η - how many words or forms could it be?

Post Reply
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

η - how many words or forms could it be?

Post by Stephen Hughes »

How many words or forms could just the single letter eta (η) be?

(with spiritus asper or lenis (or none), with either accent or none, or with the iota subscriptum or without).

What are the rules of syntax that would make one, not the other, probable, if one could only read the anadorned letter itself without diacritics?
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
cwconrad
Posts: 2112
Joined: May 5th, 2011, 5:52 pm
Location: Burnsville, NC 28714
Contact:

Re: η - how many words or forms could it be?

Post by cwconrad »

Stephen Hughes wrote:How many words or forms could just the single letter eta (η) be?

(with spiritus asper or lenis (or none), with either accent or none, or with the iota subscriptum or without).

What are the rules of syntax that would make one, not the other, probable, if one could only read the anadorned letter itself without diacritics?
LSJ offers 6 possibilities, of which I think only three are serious candidates:
ἤ: conjunction
ἡ: article
ἦ , 3 sg. of ἠμί (say) in the formulaic ἦ δ’ ὅς
οὔτοι ἀπ’ ἀρχῆς πάντα θεοὶ θνητοῖς ὑπέδειξαν,
ἀλλὰ χρόνῳ ζητέοντες ἐφευρίσκουσιν ἄμεινον. (Xenophanes, Fragment 16)

Carl W. Conrad
Department of Classics, Washington University (Retired)
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Re: η - how many words or forms could it be?

Post by Stephen Hughes »

There are also;
  • the feminine dative singular of the relative pronoun ὅς too,
    or even
  • the 3rd person singular present subjunctive active of the copula εἶναι.
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
cwconrad
Posts: 2112
Joined: May 5th, 2011, 5:52 pm
Location: Burnsville, NC 28714
Contact:

Re: η - how many words or forms could it be?

Post by cwconrad »

Stephen Hughes wrote:How many words or forms could just the single letter eta (η) be?

(with spiritus asper or lenis (or none), with either accent or none, or with the iota subscriptum or without).

What are the rules of syntax that would make one, not the other, probable, if one could only read the anadorned letter itself without diacritics?
I don't know about "rules of syntax" here, but for a "seasoned" connaisseur of Greek texts, I think that the right answer would be rather Hume's "custom and habit of confident expectation." Rules of syntax, I think, are extrapolations from that sort of ἐμπειρία on its way to becoming a τέχνη.

I was reminded, however, of a curious little section of one of my favorite bits of bedside reading, Chantraine's Morphologie historique du grec -- §365-6 on the suffix η and its role as a key link in the emerging systematic construction of the Greek verb.
οὔτοι ἀπ’ ἀρχῆς πάντα θεοὶ θνητοῖς ὑπέδειξαν,
ἀλλὰ χρόνῳ ζητέοντες ἐφευρίσκουσιν ἄμεινον. (Xenophanes, Fragment 16)

Carl W. Conrad
Department of Classics, Washington University (Retired)
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Extrapolations are at best unverified truths, and at worst l

Post by Stephen Hughes »

cwconrad wrote:I don't know about "rules of syntax" here, but for a "seasoned" connaisseur of Greek texts, I think that the right answer would be rather Hume's "custom and habit of confident expectation." Rules of syntax, I think, are extrapolations from that sort of ἐμπειρία on its way to becoming a τέχνη.
Language is something extra to quantifiable mathematics, I agree. Extrapolations are at best unverifiable truths, and at worst lies hidden in the un-disprovable region - beyond where things can be tested by actual data. Working with language (as opposed to a quantifiable science) involves dealing quiet often with that grey zone.

The question arises from the OCR and the possibility that the metal-brain could anticipate from context what an unrecognisable form could be, most probably was, or even arrive at a confident best guess as to what it is. That involves making a few choices based on the limitations of the language and then elliminating some of them to arrive at the least improbable option. That positive then negative approach to syntax is how I read anyway.

I hope that the machine can be taught to "read", not just recognise characters.
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Post Reply

Return to “Syntax and Grammar”