(H)EUREKA.

This is something I’ve been wondering about for many years, and even in this age of instantly available information I can’t get a real answer, so I thought I’d turn to the Varied Reader for informed suggestions (plus the usual japery). As every schoolboy knows, when Archimedes noticed that the water level in his bath rose when he stepped into it and realized the volume of irregular objects could be calculated with precision, he exclaimed “Eureka!” Well, actually he exclaimed “Εὕρηκα,” Greek for “I have found [sc. it]!” But the Greek is, transliterated, heúrēka (the perfect of heurískō ‘I find, discover’); where has the h– gone?

Russian has Эврика (évrika), but that makes sense because it took its Greek from the Byzantines, who used essentially Modern Greek pronunciation; by the same token, it has эвристический (evristícheskii) for ‘heuristic.’ English, absurdly, has the h– in the latter case but not in the former, and it doesn’t have the Byzantine excuse: Ancient Greek words in English came via scholars with no interest in either Byzantium or the modern language. The most extended discussion of the word I’ve found is in The Merriam-Webster New Book of Word Histories, which simply says “Eureka or, as it may more accurately be transliterated from the Greek, heurēka derives from the same Greek root word as heuristic.” Which is no help at all. So: any thoughts?

Comments

  1. The OED’s early citations are about evenly divided between spellings in e- and in h-:
    [1570 DEE Math. Pref., For this, may I (with ioy) say ΕΥΡΗΚΑ.] 1603 HOLLAND Plutarch’s Mor. 590 [Archimedes] crying out, Heureca. 1658 tr. Porta’s Nat. Mag. XVIII. viii. 384 We have gone beyond Archimedes his Eureka. 1742 FIELDING J. Andrews II. xiii. (ed. 2) 267 Adams..returned overjoyed..crying out ‘Eureka’ [ed. 1 (1742) Ευρηκα; ed. 3 (1743) Heureka.]

  2. Actually, it comes from the Oxfordisation of the story’s protagonist who, as all good English gentlemen do, drops his aitches.

  3. Maybe something to do with silent H e.g. humour, honour pronunciations? These are still fairly unstable. Probably also to do with how the story was transmitted e.g modern greek via French perhaps – dropping aitches on the way – altough I’m sure its more arabic and latin. Fielding in the OED quote appears to have corrected himself (or been corrected) with the realisation that Archimedes spoke ancient greek.

  4. Maybe lost in transmission between different Greek dialects?
    Persian “hind-” lost its “H”, became “India”, etc. in most European languages because it passed via some H-less dialect of Greek.

  5. Just in case it’s not clear from the rather terse OED citation format, the Fielding spellings in Joseph Andrews were:
    1st edition (1742), Ευρηκα
    2nd edition (1742), Eureka
    3rd edition (1743), Heureka

  6. @keith100
    Do you really pronounce ‘humour’ with a silent H?

  7. Oh, perhaps you were putting it alongside ‘honour’ for comparison, in which case I apologise for having misunderstood.
    (I’m sorry, I don’t have anything to contribute to the discussion which is why I’m wittering on about irrelevances)

  8. R: Not unlikely; many Americans do not pronounce the initial h in words beginning hu-, like humor, human, etc., so they come out yumor, yuman, etc. (I’m not one of those Americans.) On the other hand, when a Briton says herb, he sounds to American ears like he’s saying Herb, as in Herbert.
    All the initial h’s in English were lost some centuries back and then some, but not all, were restored from the spelling. However, the restoration was inconsistent in different places.

  9. Maybe something to do with silent H e.g. humour, honour pronunciations?
    For this or any similar explanation to be plausible, there would have to be other such cases. I can’t think of any. Greek h- gets carried over into English in spelling, no matter how the Englishpersons involved pronounced the resulting word.

  10. mollymooly says

    Robert says French has “eurêka” and usually “heuristique” but sometimes “euristique”.

  11. “All the initial h’s in English were lost some centuries back and then some,….”
    Whoa, John. Name one native English word where the initial “h” gets dropped in any dialect that isn’t suspiciously and dangerously close to France.

  12. Could it be a confusion with Anglo-Greek “eu-” meaning “good,” or “pleasant”? E.g. euphonious/euphonium, eukaryote, etc.?
    And what about “eucharist”? Or, for that matter, “euchre”?

  13. I forgot “euphemism,” which is a word most people actually use, unlike the examples I gave. Also, “confusion” should probably read “conflation,” but the latter is not a word I use comfortably.

  14. There odometer, too, but I’m guessing French makes that easier to account for.

  15. OED says (when Etymology button clicked) ‘The correct spelling heureka is rare’. However, despite calling it the correct spelling, there isn’t an entry for heureka.
    The German is Heureka.

  16. Bill Walderman says

    Archimedes was Syracusan and probably spoke a Doric dialect, and would have actually said something closer to “heuraka.” Psilosis–the loss of the initial rough breathing (the “h” sound) had already occurred in Asia Minor Greek dialects as early as the archaic period (Sappho writing in Lesbian Aeolic and Herodotus in Ionic, for example). Eventually it must have happened in Koine, too, and it’s characteristic of modern Greek. Maybe it had happened in Sicily by Archimedes’ time. But I suspect the source that recorded his exclamation probably recorded the standard Attic form with the rough breathing and eta–which doesn’t shed light on your question.

