5 Oct 2009

The etymology of Latin tofus 'tufa' isn't written in stone


Authors Liddell & Scott in A Greek-English Lexicon, first published in 1819, claimed that Greek tophiōn (τοφιών) means 'tufa quarry', attested in Tab.Heracl.1.137, then further suggested that it be traced to Latin tōfus 'tufa' which in turn was stated to probably come from an Italic dialect.[1] However, the exact dialect remains unspecified and it's unclear why the source must be an Indo-European language, let alone precisely an Italic one.

It's often claimed that tufa, an Italian borrowing inherited from Latin tōfus, is originally from Greek tóphos (τόφος).[2] Yet *tophos is apparently unattested and theorized on tophiōn[3] which leads us to a reminder that conscientious authors must give proper courtesy to their readers by meticulously placing asterisks before any conjectural constructs to make clear distinction between fact and theory. It's possible that the spelling variant tōphus, alongside the other form tōfus, was introduced into Latin through hypercorrection and folk etymology with an imagined Greek source.[4]

At this point it should be known that Etruscan tupi is attested in the Tarquinian Tomb of Orcus (TLE 89) in the phrase Tupi Sispeś next to an image of a man carrying a boulder. It's no stretch of the imagination to read it as 'Rock (Tupi) of Sisyphos (Sis(u)pe-ś)' because of its obvious connection to the Greek myth of a sinner who in death was sentenced by the gods to Tartarus (the lower underworld) and doomed to push a monumental boulder up a mountain forever. Sadly, despite all the academic accolades of co-authors Etruscanist Larissa Bonfante and British Museum Head of Italian Collections Judith Swaddling, all the two experts can cook up in their 2006 book Etruscan Myths is a less-than-accurate translation, 'the *crime (or punishment)* of Sisyphos'[5], which simply overlooks the above facts and which thereby frustratingly obscures a source for these Latin and Greek words whose origins are otherwise unknown.

As I've remarked before on my blog, Etruscan p consistently shows lenition to a bilabial fricative /ɸ/ whenever it neighbours the high rounded back vowel u. Surely this phonological quirk is from whence the fricative ef and the aspirate stop phi of the respective Latin and Greek reflexes owe their origins. So it looks like we have a simple solution here. Alas, much like Sisyphos, I suppose we linguistic-obsessed souls are doomed to eternally strive for the heights of etymology with a boulder of unknowns strapped to our back. Ah, but what a fun and glorious torment life's mysteries are!


NOTES
[1] Perseus Digital Library, excerpted from Liddell/Scott, A Greek-English Lexicon (1940), 9th edition: Greek tophiōn (τοφιών) (see link).
[2] Skinner, The origin of medical terms (1961), 2nd edition, p.406 (see link): "Latin - tophus or tofus, from the Greek τόφος, a loose, porous, kind of stone (Hebrew, toph)." Note that Skinner mistakes Hebrew toph (תֹּף) as 'stone' instead of 'tambourine, drum'. The English term toph stone is rather from French tuf, again of Latin origin like Italian tufo/tufa, as properly explained a hundred years earlier in Arthur, Treatise on Architecture, Including the Arts of Construction, Building, Stone-Masonry, Arch, Carpentry, Roof, Joinery, and Strength of Materials (1867), p.123 (see link); Haubrich, Medical meanings: a glossary of word origins (2002), 2nd edition, p.242 (see link): "tophus is a Latinized version of the Greek tophos, 'a porous volcanic stone'".
[3] Valpy, The Etymology of the words of the Greek language in alphabetical order, with the omissions generally of plants and sometimes of the more uncommon animals (1860), p.171 (see link).
[4] Diab, Lexicon of orthopaedic etymology (1999), p.353 (see link): "NB: the spelling tophus perhaps was introduced into Latin as the more learned form, as though it were of Greek origin."
[5] Bonfante/Swaddling, Etruscan myths (2006), p.32 (see link).

22 comments:

  1. Pretty interesting. Do you have any thoughts about the final vowel?

    ReplyDelete
  2. Hmmm, ambiguity.

    Assuming you mean the final vowel -i of Etruscan as it entered into Latin as nominative -us, that's easy. Etruscans could have come to pronounce the word as something like /'to.ɸɪ/ ~ /-ɛ/ by the time it was inscribed in the Tomb of Orcus II. (The final unstressed vowel was certainly not a full high front /i/.) Since, unlike Etruscan, Latin marks its masculine nominatives with -s like all faithful PIE languages do, the result of tōfus is unproblematic.

