Oops. Me paenitet.
And I thought you got mine! Damn this infernal cage of lifeless letters! within which my soul’s pulse is hidden away. 
All’s good, friend, pardon my tongue-in-cheek accusations and poor taste. I wish we could debate these matters vocally — would you like to try Skype some time? I think a lot is getting lost it transcription.
As to what I’ve said in previous debates — let’s just say I’ve grown. If I said Nuntii Latini’s convention was “in error,” then what I will say now is I find their convention annoying and un-Classical. It’s all opinion anyway, based on a variety of evidence, historical and otherwise. But matters of opinion are not all trivial — opinions on war, for example, or justice.
And by the way, your speech’s proximity probably still is closest to Shakespeare among the dialects. 
This is definitely a new Luke (apart from the ‘if’, mind you). However, I liked the old one and, because I agree totally* with your statement, I have to wonder if that will make for good debate. I hope so, because one can grow by slugging it out. (What doesn’t kill you makes you stronger?)
Certè, novus Lucus es (“si” conjunctio divisim, nota). Anteriore autem frui et, cum verbis tuis consentiens, me rogo utrum controversiam meliorem futuram esse. Ita spero, quòd pugnando nos crescere permittamini. (‘Quod non ruit, munit’, dicamus?)
Surely. That would be good. However, the advantage of debating textually about early writings on language is that the practice itself sensitizes one to the problems of expression in these areas, and should encourage a sympathetic attitude towards the sources. You have to think beyond what is written to what is meant by their authors. If you reread, in the context of our debate, Cicero’s comments in De Oratione (Book I, §24) about the foolishness of talking about language, I suspect you will laugh.
Volo. Bonum sit. Si verùm dicam, autem, in scribendo credibile est ut fiamus sensibiliores et affabiliores ad scribendi difficultates de loquendo apud antiquos. Ultrà quod scriptum est, debes sedulò intueri quoque quae res cogitatae sunt. Ut relegas, amabò, apud Ciceronis opus nomine De Oratione (in libro primo, parte vicesimâ quartâ) circà stultitiam quorum de loquellâ expandunt, quod te faciet ridere, ut suspicor, cum controversiam nostram videbis.
[*I edited this to correct this by underlining that I still believe the Nuntii broadcasters speak beautifully and their pronunciation is a great model for clarity in public speaking.]
Spoken like a true Irishman! 
Being of Irish descent myself, I appreciate the notion.
You guys, Adriane and Luce, really know how to confound the issue. 
It would’ve been much simpler without exchanges like “you assume”, “No, you assume that I assume”, etc. But, anyway, I’ve tried to follow both your logics, and this is what I’ve been able to gather, together with a few comentaries and questions. If I misunderstood the position of either of you (or if I forgot to add to your arguments), please let me know.
Thesis: The final -m is a nasalization of the preceding vowel.
Evidence:
-
Elision in poetry and early Roman inscriptions (e.g. scriptust for scriptum est). [But you do concede that elision all the time can lead to confusion and is not preferred in careful enunciation?]
-
Syllables ending in -m count as long, for example, Ītaliam f?t? = Ītaliã f?t?, and diee = diem.
-
Sanskrit also just nasalized the accusative.
-
The nasal -m, because of its weakness, could change to -n or other consonants (including m, followed by a plosive).
Counter-thesis: The final -m is a very weak and quick closed M.
Evidence:
-
Quintilian, who says that it is hardly expressed, not extinguished, but merely obscured, and that it functions as a mark of distinction between the two vowels to prevent them from combining. [Could this also work for nasalization? Or is a nasal m always elided when followed by a vowel?]
-
Pompeius and Servius. Both of them speak of myotacism as the careless pronunciation of the final -m as an initial M, which indicates that the final -m was indeed made with closed lips and one could incur in this abuse if not careful. [However, was myotacism a problem in Classical times or just a barbarism of Late Latin? If the latter, then myotacism could just mean that the barbarians were pronouncing a written final -m and adding it to the following word, nothing more.] Two solutions are proposed [To whom, I wonder. To native Romans or to Barbarians?]: suspensio or exclusio. Both Pompeius and Servius favor suspensio, which means adding a stop after the closed lip -m. [Now, if Quintilian meant a nasal -m in the quote above, could not suspensio also refer to nasalization, since it could also serve as a mark of distinction?]