  17. David Derbes says

    I think the rough breathing was simply not understood by someone (or several someones.) It’s one thing to transliterate the Greek epsilon to e and so on; but if you don’t know Greek, you miss that the rough breathing corresponds to h, and transliterate (rough breathing) epsilon to “e”. The interesting thing is that an initial rho with a rough breathing is always transliterated as “rh”. As another poster suggests, the same thing seems to have happened with ‘hodos’; it ought to be hodometer, not odometer. But (in my opinion) the person who coined the term didn’t really know Greek, and didn’t understand the (rough breathing) o, when he or she looked up “path, road” in the dictionary.

  18. Name one native English word where the initial “h” gets dropped in any dialect that isn’t suspiciously and dangerously close to France.

    From OED entry for H:

    … at the present day only a very few words, viz. heir, honest, honour, hour, with their derivatives, remain with h mute; though others, such as herb, humble, humour, were so treated very recently, and are by some people still

    “Humour” is usually (always?) h-less for me (, though I think “humorous” is more variable. OED offers both for “humour”, only the h-ful variant for “humorous”.

  19. “Humour” is usually (always?) h-less for me (,

    = “Humour” is usually (always?) h-less for me (Irish),

  20. Breffni, how do you say “Humerus” (the bone in your arm)? I’m just curious.

  21. AJP, I’d pronounce the h in “humerus”; also in “humus”.

  22. Well, Archimedes predates polytonic orthography, and even once introduced, you don’t see it in papyri until the 2nd century A.D., and it was by no means standard until Byzantine Greek. The word εὕρηκα may once have been pronounced heúrēka, but it was typically written ΕΥΡΗΚΑ, without the spiritus asper. The h has not so much gone missing; it most likely wouldn’t’ve been written in the original.

  23. The h has not so much gone missing; it most likely wouldn’t’ve been written in the original.
    Again, this would apply to any Ancient Greek word. Any convincing explanation has to account for why this specific form lost its h-. And I’m quite sure Bill’s Asia Minor Greek dialects are also irrelevant; whatever the explanation, it’s to be found in 17th-century England, not ancient Ionia.
    I think the rough breathing was simply not understood by someone
    I’m sure the rough breathing was not understood by lots of people (and even more people today), but how many such people were involved in transmitting Greek words in the 17th century? That may well apply in the case of odometer, where the h- was dropped in French (the first edition of the OED has it under hodometer). (Oddly, the TLFi has hodomètre as the main listing, even though other French dictionaries give odomètre.)

  24. Thanks.

  25. J.W. Brewer says

    California not only has a city named Eureka but another named Yreka. Wikipedia gives several different (and perhaps incompatible) accounts for the name of the latter.
    Expanding on a point made above, I don’t have a good sense of the proportion of initial rough breathing to smooth breathing in the Greek lexicon for words beginning epsilon-upsilon (my “Little Liddell” doesn’t sort by smooth v. rough and I don’t have the time to conduct a census), but it seems noteworthy that “heuristic” and its derivatives may be the *only* commonly-used Greek-derived word(s) in English that sounds like that. At least that’s the only heu-initial Greek-derived word in my old Am. Her. Dict. (“heulandite” having a non-Greek etymology), by comparison to about three columns worth of eu-initial Greek-derived words.

  26. That’s an excellent point (made earlier by HP, but I neglected to comment on it then). I think that’s the most likely explanation: contamination from all the eu- forms.

  27. Mediaeval and modern Latin just wasn’t at all consistent about transcribing the rough breathing as h; from my medical-jargon notes, all the following are exceptions to it (and note that some of them involve whole classes of words familiar to everyone, not only medics):
    Enanthem
    Ancon
    Apophysis, indeed everything starting with apo-
    Epiploön, the omentum or caul; indeed, everything starting with epi-
    Ophthalmia
    Anopsia
    Angioma (and everything starting with angi-)
    Azygos
    Adenoma
    Atheroma
    Ependyma
    Onychogryphosis
    Atopy, atopia
    Omohyoid
    Atelectasis (and I realise now that basically every word with a privative ἁ- in Greek should have h)
    Ethmoid
    Entropion (and everything else with initial ἑν-, in; eg, entropy)
    Ectropion (and everything else starting with initial ἑκ, out; eg. ectopic)
    Icterus
    Aphtha

  28. Art thou serious?
    en- and ek- are good old prepositions and I don’t see by which magic they become aspirated.

  29. The answer is so simple and obvious that it doesn’t seem like an answer; indeed, several of the comments above have included it as an assumption (or indeed correctly posited it, though perhaps not drawing enough attention to it as the answer): the rough breathing sign is a diacritic; therefore in a transliteration (as opposed to romanization) it would be omitted. That’s all.
    Aidan Kehoe’s list is very instructive, but perhaps overly inclusive; many, if not most of those words come from roots without rough breathing (apo-, epi-, and all the alpha-privative ones jump out).

  30. The answer is so simple and obvious that it doesn’t seem like an answer
    Furthermore, it’s not an answer. If it were, no English words derived from Greek would start with h-, as I have said repeatedly above.
    And Aidan, you seem not to be clear on what a rough breathing is; on a quick overview, I don’t think any of those has one.

  31. It is an answer, albeit an unsatisfying one. The Greek-origin words that start with “h” were usually imported through Latin (which was pretty consistent about rendering rough breathing with an “h”) or were coined in the modern era by people who knew Greek to some extent and who would therefore be faithful to the rough breathings where they occurred. “Eureka” doesn’t quite fit either pattern; people saw it in Greek letters (as in the 1570 example cited in the OED and the first edition of Joseph Andrews noted above) and derived the new English word by transliterating it.
    I once asked my Greek professor precisely this question, and he said, “Because they left it off.” I too thought this was a nonanswer, but eventually I figured out what he meant.

