I’m a Latin language novice learning in isolation and I’ve only just discovered this forum. I’ve been looking at 16th and 17th century school books that focus on conversational latin (at Lilly, Erasmus and Corderius, in particular) and, from them, it would appear not unusual to stress a word’s final syllable, particularly with adverbs at the end of a sentence. These authorities often mark a final syllable with an acute or a grave or a circumflex, and Lilly’s Prosodia is clear about what these accents mean. However, when ‘a’ is marked with a circumflex in a first declension noun ablative case, most people I have read suggest that this is simply to differentiate the case from the nominative and does not indicate syllable stress. There is no implication that the sign is meant to be taken literally as a stress accent representing a rise and fall in pitch, as the Latin grammarians say a circumflex should. However, why not? Am I completely wrong in believing that a final stress on an ablative “a” could make a lot of sense?
From everything I’ve read about Latin accent and quantity - which, admittedly, is hardly everything there is to read - the last syllable only receives the stress in monosyllabic words. That is, when there is more than one syllable, the last syllable NEVER receives the stress.
I think the question has been asked before in this forum about the circumflex accent in older texts. I think it is, as you suggest, to distinguish the singular ablative of the 1st declension from the singular nominative (sometimes a thorny interpretive problem). I highly doubt that it was ever used to mark stress, since stress in Latin is not arbitrary but rather dependent on regular rules. As long as you know the quantity of a word’s vowels, you can determine its accent.
But there are others here with more experience than I in the pronunciation of Latin.
Best luck,
David
PS - Have you enjoyed Lilly and Erasmus so far? I’ve never done much reading in their conversational guides, though I have heard of them. Do you recommend them?
the last syllable NEVER receives the stress.
Well, yes, almost never. Priscian says:
“Ratio namque distinguendi legem accentuum saepe conturbat. Siquis pronuntians dicat poné et ergó, quod apud Latinos in ultima syllaba nisi discretionis causa accentus poni non potest: ex hoc est quod diximus poné et ergó. Ideo poné dicimus ne putetur verbum esse imperativi modi, hoc est póne; ergó ideo dicimus ne putetur coniunctio rationalis, quod est érgo.”
I think it is, as you suggest, to distinguish the singular ablative of the 1st declension from the singular nominative (sometimes a thorny interpretive problem). I highly doubt that it was ever used to mark stress, since stress in Latin is not arbitrary but rather dependent on regular rules. As long as you know the quantity of a word’s vowels, you can determine its accent.
I definitely agree with that. A circumflex accent was (and is) often used to mark a long syllable, especially in the ablative case (first and second declension mainly).
Thanks for the comments. Allen (Vox Latina) notes words such as nostrás, illÃc, adhúc, addúc, tantón stressed on the last syllable (due to loss of a former final vowel --nostrátis, illÃce, adhúce, addúce, tantóne), in syncopated verb endings -át, -Ãt, from -ávit and -Ãvit, and stressed final vowels in words with a subsequent enclitic (-que, -ne). The early-modern grammarians mention these and more. When the matter of accenting in early books is discussed, it is usually said that the practice was unnecessary because the reader already knew how to pronounce the words. I believe this is a wrong interpretation of the purpose of accenting (at least in the early-modern period). I believe it was less to do with word pronounciation that with word stress and de-stress within the sentence, and that it directed the reader to a musical interpretation of the sentence’s rhetorical object, as opposed to its literal meaning. It helps very much if you read aloud while following the accent directions and using these rules: an acute accent represents a pitch rise, a grave a pitch fall, and a circumflex a rise followed by an immediate fall back to the median level (exactly as Lilly and others say they should). This is above and beyond the pitch change required to accent a syllable within a word. Very slight changes (perhaps a tone) suffice to accent a syllable, but the accents in these writings represent the highest and lowest notes of your normal speaking voice–those we use for sentence colour and emphasis. This explains why the accents appear so selectively and where they appear. Read like this and you will hear the text come to life (and this was before mechanical recording devices). So, in answer to David’s question “are the colloquies of Lilly, Erasmus, Comenius, Corderius, etc. worth reading?”, they are indeed, and not just for what these writers say but for the SOUND of them saying it in the original editions. That’s the context within which I’m looking at circumflexed final-‘a’ and, yes, it does disambiguate the word’s sense, but it seems very hard to suddenly change your interpretation of a circumflex when you’ve just got into your stride with the author’s music and rhythm (particularly when that circumflex appears elsewhere in the sentence over other vowels where it can have only a pitch purpose) and where acute and grave accents appear over other final vowels.