-
Priscian says that an -m would not suffer a long (by nature) vowel, but nasalizing the last vowel actually makes it long.
Am I correct, up to here? Let me know.
Valete!
P.S.: By the way, Luce, I did spot an apparent contradiction when you tried to deal with myotacism. You call Pompeius’ solution against myotacism a hypercorrection and unlatin. But isn’t myotacism a Barbarism and un-Latin? If so, then how is it a hypercorrection to avoid it?
Amadeus, this is a fabulous synthesis, and is much appreciated! since we badly need to get back on track with this discussion.
I do not. Allen actually notes (in those lovely passages from Vox Latina pp. 78-82 Adrian had me read ) that in the hundreds of elisions/contractions in Vergil, for example, only two are ambiguous, and neither change the meaning of the line. I’m not sure of the state of elision/contraction in Mexican Spanish, but in Italian it is frequent, and necessary unless you want to sound like a barbarian, and I never found myself confused. The only challenge, at first, was recognising individual words, since they seemed to flow so cleanly together; but once I became fluent, this became a non-issue.
Other evidence: older, preclassical inscriptions, in which no final '-m’s are written (meaning that there was never a “loss” of a closed final ‘-m’ — it never exsisted in the first place).
P.S.: By the way, Luce, I did spot an apparent contradiction when you tried to deal with myotacism. You call Pompeius’ solution against myotacism a hypercorrection and unlatin. But isn’t myotacism a Barbarism and un-Latin? If so, then how is it a hypercorrection to avoid it?
I see your confusion. I think this may come from my original misunderstanding of the passage, and my present lack of clarity on what Pompeius really means, since it is not clear. It seems, anyway, that he was advocating it as ‘preferable’ — advocation of the pronunciation of final -m closed rather than a nasalized vowel (or no difference from a pure vowel) is hypercorrective, since it demonstrates fundamental misunderstanding of what that letter means phonetically (i.e. nasalization of the preceding vowel).
Ah! I just remembered something:
Final ‘-n’ in Japanese is quite similar to the final ‘-m’ of Latin, and is written ã‚“ in hiragana, ン in katakana.
Here’s a little on it:
http://en.wikipedia.org/wiki/ã‚“
(The forum’s URL reading technology doesn’t seem to like the hiragana in the address; search for “N kana” in wiki to find the page.)
[Å‹] (before /k/, /É¡/ or /m/)
[m] (before /b/ or /p/)
[n] (before /d/, /n/, or /t/)
[Å©] (between /a/ and /o/ or before /s/)
[Ä©] (between /i/ and /o/)
[É´] (at the end of an utterance)
It even makes the final syllable long! with two morae!
Does Japanese do elision/contration? I only had a year of the language, not enough for such details, I fear, but enough to recall the striking resemblance to Latin phonology!
The [N] seems to be a fascinating sound:
http://en.wikipedia.org/wiki/Uvular_nasal
So between vowels it is merely a nasalized version of the preceding vowel, but sandhi occurs just as in Latin before consonants.
Could we have found a plausible hodiern candidate that could teach us about Classical Latin ‘-m’?
I believe we have.
Salve Amadee,
You assume it would have been! For me it’s more fun because I have to translate into Latin everything I say in my additional one-player game. (Sadly, it’s one player because no ones has time to correct my errors and I wish they would. No-one has tried my ‘divitiae’ enigma in the other thread, either, and that makes me feel silly! Please someone do! 
)
Opinaris graciliùs! Meâ parte, quod mihi oportet omnia in Latino vertere regulis ludi mei quem solus ludo, illae deflectiones cum maximè placent, (Tristis sum ut solus ludo et nemo emendando peccatorum meorum grammaticorum tempus dat. Aliter velim. Nemo etiam qui aenigmam meam de divitiis in filo alio conatus est. Quod me ineptum habere facit! Amabò vos, conemini. )
I’ll respond to your post properly later.
Tuae epistolae continuò meliùs redibo.
Ave, amice.
It is to the credit of Vergil’s art that he only admits elision where there is negligible possibility of misunderstanding. Elision and Hiatus were discussed in Greece as far back as the 5th century BCE, beginning with Isocrates. To say there was no issue about elision and clarity of speech because elisions will almost always be clear to native speakers is a bit dismissive of the historical debate.
