Some greek alphabet questions

[zompist]

it explains the somewhat odd use of γ for ŋ before velars.

Mm? How would γ having a value of /ŋ/ before a labial nasal help explain it having that value before velar stops? I'd think you meant that /gm/ assimilated to /ŋm/ so that γ spread to the pre-velar nasal, but it seems like that would negate the irregular verb part, because if /ŋ/ and /g/ were spelled the same, there wouldn't be any visible orthographic difference.

I was trying to avoid a lot of typing— if you want the whole argument, find the book. But here's his verb example (sorry, can't be bothered to use Greek characters):

present / 3s perf / 1s perf

1. lég-omai / lélek-tai / léleg-mai
2. phthégg-omai / ephthégk-tai / ephtheg-mai

Compared to Verb 1, Verb 2 has a nasal ŋ before the final consonant of the root in the first two forms. If <gm> is /gm/, this nasal is lost in the 1s perf. If it's /ŋm/, then verb 2 always retains its nasal.

(As he points out, to be completely regular we'd expect ephthegg-mai; but that would have /ŋŋm/ which gets simplified to /ŋm/.)

[finlay]

Do you mean /ŋm/ or [ŋm]? (It's just I didn't think Greek had /ŋ/ but I might just be wrong about that)

[zompist]

I mean /ŋm/ as Allen considers /ŋ/ a phoneme— but it's an extremely marginal one. There are a few words that contrast /gg/ with /ŋg/.

(A language where /ŋ/ is barely phonemic might sound quite unusual if we weren't using one...)

[vec]

(Icelandic is like that too. /ŋ/ only contrasts phonemically with /n/ before d and l.)

[Qwynegold]

/l/ > /ts/ is quite a weird sound change

[l] > [ɬ] > [tɬ] > [ts] wouldn't be all that weird, actually...

Is that l > ɬ an unconditioned change?

[Zhen Lin]

It's quite possible that I'd have to look on a Thai website, which I wouldn't be able to read anyway.

Thai is another one of those scripts. The sort used in advertising in Thailand looks nothing like the usual printed forms!

[hwhatting]

/l/ > /ts/ is quite a weird sound change

[l] > [ɬ] > [tɬ] > [ts] wouldn't be all that weird, actually...

Is that l > ɬ an unconditioned change?

Don't think so - standard Modern Greek still has /l/. This seems to be a dialectal development.

[Io]

The funny thing about Greek fonts is that I saw a much, much greater scale of variation when I was actually in Greece than I've been able to find on the internet. For instance, eta, which always has a tail in the fonts I have on my computer, often looks exactly like 'n', and lowercase chi and zeta are very often rendered as 'x' and 'z' – again, without the tails that they tend to have on my computer. Capital omega often looks like an O with a line underneath instead of, well, like an omega...

I think at least some of that could be put down to western influence or something.

You should see some medieval Greek manuscripts, they are pretty much impossible to read, also if you've ever paid attention to church inscriptions in Greece you've probably noticed that Σ is С, as for Ω being an O with an under score that's pretty normal, actually that's how it's usually written in handwriting. Personally, I started writing η as 'n' because I don't like how it turns out when I write it and eventhe natives write it like this.

One of the weirdest letters in handwriting is λ (link), as for υ and ν as others have already said they're written the way they look, you don't write ypsilon with an descender like Latin 'u', I'll try to get later some examples.

>In handwriting and various decorative fonts they're extremely commonplace

I really don't remember ever seeing in handwriting Л written in the ugly limped-П way, capital Д is either like printed delta (Δ) or more commonly like the handwritten capital delta.

[Nannalu]

I've seen, on the cover of a book presumably in Russian, cyrillic <Д> written as <Δ> and cyrillic <Л> written as <Λ>. I can't say I've ever seen these variations online though, so I'm not surprised in the variations of greek you've seen.

Russian <Д> is written like Delta in cursive and <Λ> is the cursive form of Cyrillic <Л> and it's a little lambda for lowercase - so there's nothing wrong there.

[makvas]

I mean /ŋm/ as Allen considers /ŋ/ a phoneme— but it's an extremely marginal one. There are a few words that contrast /gg/ with /ŋg/.

(A language where /ŋ/ is barely phonemic might sound quite unusual if we weren't using one...)