  32. Bill Walderman says

    OK, I think I have a plausible hypothesis. The story is apparently recorded in Vitruvius’ work De Architectura, which was written in Latin, not Greek and was preserved in Western Europe.
    http://en.wikipedia.org/wiki/Eureka_(word)
    The textual tradition of Vitruvius from antiquity through the middle ages would probably have preserved the Greek word in capital letters without diacritics: EYRHKA, since as someone else noted, the diacritics weren’t added to Greek texts (apart from Homer) until the Byzantine period. (This is typical of Greek words preserved in Latin texts, which incidentally often got garbled in the course of transmission by copyists who didn’t know Greek at all.) The word likely made its way into English from a text (either Latin or an early English translation) which reproduced the word or an English transliteration thereof without the Greek diacritics, as it was found in the MSS (and probably the early printed editions as well) of Vitruvius. In languages such as German–probably languages that have the /h/ phoneme–the initial rough breathing (“h”) has been restored by scholars, but that doesn’t seem to have happened in English. Again, this is a hypothesis, but it does attempt to address the question.

  33. Bill Walderman says

    “but that doesn’t seem to have happened in English.” Or at least didn’t happen consistently in English, and the form without the initial h won out.

  34. Bill Walderman says

    “but that doesn’t seem to have happened in English.”
    Or at least didn’t happen consistently in English, and the form without the h seems to have won out. One other point: both Greek and Latin manuscripts in antiquity were written entirely in majuscule (i.e., capital) letters. Greek minuscule hands didn’t emerge until about the 9th century, if I’m not mistaken. So Greek words in Latin texts would originally have been written entirely in what came to be known as capital letters (“capital” because once minuscule letters developed, majuscule letters were used exclusive at the beginning of chapters (“capita”)).

  35. By dropping Haitches as they be hard to mouth for some [hoy ?], does that explain why Iberians used a Jay to voice the missing “H” as there be no j in the lingo of conquerors of Espania??????

  36. people saw it in Greek letters (as in the 1570 example cited in the OED and the first edition of Joseph Andrews noted above) and derived the new English word by transliterating it.
    OK, I see what you mean now, and Bill expands on it convincingly. I guess the EYRHKA hypothesis is at least as good as the eu-words hypothesis. Apparently I was overestimating the Hellenism of the average 17th-century EYRHKA adopter.

  37. “All the initial h’s in English were lost some centuries back and then some,….”
    I believe that this is true of H in non-stressed syllables, as in Eureka, but not true of Hs in stressed syllables.
    So if English ever pronounced the H in this form, it was immediately lost with all the other Hs in non-stressed syllables.
    Words like humor, honest, herb, etc. are all irrelevant, as being in a different linguistic environment.
    I also believe that this is the explanation for the crazy British “an historian” where they have re-instated the H but kept the “an” which should not occur before an H that is pronounced.

  38. Bill Walderman says

    “Greek words preserved in Latin texts . . . often got garbled in the course of transmission by copyists who didn’t know Greek at all”
    Just like me — I should have written EYPHKA, not EURHKA.

  39. mollymooly says

    @Breffni:
    I’m guessing your childhood friends were /hju/ h-droppers from whom you learnt “humour” but not “humus”. I have a few words with /w/ instead of /hw/ for this reason.

  40. FWIW, I was taught/modeled to say “an” in front of H- on an unstressed syllable, so “an historic occasion” but “a history”. I know I still do it, because I get called on it occasionally, but I don’t notice it and it doesn’t bother me that others don’t.

  41. Latin (which was pretty consistent about rendering rough breathing with an “h”)
    There are plenty of cases where the rough breather wasn’t ‘rendered’ into Latin – because a cognate word was already in Latin.
    Gr. [h]erpw; Lat. serp[long]o
    Gr. [h]ex; Lat. sex
    Gr. [h]uper; Lat. super
    Gr. [h]upo; Lat. sub
    Evidence of a morphological evolution in one (or both) of these languages that pre-dates their (or one of their) having been written?
    (serp[long]o is one of many examples of the Latin word being, or appearing to be, ‘closer’ to the Sanskrit word than to the Latin: Skt. [root]s[ri]p to creep, slither, crawl. — a reason for a (quite) tentative hypothesis that, in this morphological category, it was the Greek (anti-psilotic) rough breather that was the innovation, rather than the Latin “s”.)
    English “serpent” is borrowed from the French daughters of Lat. serp[long]o, but English “six” is a descendant of a genuine Old English cognate of both Gr. [h]ex and Lat. sex, all daughters of (I guess) the proto-Indo-European *6.

  42. Psilosis–the loss of the initial rough breathing
    Actually, psilosis is more interesting than this change; it also involves the de-aspiration of mid-word consonants. OED calls this mid-word change a “substitution” and gives the example [h]rapus “for” [h]rap[h]us. — Though why should one assume that the oldest “Greek” language had the aspiration, which was ‘lost’, rather than not, and the aspiration was added later?

  43. komfo,amonan says

    @deadgod: Forgive me if I’ve completely understood you, but a distinction must be drawn between words that are cognate between Latin & Greek (such as your list), & words that Latin borrowed from Greek, such as, er, hypodidascalus (picked at random from Cassell’s).
    (serp[long]o is one of many examples of the Latin word being, or appearing to be, ‘closer’ to the Sanskrit word than to the Latin: Skt. [root]s[ri]p to creep, slither, crawl. […])
    I’m thinking one of the two instances of ‘Latin’ here is erroneous.
    Anyway, Pokorny has *serp- for ‘creep, crawl’. So it would indeed seem that PIE s- became Greek h- sometimes (Wikipedia, FWIW, confirms this).

