Talk:Latin phonology and orthography/Archive 2

ARMA VIRVMQUE CANO etc
About Classical Latin orthography: did Classical writers use commas and spaces? I know that dots sometimes separated words on monuments, but how about in books? And what about other punctuation? Does anyone know whether Virgil would have written spaces and commas? Should the example be rewritten as:

ARMAVIRVMQVECANOTROIAEQVIPRIMVSABORIS ITALIAMFATOPROFVGVSLAVINAQVEVENIT LITORAMVLTVMILLEETTERRISIACTATVSETALTO VISVPERVMSAEVAEMEMOREMIVNONISOBIRAM

Ugly as that is to our eyes... garik 17:00, 4 February 2007 (UTC)


 * They used different punctuation marks than we do today, at least sometimes. In public inscriptions, they did not normally use spaces between words. I don't really know much about this, though. FilipeS 17:48, 4 February 2007 (UTC)

Diphthong AE shifted to /e:/ or to /ɛ:/?
Everyone on this page seems to agree that AE shifted at one point in time from /ai/ to /e:/. Today, many people pronounce ai as /ɛ:/. Is this an a-historic attempt to distiguish between ai and long e or is there some historical background to this?


 * That is wrong. AE shifted to /ɛ:/. FilipeS 00:16, 11 March 2007 (UTC)

Thank you! I was seriously puzzled. Follow up Q: The main text states that OE became /e:/, shouldn't it rather be /œ:/ like in modern French or like a lengthend short German "ö"? berndf 09:40, 12 March 2007 (UTC)


 * No, it really was /e:/. The sound /œ:/ is quite unusual in the Romance languages (though it does exist in French). FilipeS 18:30, 14 March 2007 (UTC)


 * And French /œ/ (sans length mark actually) comes from a fronting of not a rounding of  or .  Hence fleur/floral.   Æµ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 02:36, 15 March 2007 (UTC)


 * Interesting, I never realized French was the only Romance language which had the /œ/. @Aeusoes1: Yes, modern French does not distinguish between long and short vowels, so the length mark is unnecessary. Of course, /œ/ and /ø/ are considered variations of /ɔ/ and /o/, respectively in many languages which have these sounds (hence, e.g., German letter ö which is historically actually a stylized oe or the ø in Scandinavian languages). Thank you for the feed-back. Berndf 11:36, 16 March 2007 (UTC)

The Classic Latin pronunciation
The Classic Latin pronunciation is the same as the Modern Latvian pronunciation. Nowadays Latvian sound system is practically a copy of Latin sound system or vice versa: Roberts7 14:33, 28 March 2007 (UTC)


 * Interesting hypothesis, but not supported by any author I've ever read on the subject. Most scholars I've read also agree that Classical c was never given a soft value. (see Tore Janson A Natural History of Latin, p108 "...the letter c, which in Latin in the classical period always represented a k-sound.")  In addition, Latvian v does not match the usual pronunciation given for consonantal u in Classical Latin (see Janson, p.114).  These two points alone show that your statement is incorrect and must be discarded. --EncycloPetey 00:23, 3 April 2007 (UTC)


 * Speaking about cē. Yes, in Classical Latin it was kē, but since 500-600 AD it was tsē. The same was in Latvian. Modern Latvian c [ts] comes from Proto-Latvian k. Consonantal u was in Proto-Latvian, e.g quan 'what' (Modern Latvian ko [kuo], sometimes pronounced as [kwa]], if you meant u in qu, not in gustibus, where is normal [u]. Also Latvian lauva 'lion' (in Latin transcription it would be laua) is pronounced as [lauwa], so v pronounced as [w] exists even in modern Latvian. But I don't agree that vita could be pronounced as [wīta]. Only KV (qu) and GV (gu) was [kw] and [gw], all other V was true [u], [ū] or [v] (as in Italian). Roberts7 03:01, 5 April 2007 (UTC)


 * While it is fine to disagree with the published work of Latin experts, you should provide firm evidence to underpin your reasons for disagreeing. Otherwise, you are simply blowing smoke. --EncycloPetey 13:58, 5 April 2007 (UTC)

Z and so on
I have few remarks about the article:

1. '''/z/ was not a native Latin phoneme. The letter Z was used in Greek loanwords to represent Zeta (Ζζ), which is thought to have denoted /z/ by the time the letter was introduced into Latin. Some authorities have maintained that Latin Z may have represented /dz/ but there is no clear evidence for this.'''


 * It may be worth adding, that "z" is a double consonant for poetry (it creates position), and also ancient grammarians maintain that it is a double consonant, and mention it as such beside "x" (Priscian: "sunt etiam in consonantibus longae, ut puta duplices x et z; sicut enim longae uocales, sic hae quoque longam faciunt syllabam"). So, contrary to what article says, there is some evidence that this letter did not represent simple /z/, but rather it was a consonant cluster, like x (nobody says, IIRC, what are the consonants in the cluster, but "dz" is the most obvious choice IMHO).


 * Sidney Allen deals with the pronunciation of the greek Zeta in his Vox Graeca and Vox Latina: according to him, the early value of Z was [dz] or [zd] (through metathesis). However, this changed to [z] already during the 4th century BC, (i.e. very early from a Latin perspective.) "Such a [z] would presumably have arisen from an earlier [dz], and after short vowels at least the original quantitative pattern is likely to have been preserved by gemination, i.e. [zz]". (Vox Graeca, p. 58.). This is also the pronunciation he proposes for intervocalic Z in Latin.


 * On your following points I quite agree. Alatius 13:26, 14 April 2007 (UTC)

2. Latin had no aspirated consonants and so these digraphs tended to be pronounced like F, T, and C/K (except by the most careful speakers).


 * "PH" was also written (and so pronounced too, at first place) as simple "P", hence spellings like "Bosporus", "purpura" (from "porphyra") and such.

3. The diphthongs of AE and OE generally became monophthongs, /ɛː/ and /eː/ respectively, after the period of the Roman Republic.


 * This is true, but very far from being precise: "after the period of the Roman Republic" may equally mean 20 BC or 12th century AD. This should be clarified IMHO, that the first signs of the monophthongization, limited to some rural areas (and noted as a curiosum by Roman scholars of the period) are attested in the 1st century BC, and the process did not complete before the end of the 3rd century AD (Grandgent: "it may be called regular by the fifth [!] century").

4. Q clarified minimal pairs between /k/ and /kʷ/, making it possible to distinguish between disyllabic cui /ˈkui/ and monosyllabic qui /kʷiː/.


 * I think both words - cui and qui - are monosyllabic. The difference is, that cui has vocalic "u" and semivocalic "i" (so "ui" effectively forms a diphthong here), whereas "qui" consists of the consonant "qu" and a vowel "i".

82.210.159.30 12:12, 13 April 2007 (UTC)

Prounciation of H and S in the US
As an American who has lived in Rome, I have long been aware that the US pronunciation of "Church Latin" (as we put it) is roughly the same as Italian.

However, I was surprised to see here the following (which has been copied to other sites as authoritative):

"H is silent except in two words: mihi and nihil, where it is pronounced as [k]

S may represent a voiced [z] between vowels. "

In the US (and Canada?), "Church Latin" is taught with the "H" pronounced as in English, not in Italian (except where it hardens the preceding C in Greek words). Thus, "habet" is pronounced not too dissimilarly to Tolkien's "hobbit" ;-). Furthermore, we never pronounced "mihi" as "miki", although we were aware that Classical Latin sometimes did. And, honestly, I never heard of "nikil" - even my Classically-trained Ph.D. instructor didn't say that.

In terms of the "S", while Italian voices an intervocalic -s-, the usual way that Church Latin is taught in American English does not voice this "s". Thus, we would pronounce "casa" (assuming it were a Latin word - work with me on this) as the Italians pronounce "cassa", not "casa".

I see that there is discussion that Ecclesiastical Latin varies in pronunciation from area to area (quite true). But given that there appears to be more interest in the resurgence of Church Latin in the US than in Europe (see current news reports on the possible return of the Tridentine Mass by Pope Benedict XVI, which is being given more play in the US than Europe - e.g. AP news story), would it not make sense to make this notation as an example of a localization of Church Latin?

