2010 @ 8:49 am

    Yeah, but I think 'graffiti' is usually used to refer to a whole bunch of them as opposed to one single graffito, maybe similar to 'etchings' or something like that in English. I also find 'graffito' weird – it would be like talking about one spaghetto. A sandwich, on the other hand, is a clear singular entity, so the '-i' ending just sounds wrong – asking for a panini sounds about as bad as asking for a sandwiches!

    But I lived in Italy, and I'm sure if I didn't speak Italian it wouldn't bother me at all. It's a fine line between ostentation and respect for a foreign language I suppose… With Latin, the situation seems different, as most people don't really speak it in their daily lives.

  52. Freddy Hill said,

    October 5, 2010 @ 8:51 am

    Is this an early eggcorn? Mishearing syttiba as a form of syllabe?

  53. Vincent Daly said,

    October 5, 2010 @ 9:04 am

    Along with omnibus, ignoramus is another one that isn't a nominative singular noun at all.

    Prospectus is fourth declension (u-stem) so prospecti isn't good Latin.

    But in any case, I'm with J. Goard.

  54. George said,

    October 5, 2010 @ 9:08 am

    I think the determination should be the language one is speaking? If it is Latin, then the plural should comply with the grammatical form in Latin. if Italian, then Italian. However, if one is speaking English using a loan word (and often not recognized as such or accurately identified to the language of origin), then English plurals should be perfectly acceptable.

    In any event, to be super correct, we should insist on PIE forms if the word was borrowed from an Indo-European source.

  55. Rodger C said,

    October 5, 2010 @ 9:22 am

    @Pflaumbaum: Thanks, so it is. Being a neuter o-stem in -us leaves its plural rather a mystery, and I can't find a cite on short notice. No wonder the English is viruses.

  56. Rodger C said,

    October 5, 2010 @ 9:24 am

    @Vincent Daly: Prospectuses! That's the other one I was trying to thnk of in the first place.

  57. Andrew (not the same one) said,

    October 5, 2010 @ 9:41 am

    zoetrope: Yes, but one rarely has reason to refer to a single spaghetto, so it's unsurprising the word hasn't taken root in English, whereas one quite often sees a single graffito. If that isn't the singular of 'graffiti', what is? I think that in practice the answer is probably 'piece of graffiti', but having 'graffito' to hand would be useful.

    Regarding 'bi' as plural of 'bus', there is the poem by A.D. Godley.
    But as this is macaronic, i.e. partly in Latin, it shouldn't be taken as determining what the plural should be in English.

  58. J.W. Brewer said,

    October 5, 2010 @ 9:43 am

    Computer-science types have been known to form jokey irregular plurals on non-Latinate patterns as well: http://en.wiktionary.org/wiki/VAXen. The interesting thing is perhaps that some common Latinate morphological patterns are productive in English (at least for pretentious/jocose purposes) but others aren't. Hard to prove a negative but I can't off-hand think of an example of a Latinish-sounding word or new coinage ending in -a which has been pluralized, even jocosely, in -ae where that was not the actual Latin (perhaps Late Latin) plural. And I can't recall, for example, ever personally hearing/seeing the correct-in-Latin spatulae as the plural of spatula in English, although googling shows it's not unknown, perhaps especially in the specialized context of gecko anatomy. Since I assume no one who ever actually studied Latin learned second declension nouns w/o also learning first declension nouns, it's curious-if-true that -us/-i should be productive in a way that -a/-ae (perhaps) isn't.

  59. J.W. Brewer said,

    October 5, 2010 @ 9:46 am

    Oh, and on the Italianate words, "spaghetti" and "graffiti" in English are probably best understood as mass nouns rather than count nouns, typically agreeing with singular verbs.

  60. Jason Eisner said,

    October 5, 2010 @ 9:49 am

    @Garrett Wolman (and @groki, @Aaron Toivo):

    Later versions of the Jargon File (descended from your AIWORD) do list a few more examples of irregular plurals in hackish. Look near the bottom of their overgeneralization page.

  61. Pflaumbaum said,

    October 5, 2010 @ 9:53 am

    @ Roger – yes, I'm not sure 'virus' is attested in any other case apart from nom. and acc.

    Re graffiti, I'm not sure it is a mass noun. I've heard 'a graffiti' and read 'a graffito' (though I've never heard the latter outside discussions like this one). Whereas it would be 'a piece of/stick of spagghetti'.