  44. We can’t be sure the OED has a continuous early history. But look at the quotations given in the first comment above. First is ΕΥΡΗΚΑ in Greek. Then Heureca, a full Latin transliteration (note the c as well). Then Eureka, in a translation of Giambattista della Porta’s Magia Naturalis (transcribed here). Compare the Latin original: 1597 in GB, 1589 in BUCM (Ir a pag. 302); it has εὕρηκα and ὑπερεύρηκα untranslated. Furthermore, note that in the earlier edition the accents are different: ἕυρηκα and ὕπερευρηκα.

  45. Admirable well covered, and people here had a better response than I’d have come up with. Just a quibble with Bill W: I don’t think Archimedes would have said heura:ka. Proto-Greek /a:/ went to Ionic /e:/, so Doric often has /a:/ where Ionic has /e:/. But Proto-Greek did already have /e:/ — so etas do exist in Doric. In fact, when Pythagoreans try to write in Doric in the Hellenistic period, their overuse of /a:/ for /e:/ is a giveaway that they’re just faking their Doric.
    In the case of εὑρίσκω, the /e:/ of εὕρηκα looks to be Proto-Greek, as hinted by the alternation with a short /e/. (Sophocles has εὑρέθην “I was found”, and the short /e/ is what survived as Modern Greek βρέθηκα). We do have attested instances for heura:- (εὑρᾶναι εὑρᾶσθαι) , but they’re all Byzantine.

  46. J.W. Brewer:
    >Expanding on a point made above, I don’t have a good sense of the proportion of initial rough breathing to smooth breathing in the Greek lexicon for words beginning epsilon-upsilon (my “Little Liddell” doesn’t sort by smooth v. rough and I don’t have the time to conduct a census),
    Here’s my census from the TLG lemmatiser that I work on:
    Greek lemmata starting with eu- : 3220
    Greek lemmata starting with heu- : 43
    The only lemmata starting with heu- that are not derived from εὑρίσκω “find” are the verbs εὕδω “sleep”, and εὕω (and derived εὕστρα and εὑστόν) “singe”.
    And given the productivity of the eu- “good” prefix, analogy would have been an awfully strong factor.

  47. mollymooly, my whole family says “humour” with /ju/, so I presumably got it from my parents. In fact, I’ve always thought of it as standard, at least in Ireland. I take it you have /hju/ in “humour”?
    “Humus” and “humerus”, if I’ve ever said them aloud, are probably spelling pronunciations. “Human” is definitely /hju-/.
    I have a friend who used to say “weapons” and “Weetabix” with /hw/. Possibly he still does; neither subject comes up as often as when we were ten.

  48. There are plenty of cases where the rough breather wasn’t ‘rendered’ into Latin – because a cognate word was already in Latin.
    That’s like saying “We’re all African-Americans, because all humans came out of Africa to begin with!” While unquestionably true, it’s also irrelevant. We’re talking about borrowings, not cognates.
    The only lemmata starting with heu- that are not derived from εὑρίσκω “find” are the verbs εὕδω “sleep”, and εὕω (and derived εὕστρα and εὑστόν) “singe”.
    Thanks much for that!

  49. J.W. Brewer says

    I also appreciate Nick N.’s info on how lopsided the situation is/was in Greek as well. The semantics of the heu-initial lemmata remind me of Tennyson’s famous line “To singe, to sleep, to find, and not to yield.” Or something like that.

  50. Only a complete irrelevancy. There are a lot of places in the US called Eureka.

  51. And Aidan, you seem not to be clear on what a rough breathing is;

    Quite; hahah, that was embarrassing.

  52. J.W. Brewer, you have no poetic soul. Obviously the line must be “To find, to sleep, to singe, and not to yield.” Anything else is mere dissonance and cacophony.

  53. hahah, that was embarrassing.
    Don’t worry, we take turns being embarrassed around here.

  54. mollymooly says

    I take it you have /hju/ in “humour”?

    Yes. For me the choice dor an unfamiliar word with huCV- is /hju/ [as in “huge”] or /hu/ [as in “hula”], never /ju/. I associate /ju/ with Cork, but for “human” as much as “humour”.
    The eminently logical /hw/ in “Weetabix” is only precluded by the accent-specific krazee spelling.

  55. Zhoen is back!

  56. a distinction must be drawn between words that are cognate between Latin & Greek […] & words that Latin borrowed from Greek
    While unquestionably true, it’s also irrelevant. We’re talking about borrowings, not cognates.

    The point I made – that Greek rough breathers ‘appear’ in Latin in a cognate form (s) as well as having been borrowed (h) – depended on this “distinction”:
    where the rough breather wasn’t ‘rendered’ into Latin – because a cognate word was already in Latin.
    Clearly (to me), I wasn’t arguing with or even qualifying a position, but rather was going in a related but new (and to me: interesting) direction – though not as “new” to the conversation as the African origins of all Americans.

