I did not look into either context. I referred to a hypothetical case in which one intends to say that someone “erred greatly” by committing just one error.
I don’t know what a Greek would have said if he were referring to a single error specifically. Something like μεγάλην ἁμαρτίαν τινα ἁμαρτάνειν?
I believe they might well say “μεγάλ᾽ ἁμαρτάνειν” even in this case.
Probably, in prose. Although Sophocles has “τὴν ἁμαρτίαν αἰσχρὰν ἁμαρτὼν ἀναλαβεῖν πειράσομαι.” I don’t see how that makes Smyth’s gloss wrong in the Demosthenes context though. Or did you mean something else?
Smyth says it is “substantive”, period. This statement, in its unqualified form, is wrong, as you seem to have agreed. μεγαλα in “μεγάλ᾽ ἁμαρτάνειν”, when referring to a single “error”, is not a substantive.
It’s formally a substantive at the very least. And Smyth does qualify his statement with a “usually.” And I’m still not convinced that your hypothetical would occur in real Greek speech, where the speaker actually meant to emphasize the singleness of the object.
It’s formally a substantive at the very least.
Formally, it is an adjective. I distinguished from the very start grammatical form and semantics (which alone interests me).
No, I did not mean a case when there is a need to emphasize the singleness; I meant a case where singleness is presupposed as known.
Formally, it’s a substantive, appearing next to a transitive verb, etc. A formal adjective would appear next to a substantive it was qualifying.
But if semantics is all that interests you, where is there any semantic question that you’re asking?
I thought by “substantive” you mean a noun. If you mean any part of speech used in place of a noun, then you are not speaking of a pure grammatical form.
where is there any semantic question that you’re asking
There are more than one of them, and you can easily find them all by searching for the lemma “semantic” in this thread.
Formally = mechanically = symbolically. Not “in the form of”. To prove something formally is to deliver a mechanical/symbolic proof, for example.
As I said earlier, I distinguish between grammatical form (not just any old form) and semantics.
How is what you distinguish relevant to the question of what Smyth meant and whether he is correct or not? I don’t understand why you’re bringing it up.
I think I said more than once why Smyth’s statement appears as self-contradictory to me. If you articulate what exactly in what I said specifically about his statement is unclear or seems wrong to you, I would be happy to respond.
OK, let me try to clear some of this up, largely but not wholly ignoring the last 20 or so rat-a-tat posts. (The number seems to be increasing as I write.)
- On the αν, a couple of points. First, as I think you realize, Tugodum, αν has no semantic force of its own, but affects the semantics of the verb. The meaning resides in the verb+αν combo. Second, “its semantic force in this particular sentence”: since λεγειν is infinitive, the Greek itself doesn’t reveal exactly what form it would take if it were finite. λεγ- tells us it has to be either present or imperfect, and αν tells us it has to be indicative or optative. The syntax (δοκῶ with bare infin.) tells us that it should be understood as 1st person singular. All that, combined with the context, indicates that either ελεγον or λεγοιμι would be appropriate, i.e. the meaning could be (roughly) “I think I’d be saying …” (ελεγον αν) or “I think I’d say …” (λέγοιμ’ αν).
I don’t address the following λέγων, on which I already commented.
Here’s an earlier post of mine on conditionals, intended to dispel the fog that tends to surround them:
http://discourse.textkit.com/t/conditionals-a-guide/14120/10
You or others may possibly find it useful.
- On επιεικη λεγειν. You can call επιεικη adverbial in that larger sense (as qualifying a verb) if you want, but it does not function as an adverb, and it doesn’t mean quite what επιεικως λεγειν would (“to speak in a reasonable manner”). επιεικη, just like αληθη in αληθη λεγειν, is an inner aka internal accusative (and likewise neut.pl.). I wouldn’t call it a cognate accusative (unlike e.g. λογον λεγειν) but it has the same construction.
What Smyth 1573 says may confuse (a lot of what Smyth says may confuse!), but cutting out his comparison of μεγαλ’ αμαρτανειν with μεγιστα αμαρτηματα αμαρτανουσι only makes confusion worse. His saying that the adjective in e.g. μεγαλ’ αμαρτανειν is “treated as” a substantive doesn’t mean that it is one, any more than his translating the phrase as “to commit grave errors” does. As to his saying it’s “ordinarily plural” in prose, well, it is: witness αληθη λεγειν, επιεικη λεγειν, μεγαλ’αμαρτανειν, etc. etc., while plain αληθες or επιεικες λεγειν or μεγ’αμαρτανειν would be out of the ordinary. Of course an internal accusative can be singular: λεγειν τι “to say something (something substantial),” αληθες τι λεγειν “to utter a truth,” επιεικες τι λεγειν “to say something reasonable,” αμαρτανειν τι “to commit an error,” ἓν αμαρτανειν “to commit a single error,” while “to commit just one grave error” would likely be μεγαλων αμαρτηματων (μονον) ἓν ἁμαρτανειν.
Tugodum, you say you were taught to translate a construction like μεγαλ’ αμαρτανειν as “to err greatly.” Leaving aside that “err” is not in current use (except in translations from Greek or Latin!), translation can conceal as much as it reveals. In this and comparable cases, it conceals the fact that μεγαλα is not an adverb (as you are well aware) but an inner accusative (as perhaps you are also aware). Smyth’s “to commit grave errors” is better, or “to make serious mistakes." The construction is the same as e.g. πολλὰ αμαρτανειν, to make lots of mistakes, as distinct from πολλαχως αμαρτανειν, to be wrong in all sorts of ways.
