Acute accents on consecutive syllables

[Timothée]

There was discussion on another thread about the possibility of acute accents on consecutive syllables. Are consecutive acutes possible or not? This comes into question with enclitics.

From this question are possibly to be distinguished (or maybe everything is, on a deeper level, connected?) cases such as κάρτά νιν, πάντά μοι, ἄλλός τις, ἔνθά ποτε, and even ἔστί τις (these examples from West 1998:XXXII, Probert 2003:149), where the first acute represents basically a properispomenon, because it works as a diphthong (discussion on Probert 2003:148—150, with a few references). In these words the properispomenon syllable ends mostly in a liquid or nasal (but apparently also sibilant!), and they are of trochaic nature. Even the aorist infinitive γενέσθαί τε is mentioned, an extension to the previous, as it won’t quite scan as a trochee in poetry.

The culprits for this moot question are especially Herodian and Apollonius Dyscolus. Herodian also mentions the case of an enclitic starting with σφ-, resulting in somewhat similar results. Thus we have ἄρά σφιν, ἵνά σφισι and ἔτί σφιν. (Probert 2003:150.) Modern editors tend to disregard this rule (pace West, on Iliad 6,367).

Now we come to the consecutive enclitics. This subsection could be started with ἥδε, a feminine deictic (-δε) pronoun of a demonstrative pronoun which became the article. This might then become ἥδέ γε, as it is actually followed by two enclitics. However, Herodian prefers ἧδε, according to which ἧδέ γε will be perfectly regular. I don’t know much about Herodian’s knowledge, but do wonder if the original enclitic deictic particle was obscured in his mind, and he simply considered it as a single word. (I may well be underestimating Herodian here—there could be more to ἧδε than that.) West notes that some Aeschylus manuscripts even write οὔτέ τι, μήτέ σε, cases quite similar to ἥδε.

What to do then with the cases like ἤ νύ σέ που (Iliad 5,812) and εἴ πού τίς τινα (Thucydides 4,47,3)? (Examples provided by West and Probert.) They certainly seem counterintuitive. It also seems unlikely that the pitch can arise and arise and arise through that series of acutes. Is it most logical to think that every word was started and ended roughly on the same pitch, for instance in that Iliad example, ἤ, νύ and σέ? Or may there be other suprasegmental features (I would think that something suprasegmental is there, but even if it is so that may be non-affecting to the pitch as such)? I admit being on somewhat thin ice here.

Herodian and Apollonius Dyscolus permit the multiaccented enclitic acute series. West mentions (1998:XXXII) the example of expressions like τί λέγεις, which may generally be approached via divergent interrogative τί, but the result is the same. Furthermore, West gives a few examples from Vedic Sanskrit, a language quite close to Greek particularly in its accentuation (also in other features). I only know Classical Sanskrit, the accentuation of which is different from Vedic (and similar to Classical Latin, though it goes one syllable closer to the beginning of the word than in Latin), so I cannot quite judge myself, but it would seem quite valid. West’s examples from R̥gveda are sá devā́ṁ éhá vakṣati (RV 1,1,2) and ná sá svó dákṣo (RV 7,86,6). West remarks, “[Graeci] non minus enuntiare [triplicem acutum] poterant quam Brachmanes antiquissimi.” (‘It was no less possible for the Greeks to pronounce three consecutive acutes than it was to the ancient Brahmans.’)

I would be interested to hear your views and opinions on this issue.