Ad gloriam artis Virgilii est ut elisionem non permittit ubi minima sinistra rei interpretatio sit. Tam diù quam Isocrate saeculo quinto ante aevum communem, hiatus elisioque disputabantur. Quod dicas de elisionis usu et comprehensione automaticâ quorum indigenè linguam loquuntur, id contentionem historicam potiùs curtè amandat.
I know you could critique this yourself and present strong counter arguments, Luke, if you were inclined.
Te ipse, Luce, rem analyzare possis, ut scio, et argumentationes persuasibiles adversùm praebere, si propensus sit.
Salve Amadee
You didn’t mention Diomedes and Consentius, who provide evidence for the terminal-M sound, and Velius Longus, who provides evidence against. Longus provides strong evidence for the lack of M and for a nasalized-vowel. Note Longus mentions the grammarian Marcus Verrius Flaccus, tutor to Augustus’s grandsons. His De Orthographia is lost but De Verborum Significatu survives in an abridgement by Festus (itself partial) in the first century CE. Festus paraphrases Flaccus as saying (§50), “Quite a few believed synaloepha (elision) also was involved in such writing [*], as did Verrius Flaccus; that wherever a word finished with the letter M and the next began with a vowel, that not a whole M but only the first part should be written in order to indicate that it should not be pronounced.” As you see, Velius Longus repeats this verbatim.
Diomedis Consentiique mentionem omittis. Argumenta brevia M soni dant. Velii Longi verba adnuunt ut M littera in fine dictionis non sonitur. Ut ostendit haec citatio, verbatim mutatus est Velius a Festo ubi Verrii Flacci sententias dat: “Nonnulli synaloephas quoque observandas circa talem scriptionem existimaverunt, sicut Verrius Flaccus; ut ubicumque prima vox M littera finiretur, sequens a vocali inciperet, M non tota, sed pars illius prior tantum scriberetur, ut appareret exprimi non debere.” Magister erat Flaccus grammaticus nepotibus imperatori augusti.
Quintilian gives further evidence for the sound of word-ending M being made with the mouth-closed:
Quintiliano hoc testimonium M litterae terminantis dat, quod sonum verum ore clauso sonari ostendit:
Also, you have two more bits for the sound of final-M. Martianus’s tongue-twister is especially nice.
Etiam, haec duo testimonia sonandi M terminantis. Praecipuè dictum bonum Martiani nota, qui linguam torquet:
<?xml version="1.0"?>
This is great evidence, Adriane! I shall have to go over it more slowly, but later. I’ve been very busy with work and Greek and Rhetoric, and there’s just not enough time. 
Vale!
Thanks, Luke. I don’t know better than you, but I think that we all could know better than we think we ourselves know!
Gratias tibi, Luci (quia ‘Lucius’ scribis, et non ‘Lucus’). Meliùs quà m te non scio, at nos omnes meliùs scrire possumus quà m quod nos ipsos scire putamus, ut opinor!
You’re right, and yet…
Consider the nature of documentary evidence and the nature of the world and the contexts in which we apply that evidence. In geometry, two single points provide evidence with certainty only for a straight line, but they are still evidence for an infinity variety of other shapes, known only with uncertainty. What I’ve been poking at is ‘certainty’, because when one believes in it, one stops looking hard at things and imagines everything as evidence for one’s views. And because dominant groups influence the historical availability of documentary and artifactual evidence, what may appear evidence of change in time may be the result of power shifts between groups with different accent tendencies coexisting in time. That’s my thesis,-- which, no doubt, is only partly true, if at all.
Rectè dicis, atqui…
Naturam testimonii documentarii cogites, coque mundi contextuumque in quibus acquiruntur. Cum geometriâ, duobus punctis sepositis linea recta solùm scitur ut compertum; incertè autem haec puncta aliarum numerum infinitum figurarum indicant. Certa rei ratio est quam pungebam, quoniam in eâ credendo alia intensè quaerere desines, et omnia vides testimonia rerum quas opinaris. Ultrà , gregibus dominantibus quae testimonia documentalia artefactorumque maneant afficientibus, aliquod argumentum ad tempore mutandum pertinentem videatur, potiùs sit ad effectum mutatione auctoritatis emanentem inter populos, qui accentos varios sed circuminsessiones habeant. Ecce thesis mea,–quae, sinè dubito, parùm vera est, si ullo modo vera.