Interesting how that is; /ŋ/ could also be considered barely phonemic in Mandarin (it is the only consonant that can occur as coda but can never occur initially). And in some dialects, I believe it doesn't contrast with /n/. I wonder how common this is overall.

[Io]

Greek from 756 and 1321 respectively:





Transliterations:

http://farm7.static.flickr.com/6189/606 ... a50a_z.jpg

http://farm7.static.flickr.com/6077/606 ... 4d34_z.jpg

[Io]

>/ts/

We've had a discussion about it before. I was just checking the etymology for 'tsiganin' (Gypsy) and I found out it was αθίγγανος, from α- and θιγγάνω (touch) > untouchables. Originally in Byzantine times it was used to denote members of a Manichaeic sect who moved from Phrygia to Byzantia and somehow they became identified with the Gypsies or something like that. So it's either /θ/ > /ts/ or /θ/ > /s/ > /ts/.

In the etymo dict I decided to have a look at other words starting with τσ, for example there is τσαγγός which comes from ταγγός, but its etymology is unknown/uncertain, which is the case with pretty much every non-foreign τσ- word I've checked, usually the given likely explanations are from ξ, σ, ψ, θ. By the way there is also one word in Bulgarian where ps- has become ts-, pse (dog) > tse.

And if anyone still cares about the question in the OP here is a sample from a letter:

[Exez]

A lot of /ts/ in names seems to occur as a replacement for /l/ ... e.g. Speliotis appears also as Spetsiotis, Karakolios as Karakotsios, and so on.

Although I am not an expert of ancient Greek, etymology or the greek dialects, I would hardly believe that /ts/ is a reflex of /l/. Speliotis (spelia = cave) is a distinct surname from Spetsiotis (Spetses = an island). I am not aware of the other two names but 'kolios' is the name of a fish, while 'kots-' is not related to it at all

It is correct that they are mostly from loanwords, from Turkish to English. It is also correct that the sound did not exist in Classical Greek

At least in a few cases, /ts/ is possibly derived from /tt/ or /t/

Also notice that there is also the sound/diphthong /tz/, and of course there is no letter for it

[Travis B.]

"Would hardly believe" is not enough here.

Just because two forms are presently considered differently does not mean that they are not identical in etymology, or that they do not reflect simply different paths taken phonologically.

In this case, take, for instance, the name of the (Dorian) Greek language Tsakonian (Τσακώνικα)... whose etymology is, surprise surprise, Lakonia (Λακωνία)...

[Exez]

'I would hardly believe' is not based on the analysis of the above words. It is based on my native-level knowledge of Greek, study of ancient Greek and first-hand experience of some Greek dialects. I must note that my knowledge is not scholarly, but if there was an evolution of /l/ -> /ts/ I would have known it. I don't claim omniscience but I have not read, seen (or felt) any such tendency, anywhere.

Lakonia > Tsakonia sounds strange to me and there must have been a special case for this /L/ to evolve to /TS/. According to Wikipedia, the name derives from 'exo Lakonia' (but it is unsourced)

Also, Tsakonian is not identical to what we call 'Greek language'. Even if /l/ -> /ts/ was regular in Tsakonian, I would not base a connection between 'Speliotis' and 'Spetsiotis' on Tsakonian evidence.

[Travis B.]

Okay, for starters, simply knowing a language, or even multiple varieties of it, or even past varieties of it, does not equate knowing its linguistics, especially crossdialectically*. So please, do not bring any sort of personal "intuition" about these matters into things here.

I myself am not that familiar with Greek linguistics myself but this is a specific topic I have seen discussed at various times before - that various Greek varieties have historically shifted /l/ to /ts/ in some fashion, most likely conditional and/or irregular. (The case of Tsakonia from Lakonia was just a non-personal name example that came to mind, and does not say anything about the Tsakonian language itself.)

As for a particular phonetic path from /l/ to /ts/, a change that does seen unlikely intuitively, what I have seen suggested is something like l > ɬ > tɬ > ts, a sequence that when seen in terms of each individual step is far more plausible-sounding.

* E.g. there are many things in English historical linguistics that a present-day person very familiar with multiple distinct English varieties would typically not have the least clue about, and would not be able to intuit a thing about, without a specific background in English historical linguistics, something that a native-level intuition about English simply cannot replace.