    Assuming you mean the final vowel of Italian tufa from tōfus however, that's a different tale which lies outside the pertinence of the origin of this Etrusco-Latin word and my firm competence. The direct reflex of Latin tōfus is Italian masculine il tufo, alongside French le tuf. However for whatever odd reason, Romance languages seem to have developed a secondary feminine with apparently identical meaning: Italian la tufa, French la tuffe. No mention of a corresponding Latin *tōfa is found in Perseus Online. I can't say for sure why this has happened but I'll just blame it on the peculiarities of later Vulgar Latin until I do. ;o)

    ReplyDelete
  3. News just in... French tuffe is attributed to Italian tufa (Serjeantson, A history of foreign words in English (1961), ed.2, p.192). So no Vulgar Latin craziness afterall and ergo no Latin feminine alternative *tōfa to need worry about.

    ReplyDelete
  4. You say "Since, unlike Etruscan, Latin marks its masculine nominatives with -s like all faithful PIE languages do, the result of tōfus is unproblematic." but this is not correct. Classical Latin masculine nouns had a variety of endings. Subulo, lanista and hister, for instance, are masculine nouns borrowed from Etruscan.

    ReplyDelete
  5. Ardagastus: "You say 'Since, unlike Etruscan, Latin marks its masculine nominatives with -s like all faithful PIE languages do, the result of tōfus is unproblematic.' but this is not correct."

    No, my generalized observation remains correct. This commonmost tendency in Latin for masculines to end in -us hasn't denied the other comparatively less common terminations you mention, however we're not just talking about Latin grammar but more specifically how Latin handles loanwords from non-IE languages which lack nominative endings.

    Many Etruscan names show -e where we find Latin -us such as Etruscan Marce and Latin Marcus.

    Also another Latin word, triumphus, is already suspected by etymologists to be an Etruscan loan (presumably via *triampe), amazingly parallel to my suggestion of tupi > tōphus ~ tōfus (Adams, Bilingualism and the Latin language (2003), pp.163-4).

    "Subulo, lanista and hister, for instance, are masculine nouns borrowed from Etruscan."

    So far this is just hearsay based on unverifiable glosses by classical writers. None of these words are yet found in Etruscan texts so it's unwise to take these claims for granted. Even so, none of these three words are likely to have ended in -e or -i in Etruscan as with *triampe and tupi.

    ReplyDelete
  6. No, my generalized observation remains correct.
    I don't know what your "generalized observation" is but "Latin marks its masculine nominatives with -s" is an incorrect statement by the standards of any Latin handbook.

    however we're not just talking about Latin grammar but more specifically how Latin handles loanwords from non-IE languages which lack nominative endings
    Latin is a language having five declensions, several more desinences, and also having its own peculiarities in how it fits foreign nouns in these declensions. For example, check Ammianus Marcellinus, Res Gestae, 17.12.11, where three Sarmatian names are Latinized as Rumo (3rd decl.), Zinafer (2nd decl.) and Fragiledus (2nd decl.)
    I wonder though when you mention that "commonmost tendency" if you do have a list with all words and names borrowed by Latin from other ancient languages. If so I wonder how many masculine loanwords end in -us and how many have other endings?

    Many Etruscan names show -e where we find Latin -us such as Etruscan Marce and Latin Marcus.
    However the word you were discussing is Tupi not Tupe. Even if the stress was on the first syllabe, I see no reason why should we regard the last vowel as 'e'.

    So far this is just hearsay based on unverifiable glosses by classical writers. None of these words are yet found in Etruscan texts so it's unwise to take these claims for granted.
    No offense, I find far more reliable the glosses of ancient Latin authors who might have even heard Etruscan in their lifetime than the speculations of a modern blogger based on formal resemblances.

    ReplyDelete
  7. Ardagastus: "[...] but 'Latin marks its masculine nominatives with -s' is an incorrect statement by the standards of any Latin handbook."

    Collins, A primer of ecclesiastical Latin (1988), p.13: "Masculine nouns of this declension are of two types, those with a nominative ending (-us) and those without a nominative ending; [...]". My statement is correct and you quibble too much.

    "However the word you were discussing is Tupi not Tupe. Even if the stress was on the first syllabe, I see no reason why should we regard the last vowel as 'e'."

    I can tell you haven't even bothered to read the basics on the Etruscan language. Being skeptical of something one doesn't understand is called dogmatic skepticism and it isn't a virtue.