Salve, Luce!
In response to your argument about elision not being confusing, I must say that in Spanish elision is not very common, or, at least, I can’t find much information about this phenomenon in my language. So that’s probably why I feel a little uncomfortable with systematically (not occasionally) eliding in Latin: I fear I won’t be understood when (if ever) I speak Latine to someone.
Salve, Adriane!
The main reason I didn’t mention Consentius is because he, unlike the others who speak of myotacism, didn’t offer, IMO, a better solution to the problem. He just says to put another word not beginning with a vowel between the final -m and the initial vowel. Well what do we do when we can’t do that?
As for Velius Longus, indeed he does suggest that the final -m is not a true M (m terminat nectamen in enuntiatione apparet", “ut apparet exprimi non debere”), but I was unsure what you were trying to say by quoting him. Anyway, he belongs in Lucus’ column.
So here, again, is a summary of the discussion so far. (Ignoscite mihi si molesta haec methodus est.) The main issue of contention seems to revolve around myotacism.
Thesis: The final -m is a nasalization of the preceding vowel.
Evidence:
- Elision in poetry and early Roman inscriptions (e.g. scriptust for scriptum est).
- Objection: Elision could lead to misunderstanding or to a loss of clear speech, an issue which had been discussed by the Greeks (the latter, according to Adrianus).
-
Syllables ending in -m count as long, for example, Ītaliam f?t? = Ītaliã f?t?, and diee = diem.
-
Sanskrit also just nasalized the accusative.
-
The nasal -m, because of its weakness, could change to -n or other consonants (including m, followed by a plosive).
-
The grammarian Velius Longus says that the final -m should be written half-way to indicate that it should not be
writtenspoken.
Counter-thesis: The final -m is a very weak and quick closed M.
Evidence: -
Quintilian says that it is hardly expressed, not extinguished, but merely obscured, and that it functions as a mark of distinction between the two vowels to prevent them from combining.
- Objection: This could also be evidence for nasalization.
He also says that the final -m is like the mooing of a cow.
- Question: Does a cow actually say “moo” or just “oo”?
- Pompeius, Consentius, Servius, and Martianus, all Late Latin grammarians speak of myotacism as the careless pronunciation of the final -m as an initial M when followed by a vowel (liaison).
- Question: Is myotacism a barbarism or a problem even for native Romans? Was it a problem back in Classical times? Is it a problem at all (and therefore its correction a hypercorrection)?
Objection: Liaison is the pronunciation of the usually silent final consonant of a word when followed by a word beginning with a vowel. So the inference that the final -m is usually pronounced is wrong.
Two solutions are proposed to avoid liaison: suspensio or exclusio. Both Pompeius and Servius favor suspensio to avoid hiatus.
- Objection: Suspension (a stop) could also work for nasalization.
- Priscian says that an -m would not suffer a long (by nature) vowel, but nasalizing the last vowel actually makes it long.
This is what I think, for what it’s worth in all its naivety.
N and M are similar because of their nasal qualities. What distinguishes them is that M uses the lips to end or begin, and N uses the tongue. And if you close your lips also at the end of the N sound it acquires a hybrid M sound (same with M and tongue).
What you have written above assumes M is pronounced N in Latin, except before letters P and B. That’s why you believe it similar to Japanese ã‚“, which sounds like N, except before P and B when it sounds like M (and otherwise like different sounds also in some other places).
Now note that the labials P and B involve closed lips. When you speak in an accelerated but natural way, N before P or B sounds like an M because the lips close at the end of the N sound. This applies in Japanese, Latin, everywhere. Go to the OED and look for words beginning ANP- and you will find that all words once so spelt and pronounced have become AMP- in English. There is only one English ANB- word ‘anbury’ and, with it, ‘ambury’ is noted as a phonetic variant and valid spelling.
