25 Feb 2010

Capua, fields, Indo-Europeanists, etc.

The following is a good example of Indo-Europeanists gone mad, methinks. Whenever it comes to European vocabulary, specialists of Indo-European languages are in there like a dirty shirt trying to etymologize it automatically through some concocted Indo-European root. Some roots are valid naturally, but others are suspicious and crave our skepticism.

Specialization can sometimes lead to shortsightedness. We can see why not all words we find in European languages are necessarily inherited from a language spoken 6000 years ago. There are many sources of vocabulary in any language and to assume a priori that words are automatically inherited is the same fallacious tactic used by zealous Nostraticists to claim an overabundant number of terms have some great heritage that never existed. It's much more likely that any given word, when going back six millenia, is a borrowing rather than an untouched, ancient word. Yet IEists, because of their passionate obsession, risk being ignorant of the contribution that must have also been made by Semitic, Egyptian, Basque, Aegean and other language groups. They may often stick to what they know, namely Proto-Indo-European, and attribute everything to it alone.

Enter the etymology of Capua, a city in Italy which was originally controlled by Etruscans. The Latin name was identical and called in Greek Καπύη (Kapuē). Charlton Lewis, to which this reference in Perseus owes origin, went straight to the punch and pointed the reader to Latin campus meaning 'field'. When we next look up campus, we're led to a further connection with Greek κῆπος (Doric κᾶπος).

This is interesting because further investigation reveals contradictions about the etymology of these words among IEists. Mallory and Adams have attributed κῆπος to an Indo-European root *ḱāpos 'garden'[1] while Pokorny had claimed a different root *kam-p- 'to bend' for the Latin reflex while dragging in another Greek word καμπή 'a winding (of a river)' as alleged cognate. I believe both reconstructions are likely incorrect. Pokorny's is easiest to dismiss since the relationship between campus and κῆπος is direct, phonetically and semantically, yet his etymology would absurdly contradict this in favour of 'bending' the meaning of these words to suit a nuance that isn't attested. M&A's reconstruction at least acknowledges the core meaning in these words, 'cultivated land', but then, despite this, the connection with Capua and Campania put forth in Lewis' work is left unaddressed but worthy of a solution. We're also offered a highly unlikely PIE root with a long low vowel which is odd considering that even short *a is rarely attested in PIE. Should we additionally assume a laryngeal before *p? Or perhaps should we accept the problems of these attempts to source these words to PIE and try a fresh approach?

Let's look outside the PIE bubble for argument's sake and try this from an Aegean perspective. I want to suggest an Etruscan word *capa and assign it the meaning of 'field'. This word then can serve as a direct source for Latin campus with homorganic nasal resonant -m- added. As an inanimate noun, the Etruscan plural is predictably *capava 'fields'. Coincidently, this is precisely the name of Capua in Etruscan, attested in the locative case as Capue (TLE 890) (from Old Etruscan *Capava-i). The citation form, *Capava, would have effectively meant 'The Fields', referring to the scenic topology of the area. After solving the meaning and source of Capua, we can push on to reconstruct the Etruscan name for Capua's surrounding region, *Capavana 'Campania', economically solving for the source of the Latin name too. Finally, this opens doors in regards to the exact source of the Greek words for 'garden', ie. 'Pelasgian' substrate, an idea already suspected by Robert Beekes.


NOTES
[1] Mallory/Adams, The Oxford introduction to Proto-Indo-European and the Proto-Indo-European world (2006), p.163 (see link).

19 comments:

  1. In defense of other IE-ists, the reconstructions in that Mallory and Adams book are *terrible*, and I've often found the etymological material in the Lewis and Short Lexicon to be speculative and unsystematic, but hell, it was made before the Neogrammarians, so it's hardly unexpected.

    I think Beekes is probably onto a lot of good stuff with his Pre-Greek substrate research.

    ReplyDelete
  2. The fact that long a in PIE is rare does not preclude the existence of that root, especially not if *ḱēh2p- is presumed instead; a lengthened e in some roots is fairly well-attested.

