Mare, mǎ ("horse"), etc.

« previous post | next post »

[This is a guest post by Robert Hymes]

I just happened to be reading your Language Log post from April, “Of horseriding and Old Sinitic reconstructions.” I too have always been sympathetic to the possibility of a mare-馬 connection, which I’ve tended to assume would have happened through a Chinese borrowing from Indo-European either directly or mediatedly, though as you point out the problem of the “mare” root’s presence in only Germanic and Celtic is, well, a problem.

But another point you mention — “1. ‘mare’ refers to the female of the species” — is not actually a problem. This is because the reflexes of the Indo-European words for ALL THREE of the non-horse large domesticated animals in the Indo-European bestiary are used in Germanic (well, at least the descendant languages) only for the female of the species. Thus from IE *owi- (originally *h2owi-), “sheep,” we get only “ewe,” and have supplied the general species name from the Germanic-only “sheep.” From IE *gwo-, “bovine animal” (we don’t even HAVE a singular noun for the whole species regardless of gender), we get only “cow,” and have taken our word for the male from an entirely different root and our word for the species collectively from the Latin for “movable property.” From IE *su-, “pig, swine,” we get only “sow,” and have taken the name for the whole species from god knows where. So assuming something like *marko really was an IE root for “horse” (alongside *ekwo which Germanic largely lost), it would be consistent with larger intra-Germanic patterns for it to be applied only to the female. I note that our word for the whole species, “horse,” is another Germanic-only innovation or substrate borrowing like “sheep” and “pig.” There’s a pattern here, and “mare” fits right in.


  1. pamela said,

    November 17, 2019 @ 4:32 pm

    that's fascinating. is there a standing explanation for that? other than in any breeding operation the females will vastly outnumber the males?

  2. cameron said,

    November 17, 2019 @ 5:38 pm

    If the term originally referred to a horse that was suitable for riding, it makes sense that it would originally have referred to females only. It was probably many, many, generations of breeding before they produced stallions that would suffer a rider.

  3. Andreas Johansson said,

    November 18, 2019 @ 1:02 am

    Isn't "swine" precisely a sex-neutral reflex of *su-?

  4. Scott Mauldin said,

    November 18, 2019 @ 2:56 am

    @Pamela and @Cameron

    It's my understanding that the Mongols preferred to ride female horses. I would imagine that whatever reasons they had might have been shared by whichever Indo-European steppe peoples spread the words and practices related to horseriding around Eurasia.

  5. Keith said,

    November 18, 2019 @ 3:30 am

    If I understand correctly, the problem with stallions is aggression: that you cannot put several of them in an enclosure together, they will fight. And any mare in season will attract the attentions of any nearby stallion, who will escape from his enclosure and attempt to mount her (and he won't much care if there is a rider on her back at the time).

    Mares may fight each other to establish a hierarchy, but with less violence and risk of injury.

    So at some point in the history nomadic, horse-riding people will probably have decided that mares were easier to manage than stallions. Male foals might have been allowed to grow until near sexual maturity, then I imagine would have been slaughtered to use their meat, hide, sinew and bone.

    The Scythes and Huns are credited with developing the technique of castrating young stallions. The French word for a gelding is "hongre", derived from the same root as the English word for "Hungary".

  6. Philip Spaelti said,

    November 18, 2019 @ 5:42 am

    For those of you arguing that 'it makes sense' that a female specific term might be generalized to all horses ('because riding'), you might note that you are arguing the exact opposite of what Hymes is suggesting. Hymes is arguing that generic sex-neutral PIE terms become female specific ones in Germanic.

    It might also be noted, that languages develop sex-specific terms for a species only after domestication.

  7. Rodger C said,

    November 18, 2019 @ 7:35 am

    languages develop sex-specific terms for a species only after domestication

    "Buck" and "doe"?

  8. Rodger C said,

    November 18, 2019 @ 7:35 am

    Isn't "swine" precisely a sex-neutral reflex of *su-?

    Sure, but i think Hymes was referring to the more common "pig."

  9. Andreas Johansson said,

    November 18, 2019 @ 2:00 pm

    Sure, but i think Hymes was referring to the more common "pig."

    Sure – but he claimed we get only "sow" from *su-.

  10. Jerry Friedman said,

    November 18, 2019 @ 4:45 pm

    Rodger C: Actually, "doe" may be a good example (its etymology before Old English is obscure), but "buck" might not be, since it originally meant a male goat, according to etymonline. The OED is less confident. "Originally two words, Old English buc and bucca, which became indistinguishable in form after 11th cent. So far as the evidence goes, Old English buc was used for the male deer, and bucca for the he-goat, but the instances are so few that it is far from certain that the words were thus distinguished in meaning. Old English buc = Middle Dutch boc, Dutch bok, Old High German bocch (Middle High German boc, modern German bock), Old Norse bukkr (Swedish bock, Danish buk), all meaning primarily ‘he-goat’, though in each of the modern languages applied to male animals of the deer kind (in Danish also to the ram) < Old Germanic *bukko-z."

    There's "lioness" and "tigress", but I don't know how long people have been taming them. "Vixen" should work, though—I don't think foxes get tamed much. In cooking, female fish are or used to be called "spawners", and males, "milters".

  11. Philip Anderson said,

    November 18, 2019 @ 5:24 pm

    As Philip Spaelti, it’s the modern Germanic words that are feminine, not the older forms. Old English ‘mearh’ was I think generic, and modern Welsh ‘march’ means a stallion!

    Words for animals in IE languages seem unstable, very likely to become specialised with a new, often borrowed, word being adopted for the species. IE ekwo- survives in hippos, and Welsh ‘ebol’ (foal), but Welsh, like the Romance languages, adopted caballus (as ceffyl) as the generic term.

    It’s interesting that the old names tend to survive in Germanic as feminine, but this process isn’t complete hence swine. And English hound, cognate with hund and canis, has specialised to a hunting dog.

  12. pamela said,

    November 18, 2019 @ 11:07 pm

    the thing Bob is referring to of deriving generic terms that have a female referent (or vice versa) probably has nothing to do with riding stallions or not; I don't think neollithics were riding pigs or sheep (not as adults, anyway). but they were all being bred on a large scale and in that case the majority of animals were female. the males, on the other hand, only needed to be designated by special words. that would be my first guess.

  13. Chris Button said,

    November 18, 2019 @ 11:18 pm

    It might also be noted, that languages develop sex-specific terms for a species only after domestication.

    Actually that fits quite nicely with the respective suggestions by Mallory & Adams and Pulleyblank that the eclipsing of the *markos form might have had something to do with wild versus domesticated horses or war horses versus a more generic term for horses.

  14. Rodger C said,

    November 19, 2019 @ 7:50 am

    @Andreas Johansson: Thanks, I'd missed that implication. For that matter, "hog," if Celtic, might also contain *su-, mightn't it? (Though Wiktionary gives a Germanic etymology I've never seen before.)

  15. Robert Hymes said,

    November 19, 2019 @ 10:54 pm

    I was under the impression that “swine” was a borrowing from a Latin suin- or something like that, parallel to “bovine” and so on. If that’s wrong, then in the case of pigs the transfer to specifically female meaning, as others have suggested, was incomplete. Though I might still say that the “-in-“ suffix on “swine” gives it the possible look of an adjectival derivative in any case, so I would wonder whether after “su-“ was confined to the female, the adjectival form then shifted its function to fill the gap.

RSS feed for comments on this post