I’m afraid I’m a little out of my league here, adrianus. My opinion, pace your excellent observations regarding prosody and intonation, is that the ablative singular of a 1st declension noun will NEVER have the stress. (Or, ought I to say, ALMOST never, Matthaei?) So how do we account for the grave/circumflex accents used in a variety of ways in these neolatin colloquies? Perhaps a useful analogy is the apostrophe in English, which can obviously be used to indicate both possession (“Sid’s”) as well as contraction (“it’s”). Despite the convergence of usage, there is NEVER (well, ALMOST never) difficulty in interpretation. Perhaps Erasmus, Lilly, et. al. also felt there would be no difficulty here.
Best wishes,
David
PS - Matthaei (if this is indeed the correct vocative form for 2nd declension nouns in -eus), thank you very much for the correction. Once again, I look askance at those introductory texts that oversimplify for the sake of easing a beginner into the language. But I do understand the pedagogical purpose, so I’m not too aggrieved.
Lilly says that the normal rules of stress are broken in five instances: (1) differentiation, (2) transposition (3) attraction (4) concision and (5) idiom, and this is repeated by grammarians and Latin teachers throughout the 16th and 17th centuries (my sources are mostly English). Greenwood (1590) adds a 6th: asking questions (which opens up the possiblility of stressing any final syllable at the end of a question --and some evidence suggests this could be european-wide). W.T. (the writer only gives his initials), commenting on Lilly in 1696 and specifically referring to exceptional transposition of accent, mentions ablatives “Some [words] are circumflected in the Ultima for difference sake; as, Ablatives, Poetâ, gloriâ, with these, nostrâs, vestrâs, cujâs, Arpinâs, Ravennâs, &c. Some in the penultima; as, in such syncopated words, Amâsse, decrêsse, Deûm pro deorum.” I’m not hypothesising that every ablative-‘a’ need be stressed but just that in speech it could be (and often was) stressed for the sake of clarity (or differentiation). Still unconvinced? I’m trying hard to convince myself but I like your ‘apostrophe’ argument.
My 17th-century historical study was done without Latin. Better late than never, I wanted to learn it, but using the methods and techniques taught in that period. I thought this would be interesting in tying up loose ends in the “Living Latin” tradition, by starting when immersive Latin teaching was on the way out. Latin changes over time but, unusually, it has been most successful in at least trying to keep to its classical roots. So, in spite of the changes, I think that 16th and 17th century authorities should be listened to, particularly when they talk about Latin, not so much in classical times, but as it was, or was recommended to be, spoken in their own day.
I think that 16th and 17th century authorities should be listened to
Sure, but Diomedes Grammaticus, even if he didn’t live during the “Golden Age”, had the possibility to have a closer look to latin language than anyone in the 17th century, didn’t he? 
not so much in classical times, but as it was, or was recommended to be, spoken in their own day.
Here you can make a choice, and it is completely up to you. 
Talking about the early grammarians (Diomedes and Priscian and Aulus Gellius and others) in The Roman Pronunciation of Latin (www.fullbooks.com/The-Roman-Pronunciation-of-Latin.html), Lord says “In the matter of exceptions to the rule that accent does not fall on the ultimate, we find a somewhat wide divergence of opinion among the grammarians.” Don’t worry too much about it, he says, “For as Quintilian well says: ‘Nam ut color oculorum indicio, sapor palati, odor narium dinoscitur, ita sonus aurium arbitrio subjectus est.’” And, of course, Diomedes (whom Priscian seeks to improve upon by noting some of those exceptions to the rules) can’t be used to evidence how latin was spoken (albeit as a second language) in the early-modern period, which was the point I was making.
Okay, I’m sorry I didn’t get into this discussion for a while; I’ve been terribly busy.
Anyway, the accents you see, Adriane care, such as “optimé,” are not accents, but known as “apices,” singular “apex.” This in fact is an ancient Roman handwriting (and sometimes inscriptural) bit of punctuation, meant to indicate long vowels where a vowel’s inhaerent length is either unclear or hidden.
For example: “c?re Adri?ne, optime amīce.” Compare this with the adverb: “optimē fēcistī!” Without our modern macrons, the adverbial or vocative nature of “optime” is obscured. The Romans would sometimes indicate these long final 'e’s with the apex: “optimé,” meaning not accent or stress, but length. This was a practice sometimes adopted by the revivalists of the Renaissance and also in the Mediaeval aera.