  57. Bill Walderman says

    “why should one assume that the oldest “Greek” language had the aspiration, which was ‘lost’, rather than not, and the aspiration was added later?”
    Comparative evidence makes it clear that the aspirated and unaspirated occlusives were derived from different phonemes in proto-Indoeuropean and later merged in certain Greek dialects, particularly Ionic, but the merger didn’t occur in most Greek dialects–in fact, the distinction has generally been preserved in down to the present, although the aspirated occlusives have become fricatives.
    “I don’t think Archimedes would have said heura:ka. Proto-Greek /a:/ went to Ionic /e:/, so Doric often has /a:/ where Ionic has /e:/. But Proto-Greek did already have /e:/ — so etas do exist in Doric.”
    That’s absolutely correct, and it’s possible that Archimedes (if the anecdote is true, which is doubtful) pronounced the word with an eta. But who knows what the perfect of heursiko: would have been in his dialect? I don’t think the -e:ka ending of the perfect is proto-Greek, however. Perfects in -ka were a relatively recent development in Greek, and by that time epsilon (a short closed e) alternated with epsilon iota (a “spurious” diphthong which wasn’t a diphthong at all, but rather a long closed e), not eta (which was an open e, probably closer to “a” in Eng. “at”).
    I couldn’t find an explanation for the perfect form heure:ka in any of the sources at my disposal, but I suspect it’s a form built by analogy with the contract verbs in -ao: and -eo:, which formed their perfects in -e:ka. So you can’t really tell whether there is an underlying alpha or an underlying epsilon.
    It’s also true, as Nick points out, that anyone wanting to sound “Doric” would simply substitute alpha for eta wholesale wherever it occurred, without necessarily having accurate linguistic knowledge of any particular Doric dialect. For example the choruses of Greek tragedies are in a kind of Doricized Attic, with a few of the more salient features of Doric superposed on Attic words (because the genre of choral song had to be Doric). The dialogues, on the other hand, are written in a lightly Ionicized Attic, because iambic verse had to be Ionic. Each genre had its own appropriate dialect.
    On the other hand, I wonder whether we really know how much of this is due to the original tragedians themselves, and how much to Alexandrian scholars of the 3rd century BCE or Byzantine scholars of the 13th century CE, who maybe had no qualms about changing texts to fit their notions of what Doric or Ionic should look like.

  58. Bill Walderman says

    Another thought: I wonder whether the Archimedes eureka anecdote didn’t circulate much earlier than the 17th century. Vitruvius was avidly read in the Carolingian era and there is a rich manuscript tradition from that period on. There’s an important 11th century manuscript–the source of many of the later medieval mss.–that was in Canterbury by the 15th century, and may have been in England since the 12th century; indeed, it may have originated there. (I get this from Reynolds, ed., “Texts and Transmission,” Oxford 1983, p. 443.) The Archimedes anecdote might well have been widely known before the Renaissance, and educated people probably knew just enough Greek–maybe only the alphabet–for the word to have entered the language without the initial h.

  59. The assembled have overlooked what seems obviously (to me) to be the most obvious reason: we say the word without the /h/ because that’s how he said it. I mean, it stands to reason, doesn’t it? Occam’s Hrazor and all that.

  60. Bill Walderman says

    “we say the word without the /h/ because that’s how he said it. I mean, it stands to reason, doesn’t it?”
    Not really. We’re dealing with the English pronunciation of a Greek word found in a Latin text. The Latin text was written more than 1,000 years before the word found its way into English and the Latin text recounts a probably apocryphal anecdote about an event that supposedly occurred 200 years earlier. We can be pretty certain that the Greek word was pronounced with an /h/ in most but not all Greek dialects both at the time when the event supposedly happened and at the time the Latin text originated.

  61. Bill, I’m pretty sure Z.D. was ‘avin’ a little joke.

  62. Bill Walderman says

    Joke? What’s that?
    The pronunciation of a 2,200 year old word is at stake. How can anyone have a joke at a time like this?

  63. There are some early recordings of Archimedes.

  64. That was when he was doing clubs in Hamburg, right? I like the raw, untutored sound.

  65. Terry Collmann says

    Shouldn’t that be Amburg?

  66. David Marjanović says

    Except for Scotland and the extreme north of England, English seems to have been completely aitchless around the 14th to 17th centuries or something. The King James Bible puts an or mine in front of every single word that starts with h, including hand and hundred, and Sir Walter Scott (I’m told) made fun of “the eclipsed manner of the Queen’s English”. The upper class, and perhaps the religious nuts who emigrated to America, then restored the /h/ from the spelling and presumably the Scots.
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    Off the top of my head, the (Proto-)Greek /h/, which only occurs at the beginnings of words, has several sources:
    – From *h1 followed by /j/.
    – From /s/. This is a fairly common sound change that has more recently happened, for instance, in Yakut (“Sakha”… rather Hakha) and the Innviertel dialect of western Upper Austria.
    – Out of nowhere, for prosodic or whatever reasons that I don’t understand.

  67. Comparative evidence makes it clear that the aspirated and unaspirated occlusions were derived from different phonemes in proto-Indoeuropean and later merged
    I’d be interested in reading a source (I mean: I’d read it; I’m asking for a citation) for this.
    the (proto-)Greek /h/, which only occurs at the beginnings of words
    It’s attested (much later than proto-Greek, of course) in consonant aspiration of prefixed prepositions’ final sounds – therefore, in the middle of ulcerous (?) verbs. I’m thinking of Psappho’s ka[th]eudw, the result of prefixing kata to [h]eudw (this last mentioned above as an unusual example of a non-[h]eurein-connected aspirated initial eu. . .).
    Is it so sure that some groups of intervocalic aspirates in the earliest ‘Greek’ words aren’t similar aspirations of mid-word unaspirated consonants due to an initial /h/ having been ‘pushed’ into the middle of a word?

