I am currently reading Ørberg’s Lingua Latina Per Se Illustrata, where he thankfully makes use of the macron to distinguish long vowels form short ones. However, and I have seen this elsewhere as well, he writes videt without a macron on the i, which should mean that it is a short i, as in English it or bit. But all I ever hear when people say videt, is a long i as in English beer or sheer.

The same goes for the a in pater. Ørberg writes it without a macron, so it should be short. But that does sound very odd since I have only ever heard it with a long a. There are other examples like the i in villa or the o in vocat.

So are those vowels pronounced short in classical Latin, as their missing macron suggests, or are they pronounced long despite of their missing macrons? And if so, is there a general rule behind that?

  • I also normally hear ordinary people uttering pater with long a. And I am not mistaking quality for quantity, my native language has phonemic length and also a similar difference in quality of long and short i like Latin. The word was even borrowed as páter (and the prayer as páteř), with a long a, to my language, already in the middle ages. – Vladimir F Jan 19 at 9:12
  • @VladimirF What is your native language ("my native language has phonemic length")? – Alex B. Jan 20 at 15:50
  • @AlexB. It is Czech. But I think Hungarian also borrowed the word as páter with long á. – Vladimir F Jan 20 at 16:46

The i in videt is short. The length of a vowel in classical Latin pronunciation is defined by its duration—its "quantity"—as opposed to its "quality", i.e. the nature of the sound: its waveform or timbre. But first let's have a look at short-vowel quality, since that seems to be the focus of your question. Then we'll come back to rhythm and hopefully it will all make sense.

Phonetic pressure on short vowels

Sidney Allen, in Vox Latina, says "in the case of the close and mid vowels (i and u, e and o) the long appear to have been appreciably closer than the short." That is, the tongue is higher ("closer" to the roof of the mouth) in the long forms of these vowels. (See chapter 2 for details about the evidence for this.)

However, the quality of the long vowel was felt to be the primary "nature" of the vowel. If you asked a Roman to clearly make the sound of I, you would hear ī, not ĭ. The reason that Roman speakers gave the short vowel a distinct quality was not to distinguish the short vowel from the long vowel—the rhythm did that—but because of phonetic pressure. Since a short vowel is said very quickly, there is a natural tendency not to lift the tongue as high as when the vowel gets its "full" duration. This is why a is unaffected: the tongue is low ("open") in the long vowel. The change in vowel quality in the short forms of the other vowels is just a weaker kind of vowel reduction—the tendency, seen in English and sometimes in German, to pronounce short unstressed vowels as a schwa.

Since the change in vowel quality is due to phonetic pressure from the preceding and following phonemes and the short time for the vowel, we should expect that the quality doesn't change in all contexts. Indeed Sidney Allen reports that in diēs, the short i had the same quality as long ī. This is because before e, there is no phonetic pressure to alter the quality of ĭ.

From all this, I understand that when a vowel is short, the long and short qualities are heard as allophones—perceived as the same vowel. So, it's perfectly fine to use the long-vowel quality in a short vowel. In fact, I think it's wise for a beginner to do this. The long quality is the perceptual anchor, and beginners should start by burning in the perceptual anchors. You can make your pronunciation more lax when you've mastered it—but see below for why you might never want to.

Rhythm is fundamental

When you say pater et māter, the syllables should come out in this rhythm:

♪♪♩ ♩ ♩

The syllables are: pă-tĕ-rĕt-mā-tĕr. Ørberg will explain this, but not until chapter 34. It's worth knowing right when you begin, because mastering the rhythm of a language early makes it easier to learn.

Similarly, Mārcus nōn videt Quīntum should have this rhythm, regardless of the quality that you give the short i's:

♩ ♩ ♩ ♪♩ ♩ ♩

That might be a good one to practice, because the only short syllable is vĭ-.

Quis mē vocat? has this rhythm:

♩ ♩ ♪♩

Ørberg will hit you with some minimal pairs distinguished only by vowel length, too, like malum (bad, evil) and mālum (apple).