  62. zoetrope said,

    October 5, 2010 @ 9:55 am

    @Andrew: well, I think for a lot of people 'graffiti' is the singular, but I agree that 'graffito' or 'a piece of graffiti' or 'some graffiti' (even to talk about one) are all fine. But is 'graffitis' all right as a plural? If it follows the 'panini' pattern in English then it should be, but graffiti seems a bit more mass-nouny than panini, so I'm not sure.

  63. Mr Punch said,

    October 5, 2010 @ 9:56 am

    "Syllabi" is the usual plural in my experience.

    There's a lot of discussion here of an eggcorn-like phenomenon that I think of as "classicfication" – treating words that sound like Latin or (occasionally) Greek as if they were. I think of "stewardi" ("corrected" to -dae) and "monokini" — well, perhaps we had more classical educations four decades ago.

  64. Rodger C said,

    October 5, 2010 @ 9:58 am

    I seem to associate "a graffito" with archeologists, who after all introduced the term into English.

  65. Tim Silverman said,

    October 5, 2010 @ 10:04 am

    @richard howland-bolton: in mathematics (specifically Hopf algebras) there's a back-formed singular "antipode", which unnerved me when I first encountered it.

  66. Gabe Ormsby said,

    October 5, 2010 @ 10:04 am

    @George – Regarding the "language one is speaking" approach: Reminds me of my all-to-brief semester in Germany in the mid 90s. Spent a lot of time in Italian restaurants with German eaters, and thus picked up a German plural of 'pizza.' 1 'Pizza', 2+ 'Pizzen'. I say 'a' plural form because embedded in the geeky student crowd, I wouldn't have known whether it was an in-joke like 'Kleenices' or a standard form.

  67. Pflaumbaum said,

    October 5, 2010 @ 10:22 am

    Isn't 'pizzen' in German the equivalent of 'pizzas' in English – the regular plural form?

  68. language hat said,

    October 5, 2010 @ 10:22 am

    Re octopus, the 'correct' etymological plural would be 'octopodes', but anyone who says that deserves to be strangled eight times.

    Since even the Greeks sometimes got confused and treated it as second declension, I think we can safely let that bit of pedantry molder in the obscurity it deserves.

  69. Gabe Ormsby said,

    October 5, 2010 @ 10:26 am

    @Pflaumbaum Yep, '-(e)n' is one of several regular plural forms in German. Had the noun 'Pizza' arisen first in German, it would very likely pluralize to 'Pizzen.'

  70. Pflaumbaum said,

    October 5, 2010 @ 10:41 am

    @ language hat –

    They may have got confused about polypous, but I'm not sure they did about


    Though it seems that ὀκτώπους – which is the form the Romans took – is attested once with an acc. plural in -ους, apparently meaning 'scorpion':


    (sorry I don't know how to embed links)

  71. Mr Fnortner said,

    October 5, 2010 @ 10:47 am

    Today, on this Guardian site: http://www.guardian.co.uk/environment/gallery/2010/oct/04/marine-species-british-antarctic-survey#/?picture=367290280&index=3 is the caption: Octopus are one of the more common species found in the waters of Antarctica. I don't believe octopus has collective noun status like the words sheep and fish but is more like horse and dog. Wouldn't this suggest the singular is octopu, at least to the Guardian?

  72. Mr Fnortner said,

    October 5, 2010 @ 10:53 am

    A few years back, a maker of pantyhose ran a campaign with the line: Pantyhose that fit! Shouldn't that have been pantyhoes? And just what is a pantyho? I know this is a windmill that doesn't need jousting–Google returns 61,500 results for …fit and only 50 for …fits. Still, it annoys me in an odd way.

  73. Dan T. said,

    October 5, 2010 @ 10:56 am

    What is a single piece of ravioli to be referred to as? A "raviole"?

    When you spend the afternoon sitting at the bus terminal, are you watching "buses" arrive and depart, or perhaps "bi"?