  68. I like the raw, untutored sound.
    That’s the sound of the bath water.

  69. Bill Walderman says

    “I’d be interested in reading a source (I mean: I’d read it; I’m asking for a citation) for this.”
    http://www.amazon.com/New-Comparative-Grammar-Greek-Latin/dp/0195373367/ref=sr_1_1?ie=UTF8&s=books&qid=1264772592&sr=1-1

  70. andrew ollett says

    1) there seems to be no consensus on the etymology of εὕρισκω and consequently the original source of its aspiration. chantraine favors a root *wer (probably more up-to-date *werh1), so the stem *heur- either comes from the addition of a “prothetic” vowel and analogical aspiration, or from a reduplicated stem *we-wr with dissimilatory loss of the first *w. the second finds some corroboration in old irish (apparently) -fūar “i found”.

  71. andrew ollett says

    apparently the rest of my comment got lost. so here it is:
    it’s worth noting that greek initial /h/ can also come from PIE */j-/ (as in ὅς), not just */Hj-/. and as to deadgod’s question: some greek aspirates come from unaspirated consonants immediately before /s/ (as in αἰχμή “spear”, spelled a3-ka-sa-ma /aiksmas/ in linear B).
    2) on the basis of -ε- in some forms (εὑρέθην etc), and because the old irish cognates (and old church slavonic ob-rětǔ “i found”, apparently) suggest a stem in *wrē, the eta of εὕρηκα probably represents a proto-greek */ē/, PIE */eh1/. so syracusan doric (which also has kappa perfects, which are a greek innovation but a pan-greek innovation) would have had εὕρηκα. and in fact we find forms with eta in archimedes’ writings (some of which, however, have been ionicized or koinized). it’s worth nothing, as an aside, that the works transmitted in some semblance of doric have the rough breathing (i.e., none of archimedes’ writings are psilotic).
    3) breathings and accents were somewhat haphazard, and i would not be surprised if vitruvius didn’t write them. i would be less surprised if the scribes failed to write them. and i would be even less surprised if the pronunciation “yoo-ree-ka” is entirely due to the fact the rough breathing was ignored by the (mostly greekless) people who made this story popular.

  72. Thanks, Bill and Andrew.

  73. Bill Walderman says

    Thanks, Andrew, you’ve vindicated Nick, and I concede the point. Plato has the noun heuresis with epsilon–more evidence of an original long e.
    One puzzling thing about heure:ka is that it doesn’t have reduplication, although apparently an alternative form with eta in the first syllable, he:ure:ka, is attested.
    “the works transmitted in some semblance of doric have the rough breathing (i.e., none of archimedes’ writings are psilotic).”
    Weren’t the breathings in the works of Archimedes transmitted to us added in the Byzantine era, when copyists would not have accurate knowledge of whether Archimedes was or wasn’t suffering from psilosis? (Although I suspect he wasn’t.)
    But Vitruvius, writing in Latin in the 1st c. BCE would definitely not have used any diacritical marks–he would have just written the word in capitals–and the western Latin tradition in which his work was transmitted would not have added the diacritics. And, as I suggested earlier, it’s possible that the story was in circulation among educated people with virtually no knowledge of Greek, and the word eureka began its entry into the English language, even before the Renaissance.

  74. Vitruvius, writing in Latin in the 1st c. BCE would definitely not have used any diacritical marks
    I thought the polytonic marking system was an invention of Aristophanes of Byzantium, in Alexandria, from ca. 200 BC. Had it not diffused around the Greek-literate world to the Latin west, what, a hundred-plus years later? so that, if the Archimedes anecdote was ‘Vitruvian’, it carried the EURHKA, or EUREKA, as a Latin ‘phrase’ with it from its (western, non-diacritically-marked) origin (as a phrase generator) right from Vitruvius’s less Greek-current milieu?
    That would make it seem greatly likely that ‘heureka‘ (in modern scripts) was a Renaissance re-birth (in small) of a long lost bit of ancient erudition – indeed, ‘lost’ before the knowing was even possible when the knowledge had ‘disappeared’!

  75. That Platonic use of [h]euresis – the one at Rep. 336e – occurs at a most interesting moment.

    Socrates and Polemarchus have been trying to define “justice” and have reached elenchtic obstructions to their progress – ‘no, that can’t be it’, of “benefiting one’s friends and harming one’s enemies”. Polemarchus has agreed that, together, they’ll “fight” anyone who says that “our wise and blessed men” would say such a thing.
    Thrasymachus has been listening with growing impatience, barely restrained by his neighbors (it is a gathering at a private residence), and finally bursts into the circle (of Socrates’s audience – Socrates is good at drawing people around him) and shouts at Polemarchus and Socrates that Socrates just asks questions without answering them, and, spoken threateningly, that it’s his (Socrates’s) turn to say what he thinks “justice” is.
    (Always remembering: the Republic itself begins with Polemarchus and his friends (explicitly including one of Plato’s brothers) ‘forcing‘ – jocularly – Socrates and Glaucon (another of Plato’s brothers) to attend the ‘party’.)
    My translation:
    And I, having heard [this], was struck, and, looking at him, was afraid, and I suppose that, if I had not been looking right at him before he [turned directly to] me, I would have been speechless. Now – when he had begun to be enraged by the conversation, I had first noticed him, so that I was able to begin to answer him, and, trembling, I said, “Oh Thrasymachus, don’t be harsh with us! For if we have missed the mark in the speculation of [our] conversation – both I and this [guy] – know well that we miss the mark unwillingly! Surely you don’t think that, if we were searching for gold, it would ever be against our will to defer to each other in the search and [so] to destroy [our] discovery {[h]euresin} of it, but, searching for justice – a deed more to-be-valued than a lot of gold – then truly mindlessly [would we be willing] to submit to each other and not to be earnest about making this [deed] appear. Think it so [ie. that we miss the mark unwillingly], oh friend. But I don’t think we could [do such a thing]. Therefore, for us to be pitied by you clever/fearsome [people] would be better than [for us] to be treated harshly.”
    Thrasymachus responds sardonically and says that [h]auth ‘keinh [h]h eiw[th]uia eirwneia Sokratous (‘this [is] that customary irony of Socrates’s’) that he’d predicted would come about to his companions, namely that Socrates eirwneusoio (‘would trifle; would pretend’).