[cb]

Hi, I doubt we'll ever know how these sequences sounded. However, if these successive acutes aren't simply orthographic conventions but transmit actual historical speech patterns, then going off Devine & Stephens, I doubt that a sequence of acutes would go on successively rising in pitch: that's what graves would instead indicate. According to this theory acutes don't mean you lurch up in that syllable, which would be odd-sounding I think, but rather that you've hit the peak and should drop down on the next one (which is usually unaccented!). Devine & Stephens found that multiple pitch peaks (e.g. multiple acutes) in the same phonic clause have different treatments depending on whether the peaks occur in lexical or non-lexical words. For something like ἤ νύ σέ που I assume the first three words would all be at the same approximate pitch level, a pitch plateau. It's speculation though. Cheers, Chad

[mwh]

I’d always followed W.S. Barrett’s appendix on enclitics in his edition of Euripides’ Hippolytus (he'd have prescribed ἤ νυ σέ που, against Herodian), since that seemed to make most phonological sense, all the more so after I read Wackernagel. That may well be what West was responding to, and now that I look at it again some of it does seem less than cogent. Barrett didn't bother to justify the rule against two consecutive acutes. I wonder if Homer is a special exception. I have neither West nor Probert to hand, so I can’t check what they say, but I’d expect them to be right, especially if they’re in agreement. For all accentual matters I’d trust Probert unless I had good reason not to. I used to use Chandler, who not only gives the testimonia but is great fun—he doesn’t hide his disdain for the subject, nor for the ancient grammarians—but thanks to Probert he’s to be consulted only out of antiquarian curiosity now.

The fact that Herodian lays it down that a Homeric series such as ἤ νύ σέ is OK (ακωλυτον) suggests that the issue was controversial, that others prohibited it. Grammarians were always disagreeing with one another.* His own pronunciation will of course have been nothing like Homer’s. I strongly doubt that his take on consecutive enclitics has any real authority, let alone that it tells us anything about delivery in archaic times. Like his father and like all grammarians, his approach was theoretical, and he shows little or no interest in historical praxis.

Offhand I’d have thought that accentuations such as κάρτά νιν, and άρά σφιν likewise, and οὔτέ τι likewise again, merely reflected the fact that νιν and σφιν and τι etc. are enclitic, with no implication beyond that. (Similarly with ἤ νύ σέ που etc.) Scribes or their correctors knew how to accent καρτα, and they knew that νιν was enclitic. Such accentuations are not at all common anyway. (There’s much inconsistency in accentual practice as well as in grammatical doctrine.) On the face of it I don’t see any reason to postulate κάρτά properispomenon (I’m not even sure what that means), and certainly not without postulating the same for άρά. Aren’t they all one and the same phenomenon, and simply represent a different graphic system? But again, I don’t have Probert to consult, so I could be talking through my hat.

I’d separate enclitic series from interrogative τί λέγεις etc., though of course all accentual matters are ultimately connected.
I agree with what Chad says (including his “if”).

* About Homeric readings, needless to say (e.g. at Il.19.384 they argued over ἑοῦ, ἕο, οἱ; two lines later over ὥστε, αὖτε, εὖτε), but also about accentuation, e.g. ταρφειαι at 357: like ταχεῖαι, like ὄρειαι, like πυκιναί. Each had its adherents.

[jeidsath]

I object to this characterization of Probert. I have her book (one of my few books not boxed up at the moment), but it hardly replaces Chandler, unless we're talking about its use for copyediting. Probert is an almost entirely proscriptive guide, and points to Chandler, Wackernagel, etc., at every opportunity for any discussion. Her 2006 book is original research (in Part II), and very interesting, but not any sort of comprehensive guide, like her 2003 book. As the introduction states, Probert 2003 is intended as a workbook (and is very good at what it is), and she specifically says to use Chandler if you want a reference book. This statement of Probert's goes too far, in my opinion; her book works very well as a reference to the existing literature.

But so in this example. On pg. 152 Probert gives the rule of Herodian, says that some modern editors have disagreed, and points to others for all discussion.

--

In Japanese, the particles function very similarly to enclitics in regards to accentuation. An un-accentuated word followed by a particle makes the un-accentuated word take the accent on the final syllable, and pitch falls on the particle. To a Japanese, this sounds as if they are being pronounced as a single word. For sequences of two particles, the Japanese rule certainly follows the rule of Herodian, with に getting the accent. Here is the only sequence of three particles that I can find on the web, and I wish that it were part of a full utterance:

https://forvo.com/word/%E3%81%BE%E3%81% ... %E3%81%AF/

But to me that sounds like Herodian's rule for Greek. This proves nothing about how Greek was accented, but it does make it more plausible that the grammarians were describing a real linguistic phenomenon that they heard in speech.

