Vowel quality vs. quantity

Hello,

The Wikipedia page about the pronounciation of Classical Latin vowels talks about a distiction not only between short and long, but also between degrees of openness, such that the short U had the same quality as long O (except it was short) and the same goes for short I and long E which supposedly only differed in length.

While this phenomenon does not seem to be disputed, other sources (teachers of Medieval French among others) maintain that CL only had the long/short distinction, which only progressively transitionned into a open/closed distinction, and only in late latin.

So I'm wondering about the chronology here. Did classical latin already have that double distinction, or is it a later, transitory stage that ultimately led to the lost of vowels quantities which came to be replaced by vowel openness in Romance languages?

Thanks.
 

Godmy

Sīmia Illustris
Hey, interprete,
so first about the sources. When it comes to the restored pronunciation, any older source is practically useless and also it has very little to do with the traditional/medieval national pronunciations of Latin in respective European countries (like France). You have to use an up-to-date resource since the comparative linguistics and the knowledge of the proto-language the Latin has descended from (the Proto-Indo-European) gets better and better almost with each decade. The Wikipedia uses the up-to-date sources, but you have to read them carefully. The best for you is to approach the source itself like is for example the still quite modern book from the linguist Sidney Allen, called Vox Latina.

Vox Latina reconstructs the vowels thus and it is widely accepted both among latinist and historical linguists/phoneticians, so you can go by this:

latinvowels.PNG
Now, you have to imagine (in case you or a potential reader of this thread are/is not a phonetician) that it's a human head/mouth from the profile with the eyes/teeth being on the left and the neck/pharynx on the right. The tongue in the mouth travels either up or down or forwards or backwards and the place the tongue stops in is what pretty much makes a vowel in this diagram (then it is also whether you round your lips or not round your lips, one can round their lips for any tongue position generally in phonetics, not Latin of course).

So, as you can see, the short u is still more closed than the long ō, so the information about their same quality you've got from the Wiki is either badly interpreted on your side or simplified on theirs.

Further on, you can see that it should be just a and ā that differ solely by length with the identical quality.

All long vowels, of course, must be enounced as long (by duration, not by diphthongization as one does e.g. in the English word café <- that's "no, no"), no matter whether you get the quality exactly correct or not, the length is still pretty important in the classical Latin and should not under any circumstances be neglected (more in the thread in this section "Let us hear and analyze our Latin pronunciation").

___________
My native language, Czech, with the exception of i/ī (where we do it just like here) differs all short and long vowel pairs solely based on their length, the quality being the same, although there is an argument in phonetics that especially high/closed vowels tend to "tense up" as they are lengthened and naturally move higher (in an arbitrary language), so one could argue that Czech wouldn't in practice differ from Latin that much anyway...

Anyway, I hope I answered your questions, if you need more explanation, tell me.
 
Many thanks Godmy for your detailed answer. Just coming back to a few points:


Hey, interprete,
so first about the sources. When it comes to the restored pronunciation, any older source is practically useless and also it has very little to do with the traditional/medieval national pronunciations of Latin in respective European countries (like France).
You misunderstood me, because I wasn't very clear. I mean current teachers of old French, in this 21st century, in French universities, rely a lot on latin phonology to explain how old French words derive from classical Latin though vulgar/late Latin (not sure about the terminology used in English here) especially when it comes to the evolution of vowels from classical Latin all the way to old French.
And so from the courses I've found online, according to them classical latin only had the short/long distinction, not the vowel rounding distinction which (still according to them) only appeared in later latin and, after a period of coexistence, eventually completely replaced the long/short distinction, as can be seen in early (and later) romance languages.

So, as you can see, the short u is still more closed than the long ō, so the information about their same quality you've got from the Wiki is either badly interpreted on your side or simplified on theirs.
Wikipedia substantiates this claim by pointing out orthographical errors dating back to classical latin which strongly suggests that the confusions in spelling are due to the sounds being the same, their argument being that if the vowels were all different, it would have been odd to mix them up. Here are the cases mentioned on the Wiki page:

Short mid vowels (/e o/) and close vowels (/i u/) were pronounced with a different quality from their long counterparts, being also more open: [ɛ], [ɔ], [ɪ] and [ʊ]. This opening made the short vowels i u [ɪ ʊ] similar in quality to long é ó [eː oː] respectively. i é and u ó were often written in place of each other in inscriptions:[35]
  • trebibos for tribibus [ˈtrɪ.bɪ.bʊs]
  • minsis for mensis [ˈmẽː.sɪs]
  • sob for sub [sʊb]
  • punere for pōnere [ˈpoː.næ.rɛ]
What do you make of this?

