On the IE root of άμπελος (vine)

sotos

Senior Member
Greek
Some online etymological dictionaries claim that AGr. άμπελος (vine) comes possibly from a "pre-Greek substratum". Am I the only one to see the similarity between the IE root *ab(e)l- (apple etc) and Gr. "ampelos"? Notice the Hittite "mahla" (cognate to malus) = "branch of grapevine" and the spread of *ab(e)l to semitic languages with the wider meaning of "fruit". (see VACLA V BLAZE K INDO-EUROPEAN „APPLE(S), p. 18. ). Is there any reason to exclude *abel- as the root of ampelos? Thanks.
 
  • There's also the theory that it comes from the v. «εἰλέω» eiléō --> to roll, turn, wind, revolve (PIE *u̯el- to turn, wind cf. OIr. fillim, to fold, return, Proto-Germanic *waltijan > Ger. wälzen, Eng. welt). Of course, the prefix «ἄμπ-» ắmp- is inexplicable (although some suggest it's from the preposition «ἀμφί» ămpʰí-).
     

    Penyafort

    Senior Member
    Catalan (Catalonia), Spanish (Spain)
    How common is it for Ancient Greek to develop a nasal in that context? If that is the only case, then it can certainly be 'arguable'.
     

    Linnets

    Senior Member
    I saw that. But why the existence of a common root for several phytonyms (apple, ampelos, avellana etc) is "arguable"?
    First of all avellāna ('hazelnut') comes from Abellānus 'from Abella' (an Oscan town in Campania). It is possible however that Abella in turn could mean 'apple-bearing' (there are a lot of apple orchards in that zone). As for apple and ἄμπελος are concerned, they're probably unrelated; it's more likely that Greek ἄμπελος and Latin pampinus ('vine leaf') come from a Mediterranean stem *amp- also believed by some (Devoto) to be the source of Italian lampone ('raspberry').
     

    Delvo

    Senior Member
    American English
    I saw that. But why the existence of a common root for several phytonyms (apple, ampelos, avellana etc) is "arguable"?
    The PIE word you can reconstruct for it, *h₂ébōl, doesn't look like a PIE word.

    1. They don't usually have two syllables with the vowels separated by only one consonant, or with the second syllable not being recognizable as a familiar PIE ending/suffix.

    2. The loose similarity between "bol" and the other set of words that have been mentioned here, like Latin "mal", suggests a relationship, but, if there is one, it's not one that followed from PIE to later IE languages by the usual sound-shift rules; there's no "b→m" rule or "o→a" rule in Italic languages, for example. So something else must have happened, like different IE branches importing these words at different times, possibly in different forms from different neighbors.

    3. There are practically no known occurrences of a PIE *b; it's so rare that its rarity makes any proposed examples seem out-of-place. Later IE languages' sounds that might otherwise be expected to have come from it normally turn out to have come from something else instead.
    • In Indic languages, some sets of neighboring sounds could de-aspirate a *bʰ.
    • In Greek, the typical origin of /b/ was from PIE *ɡʷ.
    • In rural dialects of Latin, it could come from PIE *ɡʷ, as in "bos/bovinus" (which was then imported to the speech of the city), but Latin is mostly the language of the city, where a more common origin for /b/ was earlier /d/ followed by a labial, as in "duellum→bellum", "duonus→bonus", and "dui(s)→bi(s)". In the prefix "ad", it was voiced from a PIE *p.
    • In Proto-Germanic it would have become *p, but the original PIE hole in the pattern follows Grimm's Law too, so PG has a hole in its pattern at *p. English has lots of words with "p" now, but they're just about universally imported from other languages.
    ...and so on.

    * * * * *

    Another reason that I think applies but am not completely sure of is that converting a voiced plosive to a nasal followed by an unvoiced plosive is not a normal sound shift in Ancient Greek. Just the fact that they seem similar isn't enough; the conversion would need to be consistent with other phonetic conversions that happened in the same language at about the same time, and I don't believe this type of conversion was happening in Greek at that time.
     

    sotos

    Senior Member
    Greek
    The PIE word you can reconstruct for it, *h₂ébōl, doesn't look like a PIE word.