Sometimes we have to move beyond what we were taught.
Joel, you say you post here so that you can be corrected. That’s an excellent strategy for your own learning if you are then corrected, but otherwise it risks merely spreading misinformation, something I think we should try hard to avoid. Often I do correct you, as I did in this case, but I can’t correct you all the time. 
Anyway, I hope what I’ve written above will help you understand. You don’t need to look for a “set phrase,” merely look at the construction (bare adj. w/ verb). In the extract you quoted, ..τάχα δόξει λόγον τινὰ ἐπιεικῆ λέγειν· ἐπιεικῆ δὲ εἶπον,…, the first επιεικη of course agrees with λογον, a cognate accusative, while επιεικη λεγειν without λογον τινα would be neut.pl. (Cf. Smyth 1573 part-quoted by Tugodum above.) The follow-up επιεικη I presume simply puts the first in quotes, as it were: “I said ‘reasonable’ …”. But I don’t know the context.
If anyone has any serious questions about any of this, feel free to ask, and I may reply tomorrow or later. But I’d hope that this is adequate elucidation, and might be useful for more than just the two of you.
Thanks, mwh! Very informative as usual. But please let me double check. I took it that:
1.a (this is strictly regarding English) “I’d be saying …” is counterfactual, meaning: “I would be saying …, but am not, in fact, saying …”;
1.b you allow for the possibility of such a counterfactual meaning in 34 d1-3;
2.a “to commit grave errors” renders the meaning of μεγαλ’ αμαρτανειν more accurately than "“to err gravely”;
2.b by analogy, “to say reasonable things” (as was jeidsath’s initial suggestion) renders επιεκη λεγειν more accurately than “to speak reasonably”, and “to say right things” renders αληθη λεγειν more accurately than “to speak rightly”.
Did I get you right? Besides I have two other questions:
3 deStrycker & Slings in their commentary on 24 a7 say “ὅτι ἀληθῆ λέγω, ‘that I am right’, different from ὅτι τἀληθῆ λέγω, ‘that I speak the truth’”, and on 33 b8–“ὅτι ἀληθῆ λέγει: ‘is not right’ (see note on 24 a7).” Am I understanding correctly that they disagree with you, as they insist on adverbial rendering of ἀληθῆ without article?
4 You seem to be saying that accusatives in μεγαλ’ αμαρτανειν, επιεκη λεγειν, αληθη λεγειν are all “internal.” I have to admit, I do not understand why. I am used to thinking that an accusative is internal (in relation to the verb it complements) if it constitutes a pleonasm (as is clearly the case with cognate accusatives); but I see nothing pleonastic in the indicated examples, so would be grateful if you could clarify to me this broader meaning of “internal accusative” for which I’ve been looking in vain in Smyth.
Tugodum, In response to my post, I think the longest I’ve ever written, you focus on “rendering”?!
On your 1a: When someone remarks “In this weather I’d kill for an ice cream” I don’t think of the remark as meaning “I’d kill for an ice cream but am not, in fact, killing for one.” If it somehow helps you to think of it so, go ahead, I’ll not contradict you. But I thought you were interested in semantics. I advise learning something of pragmatics.
On your 3, No.
On your 4, your only substantial question, you need to revise your thinking. Smyth does not do a good job of defining or explaining what he calls an internal object (more often referred to as an inner accusative), but he offers as an example ὁ ανηρ τυπτει πολλας πληγας (1554a, cf. 1555, 1563-77). We can think of that as quasi-cognate (πλησσει πολλας πληγας would be cognate in the strictest sense). All cognate accusatives are internal ones, but not all internal accusatives are cognate ones.
τυπτει με πολλα: με external accusative (what is normally called a direct object), πολλα internal. Similarly ποιεῖ με κακά, etc. etc. No external acc. is needed. τυπτει πολλα he strikes many blows, μεγαλα αμαρτειν to make many mistakes, τουτο αμαρτειν to make this mistake—the accusative is internal to the verb, and there’s no pleonasm involved.
I think that’s enough from me, unless someone else wants to come in.
As regards 1.a, I was just asking (as I explicitly specified) for help with my English (which is not my native language). In particular, it is hard for me to grasp the difference in meaning in English between “I think I’d be saying …” (ελεγον αν) and “I think I’d say …” (λέγοιμ’ αν).
Thanks for your examples as regards 4. Unfortunately, examples by themselves are not sufficient for me to get the concept (and thus to “revise my thinking”). I would be grateful if someone should give–or refer me to–an explicit and clear definition of internal accusative.
I honestly do not understand why you took offense at my #2; at any rate, my apologies for the unintended annoyance.
Also, I do not understand your “No” to my 3. If deStrycker & Slings had taken ἀληθῆ in “ἀληθῆ λέγω” as an internal accusative/object they would not have had grounds to refuse (as they explicitly did) to translate it as a noun (“truth”).
This is getting us nowhere, and could go on for ever. I’m sorry, Tugodum, my efforts at elucidation are at an end. I hope what I’ve written may be of help to someone, if not to you.