William J. &#39;Bill&#39; McCalpin 03:14, 27 May 2007 (UTC)
 * The recommended pronunciation of Church Latin is that indicated in Liber Usualis, pages xxxvj-xxxix. Lima (talk) 08:03, 20 July 2008 (UTC)

short vowels
What the *&^%$ is going on? The vowel table shows for short vowels, but in all the transcriptions actually used on the page /u, i, e/ are used. Which is it? Can we please have some references cited so we can trust where this information is coming from. Widsith 09:12, 27 June 2007 (UTC)


 * The notation in the article is in need of being standardized. See above, and please moderate your language. . FilipeS 14:07, 27 June 2007 (UTC)


 * All right, I made a reversion to the earlier vowel chart but with a note on possible phonetic attributes. In the scope of phonemic / slashes /, in addition to being less controversial, it is more consistent with the literature on Latin to have the same symbol for both the long and the short vowel variants.  I also put a cite check on the attributes so that Encyclopetey or whoever can put a source on that.  I also changed the diphthongs back but if anyone would like to put a note on the phonetic attributes (with sourcing, of course) then keep the /slashes/ vs [ brackets ] distinction in mind.  Thank you. Æµ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 18:23, 26 July 2007 (UTC)

I still can't find any authors who clearly state an opinion on the pronunciation of Classical Latin vowels. I have finally tracked down information pertaining to Vulgar Latin pronunciation, from L. R. Palmer, 1987. The Latin Language. He indicates that a change was happening in Latin in which quantity was disappearing from the spoken langauge and stress changed, but I can't find anything that clearly states what changes may or may not have occurred in quality of the vowels. Here is Palmer's table from p. 156: It looks from the text as though should appear for short-i; possibly it was omitted from the table through a typesetting error. Palmer, unfortunately, does not address the issue of borrowed Greek upsilon. --EncycloPetey 19:51, 3 September 2007 (UTC)

AE & OE again
Why is it that in most of the Germanic languages and Hungarian these Latin diphthongs have become very different vowels (~ and ~ respectively) rather than ?Cameron Nedland 23:03, 2 September 2007 (UTC)


 * Germanic languages and Hungarian don't come from Latin.  is often used as an orthographic indicator that the preceding vowel is front.  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  23:09, 2 September 2007 (UTC)


 * Well, in Latin loanwords, the old diphthongs become the unusual vowels rather that /e/. Does anyone know why?Cameron Nedland 00:09, 4 September 2007 (UTC)


 * Are you sure they're loanwords? Give a few examples.  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  07:00, 4 September 2007 (UTC)


 * Ecology in German is Ökologie, from Latin oecologia, from Greek oikologia. Most international words like this follow the same pattern.  —Preceding unsigned comment added by Cameron Nedland (talk • contribs) 16:39, 4 September 2007 (UTC)


 * Hmm... I can make guesses about it but I'm really not sure. Perhaps you should go to the help desk.  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  17:25, 4 September 2007 (UTC)


 * This sounds like a question for one of the Wiktionaries. --EncycloPetey 19:09, 4 September 2007 (UTC)

If some Germanic languages pronounce those digraphs as rounded vowels (English does not), it's due to a spelling pronunciation. FilipeS 16:36, 6 September 2007 (UTC)

/kw/ v.s. /kʷ/
Were there actually any minimal pairs between these two?Cameron Nedland 21:00, 26 October 2007 (UTC)


 * There's one in the article. Cui and qui (or CVI and QVI).  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  21:03, 26 October 2007 (UTC)


 * I thought cui was either /ku.i/ or /kui/ (supposed to be a diphthong for the 2nd one).Cameron Nedland 21:48, 26 October 2007 (UTC)


 * As far as I understand, cui is a monosyllabic word. The article says it is a falling diphthong but I'm not sure if that's true (it might be a mistake).  I don't know Latin so I can't come up with any other minimal pairs, but if you're looking for evidence for  being more than just some sort of realization of /k/ plus /w/ consider the word colloquium which can be transcribed .  If qu were underlyingly a /k/+/w/ this would be  which doesn't fit with the phonotactics of Latin or any language I've encountered.  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  22:56, 26 October 2007 (UTC)


 * Cui does indeed probably contain a falling diphtong -- in any case it does not start with (as is evident from its use in poetry: it does not lengthen the final syllable of a preceding word ending in a short vowel). As for your transcriptions of colloquium I'm sceptical about, because the Latin  does not normally occur before a consonant; that is to say, the normal pronunciation would rather be.
 * To adress the original question, I'm not aware of any occurances of inside a Latin word (though I could of course very well be mistaken). We could however combine two words and form the contrasting pair Aquinum and ac vinum. A pronunciation with  inside one word is theoretically possible through synizesis in words like docuit . I have however been unable to find an example from classical Latin poetry where such a pronunciation is demanded. Alatius 02:20, 27 October 2007 (UTC)


 * Alrighty, thanks guys.Cameron Nedland 17:59, 27 October 2007 (UTC)


 * Well, I don't speak Latin either, but I'm pretty sure, from how the word is pronounced in Romance languages, that cui would have been a (falling) diphthong . Thus, qui could well have been pronounced as a rising diphthong, that is, as I believe it still is in Italian. Both are monosyllables; I see no need to appeal to a labialized . FilipeS 18:40, 14 November 2007 (UTC)
 * W. Sidney Allen's Vox Latina, which I believe is still the Bible on Latin pronunciation (classicists may like to correct me) does treat it as a labalised [k]. However, I can't remember his argument for why this should be. garik 18:45, 14 November 2007 (UTC)

Short I and short U again
I've been having a change of heart. If there are no objections, I will edit the article to make /iː/ and /uː/ the transcription of the long vowels, and /ɪ/ and /ʊ/ that of the short vowels of classical Latin. My reasons are as follows:
 * I started to think that, if I had to explain to anyone how to pronounce a short I and a short U in classical Latin, I would certainly use the English vowels. They may not be exactly the same as those of Latin, but I can't think of a better a starting point.
 * Phonetically, there is some "wiggle room" on what IPA symbols can represent. Although cardinal /ɪ/ and /ʊ/ are near-front / near-back, these symbols are sometimes used for front/back vowels. And this is a simpler notation than /i/ and /u/ with a diacritic.
 * It's what the sources we've looked into say. Although I don't always find their arguments convincing, we must follow the sources.
 * Because it's the notation used in several well-known books about Latin written in English, it's the one that the readers are likely to be familiar with, and that future editors are most likely to use. We need to make an effort to make the phonetic notion uniform throughout the article, which it currently isn't. FilipeS (talk) 13:15, 10 December 2007 (UTC)


 * As I've said before, I oppose and  (phonemic representation) but I don't oppose  and  (phonetic representation).  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  18:29, 10 December 2007 (UTC)


 * But they were different phonemes! Why do you oppose it? FilipeS (talk) 19:29, 10 December 2007 (UTC)


 * From my experience, it is more consistent with the literature on Latin to have the same symbol for both the long and the short vowel variants. Transcribing them as  and  shouldn't give people the impression that they're the same phoneme. Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  19:55, 10 December 2007 (UTC)


 * If you transcribe them as /i:/ and /ɪ/, people are not likely to think it's the same phoneme... On the other hand, if we're going to use what is traditional in the literature on Latin, then why not just write i/ĭ and ī? FilipeS (talk) 21:23, 10 December 2007 (UTC)
 * Yes, so neither of what we are proposing will create such confusion. i/ĭ and ī are not IPA.  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  23:38, 10 December 2007 (UTC)
 * Well, does it have to be IPA? (That's not a rhetorical question; I'm seriously asking.) IPA should of course be used for the phonetic transcriptions, but a phoneme is by definition an abstract entity, and can be denoted by any symbol, as long as it is obvious what is meant, which is certainly the case with /ī/ etc. In view of that, I don't really have any strong opinion on how to describe the phonemes. (It could be said, though, that /ɪ/ conveys more information than is necessary: /ɪ/ is not a different phoneme from /iː/ by virtue of the quality, but due to the length.) As for the phonetics, I'm all for [ɪ] and [ʊ].Alatius (talk) 10:14, 11 December 2007 (UTC)


 * What makes you say that "/ɪ/ is not a different phoneme from /iː/ by virtue of the quality, but due to the length"? There are many languages where the quality of long vowels is just as important in distinguishing them from their short counterparts as their quantity, if not more. English, for instance. FilipeS (talk) 11:39, 11 December 2007 (UTC)


 * Yes, I see what you mean. I was mainly thinking of this (which I have quoted earlier): "The Latin short i also may well have had a closer quality (more like that of the long ī) before vowels, to judge from the Romance development of Latin dies..." (Allen, Vox Latina p. 51 ff). I.e., Allen thinks that it is possible that /ɪ/ (or "/ĭ/" or "/i/") had the allophone [i] in some positions. But in other positions, it is of course possible that a Roman would have heard [i] as /i:/, though I don't know if anyone can know that for sure. Anyhow, as I said, it doesn't matter to me; if you write /ɪ/, I'm not going to change it. Alatius (talk) 12:11, 11 December 2007 (UTC)

It looks like there's no consensus for a change of transcription. Looking at the article again, it's in pretty good shape. Unfortunately, related articles like Vulgar Latin are full of inconsistencies of transcription. And don't get me started on whether to write Latin words in italics or all-caps, and to name verbs with the infinitive or the 1st. person present tense! Oh well, no rest for the wicked... ;-) FilipeS (talk) 00:01, 15 December 2007 (UTC)


 * Don't feel isolated. We've been holding off on adding pronunciations to Latin words at Wiktionary until some consensus can be reached here to use as a "standard". --EncycloPetey (talk) 01:02, 15 December 2007 (UTC)


 * But no one is objecting, I believe, to the phonetic representations [ɪ] and [ʊ]. So the change should be made whenever the actuall pronunciation is discussed. Specifically, I'm thinking of the example of classical pronunciation at the end (Arma virumque), which currently says [e] etc. Alatius (talk) 11:06, 15 December 2007 (UTC)

Aeusoes has objected, and not without reason. I myself have some misguivings about these transcriptions, as I expressed above. And the truth is that our knowledge of how Latin was pronounced in classical times comes from secondary sources. We'll never be able to find out exactly how short "i" and "u" were pronounced. A broader transcription does circumvent the problem, by being less assertive. FilipeS (talk) 15:05, 15 December 2007 (UTC)


 * You are confusing me. When I first tried to change the transcription to [ɪ] you were the main opponent to that. Then you yourself proposed [ɪ], only to now have changed your mind again? Aeusoes explicitly said "...but I don't oppose and  (phonetic representation)." I think it doesn't hurt to provide a rather narrow transcription (as narrow as is reasonably attested); but of course, we could add a caveat about the transcription being hypothetical. Alatius (talk) 16:56, 15 December 2007 (UTC)

It's no wonder I'm confusing you. I have ambivalent feelings about this! :-) However, right now, after hearing from the three of you, rereading the article, and thinking more about the matter, I feel more inclined to leave the article as it is.