    I'm not averse to this theory per se, but I don't see why it has any particular explanatory value over reconstructing a PIE root for the two. What puzzles me here is the fact that, in your theory, Latin turned Etruscan *capa -> campos (why would this be borrowed as a second declension?) and *Capavana -> Campania (ignoring the disappearance of the -v- there) but left *Capua as such. Given that *Capua and *Capavana most likely would have been heard by the same group of people and integrated into the language at approximately the same time (ie. when relations with Campania become important) it seems odd that the nasal would be added there, yet not in the mostly-identical name of the city.

    ReplyDelete
  3. Ethan Osten: "The fact that long a in PIE is rare does not preclude the existence of that root,[...]"

    Good, now replace 'long a in PIE is rare' with 'He is invisible' and replace 'that root' with 'Jesus' and reflect on the insanity of rejecting Occam's Razor in favour of blind faith.

    It's not just about the unlikely PIE vowel but you do need to accept that this in itself is an important fact about the language. So when someone proposes an **already weakly substantiated root** (note that these alleged cognates don't even hold water) by sticking a provably rare and possibly non-existent vowel in the root, our skepticism is earned.

    "especially not if *ḱēh2p- is presumed instead"

    This is impossible given Eichner's Law (ie. laryngeals fail to colour to ). You could only mean *ḱeh2p- with short *e but the root is poorly justified regardless.

    "What puzzles me here is the fact that, in your theory, Latin turned Etruscan *capa -> campos (why would this be borrowed as a second declension?)[...]"

    If it was simply about the terminating vowel, how would you explain Latin catamitus from Etruscan catmite (ET Ta S.12) then? There's another possible example of borrowing: Etruscan tupi (TLE 89) > Latin tōfus.

    "[...] *Capavana -> Campania (ignoring the disappearance of the -v- there) but left *Capua as such."

    Since Etruscan v represents [w] (as in Latin), the phonetic difference between Etruscan ['kɑpəwa] and Latin Capua is entirely moot. And perhaps you should look up Capua at the Perseus website and read carefully what it says. Notice the adjectival forms listed: both Capuānus and Campanus.


    "[...] it seems odd that the nasal would be added there, yet not in the mostly-identical name of the city."

    Interestingly the very word for '7' in Etruscan, semφ, which is an obvious borrowing, implies a similar spontaneous addition of nasality during transfer. It can happen. Borrowings are sometimes unfaithfully reproduced in the donor language.

    ReplyDelete
  4. Mattitiahu: "In defense of other IE-ists, the reconstructions in that Mallory and Adams book are *terrible*[...]"

    I definitely agree that this is a poor source but it's impressive-looking source and backed up by impressive names and degrees, so I feel an urge to rag on it a lot. ;o)

    The dilemma here is how do we label in a scientific way which IEists are acceptable and which pure doodoo? It's too subjective, so I just focus on the statements published by what are, regardless of any feelings we may have, accepted 'IEists' with noteworthy careers.

    Adams got his from the University of Chicago, Mallory from the UCLA. As such, like or hate it, these universities have given them approval regardless of their scholarship and they are indeed 'IEists' in every sense of the word (...which only plays quite handsomely into my anti-elitist, anti-credentialist stance).

    And yes, Beekes may not have conclusive answers but I can't help but htink there are leads to new information in his work lurking about.

    ReplyDelete
  5. "Good, now replace 'long a in PIE is rare' with 'He is invisible' and replace 'that root' with 'Jesus' and reflect on the insanity of rejecting Occam's Razor in favour of blind faith."

    This is definitely the funniest thing I've read so far this decade. Congrats. :)

    Anywhoo... I agree it's difficult on the `reliability' categorization of the work of Indo-Europeanists. Lehrman reports that Cowgill was accustomed to say that "it was easy to be an Indo-Europeanist, much easier than to be say, a physicist, a chemist, or an engineer, because and Indo-Europeanist could come up with any kind of fantastical nonsense and get away with it..." (Cowgill Collected Writings:xliv).

