But that claim is that γμ was ngm, with the ng in sing.
How does it claim? It that was the case, than why did they not wrote nm, or even gm (so we can than quarrel if its even ngm)? I think n does not like to be place near plosives either. n+b gives mb and n+p makes it to mp, also n+k gives gk (the last g as ng, the others m are banalities m).
I’m confused now. Aren’t we discussing γμ (your gm I guess), in relation to the pronunciation of the word σιγμα?
Of course we are still after σιγμα pronunciation. Slowly I am proving that σιγμα could never have been pronounced sinma, neither singma. N (as well NG) before M becomes M, so G before M stood always as G, otherwise they would have written it from earliest times as MM (simplified perhaps to M). I think a clue gives also gramma, which probably comes from grap(h)ma, but p(h) before m stands not.
That reminds me of this little rule in castillian saying that ‘m’ is to be written before ‘p’ and ‘b’.
Perhaps William, belly full of pork and pints, can shed some more light on this for us. >
Well, I didn’t see the need to repeat modus.irrealis’ fine points.
N sure, but NG? Do you have any examples of that?
Do you have any example for the opposite?
I could go on and on proving that NM (and NGM) is simple un-Greek pronunciation. But not MN. Maybe it was always simna and people twisted it? 
Why wouldn’t it just stay as χ?
M is voiced, so χ has to become voiced. X turns to a voiced X, a G. This could give reasons to say that ancient-Greek’s G was never that hard-sounding G as in Latin or English, but a soft one, a voiced X. There are words with χμ, but in this case we have also the ypsilon before, working as a voiced semi-vowel (or semi-consonant?).
Edit: this is also an excellent example why one should never reject Modern Greek so easily. Why we have here GM and not XM, is easy to think if we take modern pronunciation into account. In Modern Greek the pronunciation would be τεβγμα, so we speak βγμ, no place is left for χμ. Astrong clue that ancient υ was never a u, not in such combinations as in this example.
You simply assert over and over that because it’s not in modern Greek it’s “un-Greek” tout court. This isn’t proof.
I use examples only from Ancient Greek. There are no words with NM, and whenever a n+m occurs, suddenly we have MM. There is no GM tracing back to NM. NM (and NGM, not to forget it) is simple un-Greek. either you take it granted, or not.
I leave it to you.
Here we go…
From Allen’s Vox Graeca, 3rd ed., pp35-36 (I omit references to his Vox Latina). Since not everyone will have an IPA font, I will use [N] in addition to the usual IPA [ŋ] (the ng in sing not anger).
"We have already mentioned that, in addition to the dental and bilabial nasals, there was in Greek, as in, for example, English and Latin, a velar nasal sound, occurring before velar plosive consonants, where it is represented by γ — e.g. ἄγκυ?α, ἔγχος, ?γγύς. Varro identified this with the sound of the n in Latin angulus etc., which was clearly a velar nasal (described by Nigidius Figuls as `inter litteram n et g’ and as not involving contact with the (hard) palate). The use of n to indicate this sound, as in Latin, is understandable enough, since the velar pronunciation is automatic before velar plosives; and similar spellings with ν are found in Attic inscriptions (regularly before 5 c., e.g. c. 550 ενγυσ). But the normal Greek spelling with γ for [N] ([ŋ]) is on the face of it remarkable, since it is as though we were to write e.g. English ink, finger as igk, figger. There is nothing in the nature of a velar plosive that would account for the nasalization of a preceding plosive; so that the only logical explanation for such spellings would be if γ had this nasal [N] ([ŋ]) value in some other environment where it was phonetically intelligible; from such a context the writing with γ could then have been transferred to other positions (on the principle, familiar also to some modern schools of phonology, that a given sound must always be allotted to the same phoneme).
The most obvious candidate for providing such an environment is the position before a following nasal, that is, if γμ and/or γν were pronounced [Nm, Nn] ([ŋm, ŋn]) (like the ngm, ngn in English hangman, hangnail), as in the case of Latin magnus etc.
There is in fact a tradition, preserved by Priscian (+Gl, ii, p. 30 K) as ascribed by Verro to Ion (probably of Chios), that the [N] ([ŋ]) sound represented by γ in ἄγκυ?α etc. had a special name in Greek, and that this name was ἄγμα; since the Greek names of letters are otherwise related to the sounds they represent, such a name makes sense only if it is pronounced [aGma] ([aŋma]), that is, if the γ is pronounced [N] ([ŋ]) in the position before the nasal μ."
He then goes on to give the perfect passive examples given earlier.
Interestingly, all three of the authors I consulted — Sihler, Allen, Palmer — say that in the case of -γν- the [ŋn] interpretation might be right, but it’s much less certain.
had a special name in Greek, and that this name was ἄγμα; since the Greek names of letters are otherwise related to the sounds they represent, such a name makes sense only if it is pronounced [aGma] ([aŋma]), that is, if the γ is pronounced [N] ([ŋ]) in the position before the nasal μ."
So the only clue we have is this: since it is called agma, so it had to be pronounce angma. I disagree, I’m not convinced. Do we have any mispelling for γμ (e.g. νμ, νγμ)? No, never was one. Priscian, as a genuine grammarian, wanted to remodel the langauge according to his ideal. Thanks God, it did never happen. I find it remarkable, there are other interesting mispellings even to our days: συνγνωμη versus συγγνωμη vs. συγνωμη.
Consonants: Attic Combinations:
What a shock.
And it is patently not the only clue we have — it’s merely one of several, all of which work together nicely. Why must we do this absurd epistemology tango every time this subject comes up?
Have you read Allen’s Vox Graeca? Do you have access to a copy? Would you read it if I arranged to have one get to you, if you haven’t already? Or are all linguists and grammarians, like Cretans, liars to you?
Blogger denies me access to these images.
Annis try copying their url from the Properties and then open them in a new window/tab. I did it and then for some reason they appeared on my textkit page too