    There's absolutely no question where the accent was in tupi or in *any* Etruscan term. The accent was always fixed on the first syllable (Bonfante/Bonfante, The Etruscan language: An introduction (2002), ed.2, p.81) as in many other languages like Proto-Germanic. Also, Etruscan spelling was never standardized and was based on the quirks of the scribe's perceptions of his own pronunciation. Spelling variations in Etruscan texts, particularly after Syncope c.500 BCE, are simply impossible to miss unless, again, one hasn't read anything on the topic.

    Regardless, we're dealing with the Latin interpretation of Etruscan unstressed word-final unrounded mid-to-high vowels (whether unstressed -i or -e is particularly moot given its proven strong stress accent). An ample list is already known of Etruscan nouns/names in -e with Latin equivalents in -us: Bonfante/Bonfante, The Etruscan language: An introduction (2002), ed.2, p.70.

    "No offense, I find far more reliable the glosses of ancient Latin authors who might have even heard Etruscan in their lifetime than the speculations of a modern blogger based on formal resemblances."

    Why should I take offense to transparent ad hominems? I always openly cite my references to distinguish myself from the chaff. And considering Dioscorides' glosses and his apparent confusion between Latin and Etruscan, your prejudice of one and empty faith of another is illogical. Everyone must be questioned equally because we are all fallible human beings.

    ReplyDelete
  8. My statement is correct and you quibble too much.
    No, it is not. That page refers only to 2nd declension masculine nouns (which they do not all end in -us, as even your quote states). However Latin has 5 declensions. From the same book for other desinences the masculine nouns might have check also the first declension, the 3rd declension,the 5th declension and also an overview of declensions.

    You might get angry on me for saying this, but observing your claims, your bibliography and your Google search strings is quite obvious you don't know this language. And I'm not saying to offend you, but to question a) why do you insist to defend such erroneous claims? and b) why do you etymologize the words of a language you haven't studied?

    I can tell you haven't even bothered to read the basics on the Etruscan language.
    I actually do know that Etruscan words have their stress on the first syllabe.

    Regardless, we're dealing with the Latin interpretation of Etruscan unstressed word-final unrounded mid-to-high vowels
    There is no proof whatsoever that Latin interpretation of Etruscan final i is the same as Latin interpretation of Etruscan final e, but also the preceding consonants might modify the vowel and its perception by Latin speakers, and also the nouns can be assigned to a declension or another based on some other criteria like familiar patterns to the speakers (borrowed nouns sounding like existing nouns might be assigned to same declension). It's a very complex problem which you quickly dismiss.

    whether unstressed -i or -e is particularly moot given its proven strong stress accent
    Not really. I know in several languages words ending in unstressed -i or -e, and no matter how hard I'd stress the first syllabe the ending keeps being different (especially in simple, bisyllabic CVCV words). Of course, I don't deny that in some languages, a certain type of i and a certain type of e might eventually merge in some circumstances, I don't deny that last vowel might be simply dropped in some other circumstances, but there's no proof this always happens, there's no proof - so far - this is the case here.

    Why should I take offense to transparent ad hominems?
    Are you not modern, not a blogger or not speculating?

    your prejudice of one and empty faith of another is illogical
    But there's no faith involved here. They are contemporary witnesses, you are not. It's rather illogical to believe that if an ancient writer reports a false Etruscan word, then all the ancient authors are unreliable in their testimonies.

    Everyone must be questioned equally because we are all fallible human beings.
    I agree very much with that, actually I'd question equally all those connecting tupi with tofus.

    ReplyDelete
  9. Errata: I said at some point that Latin has several more desinences and this is incorrect, I should have said 'endings'. There are not that many desinences, it is actually the null desinence which makes the nouns to expose a variety of endings.
    Few examples of Latin masculine nouns (not ending in -us, some ending in -s though): 1st decl, -a: agricola, nauta, poeta; 2nd decl, Ø: magister, liber; 3rd decl, Ø: pater, frater, senator, honor, vultur, consul, pugil, sol, pecten, flamen, rex, codex, ignis, pes, flos, mons, leo, sermo, homo.

    ReplyDelete
  10. If you are suggesting "tupi" was a generic Etruscan word for "rock", there has to be a reason why it was then restricted to a particular porous volcanic rock. Tufa blocks were used as building material -- could the original sense have been something like "stone block"?
    I'd expect a generic "rock" word would show up in toponymy. Any candidates?