    Do you see? At this remarkably densely ironic moment, the Socratic irony is performed, (presumably) is understood by the reader, and is discovered by a character in the text – a character who, like Oedipus during his investigation, does know, but doesn’t know what he knows.

  76. Bill Walderman says

    “the polytonic marking system was an invention of Aristophanes of Byzantium, in Alexandria, from ca. 200 BC”
    It’s true that Aristophanes is supposed to have invented the accents. I think the rough breathing (which was originally one-half of an H) is actually much older, appearing in some archaic inscriptions. (The smooth breathing–the other half of H–may actually have been a Byzantine invention.)
    But the diacritics weren’t systematically applied to ancient Greek texts until the Byzantine era. The accents were originally invented to aid in the reading of Homer, and in antiquity the diacritical marks, including the rough breathing, were generally limited to scholarly texts of Homer and maybe some other archaic poetic texts for which contemporary Greek speakers needed special help, and even in those texts they weren’t consistently applied. They weren’t used at all in ordinary Greek writing.
    Here’s an image of a papyrus of a text of Demosthenes’ famous speech “On the Crown”, beleived to date from the 1st c. BCE, i.e., from about the time of Vitruvius:
    http://library.princeton.edu/libraries/firestone/rbsc/aids/papyri/images/Demosthenes.jpg
    In the second line you can clearly see the word “he:/sthe:n” (“I sat,” or “I waited”), written without a rough breathing or accent, and the rest of the text has no diacritics, either. (The word actually occurs in a letter supposedly written by Philip to the Thebans cited by Demonsthenes in his speech, but today most scholars think that quoted passages of letters and legislation in ancient Greek speeches are spurious additions by later editors of the texts designed to fill out the contents of documents that were read in court but not included in the original speeches themselves.)
    And here’s another image, this time of a papyrus text of Homer dating from the 2d c. CE:
    http://library.princeton.edu/libraries/firestone/rbsc/aids/papyri/images/Homer.jpeg
    This text has no diacritics, either. Even most texts of Homer weren’t supplied with diacritics in antiquity–the diacritics were only in special editions prepared by scholars.

  77. The accents were originally invented to aid in the reading of Homer
    Yes, “reading” entailing scholarly lexical and metrical analysis – Homer (even then) needing re-birth.
    I’m a little skeptical of the supposition that Aristophanes “invented” the whole of the polytonic diacritical system (perhaps excepting the ‘half’ of the “H” which (again, supposedly) was adapted to make the soft breather).
    Aristophanes, a widely respected student of Callimachus and chief Librarian, was known for his brilliant prosodic investigations and Homeric scholarship, and the diacritical innovations made possible much closer understanding of Homeric words, grammar, and ‘feet’ . . . so . . . Aristophanes invented the diacritical marks! This supposition seems, of course, possible, perhaps likely, but also possibly coincidental, to me.
    Vitruvius was also (at least) something of a scholar, and, despite his western location and busy life, he surely could have been knowledgeable about Alexandrian Greek-script innovations – or, as you say, a pretty good chance ‘not’.
    Anyway, the recovery of the rough breather on Vitruvius’s quotation of Archimedes does seem most likely to have been a Renaissance re-birth of a sound-mark long lost in the West, regardless of whether Vitruvius himself had been unusually modern in his use of [h] or, more typically, ignorant of the mark.
    So a tentative answer to language hat’s question would be: the rough-breather mark disappeared in the West, whether Vitruvius’s original text had it or not, to the tiny extent that it ever was in the West, and despite its continuing use by scholars in the East. (The rough-breather sound is agreed not to have survived the transition from Hellenistic/koinh Greek to Byzantine Greek (??).) Its uneven ‘re-birth’ in Renaissance and modern quotations of the Greek word [h]eurhka is a mark of some particular quoter’s precision and knowledge, or dogmatic pedantry.

  78. Bill Walderman says

    “the diacritical innovations made possible much closer understanding of Homeric words, grammar, and ‘feet’ . . . so . . . Aristophanes invented the diacritical marks! This supposition seems, of course, possible, perhaps likely, but also possibly coincidental, to me.”
    You could say that the invention of the accents is traditionally attributed to A. of B. They didn’t aid so much in understanding Homeric words, grammar and meter, as in the reading and recitation of the Homeric poems. After all, most Greeks experienced Homer through hearing him read aloud, not through reading, and those who did the reading would have needed pronunciation help with some words. But as you can see from the papyrus fragment, the breathings and other diacritical marks weren’t widely used even in Greek mss. of Homer in antiquity. And even in mss. in which they were added, they weren’t added consistently and uniformly the way they are in modern editions of ancient Greek texts.
    The breathing marks (rough and smooth) were derived from the symbol H, which represented the “h” sound (voiceless laryngeal fricative or something like that) in the Attic alphabet in use in Athens until 403 BCE. (The symbol E denoted both /e/ and /e:/ in the old Attic alphabet; in 403 BCE the Athenians officially adopted the Ionian alphabet, which assigned the symbol E to /e/ and the symbol H to /e:/), although it was probably in unofficial use before then).
    I don’t think it’s strictly speaking accurate to say that the breathing marks were “lost” in the West–it’s just that, like punctuation marks and word divisions, they were not widely used in Greek writing–even in Greek-speaking areas–until the Byzantine period. Vitruvius undoubtedly knew that the word was pronounced with a rough breathing, but, just like his Greek contemporaries, he wouldn’t have written the symbol.
    But you’re right that when the word heureka appears with an “h” in modern European languages, it’s due to scholarship or pendantry, take your pick.