(Please note, though, that Aeusoeas also wrote "I oppose and  (phonemic representation)".) FilipeS (talk) 17:03, 15 December 2007 (UTC)

Long consonants
The article makes it a point of stressing that long and short vowels were different phonemes in classical Latin. Fine, but what about long and short consonants? They were different phonemes, too... FilipeS (talk) 13:17, 10 December 2007 (UTC)
 * What I've read of such consonants treats them as doubled, and from what I understand this occurs only when the written consonant is doubled. Is that what you mean?  Does that match what you're read as well? --EncycloPetey (talk) 14:47, 10 December 2007 (UTC)

Phonetically, a double consonant is the same as a long consonant, as far as I know. FilipeS (talk) 14:57, 10 December 2007 (UTC)

Y, etc. in loanwords
"was used in Greek loanwords with upsilon (ϒυ ). Latin originally had no close front rounded vowel, and speakers tended to pronounce such loanwords as (in archaic Latin) or  (in classical and late Latin) if they were unable to produce." This may give the wrong impression that Y was used in archaic Latin, which it wasn't. It should be made clearer that back then the Romans simply nativized Greek Y into Latin V (U), likely in both writing and pronunciation. It was only later, in the classical period, that it became fashionable among Roman intellectuals to write Greek words "unchanged", that is by transliterating Z, Y, X, Φ and Ρ as Z, Y, CH, PH, RH instead of adapting them to Latin phonology.

I think the article should also mention/discuss to what extent the Romans really bothered with imitating the Greek pronunciation. I find it likely, for instance, that only the elites seriously tried to pronounce Y as /y/, and that most speakers simply wrote Y and read /i/. FilipeS (talk) 23:57, 14 December 2007 (UTC)

Double glides
(1) appears at the beginning of words before a vowel, or in the middle of the words between two vowels; in the latter case the sound is doubled: iūs, cuius. Since such a doubled consonant in the middle of a word makes the preceding syllable heavy, the vowel in that syllable is traditionally marked with a macron in dictionaries, although in fact the vowel is usually short. Two questions:
 * 1) Was this /j/ not actually an allophone of the vowel /i/? Are there minimal pairs for the two?
 * 2) Wouldn't the transcription [kuijus] (or ) be more accurate? Please see the discussion at semivowel and non-syllabic vowel. FilipeS (talk) 17:23, 24 December 2007 (UTC)


 * I'm pretty sure that Latin glides were not allophones of high vowels. Because they appear in different parts of the syllable, it might be impossible to find a minimal pair.  A better question would be, can we find an instance of [i] becoming [j] or vice versa with the addition of a suffix?  That will show allophony Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  19:38, 24 December 2007 (UTC)

If they systematically appear in different parts of the syllable, isn't that a good indication that they may be allophones? FilipeS (talk) 19:52, 24 December 2007 (UTC)


 * Not by itself. English [p] and [u] appear in different parts of the syllable but there's no question that they're different phonemes.  Likewise, people rarely peg [w] as an allophone of /u/ in English.  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  21:12, 24 December 2007 (UTC)

Nevertheless, Latin speakers used the same letter for both phones. Clearly, telling the two apart was regarded as unimportant by them. FilipeS (talk) 21:29, 24 December 2007 (UTC)


 * Latin also used the same letters for long and short vowels but we know that they were two different phonemes. Either telling the two apart was unimportant or they figured any speaker of Latin would know which I was syllabic and which was not.  Really, minimal pairs are usually fine but the best way to tell if two sounds are allophones of one phoneme is if they replace each other upon suffixation.  This is how we know  and  are allophones of  in English (ride + -er → rider) and that  and  are allophones of /i/ in Russian (|| 'one' + || 'adjectival suffix' →  'first'; || 'three' + || →  'third').  Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  22:45, 24 December 2007 (UTC)

Long vowels were marked in many Latin inscriptions. See the article Apex (diacritic). Also, ancient grammarians made reference to them. Did they ever say that the letter I could represent a third sound, besides the short and the long vowel? FilipeS (talk) 16:43, 26 December 2007 (UTC)


 * I don' know what the ancient grammarians said. If I recall correctly, though, Latin /w/ came from PIE


 * gʷ(ʰ) so it's very unlikely that [w] was an allophone of long or short /u/. Indo-European sound laws Do you know if ancient grammarians ever said V could represent a third sound other than the long and short vowels? Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  19:58, 26 December 2007 (UTC)
 * You think I should ask the language help desk for help on this? Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi] 19:58, 26 December 2007 (UTC)

I don't know what's the standard procedure, but perhaps we should just wait and see whether someone else comes to weigh in on the matter. It is a very minor point. The article seems to be quite decent, regardless. FilipeS (talk) 18:04, 27 December 2007 (UTC)

POV or possible anti-Catholic
I have looked this page over and it does not say anything overtly offensive but I was very suprised to see that you have made the claim that the Reconstructed pronunciation is the one that the Romans actually used and then you made the comment that teachers are not teaching this pronunciation. The implication here is that only the way the Romans spoke before Christianity is REALLY Latin, that the Reconstructed pronunciation is this CORRECT pronunciation and that any teacher that does not teach this is wrong. I must dispute this. I know of only one reputable scholar that thinks this Reconstructed pronunciation is correct, and he is humble enough to admit that it is only a little more correct then modern but not exact because we have no way to know for certain. I know of houndreds of scholars that think that the current pronunciation is the correct one. They have different reasons for thinking this, some thinking that it was always pronounced this way, while others see no reason to try to go back in time with Latin to the Roman era as we would not want to go back in time to the Shacksperian era to learn the correct pronunciation of English. This seems to me to be once again an exampl in the culture to take whatever the Catholic Church says and disagree with them for no other reason then that you Hate Catholics. It is a fact that only a minority of Scholars adhere o this Reconstruced pronunciation. Most Latin speakers do not use this Reconstructed pronunciation and I will tell you out of hand that using this Reconstructed pronunciation makes some Latin words unpronounciable. There are several things wrong with the Reconstructed pronunciation but the biggest problem is with the vowels. They teach that a long A and a short A had two different sounds. This is just not true. A Latin A only ever has the A sound. Teaching that it can jump sound destroys the correct pronunciation of Latin.

I would say that you should seperate the pronunciation schemes into Modern pronunciation which is still used by MOST if not almost ALL Latin speakers, like myself, and the `proposed Reconstructed historical` pronunciation. We do not want to make it sound to people that we are teaching something as fact when it is just a speculation, and one that only a minority of scholars agree with.

The heart of this goes back to the reason why the student is learning Latin in the first place. If they are learning Latin to only read Cicero, then ANY pronunciation would be useful because he only wants to read. If he is learning Latin to speak to Latin speakers, then he will need to learn the Modern Catholic pronunciation otherwise he will only be half intelligable.

How do we Latin speakers learn Latin? By listening and repeating. Reading and writing come later. First you must listen then repeat, then have a conversation. Then Practice, practice, practice. Only after all of that do you open a book and learn to read and later to write. It must be taught as a language and not as a code for English. No one on this planet teaches the Reconstructed pronunicaiton this way because it will not work. The teacher will not be able to pronounce all the words on the page so the students will not be able to listen so they can not repeat. Students will not learn to speak Latin with this pronuniciation. And really, I do not see the logic of saying that Thomas Aquinas was any less of a Native Latin speaker then Cicero. Considering the amount of written work left to us and the quality of that work, I would actually hold Aquinas up much higher. It is just not right to slander and insult all Latin speakers in the world that have ever existed who did not speak as Cicero did, if indeed he did speak the way the proponants of the Reconstructed pronunciation claim, which I doubt highly. Of the thousands of inscriptions we have from ancient Roman times, more then half attest to the truth of the Catholic pronunciation and go directly against the Reconstructed pronunciation which is based almost entirely on Cicero .