Thank you for your other references, that's all very interesting. What surprises me is that (unless I misunderstood) Wikipedia presents as undisputed, information that does not seem to be unanimously taken for granted.
 

Godmy

Sīmia Illustris
Many thanks Godmy for your detailed answer.
No problem! :)


You misunderstood me, because I wasn't very clear. I mean current teachers of old French, in this 21st century, in French universities, rely a lot on latin phonology to explain how old French words derive from classical Latin though vulgar/late Latin (not sure about the terminology used in English here) especially when it comes to the evolution of vowels from classical Latin all the way to old French.
I'm not sure I misunderstood, but anyway, current teachers of Old French won't have that much information that historical linguists focused on Classical Latin will, I suppose. And at the end, you should read the primary sources, not the secondary ones, although they are very helpful too (by primary I mean books, by secondary I mean Wikipedia and generic online articles).

And so from the courses I've found online, according to them classical latin only had the short/long distinction, not the vowel rounding distinction which (still according to them) only appeared in later latin and, after a period of coexistence, eventually completely replaced the long/short distinction, as can be seen in early (and later) romance languages.
I guess you don't mean vowel rounding but vowel quality here, otherwise I understand. Well, according to the current scholarship on the classical Latin, the short and long vowels slightly differed in quality, as shown in the diagram. Eventually the length was lost, but that's not what concerns us. Old French teachers are also concerned with classical Latin or proto-Italic or Proto-IndoEuropean only secondarily as vulgar (late) latin is of their main concern (chronologically).

Wikipedia substantiates this claim by pointing out orthographical errors dating back to classical latin which strongly suggests that the confusions in spelling are due to the sounds being the same, their argument being that if the vowels were all different, it would have been odd to mix them up. Here are the cases mentioned on the Wiki page:


What do you make of this?
They certainly were similar, which the diagram I've posted shows (also the information you're mentioning are mostly from the same book I already mentioned), also there were many dialects in the ancient Italy etc... The restored pronunciation tries to focus on some universal literary city-of-Rome dialect in the first century BC (mainly) and it says that despite the similarity, there is still a difference between the two mentioned vowels. How much they are similar or the same you can see for yourself in the diagram I posted.

Thank you for your other references, that's all very interesting. What surprises me is that (unless I misunderstood) Wikipedia presents as undisputed, information that does not seem to be unanimously taken for granted.
Well, I checked the Wiki article and it mostly states Allen as its source (Vox Latina) whence I also took the diagram and... I don't think it poses any quandary for the learner about how you should pronounce it.

If you are looking simply for historical linguistic information on different variants of Latin, I suggest you read the book and look at the data in the detail. If you're looking for guidance about what the norm is, the diagram (from the same book) is what we currently consider the norm (similarity but not identity in respect to some literary city-of-Rome dialect of a certain time period, when it comes to the pairs in question).
 

Matthaeus

Vemortuicida strenuus
Besides Vox latina, I'd also suggest Palmer's The Latin Language as a good source concerning Vulgar Latin. I'm making this post, because, coincidentally enough, I'm reading a chapter on Vulgar Latin from that book at the moment.
 
Thanks once more Godmy.
After reading the Wikipedia article in light of your answer, I realize my problem is that I confused « similar » with identical. I don't want to waste your time asking follow-up questions that might find an answer in the book you recommended, but I wonder how easily differenciable those similar vowels were.
If I try to apply the diagram, I find it easy to distinguish the quality ascribed to long i, from that ascribed to short i (sorry I don't have IPA) for example, but that vowel in between the two wouldprobably give me a very hard time differenciating from the other two, let alone pronouncing it right, because they are so close.
Same for long e (which, if I got it right, equates French é) and short e (French è, please correct me if it's not the case): I can't imagine there being yet another vowel in between é and è that would be readily distinguishable in speech.
I am certainly not questioning the information but I can't help expressing my surprise. Are there extant languages that use such fine differences between their vowels?


Well, I checked the Wiki article and it mostly states Allen as its source (Vox Latina) whence I also took the diagram and... I don't think it poses any quandary for the learner about how you should pronounce it.
Well guess what, after reading (and yes, misunderstanding) that Wikipedia article, I was about to go and transcribe whole latin texts in IPA to train myself to read them, and I was going to ascribe the long E sound to short Is , and the long O sound to short Us, etc., based on the similarity mentioned in the article. So I'm glad you cleared it up for me (although I'm certainly not glad that the matter is more complicated than it seemed).

Since you are Czech, I wanted to ask you something about vowel quantities but I hope I'm not going too much off-topic: does your language make it compulsory to preserve quantities even in words that cannot be confused?
I know that vowel quantity in latin can help differenciate some verb forms, like venit and vēnit for example. But I wonder what the motive would be to have a long a at the beginning of āmbulare for example, when it doesn't help achieve any additional clarity. Do you have similar examples in your own language of words that have long vowels in odd places that people pronounce correctly even though shortening them would not lead to any confusion?