The apex was also placed over a vowel in a syllable whose quantity was already clearly long, yet the vowel’s quantity was long as well, allowing itself to be recalled instantly. For example, we write out “rēx,” while a Roman might have written “réx.” Other examples: “diréctum,” which we write “dīrēctum.” For “?ctit?,” “áctit?” — however, “dux” would receive no apex, since the ‘u’ is short in “dux;” althrough, a Roman might easily have written “noster dux dúxit.” There were also some words where the length was a very powerful factor, or in fact that the length came before a lost final syllable, e.g. “illūc” we write, and a Roman would have written “illúc.” For “adhūc,” “adhúc.” In this case, the apex seems to remind us of stress as well as length.
In short, the ancient Romans would not indicate all long vowels, just the ones where it might have been ambiguous or uncertain. All the normal accent rules of Latin remain intact, despite the various implementations of these ancient mores by Erasmus et aliis.
Salve Laureola. Your observations are very relevant, I think, for the origins of accenting notation, but even in that case I think the grammarians were saying the opposite: that these signs indicated accent (which they distinguish from vowel length), although which symbol was selected, how it was used or where it was placed, these factors were affected by the length of the vowel receiving them. These symbols were not placed to show you the vowel length but to indicate a word stress, where that stress was useful in disambiguating a word. However, I’m not looking at roman handwriting, but at dozens of 16th and 17th century school books and grammar books teaching Latin orthography, syntax, grammar and pronunciation, and they’re very explicit about the use of acutes and graves and circumflexes to signify accent/tone (not breath/spirit or vowel-length/time/macron). And they’re often simply repeating what the early grammarians said about the use of such notation to signify accent or musical tone as opposed to vowel-length or time.
For what it’s worth, this confirms Luci sententiam:
NOTE.–The Romans sometimes marked vowel length by a stroke above the letter (called an apex), as ^ ; and sometimes the vowel was doubled to indicate length
Section #10, Allen & Greenough’s New Latin Grammar
http://community.middlebury.edu/~harris/AG_1.html
And incidentally, Allen & Greenough are also aware of the circumflex: “In such cases [unusual or exceptional syllable lengths] the length of the syllable is indicated in this book by a circumflex on the vowel.”
Section #11 (http://community.middlebury.edu/~harris/AG_1.html)
It seems strange to me that they would use the circumflex in such a manner if they were familiar with its strikingly different, indeed contradictory usage by renaissance luminaries and Latin humanists.
-David
Hi, Davide & Bellum Paxque. Well, Allen & Greenough use an acute (?) to indicate an Apex and talk about other devices such as a raised ‘I’ and vowel-doubling to indicate long vowels, noting that none of these devices came into general use. However, the grammarians are clear about the roles of the accentus/tonus as indicating stress/pitch/modulation (there is a debate about the implications of each of these words but it doesn’t affect my argument) and being of three sorts: acute, grave, circumflex and what each sort means. A&G’s note is misleading. But their use of a circumflex to indicate vowel-length when differentiating homonyms is entirely consistent with 18th-20th-century practice. I can see this practice coming into teaching at the end of the 17th century. I believe that the earlier stress meanings of the accent notations had been forgotten or were initially being passed over as no longer important, precisely because Latin was no longer being taught as a living conversational language. As the 17th-century moved along, the language came to be taught as one for reading rather than for speaking everyday, and prosodic elements relating to the finer points of speech were being left out of school books. I think the primary-source evidence shows this.
Adriane, for what it’s worth, my name is not Laureola; that’s just the name of my website, which draws its name from my girlfriend. Go visit it!
If you’d like to see a modern example of the apex in some recent Latin, go to my blog and scroll all the way down, to the post about bats; I was using it there experimentally.
Luci care, sincerest apologies for the mistake.
I had already been to your website when I was looking at contributions elsewhere on the TextKit site. I bookmarked it because I think it’s a wonderful source that would benefit me greatly, when it comes to reading and learning.
I take your point about the apex in Roman manuscript (as Allen & Greenough and others say) but what’s the evidence that length is meant as opposed to accent or tone? I certainly am interested in reading it. Although, even if it applied in classical times (and was independent of what the early grammarians say about acute, grave and circumflected tones), that wouldn’t bear on the statements of later grammarians (early-modern) about what these accents meant in their period.