I am not saying that you should change this article to conform to what the Catholic Church says. I am asking you to be respectful and take the TRUTH into consideration. For as Aquinas says . No one who promotes the new pronunciation can say with 100 percent certainty that it is correct. No one who promotes this pronunciation speaks Latin on a daily basis. The Catholic pronunciation is used by the vast majortiy of Latin speakers. They use this pronunciation and take Latin onto their lips everyday, at every hour. Please be respectful in your language you use here. Do not try to make the case that the Catholic pronunciation is wrong or bad or in any way deficient. That is just being Anti-Catholic. If teachers want to use the Reconstructed pronunciation, they certainly are free to do that but I will confront anyone that says that I MUST use it because I am not convinced at all, not in the least little bit, of its authenticity. It certainly does not invalidate the Catholic pronunciation and you should not go out of your way to say that the Catholic pronunciation is not what the Roman Empire spoke and in so saying implying that the Catholic pronunciation is wrong because you do not know. You were not there. So again, be respectful and do not make such implications in you words to build up walls between people that want to learn Latin and the Catholic Church. I have taught lots of people who were not Catholic to speak Latin. And though I hope they would see the beauty of the Catholic faith, I do not force them to convert. I teach them Latin as a language. That is the only legitimate way to learn Latin. —Preceding unsigned comment added by 60.47.43.241 (talk) 01:03, 29 February 2008 (UTC)
 * First of all, see WP:AGF, any bias here is not intentional and wouldn't be surprised if some of the more frequent contributors to this article were Catholic.
 * Second, the article isn't taking the stance that pre-Christian Latin is better or "more correct." The article simply deals primarily with classical Latin pronunciation.  In the chronology of Latin speech, this is the variant spoken before Vulgar Latin was widespread.  Despite this stated focus (see the lead), there's still quite a bit on other variants in the article.
 * Third, the Latin speech today is not native speech. So the pronunciation of Latin today is not only less authoritative but actually varies amongst speakers depending on native language.
 * Finally, our understanding of classical Latin pronunciation is not wholly based on reconstruction. Rome had grammarians and some of their writings survive.  Presumably (I haven't seen any of them) there is corroboration between the reconstruction and the ancient grammarian descriptions — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  01:15, 29 February 2008 (UTC)


 * Aeusoes1, this teacher of Latin at university level seconds your thoughts. InfernoXV (talk) 18:41, 2 March 2008 (UTC)

IPA Pronunciation of "troiae" in the Virgil example
Shouldn't it be "trojai"? It's currently (ignoring length/stress marks) "troje". Grover cleveland (talk) 18:12, 15 June 2008 (UTC)

virumque
I was bold and made a slight change to the IPA transcription of the beginning of the Aeneid. Shouldn't /m/ assimilate to [ŋ] in "virumque"? I understand -que is an enclitic, but I still feel that it is more accurately transcribed as [ŋ]. Anyone who strongly disagrees can change it back.--El aprendelenguas (talk) 01:13, 12 September 2008 (UTC)
 * That might be a bit too much OR speculation. While diachronic assimilation of /m/ before velars is common, it's not necessarily the case that a language that has /n/ assimilating [+velar] will also have /m/ do the same.  After all, in English we don't pronounce "bomb crater" as "bong crater" or kumquat as "kungquat."— Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  03:33, 12 September 2008 (UTC)
 * Ah, you have a point. If you prefer the previous version, I don't mind if you switch it back.--El aprendelenguas (talk) 00:29, 13 September 2008 (UTC)

Derivative languages
"In Classical times, the people in the street did not speak the formal, Classical tongue. They spoke what is known as Vulgar Latin, which was already very different from its sibling, mainly because of simplifications in its grammar and phonology."


 * I think that this is very imprecise. What is "Classical times" here? The so called Classical antiquity? If so, the period is too long to be covered with such a sentence, because the situation changed over centuries. Or, if the "Classical times" are the times of late republic (Cicero, Caesar etc.), then the statement is simply not true. Mamurra (talk) 09:56, 19 October 2008 (UTC)


 * I believe "classical times" means 75 BCE to 200 CE. Keep in mind that "vulgar latin" itself is an amorphous entity subject to change over time even in this classical period.  I agree this could be clearer, as well as sourced. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  18:06, 19 October 2008 (UTC)


 * Ok, then. I am especially wanting to see the source, which supports the claim, that in, say, 50 BCE, the spoken Latin was very different from the litterary Latin, and yet that the differences were mainly because of simplifications in grammar. Mamurra (talk) 19:07, 19 October 2008 (UTC)

Derivative languages, part two
The section uses such a phrase: formal, classical tongue. What does this mean? Does it mean, that classical Latin is formal by definition (and archaic or late Latin are not)? Or does this mean, that besides the formal classical Latin there is yet some less formal (and grammatically simplified) classical Latin? If so, does this mean that the "formal" means the same as "non-vulgar"? Mamurra (talk) 18:15, 20 October 2008 (UTC)
 * My understanding is that the classical tongue itself is formal. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  20:21, 20 October 2008 (UTC)


 * And what is the criterion of calling it "formal"? This adjective is used to mean something, but I must admit, that I don't understand, what is meant. Mamurra (talk) 21:09, 20 October 2008 (UTC)
 * Use in formal settings? I'm not sure.  It wasn't my intention to add confusion by rephrasing things.  I'm sure that once we find sourcing we'll have to reword some things anyway.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  00:50, 21 October 2008 (UTC)
 * Formal here means that Classical Latin, at least by the time of the Roman Republic, was restricted in use to formal circumstances (speeches, official documents, literary works etc.). Vulgar Latin certainly was rather different from Classical Latin, although one could obviously debate how different something has to be before it counts as very different. The bit about differences being "mainly because of simplifications in its grammar and phonology" is dubious. Phonology, after all, is a part of grammar. And "simplification" is a poor choice of word, since relative linguistic complexity is a very uncertain area (there's no easy metric for saying that language x is simpler of more complex than language y). Between Classical and Vulgar Latin there were both grammatical and lexical differences; the grammatical differences included both syntactic and morphological differences. Phonological differences are obviously harder to identify for modern researchers, but certainly there was something analogous to RP: a pronunciation of Classical Latin that was promoted by teachers; and writers made fun of the way uneducated people pronounced things. I agree that this needs sourcing, as everything else does on Wikipedia, but it's worth adding that, "simplifications" aside, none of this is controversial. garik (talk) 15:20, 21 October 2008 (UTC)


 * Once someone defines Classical Latin as being restricted to formal circumstances, an easy conclusion can be drawn, that it was not used in everyday speech. And vice versa, if the assumption is that a language is not spoken, then it, if it is used at all, must be confined to formal settings. But such reasoning is a typical circulus in demonstrando, as the conclusion is the same as the assumption. In fact, or rather to my knowledge of facts, no primary sources exist to support the hypothesis, that everyday speech of Romans living in Rome (for Classical Latin, as we know it, is basically, in late republican period, the litterary version of the language of Rome) differred substantially from the language we know from their written works. So, secondary sources (where the wikipedia likes to refer to) cannot express this in any other way than as a hypothesis. A hypothesis, in turn, is a point of view. So, the Wiki article should not say the Classical Latin was not spoken, but rather according to prof. X, Classical Latin was not spoken(reference here). Otherwise such a statement is an overinterpretation of the facts.


 * And facts are for example such, just to use the example you cite, that it is not that "writers made fun of the way uneducated people pronounced things". It is Catullus, who made fun of ONE Arrius, unable to pronounce aspirates, and the other Petronius, who makes fun of the contrast between the barbarian Latin in mouths of _Greek_freedmen_ (i.e. foreigners, who learned Latin as secondary language and do not command it perfectly) and their pretensions to be Roman citizens. In neither case we can know, that the dubious Latinity is the actual standard spoken Latin (aka Vulgar Latin), or rather a sub-standard slang (in Petronius) or a rural dialect (in Catullus). In both cases, though, there must be some correct-speaking majority, whose linguistic taste and sense of correct grammar and pronuntiation the writer appeals to, because otherwise it would have not been any fun.


