A Turkish etymology for both α and σιχτίρ?

Language advisory

In the last obscenity-filled post on this blog, Pierre left a comment on α σιχτίρ "fuck off", which is derived from Turkish:

The Turkish is sıçdırmak ( ﺼﭽﺩﺭﻣﻕ ) with a chim, rather than a kha, and it gets "shit" right back into the context. Actually, it is a causative form and means "to make (someone? / yourself?) shit" and it appears to be imperative. My guess is that "ey sıçdır" is all Turkish and means "Go take a shit."

In fact, there really is a Turkish verb sikmek "to fuck" ("Turkic cognates include Azeri sikmək and Uzbek sikmoq"), and this discussion thread goes through the diffusion of siktir into the Balkans and Armenian.

What I did *not* know from the thread though, is that the interjection siktir in Turkish has a variant hassiktir. And that makes me look at the α in α σιχτίρ, and speculate whether it too is Turkish in origin—and not Ancient Greek, as is normally assumed.

α in α σιχτίρ or ά γαμήσου corresponds to get in get fucked or get lost. It's a hortative particle, and I think what is commonly assumed about its origin is wrong.

Modern Greek has four similar hortative particles.

• άμε is a verb in origin; it's derived from Classical ἄγωμε "let's go", but in Modern Greek is used as "go!", as an exhortation. We know that because in Early Modern Greek, ἄγωμε was used to mean "go!" instead of "let's go": (ἄγωμε μὲ τὴν κάμηλον τὴν μακροσφονδυλάτην "go with the long-neckled camel!", Entertaining Tale of Quadrupeds 768)
  
• άντε has been argued (tortuously) to be a verb in origin as well: ἄγε δή "go indeed!" The etymology makes no sense, and though the Triantaphyllides dictionary vacillates, and the other dictionaries are unrepentant, the obvious derivation is from the Turkish exclamation haydi. (I have 12 pages of an uncompleted paper arguing this; there are straightforward cognates throughout Turkic, and in Russian and Ukrainian via Tatar.) Notwithstanding its origin, άντε does act in some ways as a verb, including picking up a plural ending (άντεστε), and taking subjunctive complements (άντε να δεις "go to see, go and see")
  
• The other two particles are άι ~ άει and α. They are assumed to be related; this is the entry from the Triantaphyllides dictionary on the pair:
  

  α² & άι interj.: before exclamation (cf. άντε); depending on context and intontation expresses (a) indignation, annoyance, dismissal; "go, get": (with an imperative or να + subjunctive with a comparable meaning) α/άι πνίξου / παράτα μας / χάσου / να χαθείς "go drown! / go leave us alone! / get lost!" / || (with σε "to", article and noun) α/άι στην ευχή / στο διάβολο / στο καλό "to the blessing! / to hell! / to the good!". α/άι στη δουλειά σου "to your business": "mind your own business, continue what you are doing". || friendly reproach: ~ να χαθείς! "get lost!" (b) exhortation: Άι στο καλό, παιδί μου, και πρόσεχε "go to the good (farewell), my child, and be careful": go on, go to the good [= farewell]. || wondering about what will happen: α/άι να δούμε πώς θα τα βγάλουμε πέρα "let's see how we get through this". α/άι να δούμε τι θα γίνει "let's see what will happen". [α: truncation of άι· άι: άε < Ancient ἄγε "onwards!" (imperative of ἄγω "go") deleting intervocalic [ɣ] and diphthongised]
  

  
  (It's α², to distinguish it from the interjection "ah!", α¹.)
  

Now, unlike άμε and άντε, άι and α do not act like independent verbs. They are prefixed to imperatives, which means they are not verbs with a dependent verb: they are acting like interjections—or serial verbs. άμε can't precede an imperative; άντε can, because άμε is not an interjection and άντε is. (άι and α do precede the subjunctive as well, but the subjunctive is also used as a gentler command; so it's consistent with άι and α behaving as interjections.)

• άμε/άντε/*άι/*α να δεις ποιος είναι "go to see who it is"
  
• *άμε/άντε/άι/*α δες ποιος είναι "go, see who it is"

They also cannot appear on their own as verbs (or for that matter as interjections): they must always introduce something.

• άμε/άντε/*άι/*α "go on!"

And άι/α has no flexibility with what prepositions it can take: it cannot take από "from", meaning "go past":

• άμε/άντε/*άι/*α από την αγορά "go past the market"

The only prepositional phrase άι/α can take is σε "to" with a definite article:

• άμε/άντε/άει/α στο διάολο "go to the devil" (to hell)

Surely that means άι/α is behaving like a verb here? Well, no. If it is a verb, why the constraint on having a definite article?