  74. zoetrope said,

    October 5, 2010 @ 11:07 am

    A raviolo, at least according to wikipedia: http://it.wikipedia.org/wiki/Ravioli

  75. richard said,

    October 5, 2010 @ 11:09 am

    Just to put this in a non-Indo-European context, English regularly ignores pluralization rules when borrowing outside its own family. Consider: "honchos" (from Japanese hanchou 班長). The Japanese plural (if they bothered making one) would be made by attaching -tachi, possibly first inserting the honorific san first (honchotachi or honchosantachi). Or, another example from Javanese, "gamelans," which should be pluralized as "gamelan-gamelan" if we follow the rules of the original language. Yet the plurals in English pass unremarked.
    Clearly there is an outrage gap here that indicates discrimination against non-Indo-European languages. Fye, I say, fye, and several poxen on your housen.

  76. Moritz said,

    October 5, 2010 @ 11:15 am

    The 2010 numbers for the pl. octopus Google search:
    octopi: 217000
    octopuses: 302000
    octopods: 36200
    octopodes: 20400
    octopii: 25400

    So octopuses is now found (significantly) more often than octopi. In 2004 both were about equally frequent. The bottom two are now relatively more frequent than before, particularly octopii.

    COCA has 36x octopi, 128x octopuses, 20x octopods…

  77. Pflaumbaum said,

    October 5, 2010 @ 11:29 am

    @ richard – it's not just outside I-E languages, it's outside Latin and Greek (and occasionally French and Italian). It's a prestige thing… no-one is expected to use Hindi or Russian or Yiddish suffixes when pluralising their loan-words.

    Incidentally, does anyone know where the hell the word 'syllabub' comes from? My etymological dictionary has no idea.

  78. George said,

    October 5, 2010 @ 11:39 am

    Since my high-school Latin was deep in the last century and the declensions have long since faded away, I seem to recall that the Latin suffix /-i/ was pronounced like /-ee/ (as in bee) not /-ai/ (as in bye). So, would it be 'syllabee' or 'syllabai' if we want to speak like an educated Roman?

  79. mollymooly said,

    October 5, 2010 @ 11:41 am

    There is a hierarchy of forms, from worst to best, thus:
    1. syllabuses
    2. syllabi
    3. syllabuses

    It is important to ensure that readers can tell you are using form #3 and not form #1. In paper-based writing this is difficult, but a web-based document can hyperlink such problematic words to the relevant section of the author's personal styleguide of pedantic lexical oneupmanship. A WordPress plugin should already be available to facilitate this.

  80. Kyle said,

    October 5, 2010 @ 11:57 am

    "though I can't recall the last time I had occasion to use the plural form."

    It must be nice not having to teach multiple preps in any given semester. Meanwhile, I'm constantly using the plural.

  81. J. W. Brewer said,

    October 5, 2010 @ 12:14 pm

    On graffiti, COCA has 35 instances of "graffiti is" and 19 of "graffiti are," but all but 6 (by my hasty count) of the latter are false positives where graffiti is the last word of a plural NP, e.g. "Hispanic gang members and gang graffiti are even proliferating throughout the bucolic Shenandoah Valley." So I would agree that it's not always and everywhere used as a mass noun rather than plural count noun (albeit w/ uncertainty as to the singular), but this data is (yes, I know other idiolects would say "data are") consistent with my impressionistic sense that the mass-noun usage is the dominant one.

  82. Craig said,

    October 5, 2010 @ 12:17 pm

    This is so triggering my anti data-as-plural peeve. Must .. resist .. must .. resist ..

  83. Kyle said,

    October 5, 2010 @ 12:18 pm

    @ Aaron Toivo

    No, the correct plural form is *nix

  84. Coby Lubliner said,

    October 5, 2010 @ 12:47 pm

    Re graffiti: MWCD11 gives separate entries for graffiti (in the modern sense, with a very good usage note) and graffito, pl graffiti (in the archeological sense).

  85. J. W. Brewer said,

    October 5, 2010 @ 1:05 pm

    "Pantyhose" is surely one of those inherently plural nouns like trousers or scissors (if there's a fancy technical term for such nouns, I've forgotten it). The old more obviously inflected-as-plural "hosen" sounds to my ear entirely obsolete/archaic in English. "Pantyhosen" may be current in German, but the first few sites I googled up seem like they might be sufficiently, um, pornographische that I'm not going to click through to learn more about the usage.

  86. Nightstallion said,

    October 5, 2010 @ 1:14 pm

    I doubt anyone here would know, but in German, there's a trademark name "Wettex" which I like to pluralise to "Wettices", so count me in among the Kleenices. ;)

  87. Dan T. said,

    October 5, 2010 @ 1:29 pm

    I think "cherubim" has a plural suffix from the Semitic language group.

