29 Jan 2009

Nouns modifying nouns in Etruscan and related languages

I've been noticing a quirky grammatical pattern for a while in Etruscan but for the longest time didn't know how to interpret it. My journey started back at the beginning of my studies of this language when I was confused about why calendar dates and ages showed each element of a complex number declined in the same case ending. For example, in TLE 181, we read avil-s XX tivr-s śa-s meaning "of 20 years (and) of 6 months". In these instances, despite the meaning being clear, the grammar remains to be explained by specialists. If there is only one postposition in the English translation, why is the head noun tivr 'month' and the modifying numeral śa 'six' declined with the very same case ending to express 'of 6 months'?! The same curious phenomenon appears to exist in names where both a first and last name agree in case ending (eg. TLE 84: Larθi-ale Hulχnie-si Marce-si=c Caliaθe-si 'For Larth Fulchnie and Marce Caliathe'). If we dare to expand this phenomenon to all instances where nouns appear to modify other nouns, we help explain the complex grammatical structure of the noun phrase śacni-cś-tre-ś cilθ-ś attested in the Liber Linteus.

Now, I'm starting to wonder if this is not a special feature of all languages of the Aegean family, including Lemnian. On the Lemnos Stele, an as-yet undeciphered phrase is inscribed: aker tavarśiu vanalasial śerunai murinail. The overwhelmingly popular pet theory that has been published ad nauseum for decades is that murinail must mean "of Myrina" since Myrina is an ancient city on the Lemnos island. As admittedly intoxicating as this translation is, it has never been explained why we don't read a more grammatically acceptable *Murina-l instead. Indeed, I don't think anyone in their exuberant zest has bothered to notice their tiny grammatical faux-pas. Any acceptable analysis of this Lemnian phrase must account for the intervening iota before this perceived genitive ending -l.

So back to this grammatical pattern I'm observing in Etruscan, perhaps we can use this knowledge to help crack this Lemnian phrase in a more grammar-conscious way. Notice that we first have two possible noun phrases both inclined in the genitive-II ending: [tavarśiu vanalas]-ial and [śerunai murinai]-l. In turn, we might suspect that the second phrase is composed of yet another layer of case agreement, this time in the locative case: [śeruna]-i [murina]-i. *Vanalas may be understood as "of Venel (praenomen)".

If we piece this all together we get the following rough but promising partial translation of this Lemnian phrase: "an [aker] of Venel's [tavarśiu] (that is) before (the) [śeruna murina]."

8 comments:

  1. Wow, a sort of 'embedded' casemarking.

    Does that occur at all in other languages? I've never seen such a thing. Nevertheless it does seem to make semantical sense here.

    ReplyDelete
  2. Perhaps the seemingly redundant "-s" is some kind of reinforcement placed at the end of a noun phrase? Marking the end of the phrase?
    Could it be a relic that suggests that lists were originally usually done in threes? (Kinda grasping at thin air on that one!)

    Also, the would-be "genitive II" marking shows up as -ial on "vanalasial", but only -l on "murinail". Why? Are the two vowels dropped if there's already a vowel there?
    And is it possible that the "-a"s in "[śeruna]-i [murina]-i" are suffixes?

    ReplyDelete
  3. PhoeniX: "Does that occur at all in other languages?"

    Up until today, I wasn't sure. However I found out a new word today: "Suffixaufnahme". Try reading Suffixaufnahme [pdf] by Marcus Kracht. Old Georgian is apparently one example of such a language.

    ReplyDelete
  4. Seadog Driftwood: "Perhaps the seemingly redundant '-s' is some kind of reinforcement placed at the end of a noun phrase?"

    Nope. In Etrucan -s is a type-I genitive marker.

    Seadog Driftwood: "Also, the would-be 'genitive II' marking shows up as -ial on 'vanalasial', but only -l on 'murinail'. Why?"

    That's an excellent question. Based on internal reconstruction, we know that the genitive-I ending was once *-(a)si, the genitive-II ending was *-(a)la and the locative was *-(a)i in a Pre-Proto-Etrusco-Lemnian stage prior to an instance of syncope around 1000 BCE or so (ie. not the second syncope that occured around 500 BCE in Etruscan).