Thanks again for your time.
 

Godmy

Sīmia Illustris
Hi again!
that vowel in between the two wouldprobably give me a very hard time differenciating from the other two, let alone pronouncing it right, because they are so close.
I guess the way to start would be to pronounce ē with the same quality as e but just longer and then "tense it up" a bit, the French, to my information uses such "e" whenever they write the diacritic over it and it's still distinguishable from "i". You can start with lengthening (more similar to how Czech does it) and then try to tweak the quality (close the long ones) gradually... in time.

Edit: Yes, you just mentioned French, now I got to that part of message.^



About Czech: mostly (I'd say always) it is compulsory unless some dialects which may tend to pronounce certain endings or even specific words shorter (but that difference is rather lexical than phonetic, because it usually is a different word pronunciation, but in a different word the vowel group doesn't get shortened, although there are also some very strange dialects that tend to suppress all the lengths, but these are people living near to the Polish border, they sound quite strange to the rest of the nation). If you don't preserve the long vowels in speech (even if rapid) you sound right away as a foreigner and sometimes it makes the understanding harder, yes. Also one difficult thing for foreign speakers of Czech is to keep the accent on the beginning of the word (or on the beginning of the preposition+word group) and yet in the same time do the subsequent vowel lengths correctly, the intonation is usually quite hard to get even for other Slavs who approach Czech, even for Slovaks with whom we have otherwise mutually intelligible languages. Czech is, among the Slavic languages, phonetically an exotic creature...
 

Matthaeus

Vemortuicida strenuus
But it's true! All the people I know here make no pretension of pronouncing the vowels correctly in Latin. It sometimes drives me crazy.
 

Iáson

Cívis Illústris
But it's true! All the people I know here make no pretension of pronouncing the vowels correctly in Latin. It sometimes drives me crazy.
Don't worry, I live in an English speaking area where there is a phonemic distinction between short and long vowels, and still no-one bothers to pronounce them correctly in Latin.
 

Matthaeus

Vemortuicida strenuus
ok, making the short-long distinction is one thing, but not accenting the right one is a different matter altogether
 
Hi again!

I guess the way to start would be to pronounce ē with the same quality as e but just longer and then "tense it up" a bit, the French, to my information uses such "e" whenever they write the diacritic over it and it's still distinguishable from "i". You can start with lengthening (more similar to how Czech does it) and then try to tweak the quality (close the long ones) gradually... in time.

Edit: Yes, you just mentioned French, now I got to that part of message.^
Thanks again Godmy for taking the time to answer.
Actually my question was not about how to train myself to pronounce the vowel ɪ between i and e, or the vowel ʊ between u and o. My question was, are those differences really audible, practically speaking, if we set aside the theory? I really doubt that but I would be happy of course to be proven wrong.
I think in the best of cases any individual can train themselves to pronounce those 2 sets of three vowels so as to make them sound different from one another to his or her own ear, in isolated syllables. But in the middle of a whole sentence enunciated at normal talking speed, I really don't see that happening. Even when hearing IPA recordings (like here for example) I notice that some of the vowels are a bit off and the way they are recorded do not match any French vowel although they are supposed to... only rough approximations. Hence my serious doubt about the level of realism in those theoretical vowels ascribed to Latin.

Coming back to Czech I have another question if you don't mind. I have zero knowledge of Czech other than its name, so I checked the Wikipedia page on Czech which offers a sample recording of someone speaking Czech. And what surprised me (aside from the numerous Russian-sounding words, because I know a bit of Russian) was that I was looking for long and short vowels, yet was unable to make them out at all!
In contrast, when I listen to recordings of people speaking in Latin, even though I am a beginner and probably recongize as many words in Latin as I do in Czech based on my knowledge of Russian, I find that the vowel quantities are very (too?) obvious.

So my question is: when you have a long vowel in Czech, does it really take twice the time a short vowel does to pronounce? or is the difference much more discreet? Isn't it then a bit artificial to make long vowels in Latin last twice as long as short ones?

Thanks again.
 

Godmy

Sīmia Illustris
Well, are you suggesting that Czech has an x phonological description which doesn't correspond to practice? All native language described phonologies show only discreet units that are in practice distinct to the natives and the human ear is a pretty intricate thing, it can hear incredibly minute differences (the brain is another issue, it can start ignoring hosts of differences from about the 1st year of age, when the brain is "taught" so, due to getting used to the one and only native future phonology of the infant).

I just don't think you can mean that question seriously, I'm sorry.