  88. The Ridger said,

    October 5, 2010 @ 1:58 pm

    @Mr Fnortner: It probably suggests that the Guardian is misspelling (again), but possibly means that they think of octopus as fish, thus not having a separate plural unless you're speaking of types of octopus (cf Herring were once numerous).

  89. Pflaumbaum said,

    October 5, 2010 @ 2:07 pm

    George, it would be an '-ee'.

    But your 'educated Roman' would have pronounced the y somewhat like the 'u' in the French 'du', and a proper double 'l' as in Italian, and he might have reproduced the Greek rising pitch accent on the first syllable.

  90. George said,

    October 5, 2010 @ 2:10 pm

    @Dan: "I think "cherubim" has a plural suffix from the Semitic language group."

    Right, the plural suffix is /-im/ – one of them in Hebrew is 'keruub.' Maybe we should insist on pluralizing words of Hebrew origin with /-im/.

  91. Andrew (not the same one) said,

    October 5, 2010 @ 2:23 pm

    Regarding 'octopus'; singular-for-plural was historically used for many sorts of animals when they were seen as objects of pursuit – you can talk about 'hunting lion', for instance, or 'shooting duck'. Birdwatchers often seem to use it as well – they say 'Snipe are common hereabouts' and so on. So I don't think 'octopus are' can be ruled out, even though it isn't a fixed form like 'deer' or 'sheep'

  92. Jerry Friedman said,

    October 5, 2010 @ 2:56 pm

    Thanks for the discussions of graffiti, which I got to use today. I told my students that the verb corresponding to derivative is differentiate, and one young man said, "Where's the logic in that?

    I said, "You want logic? This is a language where the singular of graffiti is tag."

    (This was for one of the seven classes that I have syllabi or syllabuses for, since I teach and don't do research. I say syllabi more often and hear it a lot more often. Mollymooly is right on target.

    Yiddish words have their original plurals: kneydlach, kreplach, kinderlach, spatulach…

  93. Olga said,

    October 5, 2010 @ 3:27 pm

    Thank you for posting this. For years now, I have countered the condescending corrections ("Sorry, you can't know this since you're not a native speaker, but the correct plural of syllabus is syllabi, because this is a Latin word") by explaining that syllabous should be inflected in analogy to pous, resulting in the plural syllabodoi. Turns out that I was wrong, but so what? It's good enough to shut up the idiots. I will however memorize the sittaba thing; it looks like it will work even better.

  94. On Pluralizing | errata said,

    October 5, 2010 @ 3:49 pm

    […] Lieberman, today in The Language Log, tells us that the word syllabus is mocked up construction of a misreading of a Greek word sittyba, […]

  95. George said,

    October 5, 2010 @ 4:04 pm

    @ Pflaumbaum: "it would be an '-ee'."

    Thanks. So, if one insists on the Latin plural form, should an English phonological corruption be permitted? Should we stone those who say 'syllabai' or 'cactai?'

    I would guess the /-ai/ pronunciation is just one more victim of the Great Vowel Shift.

  96. Östen Dahl said,

    October 5, 2010 @ 4:33 pm

    There is a case of a "fake" Latin plural that hasn't been mentioned in the discussion, it seems: "pluralia tanta" — which assumes that "tantum" in the expression "plurale tantum" is an adjective agreeing with "plurale", when it is in fact an indeclinable adverb meaning 'only'. I have heard or seen it used by respectable linguists.

  97. Michael Tobis said,

    October 5, 2010 @ 4:49 pm

    In my family the plural of cactus is cactoosim, with emphasis on the middle syllable. (Its usage is to designate the entire southwestern arid zone where civilized people don't go.)

    Hope this helps.

  98. blahedo said,

    October 5, 2010 @ 5:20 pm

    @George "…the Latin suffix /-i/ was pronounced like /-ee/ (as in bee) not /-ai/ (as in bye). So, would it be 'syllabee' or 'syllabai' if we want to speak like an educated Roman?":

    Thank you for bringing this up, and this is further evidence for the @J. Goard camp: these are really, really English words (now), and it really is a (semi-)productive paradigm *in English*, albeit inspired (or "seeded" as J. puts it) by a highly regular paradigm in Latin. Basically, there are two pronunciation styles in active use for Latin and modern borrowings from Latin: Italianate/Ecclesiastical and Reconstructed Classical, and at this point following the traditional Anglo-Latin rules is a sign of thorough assimilation into English (or perhaps extreme eccentricity, but for common vocabulary we can rule that out).

