Accentuation of third-person pronouns

Greek Homeric Greek and Early Greek Poetry
1

I’m a little confused about the accentuation of third-person pronouns (the forms of ὁ) in Homer (functioning as nouns).

If I’m understanding correctly, then relative pronouns are always accented, ὅ, while demonstratives and adjectives are never accented, ὁ. Also, the emphatic first-person and second-person pronouns have accents, ἐμοί, while the unemphatic ones don’t, μοι.

But when these words are acting as third-person non-relative pronouns, I’m unclear. Grammars seem to list the three genders like this: ὁ, ἡ, τό, with an accent only on the neuter form, and I assume this is how they’re accentuated when used as articles in Attic. But:

τὴν δ᾽ ἐγὼ οὐ λύσω (Iliad 1.29)
I will not release her.

So here we have a third-person, non-relative, feminine pronoun, with an accent. Masculine with no accent:

ὁ γὰρ βασιλῆϊ χολωθεὶς νοῦσον ἀνὰ στρατὸν ὦρσε κακήν (Iliad 1.9)

Masculine with an accent:

ὃ γὰρ ἦλθε θοὰς ἐπὶ νῆας Ἀχαιῶν (Iliad 1.12)

Am I misinterpreting something in these examples as a non-relative pronoun when actually it’s functioning in some other way? In 1.9, I’m interpreting ὁ as a pronoun referring to Apollo, and acting as the subject of ὦρσε. In 1.12, I’m taking ὃ to be a pronoun referring to Chryses, the subject of ἦλθε.

Are there emphatic and unemphatic forms? Given the context and meaning, I don’t see any real reason why the pronoun in 1.12 should be emphatic and the one in 1.9 unemphatic.

2

West (1998) and van Thiel (2d ed. 2010) print ὃ in both ll. 9 and 12. Allen (3rd ed. 1920) prints ὁ in both lines. You can take your pick, but it’s best to be consistent. The text you’re working from seems to be inconsistent.

West addresses this point in his edition of Aeschylus (2d. ed 1998, p. XLIX). He cites the accented Vedic etymological equivalent sá, and Herodian I 471.1 L. (1879 ed. of Lenz and Ludwig?), a 2d c. CE author of treatises on accentuation.

As you’re no doubt aware, the system of diacriticals for accentuation was devised in the Hellenistic era, as an aid to reading aloud for people who were uncertain about certain words, and wasn’t applied universally to ancient texts until the Byzantine era, long after the tone accents of ancient Greek had given way to a stress-based system. It’s generally believed that the accentuation of ancient texts preserved and transmitted in Byzantine era manuscripts is on the whole accurate, but not necessarily in complete detail. There are a number of minor points of dispute among the ancient authorities, and modern editors sometimes follow different practices. Most of these differences are not something to take too seriously, in my view.

3

Chandler doesn’t really discuss this separately from prepositive articles, where he reports that manuscripts do not to accentuate them, and grammarians say to do it. (See 739.)

Galen, in a completely unrelated discussion of the accents of some particular words, says that he is merely reporting tendencies of readers, and gives some advice for what to do when you come across someone insisting on one or the other:

ἐπὶ τοῦτο γὰρ μόνον ἐπειράθην ῥεπόντων τῶν ἀναγινωσκόντων τὸ βιβλίον, ἄν τ’ εἴπῃ τις ἀξιῶν προπερισπᾷν, ὡς ἂν ἐκεῖνος ἐθελήσῃ καὶ σὺ φθέγγου, καὶ πάλιν ἂν ἑτέρῳ συντύχῃς ὀξυτονεῖν ἐθέλοντι, καὶ αὐτὸς οὕτως πρᾶττε καταφρονῶν καὶ τόνων καὶ ὀνομάτων, ὡς οὔτε πρὸς φιλοσοφίαν συντελούντων, πολύ γε μᾶλλον οὔτε πρὸς γεωμετρίαν ἢ ἀριθμητικὴν ἢ μουσικὴν ἢ ἀστρονομικὴν, ὥστε εἰ μηδεμία τέχνη δέεται πρὸς τὸ ἑαυτῆς τέλος τῆς τῶν ἐπιτρίπτων τούτων ὀνομάτων μακρολογίας, οὐ μόνον οὐ χρὴ προσίεσθαι τὸ ἐπιτήδευμα τῶν ἀνδρῶν, ἀλλὰ καὶ καταγελᾷν ὡς μάλιστα.

4

Incidentally, the forms with initial τ- are always accented. You will never see unaccented την. The forms ο, η, οι and αι, when used as articles, are never accented.

5

Yes, the last bit of the rule is exactly what Chandler quotes the grammarians as saying was incorrect. Here was their version of the rule that you’ve just given.

