31 Jul 2010

Edward Sapir and the Philistine headdress


Browsing the web as usual I came across something that captured by attention. It turns out that Edward Sapir in his 1937 article Hebrew "Helmet," A loanword, and its bearing on Indo-European phonology had reconstructed a Philistine word *kaubaɣ- 'helmet' based on the Christian story of Goliath (1 Samuel 17:5). There in Samuel, the word for 'helmet' is recorded as either kōbáʕ (כובע) or qōbáʕ (קובע), a rather curious word of non-Semitic origin whose mystery has cultivated much academic curiosity. It's certainly connected with Hittite kupahis 'headdress' although exactly how this word was transmitted between Semitic and Anatolian groups is a riddle to solve. Note easterly Hurrian kuwahi too although seeking an origin of this word in Eastern Turkey seems most unlikely despite what Puhvel suggests (see Puhvel, Hittite Etymological Dictionary: Words beginning with K (1997), p.257).

Quite eerily, I've already suggested a Minoan word for 'head' or 'summit', *kaupaθa (read Paleoglot: A hidden story behind Cybele the Earth Mother?), but I was ignorant at the time of Sapir's published hypothesis. This Minoan etymon is my attempt at better explaining (via expected Etruscan *caupaθ) the source of both Germanic *haubida- and Latin caput in a way that an over-cited Indo-European root (*)*kaput- just can't convincingly accomplish without fiddling with the phonetics. Surely, given this particular set of uncannily similar yet phonetically irregular reflexes, we should seriously be thinking about recent loan transmission rather than leaping to the comparatively more contrived assumption of some long-ago Proto-Indo-European utterance. Add to this Greek κύπτω (kúpto) 'to bend forth, stoop forth' for which Beekes fails to find a credible Indo-European etymology and suggests Pre-Greek origin, much to my bookish delight. Nonetheless, some do try even to loosely etymologize the aforementioned Philistine word in Indo-European terms with no noteworthy success.


Matching the spirit of Sapir's whimsy while in honest desire to crack this Mediterranean puzzle, I wonder sometimes about the elusive Philistine language and its origins which scream 'Aegean' all over it. Unfortunately, the Philistine civilization also still screams 'mystery' all over it and with no forthcoming answers in sight. We must remember though the mention of the *Pirastu [prst.w] in Egyptian records, one of the 'Sea Peoples' listed to have attempted a minor coup on Egypt at the end of the 2nd millennium BCE.

Now, would it be unfathomable to posit the idea that the Philistine language could be a direct descendant of Minoan (or perhaps a Minoanized language) in which the word for 'head' (and subsequently for 'headdress' through metonymy) became, for the sake of public brainstorming, *kopað in this dialect? Semites, Hittites or some other intermediary might perhaps have misheard the voiced dental fricative /ð/ as a voiced pharyngeal fricative /ʕ/ (ie. ayin) or a voiced glottal /ɦ/. Stranger things have happened in loanwords, eg. the Semitic name of the famed North African city of Carthage which entered into Greek as Καρχηδών (Karchēdṓn), into Latin as Carthāgo, and into Etruscan as *Carθaza.

13 comments:

  1. Semites, Hittites or some other intermediary might perhaps have misheard the voiced dental fricative /ð/ as a voiced pharyngeal fricative /ʕ/ (ie. ayin) or a voiced glottal /ɦ/.

    Surely /d/ or /z/ would be the expected substitute in Semitic, but the Egyptian soundlaw /d/ > /ʕ/ seems relevant here.

    I don't see how Greek /kupto/ fits in. Why loss of /a/ here but loss of /u/ in Latin?

    ReplyDelete
  2. The Egyptian soundlaw is, as far as I know, a Pre-Egyptian soundlaw and so wouldn't be significant here. I wonder if *kopatˤ would be a more apt reconstruction considering that I already have been reasoning for a while now that Minoan is from earlier *tˤ anyway.

    ReplyDelete
  3. I should address this last comment: "I don't see how Greek /kupto/ fits in. Why loss of /a/ here but loss of /u/ in Latin?"

    A sequence like [kawp] could easily be misheard as [kap] in Latin if [w] is perceived as merely the [+labial] feature of the following [p].