    ReplyDelete
  11. Ardagastus,

    I won't repeat my words. Rules of logical debate are enforced here. Failure of commenters to educate themselves on *what* is being said or *why* isn't my concern. Be aware that nothing obligates a blogger to continue publishing a commenter's obstinate, off-topic and irrelevant argument.

    The sins of your dyslogic include:

    1. Ad hominems ("No offense, I find far more reliable the glosses of ancient Latin authors who might have even heard Etruscan in their lifetime than the speculations of a modern blogger based on formal resemblances."; "Are you not modern, not a blogger or not speculating?")
    A statement's source, or a writer's spatiotemporal proximity to things described, in no way guarantees accuracy (e.g. Dioscorides' false Etruscan glosses). Only a statement's logical validity through logical means is worthy of discussion. Blogger or not, modern or not, speculation or not; these considerations are all distractions.

    (cont'd in next comment...)

    ReplyDelete
  12. (...cont'd from previous comment)

    2. Straw man arguments
    ("I don't know what your 'generalized observation' is but 'Latin marks its masculine nominatives with -s' is an incorrect statement by the standards of any Latin handbook."; "That page refers only to 2nd declension masculine nouns (which they do not all end in -us, as even your quote states).")

    You persist in misreading general statements as absolute ones by slipping in the word 'all' where I never intended nor typed. Your error is so transparent that I have difficulty believing it's genuine. Either way, it's irrelevant to the question of whether Etruscan nouns in -i can become Latin nouns in -us (particularly via their corresponding vocatives in -e) as is already hinted by the examples of Latin nominative names Titus and Marcus (with vocatives Tite and Marce) versus Etruscan Tite and Marce (functioning at once as nominative, accusative and vocative). Adapt or die.

    3. Confusing opinion for fact. ("I know in several languages words ending in unstressed -i or -e, and no matter how hard I'd stress the first syllabe the ending keeps being different (especially in simple, bisyllabic CVCV words).")
    Your statement suggests that you're ignorant not only of Etruscan but of linguistics as a whole. And why do you continue to write "syllabe" when it's properly "syllable" in English? Apocope is too commonplace in world languages to respect your position. Please read on Estonian apocope as but one example showing the falsity of your beliefs since regardless of vocal height, surely these differing vowels converged before being zeroed. Another is Friulian. Trust me, your idle observations are moot to any linguist. You're hardly disproving anything misusing your self-indulgent beliefs as counterevidence.

    ReplyDelete
  13. Daniel: "If you are suggesting 'tupi' was a generic Etruscan word for 'rock', there has to be a reason why it was then restricted to a particular porous volcanic rock."

    Yes, indeed! Keep in mind that while a generalized meaning, "rock" or "stone", applies best to TLE 89, this doesn't mean that tupi couldn't have simultaneously had a more specific connotation of volcanic rock. I think that where the myth of Sisyphos is concerned, the main point is that it's a rock of terrible burden and not that it's volcanic. I'd presume that Tartarus could be conceived of as a vast volcanic cave filled with tufa though, hence the Tartarean river, Phlegethon, the river of fire (lava).

    "Tufa blocks were used as building material -- could the original sense have been something like 'stone block'?"

    Eureka! I seriously think you've got it nailed! In fact, I'll share a potentially helpful secret with you. I've recently been sitting on an idle possibility that tupi is ultimately from Babylonian ṭuppu "(baked clay) tablet" (which in turn is a Sumerian term), yet while a semantic link between a baked clay product and a piece of volcanic rock was tempting, it only went so far. If I combine your suggestion with this etymology, things fit so much better. We can then surmise a more reasonable meaning shift of "tablet of clay" to "block of tufa". Wonderful!

    ReplyDelete
  14. @ Glen & Ardagastus:
    I can't help but feeling there are some language barriers between Ardagastus and Glen [quote: I don't know what your "generalized observation" is ]. Both apparently know enough of the latin language.

    Also, I think Ardagastus did not really write an "ad hominem". "speculating"/guessing is something we all have to do sometimes. It can be the basis for further thinking. If this further thinking leads to nothing, well, tant pis.
    I must admit, however, it looked like an "ad hominem".