 * I wrote the above just to demonstrate that this, indeed, is controversial. And always will be, because we have not enough evidence, because, as the article (or maybe Vulgar Latin, I don't remember ATM) admits, being obviously right, that the ewvidence is scarcely sparsed over ca. 12 centuries. And there always will be scholars who put in doubt opinions of other scholars, and this is what we call controversy. So, to sum up, I think that at least this one section of the present article is not NPOV, and I am unfortunately sure, that most of Vulgar Latin is not NPOV either. Mamurra (talk) 18:31, 22 October 2008 (UTC)
 * There's no use in beating a dead horse. Let's wait for people to bring sources to the table.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  08:30, 23 October 2008 (UTC)


 * I am waiting, but in meantime it costs nothing (except time) to make a reply to garik, as in his posting I can see the same overtones as in both articles. For example: "certainly there was something analogous to RP: a pronunciation of Classical Latin that was promoted by teachers". It is not certain, it is only possible, and only on the grounds of our common sense (which is unscientific). Yet, even if there was some RP in Roman schools, that says nothing about the matter, whether the pronuntiation promoted in schools was artificial, for this is not obvious. In my country, for example, the schools indeed promote some standard of the language (incl. correct pronuntiation), but at the same time this standard "dialect" is simply alive in houses of the prevailing majority of citizens, and so a child before s/he goes to school, learns it basicaly at home. And in other countries (like in Germany) the standard dialect is indeed less or more artificial and some local dialect is spoken in houses, which (the dialect) indeed less or more differs from the written language. So, depends on where you live, your common sense tells you something else about what is natural, and so this method of reasoning about the ancient Rome should be abandoned. What matters is sources. And sources tell us, that the only change in pronuntiation (of the spoken Latin), which took place in 1st century BCE, was caused not by school, but by fashion. And, interestingly, this change is also reflected in written works, specifically in metre. So, to be back to the matter, I am currently not commenting your edit or policy of waiting for sources some time, and then deleting the unsourced matter (which is probably correct most of the time). I am just telling garik (using some, maybe lenghty, argumentation), that there are some controversial things in the POV presented in the fragment of the matter we're discussing. Mamurra (talk) 10:07, 23 October 2008 (UTC)

My 'certainly' was based not on common sense (many cultures don't promote a standard), but on what I'd read about the situation. But I admit it's been a while since I did that reading, and I'm prepared to accept that the issue may be more controversial than I'd been led to believe. Though it may not be. I, sadly, don't have time now to look at the sources. Hopefully some classicist can help us out. However, I would add that the argument is not as circular as it may appear. In fact, it's not actually circular at all. It's just that there is more than one claim. The first claim is this: that there were at least two forms of Latin around during the Republic. One was restricted to formal contexts; the other was used in other contexts. The second point is that we call the first Classical Latin and the second Vulgar Latin, or Vernacular Latin. The third issue is the degree to which these two forms of Latin differed. Presumably the difference was not as great as between Modern Standard Arabic and the various colloquial Arabics; perhaps it was more like the difference between modern literary Welsh and colloquial Welsh. You, as I understand it, are challenging the claim that there even were two (or more) varieties at any one time and place (we can ignore diachronic and geographically based variation for now, which seems to be a different question); you would support the view, in other words, that any individual speaker of Latin would have spoken and written much the same Latin to servants, friends and senators, and would have used much the same Latin again in literary works, official documents and speeches. Or you may simply question the degree to which the language varied between circumstances. Or you may simply be perfectly agnostic on the issue, and be claiming that there may well have been more than one Latin, but that evidence is too poor to be certain. Am I right in one of my interpretations of your view? It's an interesting question, and I'm intrigued by the possibility that the Roman Republic and Empire might never have been diglossic at all (at least with regard to Latin). garik (talk) 11:14, 23 October 2008 (UTC)


 * What I am claiming is that the evidence is definitely too poor to state with such inevitable certainty, as it is stated in the articles, that f.e. Vulgar Latin, as in this political graffito at Pompeii, was the speech of ordinary people of the Roman Empire — different from Latin as written by the Roman elites (it is great pity that the photo is of too low quality to actually discern the letters and read the text), or that classical Latin was not used in speech. "there were at least two forms of Latin around during the Republic. One was restricted to formal contexts; the other was used in other contexts" - but one of the questions is, what is the formal context. This is where I started questioning the phrasing there. The language we know is defined in script. If we define any script (a written book, a decree, a speech, a manual, a letter, a poem) as formal usage, then this leads us to the conclusion, that the language, as we know it, wasn't used in speech, because speech is "informal". And if it wasn't used in informal speech, then it was confined to formal usages. But this seems to me at least, again, a pretty robust circulus in demonstrando, as we just have demonstrated what we had assumed. We can, I think, break it by looking at the first part of the claim you referred to: "there were at least two forms of Latin". Were they? In fact, what we have recorded in the script, is several registers of the language: poetry, rhetoricized prose, scientific prose, philosophy, comentaries, (kind of) novels, whatever. We don't know it, but from our experience with other languages we can deduce with high probability, that there was yet one register: spoken language. Now, how did the spoken language differ from the written language? We don't know, and as nobody (of Romans) says anything about it, we have to assume, that it didn't differ badly from the written Latin, right? Because if it differred, it would have been a great matter for grammarians to write treatises about. Or at least there would be some mentions of it, just as about the atticism in Greece. But what we posses - for the late republican period - are scarce mentions about two or three differences between the script and the speech, yet only in pronunciation, and that's, to my knowledge, all. So, the evidence (or rather lack of) points to the conclusion, that the spoken Latin was apparently nothing special. And this is one thing. The other thing is: all scripts we have, are they really all so formal? Sure, Virgil's poems or Cicero's speeches, they are certainly formal (although schoolbooks on rhetoric contain advices to compose speeches so that they sound natural - so it is sort of an evidence that kills off any concept about vulgar speech being very different from the litterary Latin, because the speeches, we have them, are composed in the litterary language, and at the other hand we have no other option, that to assume, that they have been compsed in agreement with the instructions of the rhetoric - so the obvious conclusion must be, that the "natural language" is pretty the same language as the written one). Anyway, let's call poems and speeches formal. And decrees. And such. But what about Cicero's letters? Why he should have used artificial language in his correspondence with his intimate friend, Atticus? Or his brother, Quintus? So maybe he wrote much the same language as he spoke. But, look at the letters, this is obviously the different register, or different style, but otherwise, grammatically and lexically, this is the same language as in his speeches and everywhere. Pompeian graffittos - I obviously haven't seen them all, but what I have seen, is pretty much standard Latin. Okay, I saw one which exhibits bad syntax, but - the only bad bit in the transcription is quare for quam; and since re and m are easily confused in the ancient cursive (the inscription is so informal as to be written in cursive), the easiest assumption is that it has been badly transcribed. Vindolanda Tablets, see numbers 248, 255, 291, 343 - these are private letters, especially 291 is interesting, because there a woman writes an invitation to her birthday party - for her sister. So this is not only private, not only within the same family, not only it is 2nd century AD, but it is also some god-forgotten province (Britain), not Rome - so one could expect at least unusual Latinity, right? Well, read the letter. It is standard Latin. All of them are. As to the diglossia, this certainly depends on the definition of diglossia. At one point we have Oaths of Strasbourg - by someone, who knows Latin, this cannot be understood. So this is a certain evidence of the diglossia. At the late empire, it is uncertain, however, if this term applies. If St. Augustine in everyday speech spoke in foro esse et in foro ire, but wrote the latter as in forum ire, is this diglossia already, or not? (or, to state it differently, if a Frenchman adds -x at the end of a word to denote plural in script, but the pronunciation doesn't change any bit, is this diglossia or not?) Even if so, this is 4th century AD - what evidence it constitutes for the 1st century BCE? None. Is there any other evidence for the 1st century BCE? Nope. So, rebus sic stantibus, quid dicam, iudices, nisi quod dixi? Mamurra (talk) 17:10, 23 October 2008 (UTC)

Why AE insted of AI
Is there any reason why AE (pronounced /aj/) was written like that instead of AI? In other places, /j/ is written I, like in IVPITER, or in the EI diphtongue. bogdan (talk) 01:32, 26 November 2008 (UTC)
 * The pronunciation in earliest Latin appears to have shifted over time, but the spelling by then had set. You can find some detailed (yet not too dense) discussion presented on pp.92-96 of James Clackson & Geoffrey Horrocks The Blackwell History of the Latin Language (2007, ISBN 978-1-4051-6209-8) --EncycloPetey (talk) 03:00, 26 November 2008 (UTC)


 * Despite that we pronounce AE as /aj/ or /ay/, one should remember, that this is an approximation: the second element of the diphthong wasn't identical to I. Besides, AE and AI differ in yet one respect: the former is (most of the time) one syllable, the other is rather two (like in GNAE-VS vs. GA-I-VS). Of course, exceptions do exist (A-ER, a Greek word, btw.) Mamurra (talk) 11:30, 26 November 2008 (UTC)

The latin "F"
There is evidence (Sturtevant, 1940 and Pulgram, 1979) to suggest that the sound transcribed with the letter "F" (and which is described in this article as a labiodental /f/) was infact bilabial (i.e. /ϕ/.) This is also the position of other Latin/Romance specialists such as Paul Lloyd and Jean-Marie Pierret. I'm modifying accordinglySzfski (talk) 12:04, 18 February 2009 (UTC)

From PIE/Proto-Italic to Late Latin
The recent reversion events makes me think we should do three things.
 * 1) Make a list of phonological changes/features from the earliest stage of Proto-Italic to Vulgar Latin in chronological order
 * 2) Divide these changes into classifications (that is, periods), something like Proto-Italic, Ancient Latin, Classical Latin, etc.  This would, of course, be based on sourcing
 * 3) Decide if we shouldn't discuss phonology as separate from orthography; this is how it's done in other phonology articles and orthography can really muddle things.

— Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi] 22:42, 1 March 2009 (UTC)
 * How many stages would we have, exactly? At its most exhaustive it would probably be something like PIE-> Late "Western" PIE-> Italo-Celtic -> Proto-Italic -> Proto-Latino-Faliscan -> Old Latin -> Classical latin -> Vulgar Latin. Since diachronic change is continuous, not segmental, we'd have a problem with delineating where the boundaries between these periods fall. For example, many of the trends thought of as markers of Vulgar Latin were already well under way in Classical Latin. (The graffiti at Pompeii bears witness to the nascent loss of phonemic quantity with the confusion of short /i/ with long /e/. Varro describes the monophthongization of /æ/ in accented syllables which already had taken place in unaccented ones. Catullus' poetry comments on the H-dropping in lower registers and the resultant hypercorrection.) Is one to think of the colloquial Latin of the Imperial period as Vulgar, then? What makes it quantifiably less "vulgar" than that of the late republican period?


 * Also, I'm not sure that such a list that included stages before Old Latin would make total sense, since it would contain items that are different not only in degree, but in kind. Comparative pyramid-like reconstructions by their very definition do not represent phonological reality in the same way that "on the ground" evidence from people like Varro does. No reputable historical linguist, for example, pretends that PIE reconstructions represent a reality, and that the unpronounceability of the resulting asterisked formulae can simply be overlooked as a symptom of the reconstructed idiom's uniqueness. To write a fable about sheep and horses in reconstructed PIE is a fun pastime, but the text likely has very little resemblance to any speech ever uttered by human mouths. In other words, reconstructed PIE (like reconstructed Proto-Italic) is not a language at all, at least not according to any defensible definition of the term. This is very different from using available data culled from actual Romans in an attempt to better understand what their speech (or their contemporaries' speech) probably sounded like. Szfski (talk) 00:19, 2 March 2009 (UTC)
 * If it's too cumbersome or tangential to begin earlier, beginning with Proto-Latino-Faliscan or Old Latin may be appropriate. I don't think we would have that much of a problem delineating period boundaries because we'd be using the boundaries that sources use.  Although language change is continuous, there is often something that people look to retrospectively that marks such boundaries.  For example, our article on the Proto-Slavic language marks the end of the Common Slavic period when *ě changed to *a after palatal consonants.
 * While there are certainly problems with reconstructing languages, it seems to me that not including what scholarship says about the development of reconstructed languages because of these problems is inappropriate considering our well developed articles on other proto-languages. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  00:37, 2 March 2009 (UTC)
 * Still problematic. There are multiple definitions of what constitutes Vulgar Latin (the Wiki article on Vulgar Latin lists the four main ones.)


 * Also, in order to give an accurate impression of diachronic change, we'll need to give a pretty good idea of the synchronic fluctuation present from the late republican period onward. As I sit and imagine what this will look like, it seems to me that this will quadruple the length of the article. Perhaps a separate article for sound changes would be in order.


 * But, to get back to the original reason for this thread... We need to establish what constitutes "classical" pronunciation and what doesn't. While the commonly accepted guidelines for "Classical Pronunciation" in schoolbooks typically prescribe a diphthong for /æ/, it's almost certain that Virgil himself did not pronounce a diphthong for /æ/ in unaccented syllables like Troiæ, and that even the pronunciation of [se:we:] for Sævæ would have at least been known to him. To say that such monophthongization is "not classical pronunciation" is odd. It's as if, in the year 3000, a professor were to tell students that the only "correct" way to pronounce 20th century English is by pronouncing "cot" differently from "caught," "furry" differently from "ferry," "fail" as different from "fell," "whine" as different from "wine" and "pen" as different from "pin," even though one or more of these mergers is/was active in almost all versions of 20th century English.


 * It's not that I don't think information on proto-languages gleaned from reconstruction isn't valid. I just disagree with the presentation of a proto-language in such a way as to suggest that it actually represents the way some group of people, somewhere, spoke. All a proto-language is is a reconstruction of a group of phones and grammar rules that would have engendered the disparate daughter languages. To assume that this reconstruction constitutes the phonetic and syntactic inventory of a single language is to assume that the reconstructed phones and rules coexisted temporally, socially and geographically (i.e. that the daughter languages don't represent the contribution of the phonetic inventories of different regional dialects, different diachronic stages of the same dialect or the sociolects of different social castes.) Ernst Pulgram points this out admirably in Proto-Indo-European Reality and Reconstruction. For example, a reconstruction of "proto-romance" on the basis of modern romance languages would yeild the non-existent */Jamajes/ as a reconstructed word for "never," with */nunka/ as a dialectic variant (whereas in actuality nunquam was the original standard word for "never," and a reflex of iam magis was originally a dialect word for "ever," whose semantic range was changed by a later innovation that affected the various daughter languages independently, and not all from one source.) Szfski (talk) 15:54, 2 March 2009 (UTC)


 * Oy. I'm now fighting this battle on two user talkpages. We need to hash this out. Szfski (talk) 05:08, 4 March 2009 (UTC)


 * Make that three. Szfski (talk) 19:45, 4 March 2009 (UTC)


 * Scratch that. Quintilian and Terentius Scaurus have convinced me of a diphthong value for /æ/ at least until the end of the second century. However, both of them explicitly state that the sound, while a diphthong is not [aj]. So, how do we represent this sound in IPA?Szfski (talk) 20:57, 4 March 2009 (UTC)


 * How about ? — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɻɛ̃ⁿdˡi]  22:09, 4 March 2009 (UTC)
 * Problem is that the second portion of the diphthong was in fact not /j/ or /i/ but closer to e. (It used to be written "ai" and was changed in the 2nd century B.C to reflect a change in pronunciation.) The secondary sources I've poured through state that the direction of assimilation favored the e not the a. The most convincing primary source seems to be Terentius Scaurus who states


 * A igitur littera præposita est u et e litteris...Et apud antiquos i littera pro ea scribebatur, ut testantur μεταπλασμοί, in quibus est eius modi syllabarum diductio, ut pictai vestis et aulai medio pro pictae et aulae. Sed magis in illis e novissima sonat. (Translation: And therefore the letter A is prefixed to U and E....and among the ancients i was written for E as evidenced by the morphological alterations including such syllabic deviations as pictai vestis and aulai medio for pictæ and aulæ. However, in those words the final sound is more nearly that of E.)


 * Please provide primary source to corroborate your last sentence. And what do you mean by "is"? In classical times, such as in Virgil's times, "ae" undoubtfully denoted a diphthong, not a monophthong "more nearly that of E". 193.0.79.14 (talk) 09:21, 5 March 2009 (UTC)
 * That sentence was a translation of a primary source "Sed magis in illis e novissima sonat." Szfski (talk) 09:57, 5 March 2009 (UTC)


 * This doesn't mean that "ae" sounds as "e". This only means that the second element of the diphthong is close to "e".
 * Second thing, this does not say anything about the quantitative correlation between these two elements, so your latest editions in the article, marking the "a" with breve over it, are rather unfounded. Quite contrary, from the words of the grammarian(s) it can be seen, that the first element was quite normal "a", while the other element was only close to "e". One can reasonably argue then, that the "a" was the prevailing element, and if one should put a quantitiy sign over either of them, it should be a macron (and over the "a"). 193.0.79.14 (talk) 11:58, 5 March 2009 (UTC)
 * See Sturtevant's page 49 of The Pronunciation of Greek and Latin as well as Allen's Vox Latina. The fact that æ was leveled in the 1st and 2nd centuries as /e/ rather than as /a/ suggests that you are quite wrong. Find me a grammarian who describes a as the "prevailing" element. Till then, I think my edit should remain as it is. Earlier in this thread, I already conceded a diphthong pronunciation until the 1st century AD. Szfski (talk) 12:16, 5 March 2009 (UTC)
 * Find me a grammarian, who does NOT describe the diphthong as if the "a" wasn't the prevailing element. Noone of them has any doubts about what is the first element, they all state that this is "a". For second element they are in doubt as the one you cited "magis in illis e novissima sonat''' (i.e. this is "more e than i", not "this is e".). 193.0.79.14 (talk) 14:31, 5 March 2009 (UTC)
 * So you think that the fact that the first element is easier to describe makes it more prominent? Kto Ci tak powiedział?
 * The fact that the first element is easier to describe for a grammarian not trained in phonology in no way means that it is more prominent, more prevalent or more salient than the second. Terentius Scaurus isn't describing the diphthongs two elements in terms of prominence. But, if the scholars I've just pointed you to seem unconvincing, then consider the fact that /æ/ merged with /e/ and not /a/ which suggests that the direction of assimilation was front, not back.Szfski (talk) 14:52, 5 March 2009 (UTC)


 * I know of no sources, which would strongly support either view. We know however, that the original state of affairs was so that there was no diphthong there, both elements were separate long vowels, as in this verse: olli respondit rex Albai Longai (... "al-ba-i lon-ga-i"). When they started to merge, they formed the diphthong, where - as we know from the grammarians - the first element remained unchanged (probably on the analogy with the rest of the declension) clearly audible and thus "easier to describe". I would say, that this was the case until the second element has finally changed into something open and close to "e". The next phase was that the "a"-element started being absorbed by the second one (quite naturally, as this is what diphthongs usually do).