    See, by the way, http://en.wikipedia.org/wiki/Traditional_English_pronunciation_of_Latin for more about the /i/-/aj/ thing and other related issues.

  99. áine ní dhonnchadha said,

    October 5, 2010 @ 6:12 pm

    My favorite plural: orang-orang hutan, for orang-hutan.

  100. Jim said,

    October 5, 2010 @ 6:45 pm

    "Doesn't explain 'zucchini' for 'zucchina' though…"

    I'm waiting for an explanation of 'zucchina' first. "Zucchin" sounds more correct for the plural of a small zucchi. And shouldn't they be 'calabazzini' or something along those lines anyway?

  101. Jim said,

    October 5, 2010 @ 6:45 pm



  102. Rodger C said,

    October 5, 2010 @ 6:53 pm

    @Jerry Friedman: I'd feared (as a (cultural) Protestant) to mention this till you said "spatulach," but I'd been thinking of the scene in Philip Roth's most famous novel where Alec Portnoy remembers when he was eight years old, and his teacher held up this implement and asked him what it was called, and he was tongue-tied because he knew his mother called it a spatula but he thought it was a Yiddish word.

    @Pflaumbaum: If "virus" only occurs in the nom and acc sing, how do we even know it's an o-stem?

  103. Jerry Friedman said,

    October 5, 2010 @ 7:23 pm

    @Rodger C.: That scene from Portnoy's Complaint is exactly what I was thinking of. I think I should have used -kh instead of -ch, by the way.

    Incidentally, we borrowed syllabus from Modern Latin, says the OED. So why shouldn't we borrow the plural from Modern Latin? I don't see anything "fake" about the process.

  104. Ellen K. said,

    October 5, 2010 @ 7:54 pm

    Regarding zucchini, Wiktionary says it comes from the Italian zucchini, which is plural for zucchino, which is a variation of zucchina, which, etymologically, is a diminutive of zucca.

  105. Jonathan D said,

    October 5, 2010 @ 8:12 pm

    I'm comfortable with 'fora', and I've read enough 'graffito' to find it unremarkable, although I don't usually see a need for it.

    In any case, Latin got it all wrong. 'I' is singular, and 'us' is a plural.

  106. Pflaumbaum said,

    October 5, 2010 @ 8:13 pm

    Yeah good point, probably it's unattested in the plural rather than the other oblique cases… I couldn't be bothered to check the links in Lewis and Short and now the site's crashed. It is meant to be related to Gk. ἰός and Skt. víṣa-.

    Jerry Friedman – those Yiddish plurals are just examples of words that have entered the language as plurals though, haven't they, like 'spaghetti' and 'graffiti'? People don't say 'shticker' or 'shmucke' or 'bagelen'.

  107. Pflaumbaum said,

    October 5, 2010 @ 8:14 pm

    Sorry that was @ Roger C

  108. Anthea Fleming said,

    October 5, 2010 @ 9:43 pm

    The plural of platypus is generally felt to be platypuses, in speech and writing. If people say platypi, they are usually informed that the word is of Greek origin and that the tiresomely learned plural would be platypodes. I will not speculate on the plural of Ornithorhynchus paradoxus.

    In our family the plural of toothbrush is teethbrish.

    A tendency to lisp induced our young children to refer to the Frogmouth (a grotesque Australian bird) in the plural as Frogmithe.

  109. Dan T. said,

    October 5, 2010 @ 11:10 pm

    Then there was Allan Sherman's "One Hippopotami".

  110. Erik said,

    October 6, 2010 @ 2:55 am

    Never hearing the word "graffito" used before, I would have thought it to mean "the people who produce graffiti" (analogous to "mafioso") or possibly "a structure containing graffiti" (see "mustachio".)

  111. Pflaumbaum said,

    October 6, 2010 @ 6:00 am

    But 'mafioso' doesn't mean 'the people who produce mafiosi', it means a member of the mafia… it's just the singular of 'mafiosi'.

  112. mollymooly said,

    October 6, 2010 @ 10:50 am

    The plural of Jesus is Jedi.