    Greek kúpto on the other hand would be the result of [aw] being misheard as [u] in Mycenaean. It's certainly phonetically plausible.

    ReplyDelete
  4. The Egyptian soundlaw is, as far as I know, a Pre-Egyptian soundlaw and so wouldn't be significant here.

    Well, Old Kingdom, but I don't see anything to put a date on this supposed Minoan-Philistine connection.

    Also the change in this word wouldn't have to have occur'd in Egyptian itself, the change may have spred to other languages of the area. There appears to be an Ugaritic dialectal change from *ẓ (the emphatic interdental) > ɣ for instance.

    (Or simpler yet, it could be related to that change with no relation to Egyptian.)

    I already have been reasoning for a while now that Minoan is from earlier *tˤ anyway.

    Sounds like a story for another day, tho I wonder how you could reach that kind of phonetic precision. Or do you mean simply "emphatic stop", not specifically "pharyngealized"?

    A sequence like [kawp] could easily be misheard as [kap] in Latin if [w] is perceived as merely the [+labial] feature of the following [p]. Greek kúpto on the other hand would be the result of [aw] being misheard as [u] in Mycenaean. It's certainly phonetically plausible.

    I'm suspicious of any loan etymology that involves mishearing a sequence that exists in the receiving languages as well (ie. /aw(p)/), especially if it's supposed to have gone two opposite ways.

    My understanding of Greek morphophonology is far too feeble to tell if this is possible or nonsense, but I wonder if the Greek word could be a zero-grade derivativ.

    ReplyDelete
  5. Tropylium: "Well, Old Kingdom, but I don't see anything to put a date on this supposed Minoan-Philistine connection."

    False, according to Loprieno, Ancient Egyptian: A linguistic introduction (1995), p.31 which places *d > in Egyptian prehistory. Ergo, chronologically implausible.

    However, that Ugaritic sound change is definitely food for thought.

    "Or do you mean simply 'emphatic stop', not specifically 'pharyngealized'?"

    In Semitic linguistics, an emphatic stop can in fact sometimes refer to a pharyngealized stop. At any rate, yes, I specifically mean 'pharyngealized' but I need to sit down and write a thorough post on that pretty soon. Hold tight.

    "I'm suspicious of any loan etymology that involves mishearing a sequence that exists in the receiving languages as well (ie. /aw(p)/), especially if it's supposed to have gone two opposite ways."

    Let's not infect rational debate with impressions based on nothing but factless suspicions and assumptions.

    All loanwords inevitably involve a mishearing to at least some extent between one language and another, so this view isn't reasonable nor material here, particularly when I've already supplied the radical example of "Carthage" in this post. You're beating a dead horse with idle opinion, even without providing a better answer of your own (which would be far more constructive than mere suspicions, I think).

    ReplyDelete
  6. Tropylium: "Well, Old Kingdom, but I don't see anything to put a date on this supposed Minoan-Philistine connection."

    False, according to Loprieno, Ancient Egyptian: A linguistic introduction (1995), p.31 which places *d > in Egyptian prehistory. Ergo, chronologically implausible.

    However, that Ugaritic sound change is definitely food for thought.

    "Or do you mean simply 'emphatic stop', not specifically 'pharyngealized'?"

    In Semitic linguistics, an emphatic stop can in fact sometimes refer to a pharyngealized stop. At any rate, yes, I specifically mean 'pharyngealized' but I need to sit down and write a thorough post on that pretty soon. Hold tight.

    "I'm suspicious of any loan etymology that involves mishearing a sequence that exists in the receiving languages as well (ie. /aw(p)/), especially if it's supposed to have gone two opposite ways."

    Let's not infect rational debate with impressions based on nothing but factless suspicions and assumptions.

    All loanwords inevitably involve a mishearing to at least some extent between one language and another, so this view isn't reasonable nor material here, particularly when I've already supplied the radical example of "Carthage" in this post. You're beating a dead horse with idle opinion, even without providing a better answer of your own (which would be far more constructive than mere suspicions, I think).

    ReplyDelete
  7. Tropylium: "Well, Old Kingdom, but I don't see anything to put a date on this supposed Minoan-Philistine connection."

