Friday, January 27, 2006

Caragounis on pronunciation of Greek

Note remarks by Chrys Caragounis on the reception of his work on the pronunciation of Greek here. See further his webpage.


  1. Feisty! Makes me want to read Silva's review even more.

  2. The author’s writing on his site is very helpful as it makes several points about pronunciation of ancient Greek. I have not read the Silva reply, but I can guess some of what he might have said.

    He discusses how ancient epigraphics undermine the Erasmian system. This is no new news for the average classicist who is more than aware of what is commonly known as “the Erasmian Error”. One hardly needs to look to manuscripts and inscriptions for this evidence. It makes sense that the Greek letter PHI can not have been pronounced as an [f] in the rendering of lat. philologia (not e.g. filologia). (N.B. however the rendering of the name Felix in Acts with the PHI which was as close to [f]).

    Similarly, the Modern Greek (erroneously labelled “historic” by Ch. Car.) – like the Erasmian system – has dissimilarities to pattern present between ancient Greek, Latin and other languages. Consider beta. In Modern Greek, it is pronounced as English [v]. Latin renders the Greek term as “baptisma”. In other words, a modern pronunciation scheme has some of the same problems as an ancient one. A massive Slavic migration shifted not only consonantal sounds but the aspectual system of the language into its modern form… not trends that were in place in antiquity.

    The author argues for a historical trend toward the modern scheme. He notes the similarity between ei and IOTA in both languages. This seems like a tricky matter from a linguistic point of view. The pronunciation of vowels is complicated. At some point, ei must have been different from both ETA and IOTA, otherwise the phoneme would have been written in a different manner. A misspelling or tendency to misspell ei as i only proves that the sounds were similar or that they had eventually become the same… For paedagogical reasons as well as practical ones it is helpful to adopt a system that allows us to communicate with others and to differentiate sounds which are alphabetically differentiated. It makes sense that the system should be based upon historical-linguistic phenomena, but the argument here seems to be that the Erasmian system is not perfect so we should take on the Modern Greek one instead.

    With respect to pronunciation, linguistically the problem is a very complex one particularly with vowels. One of the main differences between Attic and Doric Greek was the handling of what would be an ETA in Attic and a long ALPHA in Doric. Neither was correct or incorrect. Neither shows a tendency toward a later scheme towards which the language was moving. They were just different dialects that coexisted in different regions. Likewise, we could expect that there were different sets of vocalic (and consonantal) pronunciation between the educated and unlearned.

    The Erasmian system is linguistically flawed. I have, however, found it helpful for American students who do not have conscious understanding of the difference between aspirated and unaspirated [p/ph]. I also find the Latinate rendering of the vowels easy to understand… It would be more confusing to pronounce both ai and EPSILON as [e]. Historically, there was a group of speakers who did distinguish the two, so a person can not argue that the one is correct over the other.

    This is not a defence of the Erasmian Error. I think that any pronunciation scheme has two goals: it has to be paedagogically useable and it should be as historical as possible. I do not see the use of the Modern Greek system (per Car.) as either.

    Sorry for the long post...

  3. In studying the sound shifts in my Greek grammar book, it looked pretty obvious to me that the PHI was originally pronounced as an unvoiced bilabial fricative, a sound the Latins could best represent by 'ph'.
    But I haven't seen this explicitly expressed anywhere.
    Didn't Homeric Greek have another letter (similar to the sign for the female sex) which would seem to have been best represented by 'f', and was retained in Koine only as a numerical symbol?

  4. Dan,
    If I understand correctly, the difference between PHI and PI is the same as between the English words "Peak" and "sPeak". I.e. in English we we aspirate our [p] sound when it begins a syllable (put a finger in front of your mouth when you say these words). I have not read what Erasmus suggested about pronunciation, but I gather that he had a hand in generating the popular notion that PHI was to pronounced as a fricative [f]. Latin had fricatives present in its vocabulary (familia), but it does not seem that there are any native terms with a "ph" spelling... they are all borrowed from Greek.

    As far as the symbol you mentioned, Corinth had a letter "Qoppa" and Homeric pronunciation was based upon the sound of the digamma which was represented by a letter which looked like an "f" and had an English [w] sound.

  5. I'm groping in the dark here (while keeping a fierce grip on my bilabial fricative theory), but I'm thinking that Erasmus came along a bit late to change Latin spelling of Greek words. The Latins had been spelling Greek PHI as ph all along, for the reason that its sound didn't approximate any for which they had letters. I don't think that a plosive P (as in English, not carrying any semantic load) would have necessitated a special orthography.
    But maybe I'm missing something here. Are there minimal pairs in Latinized Greed (e.g. medical vocabulary) that require orthographic distinction of plosiveness in the voiceless bilabial stop?

  6. Erasmus did not change the spelling of Greek words. My point with Latin spellings of Greek is that they represent the ancient phonology of Greek. The fact that they had written PHI as ph and not as f shows that it could not have been pronounced as the latter.

    I found this page which lays out the different systems of pronunciation:

  7. Guy, have you read the section devoted to the pronunciation of Greek in Caragounis's book (The Developement of Gree etc)? The case is pretty strong, as far as I can tell.

  8. As a Chinese speaker, I could confirm that there are quite a bunch of languages in which the aspiratedness distinction play a great role.