yep
You should try the Netherlands. Dialects galore.
Does it really have to be one or the other exclusively? Perhaps both pronunciations were present, in the ancient world. Like “either” and “either”. Maybe Greek speakers didn’t even tend to notice the difference between the two. Just a speculation.
Reminds me of a famous quote, from the world of computer programming: 
Should array indices start at 0 or 1? My compromise of 0.5 was rejected without, I thought, proper consideration.
–Stan Kelly-Bootle
I’d say it’s a 3rd (or iota) class present, making the zeta from dy. Smyth sec.508 gives other examples.
This is IIRC explained by the use of reduplication. Similarly to *TITK- (TEK > TEK + TEK > TI + TEK > TI + TK > TITK > TIKT)
Thus SED + SED > SE + SED > SE + SD > SESD > [size=150]ἑζ‐[/size]
Remember that sigma is pronounded voiced before voiced consonants (most commonly beta, delta, gamma, mu)
It-s true that “s” before voiced consonants becomes (iat least n most cases) also voiced, and in some cases the previous or following consonant is ommited, but please do not pronounce it “zd” or “dz”. It’s an insult to the ears. Be sufficient speaking it “z”.
This maybe correct if we assume that both “s” and “sh” keep the present sounds and do not change. What happens if they change and (most probably) “sh” becomes “s” and we forget how “sh” sounded? That happened with gamma and beta, and to say now that gamma was “g” is totally wrong.
No.
I can, with effort, read Chaucer in Middle English. If I want his verses to scan, however, I have to do all sorts of things my native English doesn’t do. For example, “pierced” will have two syllables. If I don’t make these concessions to the changes of just under a millennium, the verse is ruined.
I’m not going throw away a reasonably sound understanding of the Greek of 500 BC in order to satisfy the ears of a speaker of Greek in 2004 AD. Especially if taking that road ruins the meter of the Greek I’m most interested in.
I can, with effort, read Chaucer in Middle English. If I want his verses to scan, however, I have to do all sorts of things my native English doesn’t do. For example, “pierced” will have two syllables. If I don’t make these concessions to the changes of just under a millennium, the verse is ruined.
Please, let me disagree one more time in this thread and ask you not to do this. We don’t have sufficient data as to how English sounded in those days, and your try will be an insult to those poets. You’ll probably speak “p” with modern English aspiration, though we don’t know if they did have such and to what degree. To their ears your “p” will sound either as “p”+“h” or soft “b”. Further they won’t be able to hear your (modern English) “i” and will mistake as (theirs) “e”. Next your “e” sounds too much simmilar to (a very short) “a” and makes things more difficult. (Someone may ask, which “a”? English has so many lol) Your “r” will be rolling or not? “c” is “k” or “s”? “K” with aspiration or without? (in other words will he hear a “k” or “g”?) What kind of "e’? And last which “d”? Stick in your modern Enlgish promunciation. Any other attempt is fruitless.
[/quote]
This depends on your definition of “sufficient.” If by sufficient you mean perfect, then you’re right. But that doesn’t mean we know nothing. I don’t see how it’s controversial to try to use the best information we have, not in the expectation that it’s perfect, but that it’s better than nothing.
and your try will be an insult to those poets.
Trying to get closer to a poet’s original words, however imperfectly, is an insult?! This makes no sense to me at all.
Any other attempt is fruitless.
No, it isn’t. And I will continue to pronounce (/ηφαιστος hehhp’haistos, and to recommend that pronunciation to people curious about ancient pronunciation, not because I think it reflects perfectly how Homer said it, but because there’s plenty of good evidence that it’s a heck of a lot closer than, say, ifestos.
I don’t understand why you’re dismissing the use of all the scholarship that has gone into this question, even if you’re not interested in using a reconstructed pronunciation yourself.
William, has your pronunciation changed along with your opinion how Zeta should be pronounced?
When I found out that Zeta was either zd or dz but certainly not z, I changed (with considerable effort) the way I pronounced it from z to dz.