    So, *-(a)si-i and *-(a)la-i yield the composite Etruscan and Lemnian suffixes for the dative, -(a)si and -(a)le. If we combine genitive-I *-asi and genitive-II *-ala together, we arrive at *-(a)si-ala (> -(a)sial).

    However if I'm correct about this latest analysis of mine of Lemnian murinail, then Pre-Proto-Etrusco-Lemnian locative *-(a)i plus genitive-II *-(a)la combine to form *-(a)i-la which is reduced to -i-l as we find here.

    Seadog Driftwood: "And is it possible that the '-a's in '[śeruna]-i [murina]-i' are suffixes?"

    No. Fantasizing about imaginary suffixes is unproductive when you could actually read some basic points about Etruscan grammar from a reasonable source like Pallottino or Bonfante. These two words are probably both given the pertinentive suffix -na and are nouns which, when combined together like this, are signifying a single concept.

    In other words, I expect that śeruna murina (śeruna being the head noun) is a noun-noun, compound-like construction much like the ones we take for granted in English such as 'money matters' or 'funeral parlor'.

    ReplyDelete
  5. Glen Gordon: Nope. In Etrucan -s is a type-I genitive marker.

    Actually, what I mean by "the seemingly redundant '-s'" was in response to "If there is only one postposition in the English translation, why is the head noun tivr 'month' and the modifying numeral śa 'six' declined with the very same case ending to express 'of 6 months'?!" The "redundant '-s'" was the one that didn't have a postposition in the English translation. Not the suffix "-s" in general. Just to avoid confusion... heh...

    Fantasizing about imaginary suffixes is unproductive when you could actually read some basic points about Etruscan grammar from a reasonable source like Pallottino or Bonfante. These two words are probably both given the pertinentive suffix -na and are nouns which, when combined together like this, are signifying a single concept.

    Well, I was close... it was part of a suffix.

    I'll be sure to check on Pallottino's work and re-read Bonfante's stuff (been a loooooooong time since I looked at it, and I remember very little of the grammar). Thanks!

    ReplyDelete
  6. Seadog Driftwood: "The 'redundant "-s"' was the one that didn't have a postposition in the English translation. Not the suffix '-s' in general. Just to avoid confusion... heh..."

    Alright, fair enough. At any rate, considering the pdf on this Suffixaufnahme and "group inflection" that I just uncovered, this -s can now be perfectly explained as a natural linguistic quirk that occurs in other languages like Sumerian and Georgian. It's better to explain something with a clear reason than to write it off as random "redundancy" afterall.

    ReplyDelete
  7. Suffixaufnahme or "Double-case agreement" is a well-known phenomenon. It is one of the features of the so-called "Anatolian linguistic area" and exists in Hurrian, Urartian, Hittite and Luvian, namely in possessive constructions.

    You might want to read:

    Wilhelm Gernot, (1995) "Suffixaufnahme in Hurrian and Urartian" Pp. 113-35 in Plank 1995, available online at:
    http://books.google.com/books?hl=pt-PT&lr=&id=0EDl8_gmmvQC&oi=fnd&pg=PA113&dq=double-case+agreement+gernot&ots=ygVQJzyaAM&sig=MZGoTCHeZustnt-rf2pWNYYTmfU

    Luraghi, Silvia (1993), “La modificazione nominale nelle lingue anatoliche”. Archivio Glottologico
    Italiano 68 (2): 145-166.

    ReplyDelete
  8. Yes, thanks for the link, Miguel. And since I know some HTML, I can make that more click-friendly for the URL-impaired who may chance upon my blog and read this in the future:

    Gernot Wilhelm, Suffixaufnahme in Hurrian and Urartian (1995)

    If I might theorize here... Such an areal feature would seem to me to have developped in a Proto-Etrusco-Cypriot stage when these dialects were still contained in Western Anatolia and Cyprus during the second millenium BCE. Very interesting indeed! I suppose I now have a very strong reason to bone up on Hurro-Urartian.

    ReplyDelete