    @ Glen:
    This final Etruscan i may in some way have become Latin us. No problem with your train of thought. There are many cases of e becoming us. Also, there are examples of Etruscan words ending in either e or i.
    But:
    If we are to believe what we see in the codified form of their language (i.e. the written texts) their vowel system was rather poorish* (as was their consonant system!**)***. Four vowels, some diphthongs, no triphthongs. Certainly the e and the i were not completely interchangable.
    In Latin the -us ending gradually disappeared from spoken language, mostly displaced by (dative) -o (at least in later centuries; unfortunately I don't know much about Latin in Etruscan times). Clearly Latin us did not sound like Etruscan us to the Etruscans, either because Etruscan us normally was a genitive, OR because the sound itself was different. Your "vocative" explanation does help, but it does not convince. I have a suspicion (needs further study) that Etruscan final -s was hardly heard. If I am right, maybe Etruscan e just sounded much like Latin us.
    I would like to have more examples of Etruscan i becoming Latin us. Do you have any?


    *
    Compare for example Frisian, one of the languages I grew up with: seven short vowels, six long ones. Some of the short ones can be lengthened. Some of the vowels can be nasalised. And there are many many diphthongs and triphthongs.

    **
    Frisian has an almost complete set of guttural/velar, dental/alveolar, labial/labiodental consonants (voiced and unvoiced plosives, voiced and unvoiced fricatives, nasals and half-vocals), an s, sh, r and l.
    There are many clusters of two, three, four and even five consonants.

    ***
    Somewhat off topic:
    Many modern languages rather easily incorporate strange sounds in loan words (e.g. Frisian voiced and unvoiced dental fricatives only in loans, mostly from English).
    This seems to have been very difficult for the Etruscans.

    ReplyDelete
  15. ZU: "Also, I think Ardagastus did not really write an 'ad hominem'."

    Not to pick unnecessarily on him but he did for the reasons I stated, according to standard definition. An ad hominem isn't a synonym for 'blatant insult' (eg. "You're a dummy and your mother stinks!"); it's more basic and insidious, involving the use of irrelevance about a person's immaterial character or actions to distract us from the logic of statements expressed.

    "Four vowels, some diphthongs, no triphthongs. Certainly the e and the i were not completely interchangable."

    Ironically we're in perfect agreement but you're unknowingly confusing certain facts together. The reduction of non-initial syllables is proven historically by the gradual spelling changes since about 500 BCE showing words apparently shrinking towards the first syllable. Claiming a reduction in non-initial -e and -i is not the same as claiming that e and i are completely interchangable *as a whole* since the phonemic distinction between the two vowels is evidently preserved in the stressed first syllable where subtler contrasts can be most salient. Many languages with fixed accents like those of Uralic have a certain inventory of 'full vowels' in the accented syllable with a restricted set of 'reduced vowels' in all unaccented positions. The richness or sparseness of a language's phonology alone doesn't mitigate against syncopation.

    "In Latin the -us ending gradually disappeared from spoken language, [...]"

    This fact says nothing on how an unmarked Etruscan noun with lax word-final -i - being used grammatically as subject, object and vocative - would be perceived by the Roman ear considering Latin vocatives in -e. I can tell you for certain that conversely Latin -o or -u could not be mistaken for -e by an Etruscan ear.

    "I have a suspicion (needs further study) that Etruscan final -s was hardly heard."

    Without facts to base it on, this is idle. To the contrary, an absence of distinction between zero and -s# would have made a distinction between nomino-accusative and genitive cases impossible and yet this distinction was maintained faithfully without hint of confusion in even the latest documents like the Liber Linteus. There seems to be too much against the idea.

    "I would like to have more examples of Etruscan i becoming Latin us. Do you have any?"

    Etruscan Ani ↔ Latin Jānus.

    "Many modern languages rather easily incorporate strange sounds in loan words (e.g. Frisian voiced and unvoiced dental fricatives only in loans, mostly from English). This seems to have been very difficult for the Etruscans."

    Not true! Consider the Etruscan use of letter phi, coding for the aspirate bilabial stop, which tends to mark many Greek loans: Φerse 'Perseus' and Φuipa 'Phoibe'. (Interestingly, come to think of it, the digraph ph is often the mark of foreignness in Latin loans from Greek, as in our discussed word tōphus.)

    It seems from what I've collected in my personal database that Etruscan words with word-initial "ph" would have been absent if it weren't for loanwords to fill in this blank. I could imagine the more learned Etruscans pronouncing such words with added aspiration while in the country, farmers might not have bothered with this cosmopolitan contrast. This would be in much the same way as the Roman handling of Greek upsilon (ie. pronounced /y/ by elites but otherwise plain ol' /u/ by commoners).

    ReplyDelete
  16. I just typed: "Interestingly, come to think of it, the digraph ph is often the mark of foreignness in Latin loans from Greek, as in our discussed word tōphus."

