Friday, February 13, 2009

Oxen and Foxes — the curious life and death of plurals

The Little Prince © 2003 by Christophe DrochonThe plural of ox is oxen, but the plural of fox is foxes. Has that ever bothered you? Ever wondered why it is? I have known people to guess that it’s because the words were drawn from two different linguistic wells, but that’s not it; both are straight from the swift-flowing waters of Old English, surviving into Modern English almost without change. So ... why is it? And have oxes and foxen ever been spotted in the wild?

Let me answer the last question first. Yes — oxes is liberally attested in the Middle English period, with its veritable explosion of different word forms and spellings, and oxes has cropped up occasionally ever since. The form foxen has occasionally been seen as a plural of fox, as in David Booth’s Analytical Dictionary of the English Language [1]. But as a plural, it would have had stiff competition from foxen = vixen “a female fox”. When it has appeared as a plural, I would guess it was under the pressure to match oxen. Or it may show a Dutch influence. The examples Booth gives — housen, foxen, eyen — have the Dutch forms, huizen, vossen, ogen.

Let’s take a closer look at the provenance of the two words. First, the obligatory litany of cognate forms. Now come on, no complaining; you knew this was coming. :)

The word ox comes from Middle English ox, from Old English oxa. Cognate are Old Norse uxi, Old Saxon ohso, Middle High German ohse, Old High German ohso, Welsh ych (pl. ychen), and Gothic aúhsa — all (well, not the Welsh) from Primitive Germanic *uhsōn.

The word fox comes from ME fox (and southern dialectal vox), from OE fox. Cognate forms include ON fóa “fox” (“vixen”, actually; the noun is feminine) — ON borrowed the OE fox, unchanged, but it was only used (and only rarely) in the metaphorical sense of a kind of fraud; Norn fūa; OS fuhs, vuhs; MHG voha, and vuhs; OHG foha, and fuhs; and Gothic faúhó — all from Primitive Germanic *fuhsaz.

Great, you’re saying, but how does this help us understand why the plural forms differ? Well, take a closer look at the OE forms, and you will notice that one of them ends in a vowel. Our ox originally ended with a vowel in all the Germanic languages, as it still does in Modern Swedish oxe, German Ochse, and Frisian okse. But in Modern English (as in Modern Dutch os) the terminal vowel has been lost. But fox had its terminal consonant from the beginning. OE fox is a strong noun, hence its plural form is foxas; but OE oxa is a weak noun, hence its plural would be … anyone? … oxan. And there you have it: that’s the reason for Modern English oxen, but foxes.

Are there other words like these? Not pox (actually = pocks, already plural). Nor tax (from Latin via Old French).

How about box? Like ox and fox, this is a direct survivor from OE box, but this word is a feminine strong noun, so its original plural, buxa, was of a third kind! Subsequently, buxa was standardized by force into boxes, just as oxen will probably someday become oxes.

And what about lox? This comes to Modern English via Yiddish, which is not Hebrew, but basically German. There is actually an all but forgotten OE antecedent in lex, leax, læx “salmon”. This was a strong noun (like fox), so its the plural would have been leaxas. But strangely enough, modern lox has no plural!

Any others? How about two more, just for fun. You are having fun, aren’t you? Modern English ax(e) is from OE æx, æcs(e); its correct plural is like box, æxa, but similarly, it has been “standardized” into the form, axes. And wax is from OE weax, wæx; a strong noun like fox, so its plural is weaxas. We’re all right there with Modern English waxes.

What does any of this have to do with Tolkien (as everything must :)? I’ll leave you with two quotations (emphasis added):
A fox passing through the wood on business of his own stopped several minutes and sniffed. ‘Hobbits!’ he thought. ‘Well, what next? I have heard of strange doings in this land, but I have seldom heard of a hobbit sleeping out of doors under a tree. Three of them! There’s something mighty queer behind this.’ He was quite right, but he never found out any more about it. — The Fellowship of the Ring (Chapter 3, “Three is Company”)
In Dasent’s words I would say: ‘We must be satisfied with the soup that is set before us, and not desire to see the bones of the ox out of which it has been boiled.’ [...] By ‘the soup’ I mean the story as it is served up by its author or teller, and by ‘the bones’ its sources or material—even when (by rare luck) those can be with certainty discovered. — “On Fairy-Stories”

[1] Booth, David. An Analytical Dictionary of the English Language. Volume 1. London: J. and C. Adlard, 1830, p. xxxiv.


  1. A noted characteristic of computer-user (and science-fiction fan) dialect is to apply non-standard grammatical rules wherever they might be amusing.