    It seems to me, at least, that the joint publications of Mallory and Adams (like this book) are more set out to reach out to the public on this obscurantist discipline presenting as much reconstructable data as possible, even some of it isn't 100% sound.

    In any case, please go ahead freely with your anti-elitist and anti-credentialist rants. I like them. :)

    ReplyDelete
  6. Glen Gordon: "Good, now replace 'long a in PIE is rare' with 'He is invisible' and replace 'that root' with 'Jesus' and reflect on the insanity of rejecting Occam's Razor in favour of blind faith."

    Amusing, but wrong. If we assume the existence of PIE *a/ā at all, then the fact that it is posited in any given root is not a priori evidence of that root being dubious. Otherwise, if the *a/ā that seems attested by certain roots is posited to be a reflex of something else, then we have to look at other laws to see whether that reconstruction is dubious or not. You can't simply say "I don't think *a existed in the language, ergo any reconstructions that suggest *a are inherently suspicious" - you have to have better reasons than that, and you know it.

    Glen Gordon: "It's not just about the unlikely PIE vowel *ā but you do need to accept that this in itself is an important fact about the language. So when someone proposes an **already weakly substantiated root** (note that these alleged cognates don't even hold water) by sticking a provably rare and possibly non-existent vowel in the root, our skepticism is earned."

    Campus and κῆπος form a reasonable pair; that drift would be dramatically less than some other roots which are absolutely normal, and they match up phonologically as well.

    At any rate, while I do agree with you that the postulation of *ā in the root is reason enough to look into whether there could be another explanation (even if one does accept a rare vowel *a in PIE), it is certainly not enough to say so conclusively, much less to dismiss the idea offhand.

    Glen Gordon: "This is impossible given Eichner's Law (ie. laryngeals fail to colour *ē to *ā). You could only mean *ḱeh2p- with short *e but the root is poorly justified regardless."

    You're quite right; my apologies, for I wrote this response in a hurry (though that is hardly an excuse for forgetting that).

    Glen Gordon: "If it was simply about the terminating vowel, how would you explain Latin catamitus from Etruscan catmite (ET Ta S.12) then? There's another possible example of borrowing: Etruscan tupi (TLE 89) > Latin tōfus."

    I did not mean that parenthetical as an objection, merely out of curiosity. Catamitus, at least, can be explained through regularization of an obviously male noun (Ganymede) into the corresponding case; it is of course well-known that masculine first declension nouns are vanishingly rare, partly due to this. Campus does not have that excuse, nor even that it was maintaining the gender of Etruscan, which has no masculine-feminine divide at all.

    As for tupi -> tofus, I regard that one as too speculative to make judgment either way; it relies on a number of assumptions that, while quite possibly valid, are still assumptions, not all of which I'm comfortable with.

    (splitting because it won't let me post?)

    ReplyDelete
  7. Glen Gordon: "Since Etruscan v represents [w] (as in Latin), the phonetic difference between Etruscan ['kɑpəwa] and Latin Capua is entirely moot. And perhaps you should look up Capua at the Perseus website and read carefully what it says. Notice the adjectival forms listed: both Capuānus and Campanus."

    Those adjectives get us nowhere. The adjective in capuānus is clearly formed from Capua; compare the identical formation of Rōma -> rōmānus, complete with the lengthened -a-. The deletion of the i in campanus can easily be derived from Campania itself, just as siculus "Sicilian" from Sicilia.

    So essentially we have two roots being displayed here; capu- and campan-. Are they related? For certain. But there is a large jump indeed between ['kɑpɑwana] (your suggested *Capavana) and ['kampania]; you have to be assuming a pattern of syncope that just isn't present in the rest of the language.

    Glen Gordon: "Interestingly the very word for '7' in Etruscan, semφ, which is an obvious borrowing, implies a similar spontaneous addition of nasality during transfer. It can happen. Borrowings are sometimes unfaithfully reproduced in the donor language."