Now in Latin, Quintilian talks about the terminal M sound as a ‘mooing’ sound and distinct from the ringing sound of a Greek terminal N (?'that ‘n’ is the velar nasal, the same at the end of the word ‘sing’", as you say?). Flaccus talks about dropping the M sound altogether. and elsewhere you can point to evidence of and N sound for an M, indeed. All are possible practices, so which happened in fact? Could not all have coexisted as dialect variants?
It requires less muscular effort and care, and different muscles, to articulate a word-terminal N sound and to keep the initial vowel of the following word clear and distinct, i.e., un-nasalized, than it does with a word-terminal M sound. The reason is that N terminates with an open mouth, --the starting position of the following vowel. The issue here is that, if the subsequent vowel is nasalized, it will sound like it belongs to the preceding consonant. Consider “-an eb-” and “-am eb-”. To terminate the -N sound and prepare for a clean, un-nasalized vowel to follow on, I must terminate the nasal vibration and drop the tongue tip from the upper teeth. To terminate the -M sound and prepare for a clean, un-nasalized vowel to follow on, I must terminate the nasal vibration and open the mouth from the closed position it assumed. The difference is N tongue movement versus M lip movement. Because the tongue is more agile than the lips, it will get into position quicker for a clean vowel to follow. The lips are slower and so there is more chance that, in smooth speech, the follow-on vowel inherits part of the nasal sound,–it gets slurred with the M and sounds like it belongs to it (myotacism). I also think for these reasons that N requires less energy or effort than M to say. Where M precedes consonants that initialize from an open-lipped position (all except B and P), it will require more effort than an N would. Consequently, given the closeness between M and N, I think that economy of effort and the natural tendency of groups to distinguish themselves in language will contribute to the adoption of an N sound for M in speech in some syllable junctures in some places in some dialects. If you are lucky enough to belong to a popular or dominant group, then you might hope to persuade (or force!) others to adopt your fashion of talking.
[Since I don’t know if what I’m saying makes sense in English, I won’t try this in Latin.
Dum nescio an quod dico anglicè intellegitur, haec verba in latino non tentabo.
Hi Amadeus. I haven’t read your post yet.
Salve Amadee. Tuam epistolam non iam legi.]
Salve Amadee
I know for a fact that cows in Ireland say “MMMMM” nasally, without any vowels! I have never met a cow from Spain, however.
In Hiberniâ, certus sum boves “MMMMM” sinè vocalibus dicere, à naribus expressum! Nunquà m autem obviavi bovi hispanicae.
If “scriptust” for “scriptum est” is evidence for omission of an M sound, then “scriptumst” for “scriptum est” in Terence and elsewhere is evidence of inclusion of an M sound.
Si “scriptust” pro “scriptum est” non sonari ostendat, sequitur “scriptumst” pro “scriptum est” apud Terentium et alios litteram sonari quidem indicat.
You’re forgetting about the evidence they provide also for the sounding of terminal M in all cases other than before a vowel.
Argumenta quae dant oblivisceris, M consonantem terminantem sonari casibus praeter eos qui ad vocales sequentes pertinent.
No. What I assume is that Latin final ‘-m’ is virtually the same sound as Japanese ん. Two ways of writing the same sound (i.e. a nasalization of the preceding vowel). That Roman-letter translitteration marks it as ‘-n’, is immaterial.
That’s why you believe it similar to Japanese ん, which sounds like N,
Except that it doesn’t sound like ‘n’.
Now note that the labials P and B involve closed lips. When you speak in an accelerated but natural way, N before P or B sounds like an M because the lips close at the end of the N sound. This applies in Japanese, Latin, everywhere. Go to the OED and look for words beginning ANP- and you will find that all words once so spelt and pronounced have become AMP- in English. There is only one English ANB- word ‘anbury’ and, with it, ‘ambury’ is noted as a phonetic variant and valid spelling.
Assimilation, sandhi, right, we’ve been over that, yes.
Now in Latin, Quintilian talks about the terminal M sound as a ‘mooing’ sound and distinct from the ringing sound of a Greek terminal N (?'that ‘n’ is the velar nasal, the same at the end of the word ‘sing’", as you say?).
Hwhat? The final ‘-n’ in Greek is identical to English final ‘-n’, IPA /n/, except where sandhi/assimilation applies. Actually the velar nasal is represented in Greek by the first of two gammas, e.g. ἀγγελος.