    I should clarify this. I mean to say that while I've become convinced of an Etruscan origin, Roman folk etymology may have instead attributed a Greek origin. Or alternatively, perhaps it's possible that digraphs like ph had become a common tactic to imply a loanword, whether from Greek or not. There are many unknowns and details here.

    ReplyDelete
  17. I have a suspicion (needs further study) that Etruscan final -s was hardly heard

    Oops! "Etruscan" should have been "Latin".

    ---

    Etruscan Ani ↔ Latin Jānus.

    Thank you!

    ---

    ... This seems to have been very difficult for the Etruscans."

    Not true!


    "very difficult" is not the same as "impossible".
    Obviously their alphabet did not allow them to represent foreign sounds well* (e.g. the lack of an o)**. It even was difficult to represent with it the Etruscan sounds (e.g.: You mentioned the use of u for a schwa). This of course does not mean Etruscans were not capable of producing foreign sounds.
    The very strong stress on the first syllable plus the absence of prepositions and prefixes must have made it even more difficult to incorporate foreign words.


    *
    Of course all languages have their problems. Frisian, Dutch and French are all relatively well equipped. Still the ones who learn Dutch or Frisian would be helped with a letter for a schwa. French spelling helps to understand the language, but learning French spelling is not that easy for those who are just francophone.
    English, almost equally well equipped, would be almost unintelligible for an Etruscan. As one can not tell the pronunciation from the spelling, words foreign to the English language are even harder to understand.

    **
    Your dictionary mentions frontac. Don't you think this is North Picene?

    ReplyDelete
  18. ZU: "Oops! "Etruscan" should have been 'Latin'."

    Ah that's more plausible... maybe-ish?

    "The very strong stress on the first syllable plus the absence of prepositions and prefixes must have made it even more difficult to incorporate foreign words."

    Yet it didn't make it the slightest bit more difficult! Etruscan abounds in foreign lexemes of various origins: Doric Greek (Atalanta < Αταλάντη), Latin (Lusce < Luscus), Umbrian (Neθunś < *Nehtuns) and Phoenician (the nomen Carθazaie 'Carthaginian' from Qart Ḥadšat 'Carthage', literally 'New City'). Don't be so narrow - Etruscan was as cosmopolitan as any language could be and incorporated a vast number of alien vocabulary despite any linguistic handicaps one may perceive in it.

    "Your dictionary mentions frontac. Don't you think this is North Picene?"

    Not in the slightest. The morphology present in the incised inscription to which you refer (see photo here) is strictly Etruscan. The Latin portion reads L(ars) Cafatius, L(ari) f(ilius), Ste(llatina). Haruspex fulguriator 'Lars Cafatius, son of Lars, (tribe) of Stellatina. Haruspex (and) fulguriator' and the Etruscan part is virtually identical: Cafates Lr., Laris. Netśvis trutnvt fronta-c 'Larth Cafate, (son) of Laris. Oracle of haruspicy and fulguriator'.

    For fronta, we should read *frunta and it's yet another early Greek loanword (βροντή). Original word-initial *p- has been softened here to f.

    ReplyDelete
  19. Thanks for your answer, and sorry for stirring up a hornet's nest. I think the Ani > Janus example more than adequately demonstrates the plausibility of the final vowel, unstressed -i vs. Lat. -us.

    ReplyDelete
  20. Stephen, you've hardly stirred up any hornet nests by asking simple questions. On the internet, they stir themselves up but don't worry, I have bugspray for that. ;o)

    ReplyDelete
  21. I should clarify this. I mean to say that while I've become convinced of an Etruscan origin, Roman folk etymology may have instead attributed a Greek origin. Or alternatively, perhaps it's possible that digraphs like ph had become a common tactic to imply a loanword, whether from Greek or not. There are many unknowns and details here.

    Although not completely relevant, this suddenly reminded me off this lovely word in Dutch thee 'tea'. Why the th? we have no phoneme expressed by this, and in fact, the language from which we loaned it from didn't have such a distinction either. It has been used purely as a device to make it look more foreign. Pretty funny.

    ReplyDelete
  22. Yep! Consider also the Japanese using special characters called katakana for spelling out foreign words while hiragana characters for native ones. Yet another example might be Ancient Egyptian which tended to spell foreign names out in full, vowels and all, as seen in the cartouche for Cleopatra, Cleopatra being of course a Greek name. Κλεοπάτρα means "Father's Glory".

    ReplyDelete