  113. Ray Dillinger said,

    October 6, 2010 @ 1:19 pm

    If we were writing in Greek, or Latin, it might be relevant whether the word actually appeared in Greek or Latin, or how Greeks or Romans might have declined it. However, we are writing in English about an English word which definitely does appear in the vocabulary of most English speakers.

    And many English speakers though not all, entirely without referent to the rules of Latin or Greek, happily pluralize any -us ending by turning it into -i. I would strongly assert that this is now the 'regular' pluralization rule in English and that '-us' endings pluralizing as '-uses' have become more the irregular case.

  114. George said,

    October 6, 2010 @ 2:14 pm

    @Ray and blahedo: I think you are right. The words are fully assimilated English words, not code switches to another language. And, this is a semi-productive English process for multi-syllable words ending in -us.

    The problem is with insisting on an /-i/ plural based on the notion that word is of Latin origin, or worse yet, dissing the good ole English /-s/ pluralization.

  115. J. W. Brewer said,

    October 6, 2010 @ 2:25 pm

    Ray Dillinger: what's the basis for your strong assertion? The last time this came up in my own life a few weeks ago, I was making fun of a friend for pluralizing "status" (in its facebook sense) as "stati," since in Latin "status" is 4th declension not 2nd. So not only a bit pretentious-sounding but (as to Latin) wrong. (One can argue that I was being pedantic, but I would respond that he was engaged in something akin to hypercorrection and thus could reasonably be called on it. This all somehow came up in the rather post-classical context of a quotation of lyrics from the "metalcore" band Atreyu.) I will admit that I had not done any empirical research in current AmE usage before making fun of him, but having just done so nunc pro tunc, it turns out that COCA has 208 hits for "statuses" to 10 for "stati." Do you want to give some examples of words in -us where -i dominates -uses to a comparable extent?

  116. Ray Dillinger said,

    October 6, 2010 @ 4:12 pm

    Well, the challenge honestly undertaken doesn't show dominance of the -i ending yet. It seems that way to me because I hear it a lot and use the pattern myself. However, while it appears enough to qualify as 'a valid pluralization' it's not the majority pattern implied by words like 'regular.'

    Measured In Google hits,

    viruses with 31.3 Mhits overwhelmingly leads virii with 186 Khits,

    octopi with 238 Khits is still trailing octopuses with 363 Khits, (octopodes has 29.1 Khits)

    platypi with 31.4 Khits also trails platypuses with 68.9 Khits, (platypodes has 3.18 Khits)

    There are cases where the -i pluralization appears to be in the lead, but the two first ones I found are suspect:

    boli with 19.9 Mhits has a massive lead over boluses with 664 Khits, but the results are suspect; 'boli' appears to have many uses besides its use as a plural of 'bolus'.

    contrapuncti with 725 Khits is solidly ahead of contrapunctuses with just 18, but that probably qualifies as an actual code switch rather than a fully-assimilated English word.

  117. language hat said,

    October 6, 2010 @ 6:44 pm

    They may have got confused about polypous, but I'm not sure they did about ὀκτάπους

    Well, of course we can't be sure, but since the written evidence is scanty, it seems reasonable to assume that if they were confused about the declension of one -pous noun they would have had the same confusion about others.

  118. onymous said,

    October 7, 2010 @ 12:50 am

    I'm glad to know I'm not alone in having used the word Kleenices.

  119. Dan T. said,

    October 7, 2010 @ 12:41 pm

    The trademark people tend not to like applying any suffixes, regular or irregular, to their brand names (e.g., "Kleenices" or "Legos"), since they'd rather the names only be used in phrases like "Kleenex® brand facial tissues".

    But if you have two Lexus® brand cars, are they "Lexuses" or "Lexi"?

  120. Dan M. said,

    October 8, 2010 @ 4:23 am

    @ J.W. Brewer,

    For what little it's worth, my mother, who is well-educated, is a mathematician, and is quite silly, often refers to using multiple "spatulae" while cooking.

  121. Matt Ezell said,

    October 8, 2010 @ 2:08 pm

    No comment on whether syllabus is a corrupted form.
    Since syllabus is from Greek not Latin, the prescriptively correct plural of syllabus is syllabontes. But no one would understand that, so we use the more familiar Latin -bi ending or plain ol' English -es plural in real life.
    Source: One of my professors from a graduate level linguistics course, CSU Long Beach.