It requires less muscular effort and care, and different muscles, to articulate a word-terminal N sound and to keep the initial vowel of the following word clear and distinct, i.e., un-nasalized, than it does with a word-terminal M sound.
That’s totally subjective. You could say also that Italians strain more muscles to keep their mouths in perfect open shapes for making their pure vowels — yet they do not regress to comparatively ‘lazy’ English vowel habits. I’m afraid that argument does not follow. One speaks as one has learned.
The reason is that N terminates with an open mouth, --the starting position of the following vowel. The issue here is that, if the subsequent vowel is nasalized, it will sound like it belongs to the preceding consonant. Consider “-an eb-” and “-am eb-”. To terminate the -N sound and prepare for a clean, un-nasalized vowel to follow on, I must terminate the nasal vibration and drop the tongue tip from the upper teeth. To terminate the -M sound and prepare for a clean, un-nasalized vowel to follow on, I must terminate the nasal vibration and open the mouth from the closed position it assumed. The difference is N tongue movement versus M lip movement. Because > the tongue is more agile than the lips> , it will get into position quicker for a clean vowel to follow. The lips are slower and so there is more chance that, in smooth speech, the follow-on vowel inherits part of the nasal sound,–it gets slurred with the M and sounds like it belongs to it (myotacism).
I’m afraid that’s also totally subjective, and not true. It’s a good sounding theory, but it’s fully a matter of opinion, and nothing to do with actual human physionomy. Also, you seem to suggest that the natural principles of liaison, which always happen except in the most halting, super-clearly spoken phrases, do not apply — but they do.
In all a good thought, but maybe I wasn’t clear enough in my initial statement about Japanese ん: that this gramme represents the same sounds as generated by Latin final ‘-m’. (By the way, Japanene ん is very special in that it only occurs at the end of words or syllables; that is, it is always final like Latin final ‘-m.’)
[Since I don’t know if what I’m saying makes sense in English, I won’t try this in Latin.
You’ve earned a break.
So cows don’t open their mouth when mooing?
If “scriptust” for “scriptum est” is evidence for omission of an M sound, then “scriptumst” for “scriptum est” in Terence and elsewhere is evidence of inclusion of an M sound.
One would be supporting evidence for nasalization, the other for sandhi, no?
Of course. The -m is pronounced when followed by the semi-consonants j & v. When followed by consonants, sandhi applies (if I understand that phenomenon correctly), and then it can change into -n and other consonants.
True. Moreover, Adrian, that they leave ‘m’ in the contraction “scriptumst” (as I do when I write “scriptumst,” “factumst”) does not mean that the ‘m’ was pronounced closed, any more than if the word is merely “scriptum.” Moreover, ‘m’ coming before ‘s’ would reduce into nasalizing the preceding vowel in any case, just as ‘n’ does before ‘f’ and ‘s’, making us certain that “scriptumst” would be pronounced scriptÅ©st.
[quote=“Amadeus”]
You’re forgetting about the evidence they provide also for the sounding of terminal M in all cases other than before a vowel.
[/quote]Of course. The -m > is > pronounced when followed by the semi-consonants j & v. When followed by consonants, sandhi applies (if I understand that phenomenon correctly), and then it can change into -n and other consonants.
Yup.
You rightly say this is an opinion, but it’s a potentially testable one. As to its having nothing to do with human physiognomy, I suppose you mean just that it’s untrue. Have you tested this hypothesis or do you just know intuitively that it is false? If the latter, surely your opinion also is just that, --an opinion. You know, of course, that in experiments of this nature involving much variation in subjects for study one only seeks statistical significance, rather than right and wrong answers. Nor am I suggesting anything which would not be discernable in good clear speech in English, for example.
Rectè dicis me opinionem offerre, atquin opinionem quae explorari possit. Quod nullo modò ad physiognomiam attinens, vis dicere thesem falsum esse, ut suspicor. Eamne probavisti vel ità intuitivè scis? Posteriori casu, nonnè tu ipse opinionem enuntias? Ut affirmanter intellegis, experimentis huius generis in quibus res inspectatae significanter variabunt, significantiam statisticam et non veritam absolutam quaeras. Neque suggero res quae loquellâ clare et benè articulatâ anglicè puta non videantur.