Why do some editors not write allative -δε as a suffix?

Greek Homeric Greek and Early Greek Poetry
1

I’m reading through the Iliad, and Perseus (along with several other sources I’ve found) consistently has it as a separate word from the noun, e.g. what is πεδίονδε ‘onto the plain’ in Il. VI.393 (Munro & Allen) is given as πεδίον δέ in Perseus.

When you click this δέ in Perseus, it refers you to the postpositive ‘but’ word (as in, μέν and δέ), which it is decidedly not.

This separate writing is also found in the maybe earlier Munro & Allen edition I found on this site: http://www.mikrosapoplous.gr/en/texts1en.htm

In more fossilized words, Perseus doesn’t separate it, e.g. χαμᾶζε (<χαμᾶσ-δε) ‘to the ground,’ or οἴκαδε ‘homeward’; but, the LSJ entry for οἶκόνδε ‘homeward’ says “better οἶκον δέ,” which boggles my mind.

As far as I can tell here, allative -δε is neither an enclitic particle (=οἶκόν δε) nor a postposition (=οἶκον δέ) but actually just a suffix (like any case ending), though not super-productive, that creates directional (allative) adverbs; therefore I’d assume οἴκονδε makes the most sense. Yet, in Il. I.606, for Perseus’ οἶκον δέ, my Munro & Allen has οἶκόνδε.

What’s the story here? Why am I seeing such discord between editions here? And what actually is -δε, grammatically, from the perspective of the poet?

2

When is a word not a word? When it’s an allative suffix, you say, and who could disagree? But this (–)δε is normally reckoned an enclitic postpositive, and who knows how to treat those? Should we write e.g. ὥσπερ or ὥς περ? It’s something of a false dilemma, I think, given ancient Greek writing practice, which doesn’t distinguish between them. It’s only modern editors (and lexicographers!) who have to decide whether something is a word in its own right or not, and interpose gaps accordingly. Lexical status is more of a sliding scale than an either/or thing.

How did the poet view it? Ηe must I suppose at some level have been aware of it as something tacked on to the end of certain words to indicate direction towards, comparable with –θι and –θεν. But if you asked him what the function of –δε was, I doubt he’d have been able to tell you. He knew how to use it all right, but it was just something he’d inherited in certain locutions. And if you asked him if it was a separate word or not he’d look at you in total bewilderment (provided he paid you any heed at all).

You point to e.g. χαμᾶζε, and e.g. οῖκόνδε does indeed seem on some sort of a par with that (though οικον is a regular accusative). LSJ doesn’t actually say “better οἶκον δέ” (I found that startling, and checked) but reports that for Apollonius Dyscolus. I haven’t looked up the cited passage, but he was an outstanding grammarian. Perhaps (I’m just guessing) he held that the δε wasn’t enclitic and should be treated like any other postpositive. There may be Herodianic scholia too.

But we, unlike Ap.Dysc., know that this –de appears to be allative in Mycenean. Unless there are occasions in Homer where the δε is separated from its noun by something else (which is conceivable in principle, but doesn’t happen?), I don’t see why it shouldn’t be treated as an allative relic in alphabetic Greek.

As for discrepancy among editors, that just reflects the confused situation. It looks as if Perseus follows an edition which followed Ap.Dysc., while the OCT didn’t. Myself, I side with you. Given the orthography of e.g. χαμαζε, inseparably merging ς+δ (if that is what in fact it does; but the analogs do, e.g. Αθήναζε), and its (pre)history as an allative, I don’t see good reason to sever this δε under any circumstances.

I don’t remember what Martin West’s practice is in his Teubner edition of the Iliad (which I don’t have to hand), nor whether he addresses the question in the preface (which is the go-to place for Homeric orthographica). If he does, heed him, not me.

3

Page XXVIII of Martin West’s Praefatio to Homerus Ilias:

Ἄϊδόσδε, δόμονδε, οἶκόνδε, Οὔλυμπόνδε pro singulis vocibus habuerant grammatici (saltem quidam) ante Apollonium Dyscolum: is binas esse contendit, Ἄϊδος δέ, δόμον δέ οἶκον δέ, Οὔλυμπον δέ (Adv. 177.23 sqq., 181.25), quod pius eius filius accepit (ad Π 445, 697; i.498.21 Lentz) quodque in multis bonisque libris servatur (Allen, Prol. 230 sq.). Mihi non persuadet; nam eadem prisca postpositio in vocibus οἴκαδε, ἐνθάδε etc. enclisin passa erat, et Athenienses certe non Ἀθήνας δέ, Μαραθῶνα δέ dicebant sed Ἀθήναζε, Μαραθῶνάδε. Nec video, qui proprium accentum retinere poturerit particula, quae plane obsolevisset vocibusque paucis solis adhuc adhaereret. Sin autem cum Apollonio non vis singulas voces binis accentibus instruere, potes mea pace Bekkerum secutus priorem supprimere, ut legas Ἀϊδόσδε, οἰκόνδε, Οὐλυμπόνδε.