    Thus, it has long been established that the plural of fan is fen, and if you have more than one VAX computer, they are VAXen.

  2. Thanken for those flashen from the past, David. Well do I remember ther herds of grazing VAXen from my college days. The plural fen, though — I can’t say I’ve ever run into that one, but I definitely grok. Part of the fun of language is learning its whys and wherefores well enough to be able to break the rules creatively.

  3. It's interesting that the modern Swedish plural is oxar, which is a completely normal plural. Has that one been standardized too?

    I am not actually sure what the plural for modern Swedish lax is - whether it's just lax or laxar. Both sound a bit strange, and I am not sure if they can be used that way. I believe that "fish-names" are most often used abstractly, as in English.

  4. Hi, Ardamir. Don’t let me pretend to be an expert on Swedish, hahae, but no, my feeling is that oxar was probably a “normal” Swedish plural form all along, rather than a form later “standardized”. The plural for the Old Norse uxi was uxar (however, there was another ON form as well: oxi, pl. øxn, a bit more like the OE forms).

    It was in the 12th century, I belive, that Old Swedish and Old Danish began to diverge from the earlier eastern dialect of the Old Norse. Having just taken a look, I can tell you that the forms oxe, pl. oxar are indeed attested in Middle Swedish (where Middle Danish has okse, pl. okser). Since the Modern Swedish form goes all the way back to the ON without interruption, I think we’re on pretty firm ground to say that uxar, oxar have been fixed and standard for a long time.

  5. Since the Modern Swedish form goes all the way back to the ON without interruption, I think we’re on pretty firm ground to say that uxar, oxar have been fixed and standard for a long time.

    To clarify, that’s Modern Icelandic uxar, Modern Swedish oxar.

  6. Great post, Jase. Somewhere, the dwerrow are rejoicing. :)

  7. Thanks! I hope they are. :)

  8. This may not make much sense to you, but it is a quintessential usage in context of the plural "fen".

  9. Thanks for the clarification, Jason.

  10. @Ardamir: Sure thing. Glad to help.

    @David: Wasn’t Bernadette Bosky at Mythcon last year? I seem to recall she gave a paper on Charles Williams. Thanks for the vintage example of fen “in the wild”. You’re right that I didn’t have quite the full frame of reference(s), though I think I picked up on enough of it. And I enjoyed reading it in any case.

  11. Yes, that's the same Bernadette Bosky. Like most of us, she plays many parts.

  12. What fun. Interesting is that in Afrikaans, the plural reverted to the form of the old singular: Fox - vos, foxen - vosse; and ox - os, oxen - osse. In house, the Dutch plural is huizen, but in Afrikaans it became huise. I don't know of any studies, but I wonder how much influence the early influx of Hugenot's had on these developments in Afrikaans away fromthe Germanic "standard".

  13. That’s really fascinating. Thanks for letting us know about those differences between Dutch and Afrikaans. A study definitely seems warranted, if none has been undertaken.

  14. Maybe you can write something on when a dialect becomes a language? It is bound to be a fuzzy distinction, but should be interesting. There is the Galician-Portugese example, the Afrikaans-Flemish-Dutch example, even the Xhosa-Zulu-Seswati-Ndebele-Shangaan clan (Nguni language group) - all quite close.

  15. It is an interesting subject. I’ll mull it over and see whether I can’t come up with some interesting or original ideas to share, but I will say that it’s not really my area of expertise. :)

  16. Hey Jason—thanks for the post, which is highly entertaining, as usual. As Calimac said, lots of computer users have extended the -en ending, although I've haven't, AFAIK, seen that usage in the wild.

    "Boxen," however, used to appear regularly on Slashdot (like me—but I eventually tired of the flamewars and repetitiveness), and the Jargon file entry defines it well.

    Maybe you can write something on when a dialect becomes a language? It is bound to be a fuzzy distinction, but should be interesting.

    CM Millward does a decent job in A Biography of the English Language, which I read last semester for a course on the history of the English language. As you note, it's not as if we've crossed the border into Canada, so linguists sometimes like to fight about whether X is a dialect or an independent language.

  17. Thanks for the comment, Jake. Nice to see you visiting. :)

  18. As to Scylding's Afrikaans and Dutch examples*, I think the -n in the Dutch plural ending is definitely on its way to becoming mute, Afrikaans has simply been a bit earlier. Also, as Afrikaans has been defining itself as a language (over the last eighty years or so) it has defined its own spelling, incorporating recent changes, whereas Dutch spelling has been a bit more conservative.

    * The Dutch equivalents of his Afrikaans examples are: vos-vossen, os-ossen, huis-huizen. The last one is exhibiting the same feature as dwarf-dwarves.