The sour milk of etymology

In 2018, I devoted three blog posts to the origin of the English words bread and loaf (see the posts for 17 October, 24 October, and 7 November 2018). Once I even discussed brisket (20 October 2010). The time has come to write something about the etymology of the word milk. Don’t hold your breath: “origin unknown,” that is, no one can say why milk is called milk, but then no one can say why water is called water either, except for connecting wet with water. It would be useful to have some semantic ties, with milk meaning “white stuff” or “nourishing product,” and water going back to a word for “thirst.” Alas, in dictionaries, one finds bare roots: melg– and wat-, which are neither obviously sound-symbolic nor sound-imitating. Yet someone in the remotest past coined those names, and they evoked certain associations! Anyway, water was always around, at least in the form of rain, while the first contact of humans with milk must have been through breast feeding, that is, long before the most ancient speakers of Indo-European learned to milk cows. (I’ll return to breast feeding at the end of this post.) Or the word may have referred to some stuff seen on plants. Yet no word for “liquid’ or “white” resembles milk.

We can say with enough confidence that once upon a time the Indo-Europeans did use milk. English speakers take it for granted that the verb milk is derived from the noun milk, but the two words may go back to different roots (see below). More relevant to English is the adjective milch (a milch cow, that is, a cow giving milk), which, according to my experience, hardly anyone understands today.

The main riddle is the great similarity of words for “milk” in numerous languages, a true embarrassment of riches. In Germanic, Gothic miluks (recorded in the fourth century), German Milch, Icelandic mjölk, and so forth occur. The faraway Tocharian word is malke. Closer to home we find Latin mulgeo and Greek amélgo “I milk,” though side by side with those words, Latin lac/lactis (as in English lactate) and Greek gála/gálactos “milk” display a different root. Apparently, related to lact– is Hittite galaktar “a sweet plant juice used in rituals.”

The situation is confusing because many words related to milk and milking sound alike but do not seem to be cognate. A classic example is German Milch “milk” and melken “to milk.” The verb seems to be akin to Greek amélgo “to draw, pull” and Latin mulgēre “to milk” (Old Irish mlegon also means “drawing, milking”). It seems almost incredible that Milch and melken should be unrelated. Even if etymological algebra draws them apart, why are those near-twins offspring of different roots?

It has even been suggested that in the history of the words for milk some sounds were at one time changed deliberately under the influence of taboo. Indeed, milk was such an important product for both consumption and ritual purposes that the word may have been modified intentionally. Ritual words are often altered for fear of offending the powers that be. But the existence of taboo in history can seldom be demonstrated, and reference to it is destined to remain an ingenious guess. As far as we can judge, milk served as the name of the product, while milking was associated with drawing.

Even if we agree that a common very ancient Indo-European word for “milk” did not exist, we have enough trouble with Germanic. The trouble comes from the Slavic side. Slavic for “milk” sounded approximately melko (Russian moloko, stress on the last syllable; Polish mleko, stress on the first syllable, and so forth.) Those who have read the post on loaf (see the beginning of this blog post) may remember that the Russian word khleb “bread” is almost certainly a borrowing from Germanic. If khleb could be taken over from Germanic, why not mleko? But we may ask: why should anyone borrow such basic words from the neighbors? As regards loaf ~ khleb, the answer is not far to seek. Bread existed in several forms (for example, leavened and unleavened; as a brick or as a flat pancake, and so forth). But milk from a cow is just milk, and borrowing seems to be unlikely. Yet nothing is certain in this area of knowledge. The Old Chinese word lac “milk” is believed to be a borrowing from eastern Indo-European. In Latin, melca “spiced milk” turned up, either a loan from Germanic (the prevailing opinion) or an illegitimate cognate.

Unexpectedly, the semantic base of the Slavic word is less opaque, because several words with the same root refer to wetness and dampness: “swamp; puddle; fog.” In Germanic, Gothic milhma “cloud” (a much-discussed noun) may be related. It appears that the Slavic root of the word for “milk” meant approximately “(a kind of) liquid.” No similar meaning suggests itself for Germanic. The few Sanskrit words that have been cited in this connection do not inspire confidence. Milk cannot be a legitimate cognate of mleko because Germanic k should correspond to non-Germanic g. For the same reason, Latin melca is a bad match for Germanic milk (by the First Consonant Shift, so often invoked in this blog, Germanic k should correspond to non-Germanic g).

The story is disappointing. Over a huge territory, people used the m-l-k word for “milk” and “milking,” which often turn out to be partly unrelated, with most of them being of obscure origin. By way of compensation, I may add that the cow played a noticeable role in Indo-European mythology. In the Rig Veda, a collection of ancient Vedic hymns, a dairy cow and its milk are often used as symbols of world fertility. (If some of the most ancient beliefs connected milk and fertility, the idea of taboo again begins to look less fanciful, because saying aloud words connected with procreation is often prohibited). According to the Scandinavian creation myth, the giant Ymir was licked from the ice by the primordial cow Auth-humla, whose name is a matter of speculation. It must also be a great comfort to our readers that the etymology of the noun cow is indeed known (no connection with milk).

Not versed in Sanskrit etymology, I would dare make a suggestion, probably made and rejected many times before me. Sanskrit dádhi “yogurt” (as pointed out by specialists, perhaps of cultic significance because of a priest’s having a name with the same root) sounds very much like Gothic daddjan “to suckle.” Daddjan is usually traced to the form dajjan, but this verb seems to be such an obvious baby word (“a word of infantile origin,” as serious sources put it) that I doubt the validity of the current etymology, which is perhaps too learned. Surprisingly, in the many words for “milk” in Indo-European, no references to baby feeding has turned up, but this is only one of the many riddles briefly touched upon above.

_{Featured image from State Library of South Australia via Wikimedia Commons, CC BY 2.0.}