The only Latin that I understand from the above is “Mihi non persuadet,” so if someone will translate, please? I will also correct any transcription errors that are pointed out.

Also, I don’t believe that Chandler covers this, but his chapter on Proclitics and Enclitics is the best introduction to the state of our ignorance on the whole subject. εἰ μὴ ἰατροὶ ἦσαν, οὐδὲν ἂν ἦν τῶν γραμματικῶν μωρότερον.

4

“At least some grammarians who preceded Apollonius Dyscolus had taken Ἄϊδόσδε, δόμονδε, οἶκόνδε, Οὔλυμπόνδε as single words. Apollonius contends they should be in two, Ἄϊδος δέ, δόμον δέ, οἶκον δέ, Οὔλυμπον δέ, and this [practice] is faithfully followed by his son and given by even many good books. I am not convinced as this same ancient postposition is given with enclisis in words like οἴκαδε, ἐνθάδε etc., and Athenians did certainly not say Ἀθήνας δέ, Μαραθῶνα δέ but Ἀθήναζε, Μαραθῶνάδε. I do not, moreover, comprehend who could have retained an accent of their own in these particles as they were clearly archaic, fossilised only in few words. If, however, one does not want to insert two accents to single words in Apollonius, one can despite me follow Bekker and suppress the previous so as to read Ἀϊδόσδε, οἰκόνδε, Οὐλυμπόνδε.”

I hope I understood West correctly.

5

Ap.Dysc.’s son was Herodian, the greatest of the Greek grammarians, whose precepts are normally regarded as authoritative. In referring to him not by name but as “pius eius filius,” “his pious son,” West is sneakily insinuating that he accepted Ap.Dysc.’s position merely out of filial piety.

Nec video, qui proprium accentum retinere poturerit particula, quae plane obsolevisset vocibusque paucis solis adhuc adhaereret.
qui is not “who” but “how.” “And I fail to see how a particle that had obviously fallen out of use, still clinging on to a few words only, could have retained an accent of its own.”

The final sentence: “But if with Apollonius you don’t want to equip single words with double accents, you can … suppress the first …”

poturerit: potuerit.

Chandler is a delight, but Philomena Probert is what we should all be using now. No fun any more.

6

εἰ μὴ ἰατροὶ ἦσαν, οὐδὲν ἂν ἦν τῶν γραμματικῶν μωρότερον.

What’s this supposed to mean exactly? :wink:

7

quodque in multis bonisque libris servatur.
libris refers to medieval manuscripts, not modern books. The Apollonian/Herodianic οἶκον δέ was adopted by the manuscript tradition. (Another reason conservative editors retain it, no doubt. To do otherwise would be to “tamper with the text.”)

εἰ μὴ ἰατροὶ ἦσαν, οὐδὲν ἂν ἦν τῶν γραμματικῶν μωρότερον.
“If there weren’t doctors, nothing would be more stupid than grammarians.” It’s a double dig, against both professions. Doctors head the stupidity list, but grammarians run them a close second.
It’s a saying found in Athenaeus’ Deipnosophists. Chandler will have known of it from Eustathius’ Iliad commentary, and it harmonizes with his own view of the ancient Greek grammarians.

8

Luckily grammar is just my hobby. :stuck_out_tongue:
Εἰ δὲ μή, μωρότατος ἂν ἦν συμπάντων ἀνθρώπων!

9

πλην των ιατρων και φαρμακοποιων! :wink:

10

καὶ μὴν ἐν τῇ χώρᾳ μου οἱ ἰατροὶ καὶ οἱ φάρμακοι ἔκτειναν δἐκα μύριους καθ’ ἔτος. οἱ δὲ γραμματικοὶ μόνον κτείνουσιν ἐλπίδας καὶ εὐχάς.

ἀλλὰ περὶ Παύλου καὶ ΜϜΗ οὐκ εἶπον. ἐκείνοι δοκoῦσι μοι σοφώτατοι εἶναι τῶν γραμματικῶν (καὶ Παῦλος τῶν ἰατρῶν).

11

When I started re-learning Greek, I started with Homer. I got the impression, by comparing with Sanskrit and early Latin, that prehistoric Greek may have an ablative *ϝοικωδ, which usually became the stray form οικω, but elsewhere that ablative case went out of use in pre-alphabetic times. But before a vowel in the next word the δ may have persisted for a while, causing a form interpreted as (ϝ)οικωδ’, which before a consonant may have become (ϝ)οικωδε. That also disappeared, but by then the -δε may have been extracted and added to the dative when used as locative, and to the accusative when used as allative.