 * I see no reason anyway to put the breve over the first element, as 1) there is no evidence whatsoever that this wasn't (or was) the prevailing element in Virgil's times, 2) this makes the false impression that the diphthong itself was short, which is of course not true ("ae", regardless of the pronunciation, is long most of the time). If there will be no further discussion, I will revert the change. PS. and 3) please don't use local languages in the discussion, even if your interlocutor understands it, there are other people who might want to fully understand, what you said, and not everyone on this planet can command Polish. Mamurra (talk) 20:06, 10 March 2009 (UTC)


 * I personally don't think that we should be in the business of modifying the page's IPA representation of Classical Pronunciation based on how people may or may not misread it. I also disagree with the notion that the second element was [j] for reasons already stated (as do Sturtevant and Allen, scholars whose work should most definitely be considered.) The breve can go. But we still need to find a better way of representing the diphthong than [aj].


 * There's also the issue of transcription in general. As Auseuoes1 pointed out, this is the speech of 2000 years ago. Perhaps we're better off not representing it in a form as exact (and as exacting) as IPA is. Perhaps we should go for phonemic, rather than phonetic, transcription.Szfski (talk) 05:25, 12 March 2009 (UTC)


 * That's right. BTW. It has IMHO never been an issue, if the AE was pronounced as [aj] in 1st century BCE, because it is rather obvious, that it wasn't: [aj] is only our, slightly archaizing, scholarly approximation (but a good one, it has two sounds in one syllable, just like AE had). At the other hand, I think that many people could have a hard time trying to differentiate between AE in Roman mouth, and [aj] (just like I myself can hear completely no difference between German EI and AJ in my own language, though I was *told* that there is some. Mamurra (talk) 09:55, 19 March 2009 (UTC)


 * Again, I repeat- there is absolutely no sense in modifying the IPA representation because of what sounds people will "have a hard time trying to differentiate." IPA is not intended to give approximations or to be user-friendly. Its primary purpose is to represent sounds such that someone familiar with it can reproduce such sounds in as exact a manner as possible. Likewise, the fact that you yourself cannot tell the difference between and what I'm assuming to be  does not mean that one should not transcribe them between phonetic brackets as if they were the same sound. Szfski (talk) 19:17, 19 March 2009 (UTC)


 * You're missing the point :) It is not a problem whether I can differentiate this sound from that. The real point (to which I agreed above) is, that exact phonetic representation lacks sense, if we don't know, what real value of the sound was. We know that it was a diphthong and that native speakers heard "ae", so that words with diphthongs written so didn't occur to them as written and pronounced differently (see Quintil. Inst. I 27); but the most fortunate approximation of that we can produce is [aj] (with long first element). As you can see yourself, there is no agreement on whether the first element was long or short, so what's the sense of trying to be accurately specific, if we really have no specific knowledge of that matter? But okay, this gonna be original research, so please provide IPA representation from a reliable secondary source, cite it accurately, and that will be probably the end of the question (until someone comes with a better source). Mamurra (talk) 22:05, 19 March 2009 (UTC)


 * Which is exactly why I'm arguing for phonemic, instead of phonetic, transcription. My other point, which seems to have gotten lost in the laberynthine whorls of this thread, is that while we don't know precisely what the phonetic value of /Æ/ was, we can be pretty sure as to what it wasn't. It definitely wasn't [ai] or [aj] after the second century BCE. Allen, Corssen and Sturtevant all attempt in circumspect ways to describe what they think this phoneme sounded like. Unfortunately, they don't give a putative IPA rendering but rather attempt to describe it in terms of their native phonetic inventories, which is obviously of no help whatsoever. I'll keep looking though.Szfski (talk) 22:22, 19 March 2009 (UTC)


 * Thereby suggesting that aj would not be an accurate realization. Szfski (talk) 22:19, 4 March 2009 (UTC)
 * Ahh, but I was using slashes, where that's more okay. Nevertheless, we're talking about a section with phonetic brackets.  Let's not be more precise than is warranted considering that this is the speech of two thousand years ago.  — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɻɛ̃ⁿdˡi]  06:28, 5 March 2009 (UTC)
 * Yes, you're right. In any event, it would seem that this thread has moved deep into TL;DR territory. Perhaps it's time I moved on to bigger and better things as Rachel Marsden did. Szfski (talk) 12:44, 5 March 2009 (UTC)


 * I noticed the breve as well. If that's to mean that it's not the primary member of the diphthong, then you'd actully want to use .  — Æµ§œš¹ <span title="Representation in the International Phonetic Alphabet (IPA)" class="IPA"> [aɪm ˈfɻɛ̃ⁿdˡi]  17:08, 5 March 2009 (UTC)

Lavina or Lavinia?
Someone has been changing "Lavina" to "Lavinia" in the quotation from Virgil. My own copy of the Aeneid clearly gives "Lavina", so does the Oxford Latin Dictionary. Can anyone propose an authority for "Lavinia"? If not I would propose to revert the changes. --rossb (talk) 17:10, 4 March 2009 (UTC)
 * Lavinia and Lavina come from two different manuscript traditions. I don't know which one is more authentic (I'd be interested in knowing if there's anyone alive who does.) But if you need authority for Lavinia, then check the Loeb Classical Library edition or the Perseus Digital Library, which uses Greenbough's edition. Both of them have "Lavinia." Szfski (talk) 19:38, 4 March 2009 (UTC)
 * Both forms in this place have ancient authority, at the moment I have no idea, what's written there in the oldest manuscripts, but Servius acknowledges both variants. He however prefers "Lavina" as correct (so it can be deduced from there that there were no sources known to him which would support the authority of "Lavinia" as genuine). Mamurra (talk) 08:34, 11 March 2009 (UTC)
 * For some reason, recent edits have left this in an inconsistent state. I propose to rectify the inconsistency, by standardising on "Lavina". --62.49.68.79 (talk) 17:30, 12 March 2009 (UTC)

Myriad systems??
Troublesome statement:
 * Myriad systems[dubious – discuss] have arisen for pronouncing the language — at least one for each language in the modern world whose speakers learn Latin[citation needed].

While the statement correctly indicates that there are millions of ways to mispronounce Latin, mispronouncements aren't systems proper, they're most usually system violations. I would actually guess that there use to be approximately one way to pronounce a Latin borrowing per language, which is not the same as pronunciation habits in Contemporary Latin ― a foreign language to everybody, having its own independent pronunciation rules. ... said: Rursus (bork²) 15:01, 21 June 2009 (UTC)


 * That couple of sentence is a conglomerate of weasel words. Toning it down and removing tags, but feel free to re-add them if you still don't find it satisfactory and undisputed. LjL (talk) 15:22, 21 June 2009 (UTC)

Rising diphthongs
In the Classical pronunciation example, all the "ae" diphthongs are represented as [ăe], but isn't this suggesting a rising diphthong (since that seems to be the IPA diacritic for "extra-short") when those diphthongs were actually always falling (either [ai̯] or [ae̯], using the kind representation that Diphthong uses)?


 * Actually, [ăe] would be two vowels with hiatus, not a diphthong. Corrected to simple [aj] per comments above. It's possible Latin distinguished [j] from [i̯], in which can we'd need to be more precise. Were aea and aia distinct, say across a word break: ae a vs. a ia, in set phrases where the words are run together? Or would they both be geminate [ajja]? kwami (talk) 18:38, 27 June 2009 (UTC)


 * Uhm, I don't know about that (IIRC Vox Latina says something about that but only in passing and without any degree of certainty), but while in the example phonetical transcription using [j] is probably alright, I'm not a fan of using it in when denoting phonemic values of the diphthongs... because it would mean you're not really analyzing them as diphthongs to begin with - after just saying they are. And if you keep in mind that they soon developed into something where there was something short of a full [i] as the falling element (rather [e] or similar), eventually resulting in monophthongization... putting [j] kind of puts the opposite idea across, i.e. that it wasn't that prone to lenition/assimilation. LjL (talk) 20:58, 27 June 2009 (UTC)


 * If we avoid /j/, how do we decide between /ai̯/ and /ae̯/? (Not an objection, just a question.)
 * I disagree that /aj/ implies they weren't prone to lenition, though you are correct that /aj/ would not be a true phonemic diphthong--but then we're back to my question above: is /aj/ distinct from /ai̯/ in Latin? kwami (talk) 22:00, 27 June 2009 (UTC)
 * Well no I guess it doesn't imply it, just kinda suggests it subtly.
 * About /ai̯/ vs /ae̯/, I'd choose /ae̯/ for the following reason: sources suggest that the change in spelling from AI to AE that occurred in the Republic reflected a change in pronunciation occurring at the same stage. So if we pick AE in the plain text (and I'm sure we would, as that's what virtually everyone uses to spell "classical" Latin nowadays), it makes sense to "match" that to a phonetic rendering for the same historical period, i.e. /ae̯/.
 * LjL (talk) 22:39, 27 June 2009 (UTC)


 * Interesting that something that subtle would be reflected in the orthography, unless maybe there was a contrast between aea [ae̯.a] and aia [aj.ja]?
 * We have a fair number of Classical Latin transcriptions in Wikipedia, and I've never seen one that used /ae̯/. If we use that here, wouldn't we want to convert the other articles too? kwami (talk) 23:39, 27 June 2009 (UTC)