• άμε/άντε/*άι/*α σε κανένα μπουντρούμι "go to a dungeon"

Now, it turns out that you can use σε phrases with a definite article on their own, as an oath, a blessing or curse, hoping that someone ends up there. You can't leave the article out if you do that: it won't be the same oath:

• στο καλό!/στο διάολο! "to the good" (farewell), "to the devil"
  
• *σε καλό!/*σε διάολο! "to good" (farewell), "to a devil"

These expressions of course imply "go!", but they have still settled into a template of requiring a definite article, and expressing wishes. άι/α is being prefixed to those established expressions: that does not mean it is acting as a real verb. You can't use άι/α before στο if an oath is *not* involved:

• άμε/άντε/άι/α στο διάολο "go to the devil"
  
• άμε/άντε/??άι/*α στο γιατρό "go to the doctor"

Which suggests that, whatever άντε was originally, it is now (also) a verb; and whatever άι used to be, it is now not behaving as a verb, but just an exclamation, a particle introducing verbs and oaths.

In fact, that's reason enough to suspect άι did not start out as a verb at all, but as an interjection—such as, say, I dunno, the Turkish interjection hay. The thing is though, there are several instances from the Historical Dictionary of Modern Greek (Modern dialect dictionary) of "missing link" particles in dialect, between the verb ἄγε and άι:

• Leucas: άγε, Lesbos: άγι, Sikinos: έγε
  
• Cyprus: άγι̮α
  
• Thessaly, Cephallenia, Cyme, Leucas, Siphnos etc. άε
  
• Maina: χάε

Still, multiple causation does happen, and the very similar exclamation hay! may have encouraged άγε > άι to be restricted to exclamation-like use.

But there's something interesting about the Triantaphyllidis definition. In all its examples but one, it uses άι and α interchangeably. All the examples I've given have been interchangeable as well, although I hesitated over ??άι/*α στο γιατρό "go to the doctor". The one time Triantaphyllidis does not use α but only άι, is in the following pair:

• *άμε/*άντε/άι/α στο καλό "oh, to the good" (euphemism for "oh, to the devil" = "well I'll be! what the deuce!")
  
• άμε/άντε/άι/*α στο καλό "go to the good" (= "farewell")

στο καλό is ambiguous between a literal blessing, and a euphemistic curse. άι can be used both the bless and to curse. α is not used to bless: only to curse. Curses such as can be found at slang.gr:

• α σιχτίρ "get fucked!"
  
• α γαμήσου "get fucked!"
  
• α να χαθείς "get lost!"
  
• α στο διάολο "go to hell!" > ασταδγιάλα ~ ασταδιάλα
  
• α να σε γαμήσω "I'm gonna fuck you!" (not as foreplay, but as one man threatening another—hence the joke reply given there by Vrastaman, "I'd rather we just stay friends".)
  

  There's a further complication that the phrase is actually used as a deferred threat: "A peculiar expression said (usually twice) instead of 'I will fuck you' during stand-offs, but usually with a desire to avoid trouble with someone who is trouble anyway. Like 'count yourself lucky', but leaving everything open, especially if the other talks back." (Halikoutis) (Pritsapirdulas adds in comments: "We also say it (1) when something is broken and we can't fix it; (2) when we react to something startling us.") But a curse it still is.

  
  
• α πάγαινε ~ α πάαινε ~ ρε α πάαινε "get lost", where Standard Greek has retained this dialectal imperative "be going" only in the context of this curse:
  

  ρε α πάαινε: an abbreviated form of the expression "pardon me sir/madam, could you possibly relocate yourselves a smidgeon in the opposite direction from me? I thank you in advance for your understanding." It is used to show beyond doubt that the speaker is not disposed to be serviceable towards his interlocutor, and that the discussion is probably coming to its definitive conclusion right about now:
  

  —Pardon me sir, could you please park a bit further on, so my car can fit in too?
  —Ρε α πάαινε, wanting to park, no less! As if you had a car back in your village, you FUCKING HILLBILLY! (acg)

So α is used consistently in contexts reminiscent of α σιχτίρ "get fucked!" I suspect now that Greeks heard both hassiktir and siktir, reanalysed the former as a sixtir, and related the a back to άι—but only in contexts to do with cursing, like the original hassiktir. This would have been helped along by the existence in Greek of α¹, the interjection "ah!"

• In the expression α γεια σου "ah your health!" ~ α μπράβο "ah bravo" = "that's more like it!", it's not immediately obvious which of the two α is involved; it could be either the pure exclamation α¹ ("aha! that's more like it!"), or the hortative α² (ά μπράβο = άντε μπράβο: "go on! that's more like it!") The restriction of α otherwise to curses makes me suspect the former, but I'm being schematic there.