    False, according to Loprieno, Ancient Egyptian: A linguistic introduction (1995), p.31 which places *d > in Egyptian prehistory. Ergo, chronologically implausible.

    However, that Ugaritic sound change is definitely food for thought.

    "Or do you mean simply 'emphatic stop', not specifically 'pharyngealized'?"

    In Semitic linguistics, an emphatic stop can in fact sometimes refer to a pharyngealized stop. At any rate, yes, I specifically mean 'pharyngealized' but I need to sit down and write a thorough post on that pretty soon. Hold tight.

    "I'm suspicious of any loan etymology that involves mishearing a sequence that exists in the receiving languages as well (ie. /aw(p)/), especially if it's supposed to have gone two opposite ways."

    Let's not infect rational debate with impressions based on nothing but factless suspicions and assumptions.

    All loanwords inevitably involve a mishearing to at least some extent between one language and another, so this view isn't reasonable nor material here, particularly when I've already supplied the radical example of "Carthage" in this post. You're beating a dead horse with idle opinion, even without providing a better answer of your own (which would be far more constructive than mere suspicions, I think).

    ReplyDelete
  8. Tropylium: "Well, Old Kingdom, but I don't see anything to put a date on this supposed Minoan-Philistine connection."

    False, according to Loprieno, Ancient Egyptian: A linguistic introduction (1995), p.31 which places *d > in Egyptian prehistory. Ergo, chronologically implausible.

    However, that Ugaritic sound change is definitely food for thought.

    "Or do you mean simply 'emphatic stop', not specifically 'pharyngealized'?"

    In Semitic linguistics, an emphatic stop can in fact sometimes refer to a pharyngealized stop. At any rate, yes, I specifically mean 'pharyngealized' but I need to sit down and write a thorough post on that pretty soon. Hold tight.

    "I'm suspicious of any loan etymology that involves mishearing a sequence that exists in the receiving languages as well (ie. /aw(p)/), especially if it's supposed to have gone two opposite ways."

    Let's not infect rational debate with impressions based on nothing but factless suspicions and assumptions.

    All loanwords inevitably involve a mishearing to at least some extent between one language and another, so this view isn't reasonable nor material here, particularly when I've already supplied the radical example of "Carthage" in this post. You're beating a dead horse with idle opinion, even without providing a better answer of your own (which would be far more constructive than mere suspicions, I think).

    ReplyDelete
  9. Tropylium: "Well, Old Kingdom, but I don't see anything to put a date on this supposed Minoan-Philistine connection."

    False, according to Loprieno, Ancient Egyptian: A linguistic introduction (1995), p.31 which places *d > in Egyptian prehistory. Ergo, chronologically implausible.

    However, that Ugaritic sound change is definitely food for thought.

    "Or do you mean simply 'emphatic stop', not specifically 'pharyngealized'?"

    In Semitic linguistics, an emphatic stop can in fact sometimes refer to a pharyngealized stop. At any rate, yes, I specifically mean 'pharyngealized' but I need to sit down and write a thorough post on that pretty soon. Hold tight.

    "I'm suspicious of any loan etymology that involves mishearing a sequence that exists in the receiving languages as well (ie. /aw(p)/), especially if it's supposed to have gone two opposite ways."

    Let's not infect rational debate with impressions based on nothing but factless suspicions and assumptions.

    All loanwords inevitably involve a mishearing to at least some extent between one language and another, so this view isn't reasonable nor material here, particularly when I've already supplied the radical example of "Carthage" in this post. You're beating a dead horse with idle opinion, even without providing a better answer of your own (which would be far more constructive than mere suspicions, I think).

    ReplyDelete
  10. Tropylium: "Well, Old Kingdom, but I don't see anything to put a date on this supposed Minoan-Philistine connection."

    False, according to Loprieno, Ancient Egyptian: A linguistic introduction (1995), p.31 which places *d > in Egyptian prehistory. Ergo, chronologically implausible.

    However, that Ugaritic sound change is definitely food for thought.

    "Or do you mean simply 'emphatic stop', not specifically 'pharyngealized'?"