 * Look, you keep hinting to this contrast between [ae̯.a] and [aj.ja] - I would be delighted to be able to say whether there is such a contrast or not, but I don't know. Honestly I kind of doubt anyone knows for sure, Vox Latina certainly seems uncertain when it comes to such really subtle things.
 * You may have a point that if the other articles (and I suspect quite a bit of the literature) use [ai], it might be better to just use that. That's a valid point also because I think none of the sources in the article actually go as far as saying that [ai] evolved into exactly [ae], but just that the tongue position was lowered, that the articulation was more relaxed, or other such vague phrashings.
 * At the end of the day, I'm pretty neutral on whether to use [ai] or [ae], there's decent arguments for both. There's one source (don't remember which right) that says it was probably pronunced "exactly" like American English [aɪ], that might be another decent in-between choice.
 * LjL (talk)

There was certainly a contrast between AEA (like in Aeaea, the island) and AIA (like in Traianus), as Romans themselves tell us that intervocalic 'i' is doubled; so it is actually 'ajaja' but 'trajjanus'. However, the intervocalic position is somehow a special case. Otherwise, where 'i' is followed by a consonant, the 'ai' cluster forms two syllables (like terra-i frugifera-i in Martial), whereas 'ae', with few exceptions (mostly loan words, like 'a-er', with long 'e') is normally one syllable. Just a note. Mamurra (talk) 11:42, 28 June 2009 (UTC)


 * So phonemic /ai̯a/ vs /aia/, /ai̯k/ vs /aik/, phonetic [aja] vs. [ajja], [ajk] vs [aik]? kwami (talk) 18:30, 28 June 2009 (UTC)

gu + vowel
I have one source that seems to indicate that gu before a vowel became /gw/ rather than the expected /gu/. However, this doesn't seem to be discussed in our article. I'd like to know because it would affect syllabification. --EncycloPetey (talk) 18:27, 9 July 2009 (UTC)


 * Why is /gu/ "expected" in that context? --LjL (talk) 20:06, 9 July 2009 (UTC)


 * Because g is /g/ and u is /u/. --EncycloPetey (talk) 20:34, 9 July 2009 (UTC)


 * Err... no? "U" is exceedingly often /w/ ("uiuus" /'wiwus/), and the article definitely mentions that. --LjL (talk) 20:38, 9 July 2009 (UTC)


 * Yes, between vowels it is, and at the beginning of a word when followed by a vowel. I'm asking about the situation where it occurs after g but before another vowel, in a situation where my sources differ as to the pronunciation.  Do you have any helpful information? --20:51, 9 July 2009 (UTC)


 * Helpful? I don't know, I thought this was about the article - which has "Usually the semi-consonant v after q or g is still printed as u rather than v, probably because in this position it did not change from /w/ to /v/ in post-classical times", which to me implies a semivocalic pronunciation after "g". It's unsourced, that's true. This claims that "gu" is pronounced /gw/ only when preceded by "n", as in "sanguis". My dictionary agrees. --LjL (talk) 21:28, 9 July 2009 (UTC)


 * Which dictionary? This does help with my case, since the specific word(s) I am dealing with have ngu in them. --EncycloPetey (talk) 22:24, 9 July 2009 (UTC)


 * Castiglioni-Mariotti, Vocabolario della lingua latina, 3rd, Loescher, 9788820166502; but it really only "agrees" in the sense that it puts neither a brevis nor a macron on the "u", which normally means the letter is not a syllable nucleus, but it isn't entirely consistent in its usage. It's got a brevis sign on "puella" though, for instance. --LjL (talk) 22:50, 9 July 2009 (UTC)


 * These are the Indo-European labio-velar series: becomes Latin qv, and  becomes Latin gv (though  I think becomes just v), so it's not surprising that v-vowel after g and q would behave differently than after p or other consonants. Or is that not the question here? kwami (talk) 06:53, 10 July 2009 (UTC)


 * I actually meant to say that the fact it's got a breve sign on "puella" but not on "sanguis" reinforces the idea it's doing it for a valid reason, even though it seems to be a bit inconsistent in other cases (mostly by just not putting any sign on vowels that are "obviously" syllable nuclei and, I guess, "obviously" short or long - obviously for the dictionary, that is, not quite for me). --LjL (talk) 13:16, 10 July 2009 (UTC)

This is not "gu + vowel" but "ngu + vowel" which "consonantizes" the "u". BTW. a dictionary putting no quantity sign on a vowel is no information, as some quantities are uncertain or simply unknown. Mamurra (talk) 12:33, 23 July 2009 (UTC)
 * I see... it does seem to do it a bit frequently, though. --LjL (talk) 13:05, 23 July 2009 (UTC)


 * Generally, dictionaries mark vowel quantities and do not mark syllable quantities, as these are usually obvious (for natural length) or follow some constant rule (for positions). Maybe this is your case. My remark above was about the difficult questions of words, in which the real vowel quantity is undeterminable. Mamurra (talk) 10:54, 28 July 2009 (UTC)

A Medieval Latin Pronunciation Problem
In the Section entitled "Medieval Latin", there is a poem and a transcription: Pange lingua gloriósi Córporis mystérium, Sanguinísque pretiósi, quem in mundi prétium fructus ventris generósi Rex effúdit géntium. When I read the poem aloud, I realized that I did not pronounce every word the same as in the transcription. Specifically, I did not shift the "ti" in "gentium" to "tsi". I read it aloud several times and "gentsium" just sounds wrong, even while "pretsiosi" and "pretsium" sound fine. What gives?

Then I remembered that in the transition to Italian, that "pretium" formalizes this sound change into "prezzo", but "gente" retained the original pure "t". So I have to wonder if there are some cases in which the existence of a preceding vowel or consonant (the "n" in "gentium" in this case) stops this "ts" change from occurring? I can't bring any academic arguments to the table, but having spoken Latin since before Vatican II (early 60s) and Italian for nearly that long, perhaps I can have the right to just question when things sound wrong... —Preceding unsigned comment added by Mccalpin (talk • contribs) 20:22, 11 August 2009 (UTC)


 * Italian gente is not derived from Latin gen. pl. gentium, but from acc. sg. gentem. So the pronunciation of the Italian gente is not related to the medieval pronunciation of gentium. If you pronounce pretium as pretsium (or precium), then there is no reason why you shouldn't pronounce gentium as gentsium (gencium). The only position (for this pronunciation manner) which prevents ti from shiftins is after 's' (e.g. in caelestium ti should be pronounced, not tsi).
 * I admire you speaking Latin for so long, but I am afraid, that your pronunciation customs have no authority, as you are not a native speaker. Mamurra (talk) 13:49, 29 August 2009 (UTC)


 * Mamurra, I thank you for your response. However, you did not address the fundamental issue - pretium did shift to prezzo in all cases, while gente did not shift to genze or gezze. Well, of course not...but are we positive that "gentium" had the same sound shift as "pretium"? That is, do you have any empirical evidence that "gentium" and "pretium" were necessarily pronounced the same way?


 * I do not mean this as an argumentative question, but as a serious question...just how do we know that "ti" was always pronounced "tsi", or are we just applying a rule that actually didn't exist in Late Latin?

William J. &#39;Bill&#39; McCalpin (talk) 05:33, 30 August 2009 (UTC)


 * I'm pretty sure that the pronunciation example titled "'Italianate' ecclesiastical pronunciation" is not meant to be a historical reconstruction, but is simply the modern Italian pronunciation. --Alatius (talk) 10:58, 30 August 2009 (UTC)


 * Well, William said "Late Latin" though; while the Italianate pronunciation section certainly is what it says (the ecclesiastical/modern Italian pronunciation), it's pretty certain some change of that sort must have happened at some stage in Latin, because it's reflected in most (if not all) Romance languages as well as Italian. --LjL (talk) 12:09, 30 August 2009 (UTC)


 * Nothing to do with having an "n" before the "t"; "gentem" simply has an "e" after "t", not a semivocalic "i" + vowel. That's what triggers the change in Italian. A hypothetical accusative singular "gentium" would most definitely become "genzio". --LjL (talk) 12:09, 30 August 2009 (UTC)

Traditional Orthography
In the "traditional orthography" for the Virgil passage there are now a number of accents. What's the rationale behind these (and I suppose I should say is there a citation for them)? I presume the acute accent is supposed to indicate the stress in the case of words with suffixed -que which the reader might otherwise be doubtful about, but if so it's clearly wrong in the case of "Lavináque" where the stress has to be on the first syllable. --rossb (talk) 11:40, 30 October 2009 (UTC)

Upper/lower case
There is a bit of a natural inaccuracy here I am trying to correct. The classical Romans used their block or their cursive; our "lowercase" (a renaissance term) was not known. The tables begin by making a point to employ block. Then in the notes we slip into lower case representations, a natural enough mistake. Where we are trying to use graphemes - the ones the Roman used - I am putting over to block if not that already. Where we are simply referring to Latin words using any script I let the lower case stand. Phonological representations of course have their own representations in lower case.Dave (talk) 16:04, 31 October 2009 (UTC)