  122. maidhc said,

    October 9, 2010 @ 3:34 am

    Add another to the "Kleenices" camp. Though I sometimes wonder if anyone gets the joke.

    But if there were a place with more than one movie theatre, each having several screens, would there be multiplices there? And if a person divorced more than once, would he have several ices?

    I think there is a good claim that "octopus" is Latin — not from classical Latin, but, as the scientific name of a species, it is New Latin. I don't know if there is any rule about when creating the name of a species, if the declension needs to be specified, but I would think that -us would default to second declension if one had to decide.

    I tend to agree with Ray Dillinger that -i has become an English plural, although I would add that its use tends to be confined to words of scientific or biological character. I am happy with "octopi" or "cacti" but not so much with "stati".

    "Indices" and "gymnasia" seem to be going out of fashion. In the days when "aviatrix" was in use, were there "aviatrices"? "Odea" turns up in the NYT crossword not infrequently, but seldom elsewhere.

    I've known people who talk about 'celli in the orchestra, and I think they try to pronounce the apostrophe as well.

    The term ceili or ceilidh has come into English, at least in certain circles, and I'd like to get people to pluralize it as ceilithe. And then after that, maybe we can give slogan the correct plural sloganeacha.

  123. Robert said,

    October 11, 2010 @ 1:41 pm

    I know there are some people who use focuses concerning optics and lenses, but foci for ellipses. There's also a disagreement in the mathematical community on whether the plural of topos is toposes or topoi. I looked at SGA 4, the work in which Grothendieck originally defined them, and he used neither, he had topos as the plural of topos.

  124. Nathan McCoy said,

    October 12, 2010 @ 6:53 am

    Fye, I say, fye, and several poxen on your housen.

    Surely you mean "poxen on your hice".

  125. Aaron Davies said,

    October 13, 2010 @ 2:12 pm

    nissan maximae…

  126. the plural of syllabus? « Daily Thrill said,

    October 15, 2010 @ 5:13 am

    […] plural of syllabus?   Leave a comment This all started from this http://languagelog.ldc.upenn.edu/nll/?p=2684 where they thrown a stone which broke my small English windows, the first thing that came into […]

  127. GerryC said,

    October 22, 2010 @ 5:36 pm

    I think maidhc makes a good point that there is justification for treating octopus as New Latin in the context of biological studies. Sciences are studied internationally so it makes sense to have a standard form whereby the plural form can be recognised by biologists who speak German, Spanish, Arabic, Gaelic or anything else.

    The same argument is valid for mathematical sciences. Radii, foci, loci and other specialised words can be learned and understood worldwide as part of the universal language of mathematics, just as all classically trained musicians understand Italian terms such as andante, fortissimo and rallentando.

    General English usage is another matter. My feeling is that there is nothing to be lost by applying /s/ or /es/ to singular forms but selective application of rules from foreign languages easily leads to confusion and confrontation.

    Forums, stadiums, cactuses, syllabuses and octopuses are clear, easily understood and logical.
    If it's not broken, don't fix it.

  128. ASG said,

    November 1, 2010 @ 4:13 pm

    A professor I had in graduate school regularly quoted a sentence in Augustine's Confessions, which he paraphrased as "angels don't need syllabi!" Today's LL post inspired me to look for the reference, and I was led to this interesting study of the word syllabus:


    Augustine wrote "ibi legunt sine syllabis temporum" — "[angels] read without temporal syllabi" — and the author of the blog post takes this to mean that Augustine was referring to the syllables of words.

    A shame, really, since I always enjoyed thinking of angels teaching courses without having to engage in the tedious business of syllabus design.

  129. Designing education #1: the double diamond | Eugenie Teasley said,

    June 7, 2013 @ 5:53 am

    […] I’ll just say—decentralise THIS. Put down the workbooks and syllabuses, and put into the hands of learners and their classroom teachers the ability to discover and define […]

  130. Around the Web Digest | Savage Minds Backup said,

    October 1, 2013 @ 1:30 pm

    […] Why invent a fake Latin plural for a fake Latin singular? Stick with the plain English “syllabuses… /KF […]

  131. George Brooks said,

    August 20, 2014 @ 2:49 pm

    For those who insist that "syllabi" is a word, let's leave aside the Latin and Greek etymology and just ask: "what is the plural of 'school bus'" ?

    That should resolve it.

RSS feed for comments on this post