    In Semitic linguistics, an emphatic stop can in fact sometimes refer to a pharyngealized stop. At any rate, yes, I specifically mean 'pharyngealized' but I need to sit down and write a thorough post on that pretty soon. Hold tight.

    ReplyDelete
  11. Tropylium: "I'm suspicious of any loan etymology that involves mishearing a sequence that exists in the receiving languages as well (ie. /aw(p)/), especially if it's supposed to have gone two opposite ways."

    Let's not infect rational debate with impressions based on nothing but factless suspicions and assumptions.

    All loanwords inevitably involve a mishearing to at least some extent between one language and another, so this view isn't reasonable nor material here, particularly when I've already supplied the radical example of "Carthage" in this post. You're beating a dead horse with idle opinion, even without providing a better answer of your own (which would be far more constructive than mere suspicions, I think).

    ReplyDelete
  12. (^Holy multipost, Batman!)

    according to Loprieno, Ancient Egyptian: A linguistic introduction (1995), p.31 which places *d > *ʕ in Egyptian prehistory.

    Ah, I see, that's clearer than Woodard's book — the one you link'd under your kithara post — which goes with simply "before Middle Egyptian" (seemingly implying "after or during Old E.")

    Let's not infect rational debate with impressions based on nothing but factless suspicions and assumptions.

    Here's a fact: by far most loanwords whose history is recorded in full, excruciating detail, such as modern-era international loans like "banana", "sushi", "telephone", "hotel", "lava" contain in most languages they're found in no random distortions of mishearing (counterexamples welcome). Even when something like "cacao" > "cocoa" happens, this turns out not to have been "mishearing" but a later change (by contamination in this case). Similarly loans such as "chamber", "champion", "chalice" don't have an initial affricate because of "mishearing" the French, but because of representing a loan before deaffrication (or from a conservativ variety, or vice versa as in the last case). Further down history we may lack a suitable resolution of data — intermediate forms may go unattested — so some apparent disrepancies will be popping up. I do see a problem however if this is taken as a mandate to draw connections even between words whose shapes and meanings match only loosely.

    While I continue to wonder what may be the story behind the Latin /a/, its quite exact correspondence of meaning with Germanic is a good reason to think they are related; ditto for your example of Carthage.

    But for /kupto/ we also need to assume some extra sense development, and explain the differing structure, which leaves open the risk of a false cognate. Words for "to bow" can and do have origins other than words meaning "head". It would be nice to have some further evidence that this does belong here.

    I could attempt sketching something: Hittite provides a potential regular pathway for getting the /u/ and Beekes' comment that
    As ku¯fÒj is in Greek isolated in its formation, it may be original as against the regular kÚptw
    [why can't they use some half-sensible transcription?] apparently explains the -pt-. But that still leaves the semantic gap between "to bow" and "helmet". We could assume eg. a parallel but unattested Aegean loan to Anatolian from the same root, but that would remain speculation just as much.

    All loanwords inevitably involve a mishearing to at least some extent between one language and another,

    No they don't. How, for example, have French parking, German Start, Swedish polis been misheard? Or if you'd like pre-literary examples, take the likes of Finnish kuningas, ruhtinas (and many less flashy cases). Any changes that have occurred in loans like this are solely to accommodate the word into nativ phonology and morphology. I see no reason to think the process of loaning function'd any differently in antiquity.

    ReplyDelete
  13. Tropylium: Here's a fact: by far most loanwords [...] contain in most languages they're found in no random distortions of mishearing"

    Opinion used as fact. Read Philip Baldi, Patterns of change, change of patterns: Linguistic change and reconstruction (1991), p.158: "By their nature, errors of perception are more likely to be random than errors of production, a fact observable in loanwords, which often undergo random phonetic substitutions."

    "I do see a problem however if this is taken as a mandate to draw connections even between words whose shapes and meanings match only loosely."

    Strawman. Although always a valid concern, your imagined mandate simply doesn't exist on Paleoglot, and I'm unwaveringly certain that I've shown my commitment to getting things right over the many years writing this blog.

    Please, Tropylium. I love criticism when well-founded. When it's merely to oppose however, it gets distractive and annoying very quickly. Do you have a valid criticism of what I've written above, based on clear references and facts? If not, hold back for a day or two and build up a good case.

    ReplyDelete