When you're speaking in the classical Latin rhythm, you will feel the phonetic pressure to pronounce ĕ, ĭ, ŏ, ŭ in a slightly lax manner. Then you will probably give in to that pressure without intending to. You don't need to do it on purpose.

Reconstructed Classical Pronunciation ≠ ancient Roman pronunciation

If you read Vox Latina or any scholarly treatment of ancient Roman speech, you will find all sorts of ways that pronunciation varied from spelling. The m in final -am, -em, -im, -om, -um was not pronounced, and the vowel was nasalized; a final vowel was often dropped when the next word started with a vowel, so Iūlia Aemiliam vocat could be pronounced Iūli' Aemiliam vocat—this is called "elision"; censor was actually pronounced cẽsor (with a nasalized e); fīliī was pronounced fīlī, at least in some centuries; Marcus and many others dropped the h in homō; calidus was pronounced caldus; and many more. But of course not everyone pronounced Latin that way; in a cosmopolitan empire of 60,000,000 people spread from Portus Cale to Ægyptus, lasting 1,200 years, there are going to be some variations in pronunciation—many of which were never documented.

Unless you are particularly interested in reproducing Roman colloquial speech, I don't think you need to worry about any of that if you're learning Reconstructed Classical Pronunciation. Reconstructed Classical Pronunciation is a modern pronunciation of Latin, just like the Italian pronunciation of Latin, the French pronunciation of Latin, the Slavic pronunciation of Latin, and all the rest. RCP is a somewhat artificial pronunciation, just as classical Latin was a somewhat artificial language. Unlike the others, RCP approximates the rhythm of classical Latin, as well as its segmental phonemes—passably well. Since the year 800, Latin has been defined by its writing, with various local standards for pronunciation, which enabled people to understand each other's speech—passably well. RCP is just one more post-800 pronunciation standard, but it's worldwide rather than local. Many people speak RCP and don't even bother with vowel quantity.

So, I conclude that it's OK to make your speech supremely clear by always using the long vowel quality even in short vowels. Catullus might mock you for it, but Catullus is long gone. If you master the rhythm, though, you'll be able to perceive the poetry. Googling for a videt in the Aeneid turned this up (Book I, line 128):

Disiectam Aeneae toto videt aequore classem.

He sees Aeneas's fleet scattered all over the sea.

which scans like this—requiring vĭ- to be short:

Dis-iec/t⁐Ae-nē/ae tō/tō vi-de/tae-quo-re / clas-sem.

If you don't follow that, wait until you finish Chapter 34.

  • 1
    That was a great answer. Since I'm not a linguist and study languages as a hobby, I must ask if by 'vowel quality' you mean the property that is captured in the vowel charts one sees for languages? Regarding what you wrote about rhythm, does that mean that the way sentences on the YouTube channel 'Latin tutorial' are read is how Romans actually spoke to one another on the streets? I am asking since I'm interested in how they spoke – Thomas Wening Jan 19 at 20:43
  • 3
    @Thomas Yes, the quality of the vowel is essentially what you’ll see in vowel charts, i.e., the difference between /a/ and /e/ is one of quality. The other, quantity, is the length, which is the difference between /a/ and /aː/ (= ā). Note that the lax/tense difference Ben mentions as being secondary to length in Latin is also present – but secondary – in German (which I’m guessing is your native language). The difference between bitte and biete is primarily the length of the first vowel, /i/ versus /iː/, but phonetically, as in English, short /i/ is actually lax /ɪ/. – Janus Bahs Jacquet Jan 19 at 21:25
  • 2
    @ThomasWening Yes, the vowel quality is what's noted in vowel charts, as Janus said. I'll edit the answer to clarify that the vowel quality is the waveform or timbre of the sound. – Ben Kovitz Jan 19 at 21:42
  • 2
    @ThomasWening How the Romans actually spoke on the street probably merits a separate question. A classics professor told me that based on comparative linguistics, we know that whatever language the Romance languages came from, it was not classical Latin. It was vulgar Latin—not what is taught in Ørberg, Wheelock, etc. I understand the difference to be like that between the formal English or German taught in schools and the vernacular picked up at home. I think someone with more expertise must give you a better answer about that. – Ben Kovitz Jan 19 at 21:52
  • 2
    @CannedMan -tĕr is a long syllable (also called "heavy") because it's a closed syllable (closed by the constant r). If the next word in the sentence begins with a vowel, though, it would indeed steal the r, leaving tĕ- a short syllable. – Ben Kovitz Jan 20 at 22:13

There are no long vowels in videt, pater, vocat in classical Latin. The vowel in the first syllable of vīlla is long.