I am not sure if I want to change again.
I don’t dismiss the linguistic science and any attempt to reconstruct sounds, but the attempt to put this pronunciations in practice and trying to read whole sentences knowing at the same time that probably it is wrong from beginning. Simple, we don’t have data to do this, and saying we may make this and that assumption and agree that “ζ” is “ζδ” is totally wrong. Especially if the same scholars also say that neighbor sounds have also a saying how a consonant should be spoken and a “ζ” is not always such. To what amountt his influence affects the sound? We will simple never know.
Following long-standing Textkit tradition, this debate has been reduced to an argument about epistemology! 
This is simply untrue. There’s all sorts of data, some of it referred to earlier in this thread. Granted, it’s not ideal fieldwork data, but we can’t just dismiss it.
and saying we may make this and that assumption and agree that “?” is “??” is totally wrong.
On the contrary. For the period of Greek I study (Homer and the archaic poets) “ζ” as “z” is totally wrong, and I know this with at least as much confidence as I know that “ξίφος” is sword.
I happen to know that both Bert and Benissimus are reading Homer, so I can either recommend a pronunciation that is known incorrect, or one that is possibly correct. I’m going to recommend the possibly correct version.
Yes, but imperfectly. I learned Erasmian pronunciation first (dz), and it’s taking a little time to overcome that.
gh only became silent after the Normans ruled Anglia.
You know, zeta could be pronounce not as d+s or s+d, but s and d at the same time. Tongue between the teeth and s-aspirate. Could work. Note: in Latin, z is a double consonant too. Anyone try to figure out the Ancient Latin of it? Hint, there is no PIE reconstruction of ‘z’. But the letter was borrowed from Phonoecian, the letter being ts/ds as in Moreh Tsaddik - Teacher of Righteousness in Hebrew.
The “sh” will probably be known for a while, it’s tracable. Skirt, Scirt, Shirt.
Well, this is exactly the way how Greeks pronounce “z”, and it does not sound like “dz” or “zd”.
Originally, Latin didn’t have “z” in its alphabet and was later adopted from the Greek one, sometimes in the second or first century BC, together with K, X and Y, which were used only for words loaned from Greek.
I happen to know that both Bert and Benissimus are reading Homer, so I can either recommend a pronunciation that is known incorrect, or one that is possibly correct. I’m going to recommend the possibly correct version.
This is for pure technical reasons impossible to do. All the scripts that we use today are re-written countless times, and every time they were refined and improve according to the tastes and phonology of those centuries, e.g how they spoke and wrote in their time. Especially if you happen to use all the acute and daseia signs etc. that are an invention of alexandrian “barbarian” grammarians, who in some cases probably didn’t speak good Greek themselves. Therefore, how Homer did speak is impossible to find out, since no written text exist from his days, the few we have are some centuries later, and the versions we use today are even more later.
Finally, the letter Z, as its frequent replacing of S before B, G, and D etc. shows 76, had a voiced s–sound like English s or z in “rose” and “zebra” respectively, not the Erasmian dz (ds) or zd (sd). The same is shown by the misspellings Seu=j (= Zeu=j, 340 B.C.); Busza&ntioi 77 instead of Buza&ntioi; e0peyh/fiszen and sunagwniszo&menoj instead of e0peyh/fizen and sunagwnizo&menoj 78. In Elis D was often substituted by Z 79. That this tendency occurred at Athens as well may be inferred from Plato, Cratylus, 418: “nu=n de\ a)nti\ … tou= … de/lta zh=ta (metastre/fousin)”. That this pronunciation of z as z was classical is shown by )Azeioi/, )Azeih=j 80, and )Azzeioi/ 81, as well as by Buza&ntioi 82 and Buzza&ntioi 83. That the z in all these cases could not have been sounded as dz or zd is shown by the resultant sound of the words, which is impossible to pronounce: A-zd-zd-e-i-o-i and Bu-zd-zd-a-nti-o-i. No doubt the Greeks pronounced them as A(z)ziü (later A(z)zií) and Bü(z)zantiü (later By(z)zántii ) respectively 84.