    I don't dispute that it can happen; I just think it very troubling when the word for a city and its surrounding region (again, presumably borrowed around the same time by the same group of speakers) gained a such a prominent spontaneous addition that was not regularized quickly by the sort of analogical you would expect. If both were posited to have gained the nasal, there would be no objection at all; it is the inconsistency.

    ReplyDelete
  8. Hello! I was just wondering, that what is this link's etymology based on, given your etymology for Capua? http://books.google.ca/books?cd=1&id=CRgiAAAAMAAJ&dq=etruscan+capeva+name&q=+capeva++#search_anchor

    ReplyDelete
  9. Paddy: "Hello! I was just wondering, that what is this link's etymology based on, given your etymology for Capua?"

    This is interesting: Cristofani, Tabula Capuana: Un calendario festivo di età arcaica‎ (1995), p.104, fn.19: "Quanto al nome di Capua COLONNIA [...] si pronuncia a favore di una forma antica capeva* plurale di cape, basata sul poleonimo capevane [...]"

    ReplyDelete
  10. Ethan,

    Please cut your comments down to below 20 pages.

    Both **kamp- and **ḱapo- appear to be rubbish not just because of dubious vocalism but because they depend on unverifiable semantic changes and on weak attestation in each individual branch.

    Comparatively, we can see that the pairing of Latin campus with Greek κῆπος, sharing both shape and semantics, is less assumptive than Pokorny's or A. & M.'s attempts so I see nothing constructive in your arguments on that regard or on your inefficient interpretations of Occam's Razor.


    "As for tupi -> tofus, I regard that one as too speculative to make judgment either way"

    You must not have read TLE 89: Tupi Sispeś 'The rock of Sisyphus'.


    "Campus does not have that excuse, nor even that it was maintaining the gender of Etruscan, which has no masculine-feminine divide at all."

    If I theorize *cape instead of *capa (as per Colonna), would you still object?

    ReplyDelete
  11. Hello Glen!

    I would like to ask a rather naive question: if *kapa did mean 'garden' in Aegean languages, then could some 'Pelasgian' toponyms, such as the river-name Kephissos be connected to it?

    As far as I know, there were at least half a dozen Kephissos rivers in ancient Greece. In the light of *kapa, I wondered if Kephissos could have meant something like 'green-ish'. Many of these Kephissos rivers are rather small, so indeed, they could have been green and stagnant over the summer period...

    The proposed derivation only assumes, that *kapa originates from an adjective meaning 'green' (that had an aspirated 'ph' in its stem), and that to the stem *kaph(a)-> *keph(a)-, a Pre-Greek (Aegean) suffix *-issos was attached (that could have replaced the original stem-vowel).

    ReplyDelete
  12. Bayndor: I would like to ask a rather naive question: if *kapa did mean 'garden' in Aegean languages, then could some 'Pelasgian' toponyms, such as the river-name Kephissos be connected to it?"

    Κηφισός Kēphisós is included in Hesychius' work, which seems to imply a Pre-Greek origin.

    Note that if Capua means 'The Fields', it directly parallels an origin posited for the toponym Ἴλιος (< Wilusa- < Hittite wellu- 'meadow').

    ReplyDelete
  13. I forgot to address Ethan's objections to spontaneous nasalization in campus from borrowing.

    I just remembered a pertinent example of this elsewhere: Latin mundus 'world'. This is considered by many experts to be a borrowing from Etruscan (see Baldi, The foundations of Latin (1999), p.166) yet the corresponding word available is muθ [LL 12.iii]. This may perhaps mean that there's subphonemic nasalization arising in certain environments in Etruscan.

    Consider also how the original Semitic name for Carthage was beautifully mangled by Greek ears into Καρχηδών Karchēdōn. Even though we prefer nice and neat borrowings, often irregular changes have nonetheless occurred, so what are we to do about that?

    ReplyDelete
  14. Nasalisation happens all the time:
    Long ago there was an Etruscan temple on the Campidoglio.