  79. Bill Walderman says

    Here’s an image of one of the most important Byzantine manuscripts of the Iliad, the Codex Venetus A, dating from the 10th century. You can see that the breathings are still half-Hs.
    http://chs75.chs.harvard.edu/manuscripts/image-viewer?folio=12r&ms=msA&image=
    The Greek minuscule hand is not easy, but not impossible, to read. Lots of ligatures. It’s the very beginning of the Iliad — Me:nin aeide thea . . .

  80. They didn’t aid so much in understanding Homeric words, grammar and meter, as in the reading and recitation of the Homeric poems.
    Well, that’s a distinction without much of a difference, isn’t it? In order to recite the poetry, its syllables have to be marched in their proper ‘feet’. And the language had changed enough by the Hellenistic period – comparing the Iliad to, say, Theocritus – that some familiarization would be needed for the reciter and audience to catch the syntax of the ‘sentences’, in addition to the words that had either changed in shape enough not to be immediately recognizable or, simply, were no longer used (for whatever reason).
    most Greeks experienced Homer through hearing him read aloud
    Yes, but as you concede, this simply defers the problem of erudition to that ‘reader aloud’.
    Also, I wonder: how much actual performing of Homer did people in the Hellenistic period enjoy? This is after the revolutions of drama and literacy, and after a multitude of aesthetic/formal and social changes marked in the histories of those institutions. People educated in Greek, about whom we naturally have the most knowledge of the history of ‘Homer’, surely
    knew at least some Homer – every big shot from Callimachus to Proclus is evidence of that. But rhapsodic performances? for most of the population, even of Alexandria or Athens?
    You know, if I went to a Shakespeare play tonight, without running through the play again today, I wouldn’t catch a quarter of what the actors were saying: it’s too dense, often the ‘sentences’ are too strangely constructed, many of the words are unfamiliar – I’d see what the characters were doing, but I’d probably miss most of the poetry of what they were saying. With all the other entertainments that arose and became popular over the centuries after Pindar, how much Homeric poetry was “recited” in later antiquity??

    I’m pretty sure the [h] is a voiceless glottal fricative: made by using the back of the tongue to control – ‘to stop’ – the rush of air up the throat and into and out the mouth. The ‘middle Liddell’ says that “[w]hen H was taken to represent [long]e [in 403 BC, as you point out], it was at the same time cut in two, so the [the right half] represented the spir. asper, [the left half] the spir. lenis; whence came the present signs for the breathings.” (This is at the beginning of the hta ‘chapter’ – a pretty useful discussion.) Perhaps more recent scholarship (than the 19th c.) pushes the soft breather’s introduction far into the future from “the archonship of Euclides” (?).
    [not] accurate to say the breathing marks were “lost” in the West
    I meant that, at the time of, say, Virgil, there were surely some Romans who understood the diacritical innovations in Alexandria, but (I’m guessing) increasingly fewer, and after Boethius? I mean that the diffusion westward of this technique was partial, limited in saturation, and evaporated eventually, whereas the scholars in the East kept the copying and commentary traditions alive, if not equaling – not remotely – the literary quality of the texts lying before them.
    Vitruvius undoubtedly knew that the word [[h]eureka] was pronounced with a rough breathing
    Really?? I thought the gist of many of the comments on this thread was that that was exactly what was in doubt! – namely, that not only was it likely that Vitruvius didn’t write (somehow) a rough-breather mark, but that he might not even have known that, classically, it had been spoken.

  81. Cross-posted there. And I see I got “right” and “left” back-frontward: it’s the left side of the “H” which was modified into the rough-breather mark, and the right side which became the soft.

  82. Bill Walderman says

    “that not only was it likely that Vitruvius didn’t write (somehow) a rough-breather mark, but that he might not even have known that, classically, it had been spoken.”
    The glottal fricative didn’t vanish until the second century CE, according to W. Sidney Allen, Vox Graeca. If that’s correct, Vitruvius would have been aware that the word was pronounced “heureka,” but wouldn’t have indicated it in his writing. Regardless of how much Greek Vitruvius knew or didn’t know, almost no contemporary Greek texts would have been supplied with rough breathings, as you can see from the image of the approximately contemporary papyrus ms. of Demosthenes. In antiquity Greek was normally written without diacritics, without spaces between the words and entirely in caps.

  83. John Emerson says

    In antiquity Greek was normally written without diacritics, without spaces between the words and entirely in caps….on butcher paper in crayon.

  84. Diberville says

    Don’t know if someone mentioned it, but all the h-less words in English are old French words long introduced in English (hoir > heir, heure > hour, honneur > honour [hence the “u”], honnête [originally with an s] > honest).
    Same goes for ex h-less words like hotel, herb, hospital and humour (from Fowler’s Modern English usage, 2nd ed.)
    Other cases of silent h existed in compound words like philhellenic and philharmonic, but the speak-as-you-spell movement has introduced an “h” there (ah, h is itself a mute-h word… as in French ;-))

Speak Your Mind

*