http://www.bsw.org/?l=72081&a=Art06.html
E-u-a-o-i-o-I
The impossibility of pronouncing the diphthongs in diaeresis (i.e. each vowel distinctly) becomes obvious also from a word such as Eu0aoi=oi (see IGA 110, 2, early VIth c. B.C.). This word, which consists of seven vowels, pronounced in the Erasmian way, would give the comical sound: ‘E-u-a-o-i-o-i’ — as if it were an exercise in vowel mnemonics. Surely the correct pronunciation was between ‘Eva-ü-ü’ and ‘Eva-í-i’.
And last I wonder if the Erasmic promunciation is of the 5th centuries, then why do you read Homer who lived some centuries prior using the Erasmic pronucniation? Did the language not change all those centuries? Do you than adopt another pronunciation when reading Plato, another when reading the Bible and even another one when reading the Church Fathers? I would become quite confused with so many ways to utter the same words
Oy. I have read this article before, and it is truly wretched. The author several times make fun of ideas without argumentation, and only shows himself ignorant of modern linguistcs.
The impossibility of pronouncing the diphthongs in diaeresis (i.e. each vowel distinctly) becomes obvious also from a word such as Eu0aoi=oi (see IGA 110, 2, early VIth c. B.C.). This word, which consists of seven vowels, pronounced in the Erasmian way, would give the comical sound: ‘E-u-a-o-i-o-i’ — as if it were an exercise in vowel mnemonics. Surely the correct pronunciation was between ‘Eva-ü-ü’ and ‘Eva-í-i’.
This is wishful thinking. Plenty of languages would have no difficulty with this. Just because it sounds comical to the author means nothing.
More:
Greek pronunciation cannot be determined by what is possible or acceptable in other languages.
Absurd. Such comparisons are the basis of all linguistics. Greek is not exempt.
Since accent as stress is integral to all speech, its existence in Greek must be as old as the language itself.
Again absurd. Plenty of languages don’t have stress accent (Japanese, say).
However, the form of the circumflex only indicated that it was the result of the contraction of two vowels, one o0cuno&menon the other baruno&menon, but it had no rising and falling tone in pronunciation — an impossibility in actual speech,
More nonsense. As any speaker of a tone language can tell you (many dialects of Chinese, other Sino-tibetan languages, etc.)
When the Greeks in time came to use the monographs θ, φ, χ in place of the digraphs, the Romans had no equivalents for these letters except for φ, hence Latin F is usually transcribed with φ!
Only after A.D.i.
(1) Stress need not exclude pitch, and in fact no pitch is conceivable without stress. (2) All Indo-European languages are based on stress accent. In Swedish, for example, which is the most ‘musical’ of the Scandinavian languages, stress-accent is clear and important. If Greek were different in this respect, it would have been unique.
This is completely false. Several slavic and baltic languages had and still have pitch accent.
(4) If the accent was essentially musical, why was it then disregarded by meter, which chose its own syllables — often unaccented — to express the pitch?
Because ancient Greek meter is based on duration. But of course the author of this paper discards vowel length… because Modern Greek doesn’t have it.
(6) Greek meter therefore must have been based on rhythm, which consisted in thesis (ictus) and arsis (fall) represented by the acute and the grave, the only proswdi/ai known in early times.
I trust the author has never read Pindar.
This paper basically proposes that ancient Greek experienced rapid variation in the a few centuries - 6th through 3rd about, it looks - and then never changed again. This would make it unique indeed.
I cannot take this paper seriously.
I do not use Erasmic. I use the work of Palmer, Allen, etc., as the basis for my pronunciation.
Do you than adopt another pronunciation when reading Plato, another when reading the Bible and even another one when reading the Church Fathers?
Well, I’m still working on Homer, Hesiod, and the archaic poets. I’ll worry about the Koine - for which the modern Greek pronunciations make a good deal more sense - when I get to that point.