    Ethan objects to the inconsistency:
    "Given that *Capua and *Capavana most likely would have been heard by the same group of people and integrated into the language at approximately the same time (ie. when relations with Campania become important) it seems odd that the nasal would be added there, yet not in the mostly-identical name of the city."

    I agree this is odd (just as odd: if the Etruscans heard a nasalisation, but did not write it: why did it disappear in Capua?).
    Unfortunately linguistic "rules" can not be depended on.
    Frustrating. We (well, I) just can not say if this (disappearing) nasalisation is important or not.

    ReplyDelete
  15. Glen Gordon: "Both **kamp- and **ḱapo- appear to be rubbish not just because of dubious vocalism but because they depend on unverifiable semantic changes and on weak attestation in each individual branch."

    Given that you leave the door open for κεπος "garden" to be derived from the same root as campus "field", is your theory not reliant on the same unverifiable semantic change? If **capa was Etruscan for either "garden" or "field," it's hardly sporting to object to reconstructing a PIE root that means one of the two on the same grounds.

    Glen Gordon: "You must not have read TLE 89: Tupi Sispeś 'The rock of Sisyphus'."

    Or "the punishment." Given that tupi is a hapax legomen, and given Sisyphus' reward, a definitive conclusion cannot be easily drawn either way.

    Glen Gordon: "If I theorize *cape instead of *capa (as per Colonna), would you still object?"

    I don't object per se; I just find it interesting.

    Glen Gordon: "This may perhaps mean that there's subphonemic nasalization arising in certain environments in Etruscan."

    Nasalization that the Latins just happened to hear loud and clear when borrowing campus and Campania, but not Capua?

    Glen Gordon: "Consider also how the original Semitic name for Carthage was beautifully mangled by Greek ears into Καρχηδών Karchēdōn."

    Really, given the Phoenician Qart-ḥadašt [k'artħadast], it should come as no surprise. Given Greek had nothing even remotely approaching ħ (a voiceless pharyngeal fricative), a kh is as good a try as the th in Latin Carthago.

    ReplyDelete
  16. Ethan Osten,

    I've told you both the flaws of the PIE roots you're obsessed with and explained the benefits of reconstructing Etruscan *cape 'field' (amended as per your 'declension objection'). You ignore all of it then argue some more by repeating yourself and having people repeat things.


    "Or 'the punishment.'"

    Whatever. "Punishment" is only based on context; "tufa" is based on context and an etymology that Latin tofus otherwise lacks.

    Read Fiesel E.-Groth P.M., Etruskisch tupi und lateinisch tofus. SE 6, 1932, 261-272.


    "Really, given the Phoenician Qart-ḥadašt [k'artħadast], it should come as no surprise."

    Treating Qart-ḥadašt > Καρχηδών as perfectly natural while flipping out over a tiny 'm' in another borrowing is transparently unreasonable. Now what might your purpose be in being so unconstructively contrarian?

    ReplyDelete
  17. Glen Gordon: "I've told you both the flaws of the PIE roots you're obsessed with and explained the benefits of reconstructing Etruscan *cape 'field' (amended as per your 'declension objection'). You ignore all of it then argue some more by repeating yourself and having people repeat things."

    Because you haven't actually addressed concerns, you've just waved your hands and said "this is normal" or cited another theory of your own that has equally weak support. You just keep making up rules ("perhaps there was some pre-nasalization") as you go along; you might as well be begging the question to begin with.

    Given how quickly you dropped the discussion regarding the absent [w] in Campania, or that regarding garden vs. field, or even the mysteriously appearing m, I don't know where you get off accusing me of ignoring things.

    Glen Gordon: "Whatever. "Punishment" is only based on context; "tufa" is based on context and an etymology that Latin tofus otherwise lacks."

    The point is that you can't just go around saying "this word in Etruscan looks like the word in Latin, and it could potentially have the same meaning, so it must be the origin." Such thoughts based on formal resemblance are inherently speculative until you have something stronger than a meaning that you've made for it to fit your argument.