I think in the best of cases any individual can train themselves to pronounce those 2 sets of three vowels so as to make them sound different from one another to his or her own ear, in isolated syllables.
But now you're not talking a native speaker, you're talking a foreign learner, that's a huge difference neurologically, partly from the reason I've just described (the brain being trained to ignore hosts of differences from about the 1st year of age and on, to accustom itself just and only to the native phonology alone). You can't compare how native speakers perceive phones and how foreign speakers can or potentially be able to.

You are asking a practical question. My answer based on my practice with foreign languages and foreign phonologies is: if you learn to pronounce the vowels correctly and the target words correctly, you start spotting natives doing it too, therefore developing the ear for it and you also start spotting some foreign speakers doing it wrong... That's my experience.

Hence my serious doubt about the level of realism in those theoretical vowels ascribed to Latin.
Isn't that arrogant to publicly doubt something you're not an expert in? Why don't you first become an expert on the issue by reading the academical books focusing on the phonetics of the Ancient Italy, then become an expert in general phonetics and in how human brain deals with native and foreign phonologies when it comes to new differences and only then start publicly imparting your doubts regarding well agreed phonetics. I'm sorry, but I just find it as a kind of assault on what most of the scholars consider scientifically correct and fully viable given everything we know both about human phonology in general and about Ancient Italy from the diachronic comparative linguistics, what we know even of the modern phonetically similar languages.

And what surprised me (aside from the numerous Russian-sounding words, because I know a bit of Russian) was that I was looking for long and short vowels, yet was unable to make them out at all!
How can you use your own perspective to make a general judgement: "I can't make the Czech long vowels out, therefore they don't exist. (I exaggerate)" You know you can always open the audio editor and measure the times to see for yourself? I already told you we (the Czechs) are very incredibly sensitive for those differences, we hear any duration aberration that happens instantly no matter how rapid the speech is. I'm not sure I use my time wisely by trying to persuade you of that. I just advise you to study general phonetics (try the books by Ladefoged) to see what all is possible, to broaden your horizons.

when you have a long vowel in Czech, does it really take twice the time a short vowel does to pronounce?
Yes

Isn't it then a bit artificial to make long vowels in Latin last twice as long as short ones?
No
________________

I'm sorry if I come as impatient, but I start feeling we're moving in circles and that becomes quickly frustrating to me at least, I suppose. Or maybe I just deal badly with others' incredulity when I feel the evidence has been amply pointed out to them.
 
Godmy, I'm sorry to see how emotional you're getting about our conversation. That was completely unwarranted. Either because my English is not clear enough, or because you've read me too quickly, you have obviously misunderstood absolutely every single point in my message, judging from the way you have distorted those points when trying to reply.

I never suggested anything at all about Czech (you imposed that impression on my words and I honestly wonder how), I merely asked questions!

I know the human ear can hear minute details. I also know those minute details are often quite unstable in many languages. In French for example several phonemes have been disappearing because they were too similar to each other, and even the natives can hear the differences between them in isolated syllables, but not in the middle of a conversation, so I fail to see how my remark was arrogant and ridiculous.

And no, doubting something I'm not an expert in is not arrogance, it's called fact-checking and critical thinking. I said I doubted, I never said I flatly rejected anything, I even said « I'd love to be proven wrong ». But I shouldn't have to be quoting myself here. You should have read that part, clearly you haven't. I'm here to learn, otherwise I wouldn't be asking questions.

Yet another thing I never said was that long vowels in Czech did not exist just because I can't hear them. Again you have completely missed my point and it's a shame to see you berating me out of you own misunderstandings. If you had taken the time to read me properly, you would have spared yourself (and me) such negative thoughts.
Oh, btw. I'm not likely to respond to any future longer post/reaction in this thread any longer. (given my frequency and willingness to post in the forum, I suspect I've devoted enough time to this issue already.)
Rest assured you won't have to reply to any further question from me.
 
Besides Vox latina, I'd also suggest Palmer's The Latin Language as a good source concerning Vulgar Latin. I'm making this post, because, coincidentally enough, I'm reading a chapter on Vulgar Latin from that book at the moment.
I too would highly recommend Palmer, as he preſents in clear and very illuminating detail how the Latin language evolved over many centuries. I have always had the greateſt admiration for his ability to explain the evolution of Latin's morphology, and would certainly ſay that he is at his beſt there, but he alſo presents a good underſtanding of the evolution of her phonology.
 
Thank you. I was recently advised to read Weiss, Outline of the Historical and Comparative Grammar of Latin, which I've just got a copy of. Not sure which of all those books most concentrates on the issues I'm mainly concerned with so I don't know which one to start with. For now I'm finishing another book about “a history of the latin language” (in French) and I'll see where it gets me.
Thanks again for the recommendations.
 
Top