Vasconic words for “milk” (2): Basque ezne, esne

ezne, esne is the only generic word for “milk” in Basque, and it refers to both human and animal milk.

Bengtson usefully lists the spread across the dialects as follows:

Meaning: 1 milk 2 sap

Araban: ezne 1

Bizkaian: ezne, esne 1, 2

Gipuzkoan: ezne, esne 1, 2

High Navarrese: ezne, esene 1

Low Navarrese: ezne, esne 1

Salazarese: ezne 1, 2

Lapurdian: esne 1

Baztanese: ezne 1

Zuberoan: eznẽ́ 1

Roncalese: ezne 1, 2

Three things to note:

(1)  The word is present and means “milk” in all dialects. In four dialects (B, G, Sal and R) it can also mean “sap”;

(2)  Five dialects (Araba, Salazar, Baztan, Z and R) exhibit only laminal /s/, one dialect (L) exhibits only apical /s/, while four dialects (B, G, AN and BN) exhibit both grades of /s/; and

(3)  One dialect (AN) has a trisyllabic variant, esene.

Bengtson reconstructs ezne, esne as *e=śene.

Bengtson presumably does this on the basis of (AN) esene. This seems reasonable. /zn/-/sn/ is an unusual consonant cluster in Basque, so where it occurs it is likely to be the result of the removal of a vowel. The extrinsic data support this surmise (see above and below).

I also believe that Bengtson is right in treating the initial /e/ in ezne, esne as a fossilised class prefix. He does this on the basis of likely cognates in non-Vasconic Dene-Caucasian languages, and I agree with him (see below).

However, I disagree with Bengtson that the apical /s/ is original. He takes this view on the basis of the sibilants present in the likely North Caucasian reflexes. My view is that the time-depth is far too great for such precision of phonetic comparison to be greatly helpful. We do, in any event, have a probable Iberian antecedent that suggests that the laminal sibilant is original.

The probable Iberian antecedent to which I refer is to be found on the Solaig Lead Foil (Solaig, Valencia), which is a roll of donors of grave goods written in the East Iberian Script. This is an anthroponym, Semerun.

Semerun is composed of the No 1 S-sign, followed by the E-sign, followed by the N-sign, followed by the No 2 M-sign, followed by the No 2 R-sign, followed by the U-sign, followed by the N-sign. The No 1 S-sign indicates the laminal sibilant. The No 2 R-sign indicates fortis /r/. The No 2 M-sign is both syllabic and multi-aspect. Placing the N-sign before the No 2 M-sign indicates that it does not incorporate a preceding vowel. The correct reading is therefore semerun (here used as an anthroponym).

Iberian semerun resembles Basque (AN-goiz) zemeron “soft cheese”.

Let us take a look at this word in more detail. Bengtson mentions it when discussing ezne, esne, but rejects the notion that it incorporates ezne, esne as its first element. His reason for so doing I believe to be ill-founded, and I will say why shortly.

Bengtson lists the following variants:

Meaning: soft cheese, cottage cheese, cream cheese

(AN-Etxalar) zenberun

Low Navarrese: zenbera, (arc) zenberauen

(BN-Salazar) zenbera

(AN-Baztan) zendereben, zenderen

(Z) zenbéra

(R) zénbra

Unfortunately, Bengtson misses two variants that are quite important:

(BN) zenbron

(AN-Goizueta) zemeron (as cited supra)

(Both the above are listed in Azkue.)

Bengtson reconstructs this word as *sen-[-bera].

Not quite. While Bengtson is right about the laminal sibilant (Solaig confirms this), he is wrong about the second part of the compound. It is of the first importance when analysing Basque words with final vowels to be certain that there is no evidence for a missing final /n/. In the case of zenbera, etc, there is evidence of a final /n/, and Solaig confirms that the final /n/ is original.

I believe that the correct reconstruction is *sen-berun, *sen-beron, and that the word is a compound comprising *e-sene “milk” and berun, beron “lead”. Basque, as does Iberian, reduces *e-sene to *sen in this compound. Basque berun “lead” (lenis /r/) exhibits fortis /r/ in Iberian, hence semerun (with fortis /r/). Iberian beron “lead” (fortis /r/) is attested as an anthroponymic compound element, as at Mogente (Valencia).

As I have mentioned, Bengtson considers that ezne, esne and the first element in zenbera are not the same word, and does so on the basis of the supposed presence of a North Caucasian cognate that exhibits a different sibilant to the much more robust North Caucasian cognates of ezne, esne (see below). However, a Basque word for “soft cheese” cannot correspond in both form and meaning to a North Caucasian word because the time-depth is far too great. Cheese is a post-agricultural product of quite recent origin. So recent, in fact, that both Germanic and Celtic use a Latin loanword.

So, let us look at the best non-Vasconic Dene-Caucasian cognates of Basque ezne, esne which Bengtson cites:

North Caucasian

Meaning: udder

Chechen: šin

Ingush šim

Batsbi šĩ

Meaning: milk

Andian š:iwu

Akhvakh s:ō

Chamalal s:ī̃w

Tindi š:ū

Karata š:iw

Botlikh š:ĩʔu, š:iʔũ

Bagvalal š:ū

Godoberi š:ĩwu

Burushaski

Meaning: milk

Hunza ṣiŋ

Nagar ṣiŋ

Yeniseian

Meaning: nipple, milk

Kottish ten, pl. tanaŋ “nipple”

Arin téŋul “milk”

Pumpokol den “milk”

(Yeniseian exhibits a reverse assibilation.)

These reflexes do tend to indicate that the initial /e/ exhibited by ezne, esne is a fossilised class-prefix and not part of the substantive word. So Bengtson’s view is confirmed on that point.

It is important to appreciate that the common antecedent of all these words referred to human milk alone. Milk was not extracted from animals until long after the common ancestral language dispersed, and indeed up to 5,000 years after Vasconic dispersed across Europe.

Does ezne, esne have cognates in other Vasconic languages?

One would expect that to be the case.

The instance that stands out is the River Shannon, the longest river in Ireland. This was attested as SENOS by Ptolemy.

Is it possible to have rivers whose names mean “milk”? Yes. There is the village of Melchbourne in Bedfordshire, whose name means “milk stream”.