    1. They don't usually have two syllables with the vowels separated by only one consonant, ...
    Thanks for this opinion. I am not a glossologists, but from the point of view of pure sciences some of the above arguments seem to work as circular logic: "It is not usual > it is excluded from the mainstream > it remains unusual".
     

    Linnets

    Senior Member
    There are practically no known occurrences of a PIE *b; it's so rare that its rarity makes any proposed examples seem out-of-place. Later IE languages' sounds that might otherwise be expected to have come from it normally turn out to have come from something else instead.
    That is what can be found on the Wiktionary, not the most reliable existing website. However, also Online Etymologic Dictionary yields *ab(e)l-.
     
    Last edited:

    Delvo

    Senior Member
    American English
    That is what can be found on the Wiktionary, not the most reliable existing website. However, also Online Etymologic Dictionary yields *ab(e)l-.
    The combination of these two sentences with "however" makes it seem as if you intended to quote a reference to the hypothetical PIE root as *h₂ébōl, instead of the part of my post which you did quote about the rarity of PIE *b. So I'll respond on both subjects.

    *h₂ébōl and *ab(e)l are the same thing. The latter doesn't specify vowel accent or length, so it doesn't contradict the accent & length markers in the first, so we can ignore those for this comparison. It's routine for a PIE vowel to appear as *e in some conjugations/declensions and *o in other conjugations/declensions, because part of how PIE grammar worked involved switching between the two (or omitting the vowel) for different grammatical roles, like the difference between "sing", "sang", "sung", and "song" in English. So, when reading a PIE word as they're written in the modern world, one must always understand that an *e could have sometimes been an *o and an *o could have sometimes been an *e. (Sometimes only one or the other survives in later IE languages in particular words, but the grammatical *e/o alternation was still active in Greek and Sanskrit grammar, and maybe early Latin too, so there is no doubt about the switching back & forth in PIE.) The usual standard for someone writing a PIE word is to write *e by default and leave it to the reader to understand that that would sometimes come out as an *o, but some authors also write *o sometimes, particularly when that's the one which later IE languages' words seem to have mostly or exclusively come from, and then it's still up to the reader to understand that it could have originally been *e sometimes as well. And, once linguists determined how to spot where laryngeals (h₁, h₂, h₃) had been, it turned out that *a was always a place where one of those *e/o spots had *h₂ next to it; when the *h₂ disappeared, it affected the adjacent vowel by turning it into *a. So *h₂e is *a, just at an earlier stage.

    On the subject of the part you actually quoted, and which another person also definitely disputed even if you didn't mean to: I don't know what somebody/anybody could think is wrong with it or why. The conspicuous gaps in the sounds of later IE languages, and what hypothetical sound they would all correlate with in PIE, are not in doubt or dispute. Greek really does have a conspicuous shortage of /b/ that correlate with anything other than *gʷ in PIE. Latin really does have a conspicuous shortage of /b/ that don't come from earlier Latin/Italic "dua/due/dui/duo" or *gʷ in certain ancient dialects or *p in often-voice-inducing environments like between vowels. Germanic languages really do have a conspicuous shortage of /p/ that aren't imported from non-Germanic languages. Indic languages really do have a conspicuous shortage of /b/ that correlate with any of these other missing sounds in other IE languages instead of something else like *bʰ in certain kinds of environments with the potential to de-aspirate (with the usual counterparts in other branches like Greek /pʰ/ and Germanic /b/). The PIE counterpart for all of these (and other) known holes in more recent IE languages ("p" in Germanic, usually "b" in the others) really would be PIE *b. This is all standard ordinary knowledge among Proto-Indo-Europeanists, probably findable in any book on PIE, not the subject of any doubt or dispute.
     
    Last edited:
    Top