[mwh]

OK, I’ll restore Chandler to my shelves with great pleasure, or would if I owned the book. He’s a great repository of antiquarian facts. It’s disappointing to hear that Probert doesn’t discuss the issue herself, but no doubt she refers to more up-to-date and linguistically informed treatments than Chandler’s. By “proscriptive” I assume you mean “prescriptive,” and that she doesn't say things like "Don't on any account follow Herodian."

[jeidsath]

My ignorance of Latin shines through once again. I'll have to get that corrected one of these days. Here are Probert's references. I will add direct links to sources as I have time, but this may have to wait a few weeks.

This is the rule prescribed by Apollonius Dyscolus and Herodian and normally followed by modern editors. Some modern scholars have doubted its plausibility and suggested other rules. For discussion, see GOETTLING, pp. 404-6, CHANDLER, pp. 280-2; VENDRYES, pp. 87-90; POSTGATE, p. 74; BALLY, p. 115; BARRETT, HIPPOLYTUS, pp. 426-7; ALLEN, p. 244; WEST, AESCHYLUS, p. xxxii.

[Hylander]

West's views are aptly stated by Timothée. For reference, here they are verbatim:

Quod ad encliticorum series attinet, doctrinam sequor Apollonii Dyscoli et Herodiani, qui omnia praeter ultimam acuebant; v. Vendryes 87-9. Aliter facit Barrett 427, pro axiomate habens, syllabas continuas non acui, nempe quod acutam demissae opponi oporteat ut acuta esse congnoscatur. Sed hoc falsum est, etenim τί λέγεις; dixisse Graecos nemo negat, et ἤ νύ σέ που (Il. 5.812 auctoribus Ap. Dysc. et Herodiano) cum triplici acuto non minus enuntiare poterant quam Brachmanes antiquissimi sá devā́ṁ éhá vakṣati (Rgv. 1.1.2) et ná sá svó dákṣo (Rgv. 7.86.6). Et in Aeschylo codices nonnulli οὔτέ τι, μήτέ σε exhibent.

I would have thought interrogative τί/τίς to be an exception, too, maybe indicative of the inflection of the sentence as a whole, but I don't know much beyond the basics about accentuation (and even less about Vedic); there are plenty of other things about ancient Greek to keep me busy.

[Timothée]

Thank you all very much for your comments. I’m still beyond my depth in many places, but am gradually getting better and taking more in.

It was slightly shorthand for me to call cases like πάντά μοι and ἄλλός τις properispomenon, but the point is that nasal or liquid can work as a second part of a diphthong. Thus πάντα works like οἶκος. In addition may come σ (ἔστί τις in Herodian < ἔστι). Ἄλλός τις, φύλλά τε, πάντά μοι work therefore like οἶκός τις (if I have not misunderstood something).

West mentions τί λέγεις only to show that our normal convention does actually allow consecutive acutes in some cases (well, in the case of τίς, τί). Vedic works as to show that it’s not physiologically impossible (my wording).

What I understand of Vedic accent (it’s not much, but the Indian tradition is on a really strong foundation) is that it was quite similar to that of (ancient) Greek and PIE. Fortson (2010) mentions in his general I-E introduction that “syllables preceding the accented syllable had low tone, and that during the pronunciation of the accented syllable the pitch rose, reaching a peak at its end and at the beginning of the next syllable, after which it fell again”. Whitney discusses the Sanskrit accent on his grammar’s pages 28—34, but I’m yet to peruse this passage. (By the bye, W. Sidney Allen wrote also on Sanskrit phonetics.)

Greek accent does produce strange fascination. Of modern Indo-European languages, Greek and Lithuanian are just about the only ones which (roughly) preserve the original accent (and Lithuanian still has pitch accent). Other I-E languages have all innovated their accentuation system, amongst which Latin (including the Italic phase) twice at least.