I thought I would post this as a typical statement made by the grammarians (this is Lilly’s latin) which is common to all periods about word stress/accent. These are the rules, and graves, acutes and circumflexes are quite clearly meant to represent stress/tone/pitch (and not vowel-length, although the vowel length can affect the choice of tone-- for example, only a long-vowel can support circumflection because of the time it takes to raise the voice and lower it straightway down again).
“Diuiditur autem Prosodia in Tonum, Spiritum, Tempus.
Tonus, est lex vel nota, qua syllaba in dictione eleuatur vel deprimitur.
Est autem tonus triplex Acutus, Grauis, Circunflexus.
Tonus acutus est, virgula recta ascendens in dextram: sic, [´]
Grauis, est virgula recta, descendens in dextram, ad hunc modum [`]
Circunflexus, est quiddam ex vtrisque conflatum, hac figura [^]
…Monosyllaba dictio breuis, aut positione longa, acuitur: vt, Mél, fél, párs, páx. Natura longa circunflectitur: vt, Spês, flôs, sôl, thüs, rûs.
…In dissyllaba dictione, si prior longa fuerit natura, posterior breuis, prior circunfletitur: vt, Lûna, Müsa. In caeteris acuitur: vt, CÃtus, látus, sólers.
…Dictio polysyllaba, si penultimam habet longam, acuit eandem: vt, Libértas, Penátes. Sin breuem habet penultimam, acuit antepenultimam: vt, Dóminus Póntifex.”
The exceptions to the rules are another matter.
Adriane, thanks for the compliment! I hope I can live up to that.
Very interesting commentary from Lilly. When did Lilly write? So far it seems like his whole description of accents is fairly identical to, perhaps immitative of Greek. It is logical, though, that the two languages would be similar in this respect, since they share a good deal in common, particularly scansion (it is extremely remarkable that the Romans were able most fully to adapt Greek versiforms into their own language, something nearly impossible for other languages). I’m curious to know what he says about the grave accent, where it goes, what it means, and whether this all is a natural aspect of Latin, or merely an imitation of Greek and artificial.
He also seems to get “p?x” wrong, which is not merely long by position, or so we know it today; the ‘a’ in “pax” should be long in any position; vide “p?ce.”
Lily (or Lilly), William, 1468?-1522 (I prefer the alternative spelling, ‘Lilly’) is the main man for English-Latin linguistic scholarship. For the next two hundred years, other grammarians are translating, construing, commenting and expanding on his royal grammar. No, it’s not artificial Latin, but Latin as it is taught and spoken in England and (as far as I can tell) throughout Europe. Nor is Lilly changing any rules but compiling the rules of antiquity. And just as the classical grammarians say an acute accent is a raising of pitch, a grave accent is a fall of pitch, and a circumflected accent is a rise followed by a fall of pitch, Lilly repeats it. You will get the same information in Varro, Quintilian, Gellius, Diomedes, Priscian or by reading Lord (The Roman Pronunciation of Latin) or Allen (Vox latina) or Sturtevant (The pronunciation of Greek and latin).
I see from works, prescriptive and otherwise–every primary source you pick up–, that the use of the grave accent (the fall of pitch) is for final syllables mostly of an adverb in the middle of a sentence (at the end of a sentence an adverb’s final syllable has a rising acute pitch) and for any preposition following a substantive it applies to (as in ‘silvam pèr’). In practice, also, I see most writers apply it to single-letter (long) prepositions (à , say), to indicate a natural de-stress (or dropping of pitch). Actually, you only really start to feel why the accenting makes sense (even where it bends the rules) by reading out loud.
By the way, I don’t think ‘páx’ is wrong. It’s a short ‘a’ in the nominative (or better, one of those cases of falling between long and short as the final ‘o’ in ‘amo’ is) so it is acuted. You do not have time when ‘pax’ is spoken normally in the context of a sentence to say ‘pâx’ (it definitely sounds odd), but circumflected ‘pâcem’, however, sounds very right with its long-‘a’ rising and falling. You can still say ‘pácem’ using the long-‘a’ and just rise and hover on the ‘a’, but the grammarians would accuse you of being rather lazy (or too modern). Note the qualification: ‘in the context of a sentence’. Word-stress can be subtly different when you’re learning it within a grammatical system than when it’s applied in speech, and grammarians become progressively bold about noting these distinctions (if only to be able to score on predecessors).