    Glen Gordon: "Read Fiesel E.-Groth P.M., Etruskisch tupi und lateinisch tofus. SE 6, 1932, 261-272."

    Given that it's not apparently in any online databases that I can find, and you're using the same citation (down to the abbreviations) found in Breyer (1993), I'm assuming you probably haven't actually read this article, and frankly, neither can I.

    Glen Gordon: "Treating Qart-ḥadašt > Καρχηδών as perfectly natural while flipping out over a tiny 'm' in another borrowing is transparently unreasonable. Now what might your purpose be in being so unconstructively contrarian?"

    No, because there's a difference between attempting to replicate sounds completely foreign to your language and trying to replicate sounds that are actually found in your language. ħ does not sound like anything in particular that the Greeks would be saying; in this light, realizing it as kh is very understandable. If Qart-hadast were in fact "Karthago" in Phoenician, and Greek transcribed it as Καρχηδών, you'd have an argument, but saying that it's not normal to represent foreign sounds idiosyncratically while saying it's perfectly commonplace to represent native sounds as something else just sounds like you don't know what you're talking about.

    ReplyDelete
  18. Ethan Osten: "Given that it's not apparently in any online databases that I can find, and you're using the same citation (down to the abbreviations) found in Breyer (1993), I'm assuming you probably haven't actually read this article, and frankly, neither can I."

    A simple google search finds many scholars referencing the same article from Studi Etruschi that you whine you can't find.

    This blog is for responsible and genuinely interested adults who have the competence to search out the necessary material for themselves, not whiners, petty arguers and obnoxious trolls.

    Next strike, you're out.


    "Given how quickly you dropped the discussion regarding the absent [w] in Campania,[...]"

    Surely you grasped this by now. Let's summarize:

    1. Capua lies in the core of Campania.
    2. Romans themselves used Campānus to mean 'Capuan' as well as Capuānus. Ager Campānus is synonymous with Campania itself.
    3. Etruscan cityname *Capua is supported by the locative form in TLE 890, after syncope.
    4. The varying gentilic forms (Capuvane [TLE 572], Capevanes [CIE 4283], Capvanes [TLE 15]) establish Old Etruscan *Capevana, and thus its root *Capeva, as it was before syncope.
    5. Capevana is one of the many Etruscan family names derived from regional names. Another is Umrina (cf. Umbria). Hardly shocking.
    6. Variation with and without -m- in these names is attested on small coins from 5th c BCE Capua: Kampano- ~ Kappano-.

    Some have even published the view that the "spontaneous -m-" is due to Oscan influence.


    "The point is that you can't just go around saying 'this word in Etruscan looks like the word in Latin, and it could potentially have the same meaning, so it must be the origin.'

    Yes, I can via explanatory power. Assuming a priori that the default answer can only be 'PUNISHMENT of Sisyphos' is baseless and just as random as a million other plausible options.

    However 'tufa block' not only fits with the inscription, image and myth but sources Latin tō[f/ph]us to Etruscan which is in turn sourced to related Near-Eastern terms preserving the original -i that Latin has apparently lost after borrowing it from Etruscan.


    "[...] ħ does not sound like anything in particular that the Greeks would be saying; in this light, realizing it as kh is very understandable."

    No, you're not being honest. Theta [tʰ] would be the closest approximation of [tħ] by far in Ancient Greek. No one in their right mind would predict chi as a result and yet, there you have it: an irregular but historically proven borrowing.

    ReplyDelete
  19. Ethan Osten,

    You've committed many trollish acts (ie. emotional rhetoric, logical fallacies, a consistant strategy of factual denials). I do hope it changes but I've rejected two of your messages so far because of it.

    I just couldn't bear to post your fruitless but predictable denials regarding the Greek coins. Anyone can see that the evidence of mp/pp alternation in these names is blunt and all of the objections you had are addressed, regardless of whether you're mentally conscious of it or not. There are further similar examples I just found in Greek like Λάππα ~ Λάμπα (now the village of Argyroupoli).

    ReplyDelete