17 July 2013

Basque mahats 'grape' (updated)

Basque mahats 'grape' derives from an earlier *baka-tś related to Latin bāca 'berry'1, with nasalization of the initial labial and lenition *-k- > -h- as a consequence of Martinet's Law, by which Paleo-Basque fortis (i.e. voiceless) stops became h or zero (mostly at word-initial but in the case of the velar stop also medially)2. The second member of the compound is a phytonym suffix *-ts found in other words such as e.g. isats 'broom'.

The Latin word has no IE etymology, but Boutkan-Kossmann have proposed a link to Berber *bqā 'blackberry, mulberry'3. In my opinion, this would be a Wanderwort also found in Kartvelian *maqˀw-/*muqˀw- 'blackberry' (Georgian maqˀv-al-, Megrel muʔ-, Svan muqˀw, Laz muqˀ-)4 and possibly also Burushaski *maɣar- 'unripe mulberry', Lezghian *niwqˀ:a(j) and Lak qˀul-nuqˀi 'strawberry'. 
1 There's also Hispanic Latin bacca 'wine' (Varron). Also Galician-Portuguese bago 'grape' (also found in parts of Leonese) derive from an unattested masculine variant.
2 Unfortunately, academic Vascologists think the only source of -h- can be a nasal *-n- (Mitxelena's Law), so they reconstruct a protoform *banatś. See R.L.Trask (2008): Etymological Dictionary of Basque (unfinished).
3 D. Boutkan & M. Kossmann (1999): Some Berber Parallels of European Substratum Words, §3.1, in JIES 27, p. 87-100.
4 Borrowed into Akhvakh muqˀ:ali.


  1. What do you think of Bengston's etymology?

    Proto-Basque: *mahanć
    Meaning: grape
    Bizkaian: maats, mats
    Gipuzkoan: mats
    High Navarrese: mats
    Low Navarrese: mahats
    Lapurdian: mahats
    Zuberoan: máhãts
    Roncalese: mats

    The higher connections are:

    Proto-Sino-Caucasian: *ʕämćṓ
    Meaning: a kind of fruit
    North Caucasian: *ʕämćō
    Sino-Tibetan: *ćh(r)iǝ̆H ( ~ ʒ́h-)
    Burushaski: *mićíl / *bićíl
    Basque: *mahanć

  2. Hi, Adyghe.

    It seems Burushaski could be related to Caucasian, although I'm not so sure about Sino-Tibetan. I'm afraid this is a flaw of many Starostin's Sino-Caucasian etymologies.

    On the other hand, Bengtson's etymology is a *pile of rubbish*. Not only he doesn't indentify the Basque phytonym suffix -ats found in e.g. isats 'broom', but he also assumes the word underwent a weird metathesis.

    Generally speaking, Bengtson's approach is flawed. First of all, he has little or none understanding about Paleo-Basque phonology and the fact than Basque has several lexicon layers apart from what can be regarded as the native core. This includes loanwords from IE, especially (but not limited to) Celtic, and also from Afrasian (Berber).

    Secondly, even in cases where a Basque word has genuine Sino-Caucasian cognates, he ignores the forms in IE languages, i.e. the "stepping stones".

  3. Doesn't the nasalization of *b indicate that h < *n?

    1. Not necessarily so, as there's a tendency for ba- to become ma- as in e.g. magina, makailao, makila. The thing is by no means all the sound changes in Basque are regular in Neogrammarians' sense (with which academic Vascologists since Mitxelena are obsessed) and sometimes are even contradictory. For example, in some words the initial labial stop becomes l-, which is the usual result of d-. Thus we can have barca > marka (B) and at the same time *banka > lanka (B).

      In my opinion, this would reflect Basque wasn't a homogeneous language but the result of the interaction of several linguistic varieties, possibly involving some degrees of creolization.

  4. Why go with a speculative etymology when a non-speculative one is readily available?

    According to Trask there's a rule about expressive words (and those dealing with bumps or humps) where *b > m before a back vowel followed by a dental or velar plosive followed by a liquid (though he gives many examples where the order of the plosive and liquid are reversed).

    1. Why go with a speculative etymology when a non-speculative one is readily available?
      If you refer to *banatś, this isn't even near to an actual etymology.

      According to Trask there's a rule about expressive words
      Trask is merely describing a bunch of words which don't conform to Mitxelena's Proto-Basque phonotaxis, and so he labels them as "expressive".

      In addition to the frequent ba- > ma-, nasalization can be triggered by a back vowel in the 2nd syllable, as a particular case of vowel harmony: beko > moko, berroilo > morroilo.