I noticed also, the common “from” form οικοθεν, and the common “at” form οικοθι. Compare the common Turkish case forms as in “köyden” and “köyde” = “from (the) village” and “in (the) village” :: Erh??? I thought that in Homeric times the ancestors of the Turks lived 3 months travel by camel away in a back corner of Central Asia and nowhere where they could influence the Greek language at the time.

12

I’m sure someone versed in Turkic linguistics may be able to explain that, but before we jump to the conclusion (for all I know possible) that there are Turkic elements in Ancient Greek, let us just say that a coincidence in sound sometimes is purely that — a coincidence. As way of example, a few weeks back I stumbled about Persian بد [bad], meaning, funnily enough, “bad”. I duly attributed to any of the many Indo-european common roots and went on with my day. However, when I later went to check it, it turned out that, in spite of all of the myriad of such common roots, this particular one is just sheer coincidence and that there is no etymological relation whatsoever between the two (identical) words, that just happened to evolve linguistically towards both the same sound and the same meaning. There are many other such cases between both related and unrelated languages. We shouldn’t rush.

13

I’d like to hear more about its use with locative. My Mycenean sucks, almost as much as my Sanskrit, but in alphabetic Greek isn’t it restricted to accusative? I’m without reference books unfortunately.

14

Hear hear. In Modern Persian we indeed have the word bad, but take the Middle Persian precursor wata. It already looks a little different. Then we have bata in Khotan Saka and bata in Sanskrit. In R̥gveda it is attested once: Yamī, the twin sister of Yama, who is the first mortal being (and thus the god of the dead), would like to have a connexion with his brother, who refuses, whereafter Yamī hauls him over the coals, saying, ‘You bata!’, i.e., “you idler!”, or something to that extent. (We should expect w- in Sanskrit, but anlauting w- and b- got mixed up early in many words, especially in those of vernacular provenance.) In Classical Sanskrit it is used in bewailing: Bata! ‘oh, woe!’, which, lo and behold, is obviously comparable with English bother! [bɒðǝ]… :smiley: English bad is a conundrum.

Take another example. Boy is fiú in Hungarian and fiu in Romanian. Now this is interesting, right? However, Hungarian fiú is a Fenno-Ugrian word, cf. Finnish poika; *p- > f- is common in Hungarian. Romanian fiu, on the other hand, has developed from Latin fī[l]iu[s].

Köyden, by the way, is also Finnish genitive of köysi ‘rope’.

One has to be very careful before jumping into conclusions as regards etymology.

15

I think you have been misguided here. The Greek -δε has—to the best of our knowledge—nothing to do with the Indo-European ablative. In Old Latin we have poplicod (Senatus consultum de Bacchanalibus) and Troiad, in Sanskrit devāt and in Avestan aṣ̌āt̰. (How wonderful it is, by the way, that Martin West even made a critical edition of the gāθās?) In Greek ablative was lost in prehistorical times, but Delphian Ϝοίκω is apparently an old (and very rare?) relic of it. (Original auslauting stops were altogether dropped in Greek.) Interestingly, LS translates it ‘at one’s own expense’, but Chantraine ‘de la maison’ (“from the house”). When Chantraine and LS differ, I tend to go with Chantraine.

Postpositive -δε is deictic (v. Frisk 354, Chantraine 255a, and Beekes 1055), used with accusative, and it actually corresponds to Russian до, German zu and English to (e/o Ablaut). There is wo-i-ko-de in Mycenaean, but it surely marks Ϝοῖκόνδε. In Avestan we interestingly have vaēsmǝnda ‘to the house’ (v. Bartholomæ 1328), which (in my understanding) is assimilated from accusative, *vaēsmǝmda. In Russian, German and English it has obviously developed into a preposition.

The enclitic -δε, found in the demonstrative ὅδε ‘this’, is most likely the same as the postpositive.

16

About my remark above about Greek ‘-θι’ and Turkish '-de- '/ ‘-da’ despite the Turks being so very far from Greece in classical times :: I have found ‘-θι’ elsewhere much nearer home :: see https://en.wikipedia.org/wiki/Etruscan_language#Morphology :: Etruscan had a locative case ending ‘-θi’.

17

Unless Crusaders picked up the word in the Middle East, and returning brought it back to Europe along with other exotica. I have seen it suggested that the heraldry word “gules” = red' was taken from Persian "gol" = rose’.

https://en.wikipedia.org/wiki/Gules