Arcadius: Αἱ εὐθεῖαι καὶ αἰτιατικαὶ τῶν ἄρθρων ὀξύνονται· ὅ, τόν, οἵ, τούς, οὕς, ἥ, ἥν, τήν, τά, ἅ. αἱ γενικαὶ δὲ καὶ δοτικαὶ περισπῶνται· τοῦ, τῷ, τοῖν, ταῖν, τῆς, τῇ, τοῦ, τῷ, τῶν, ταῖς, τοῖς. καὶ τὸ ὦ τῆς κλητικῆς περισπᾶται.

Herodianus: Πᾶν ἄρθρον ὀξύνεται, χωρὶς τῶν γενικῶν καὶ δοτικῶν· αὗται γὰρ περισπῶνται, τοῦ τῷ, τῆς τῇ, τοῖν ταῖν, τῶν, τοῖς ταῖς. καὶ τὰ τούτων δὲ ὑποτακτικά, ἀποβολῇ τοῦ τ γινόμενα, περισπῶνται. ὁμοίως καὶ τὸ κλητικὸν ὦ ἄρθρον μὲν οὐκ ἔστιν, ἀλλ’ ἐπίρρημα, περισπᾶται δὲ ὅμως καὶ ψιλοῦται τῶν ἀπὸ φωνηέντων ἀρχομένων ἄρθρων δασυνομένων «ᾧ ἐχαρισάμην». τὰ δὲ λοιπὰ πάντα ὀξύνονται, ὅ, τόν, οἵ, τούς, οὕς, ἥ, ἥν, τήν, τά, ἅ.

(Notice that Herodianus seems to conceive of his rules for articles as also applying to the ὑποτακτικά, covering them in an aside. No accent difference, just the τ removed.)

Elsewhere, Chandler says that the grammarians didn’t seem to know about proclitics at all. It all comes from manuscript tradition, apparently.

6

Thanks, Bill and Joel, that clears up my confusion. I don’t care about prescriptive rules for their own sake, in a dead language, but as a reader I do find it helpful to be able to look at ὅ and guess that it might be a relative pronoun, whereas ὁ can’t be.

The text of the Iliad I was working from was the wikisource text, which I believe is from Monro and Allen, 1920. Although Monro should be PD now in the US, there doesn’t seem to be a scan on archive.org yet. It would be interesting to know whether this seeming inconsistency was originally present in Monro, and whether there was some reason for it or whether it was just random.

7

This is the 3rd edition:

https://archive.org/details/homeriopera01home

Wikisource appears to diverge from it somewhat.

8

Ha! – I wasn’t finding it in searches because I wasn’t searching in Latin.

Hm, I’m confused now. I had a note that the wikisource text was supposed to have been done from the 1920 Monro. (Archive has the 1912 edition.) But now I can’t find where I got that information, and I’m wondering if it’s not true. I previously noticed some things that seemed to me to be errors in wikitext, such as missing quotation marks at 2.200, but in the 1912 Monro, the quotation marks are there. Monro’s accentuation of ὁ is consistent in lines 1.9 and 1.12. Whoever created the wikitext version presumably scanned and OCR’d some print edition, then corrected all the OCR errors (which would have been a huge project). It’s easy to see how errors could have crept in for stuff like ὅ/ὁ. Wikisource also has modern Greek-style quotation marks rather than Monro’s English-style ones. At line 1.8, wikisource has Τίς γάρ σφωε, but Monro has Τίς τ’ ἄρ σφωε, which seems less like an OCR error than an actual editorial difference.

9

It’s the 1920. See the copyright page, not the metadata. I expect wikisource is using the Perseus text. I was told by the Chicago Perseus director, anyway, that the ταρ in the Chicago Perseus was an editorial change from the OCT text (but not one that followed West perfectly). There is a thread somewhere from a few years back on it all, including github diffs of the different texts.

10

I see, thanks for explaining.

I realized where I got the info. Although the wikisource page for the Iliad as a whole doesn’t have info on the source, each rhapsode’s page does have a Πήγη link, which is where they say it’s the 1920 Monro.

The Perseus version is pretty different, e.g., at 1.9 and 1.26, in ways that are clearly editorial and not just OCR errors.

They seem to have changed τίς γαρ to τίς τ’ ἄρ at some point after the initial uploading of the text to wikisource. Looking at the 2015 and 2017 versions of wikisource, it looks to me like they originally used some other edition of Homer (which had γάρ at 1.8 and consistent accentuation of ὃ as a pronoun), then switched to Monro (which has τ’ ἄρ at 1.8 and consistent ὁ as a pronoun), but some changes fell through the cracks, creating differences at places like 1.12 and 2.200.