For sources like Ørberg, the general rule is very simple: macrons mark long vowels, all vowels without macrons are to be understood as short. There are occasional uncertainties or potential errors in Ørberg's use of macrons, but all of the words you mentioned as examples have fairly certain vowel lengths.

In some other sources, a macron or the absence of one is not a reliable guide to the length of a vowel. Be particularly careful with vowels that are followed by two or more consonant sounds (such vowels are often said to have "hidden quantity"). This includes vowels followed by consonant clusters and double consonants, including Z and J which by a quirk of Roman spelling were written with single letters but pronounced as double or long consonant sounds when they occurred in the middle of a word between vowels. Another general rule might be that if you see that a source uses breves (ă ĕ ĭ ŏ ǔ) as well as macrons (ā ē ī ō ū) it will not be safe to assume that unmarked vowels are short. The differences in the usage of the macron arise because Latin had a distinction between heavy and short syllables, important in poetry and for the placement of stress, that was related to but not the same as the distinction between long and short vowels.

How classical Latin vowels sounded

If you give more detail about the context in which you're hearing these words, it might be possible to comment further. Learning a dead language like Classical Latin means that you can't listen to native speakers, so anything you hear will be either just an non-native speaker's attempt at approximating classical pronunciation, or a pronunciation that does not even have that as a goal (there are multiple approaches to pronouncing Latin in present times, most of which do not have the same distinction between long and short vowels as Classical Latin).

Giving English vowel equivalents is a very imprecise way of describing the reconstructed sounds of Classical Latin. There is widespread consensus that Classical Latin distinguished vowels by means of literal phonetic duration: a short vowel is held for a shorter period of time than a long vowel. In many descendants of Latin, this durational contrast was lost, but a contrast in vowel quality exists for certain vowel pairs formerly distinguished by length. An example of how the i vs. ī contrast shows up as a contrast in vowel quality in Italian is vede [veːde] from Latin videt vs. ride [riːde] from Latin rīdet.

Based on this contrast in vowel quality in languages such as Italian, the testimony of some ancient sources that mention differences in the quality of short and long vowels, and inscriptional examples of Latin words containing short i being spelled with the letter E (or in Greek, with the letter ε), it is fairly common to reconstruct the pronunciation of Classical Latin short i in most contexts as [ɪ], a front vowel that is not quite as "high" or "close" as the corresponding long vowel ī [iː]. Another possible quality, barely distinct from [ɪ], is [e], as in modern Italian, but with a short rather than long duration (Michele Loporcaro, The Cambridge History of the Romance Languages: Volume 1, Structures, "Phonological processes", page 112).

The quality of short /a/ is typically reconstructed as [a], an open or low vowel, with the only difference from long /aː/ being duration.

The quality of short /o/ is typically reconstructed as [ɔ].

  • 2
    What is the difference between /. / and [.]? Does the first denote the length and the latter the quality? – Thomas Wening Jan 19 at 20:44
  • 3
    @Thomas /Slashes/ indicate phonemic writing, which shows the sort of ‘abstract’ inventory item that we mentally associate a sound with when speaking and listening; [brackets] indicate phonetic writing, which shows the actual quality of the sounds coming out of our mouths, regardless of which item in our inventory they represent. For example, in German, there is a phoneme /r/ which is generally pronounced [ʁ] or [ʀ] at the beginning of a syllable, but after a vowel, it’s pronounced like a very short [ɐ̯] or [ə̯] vowel: Ruhr is /ruːr/ phonemically, but [ʀuːɐ̯] phonetically. – Janus Bahs Jacquet Jan 19 at 21:31
  • 2
    @ThomasWening Simply put: If in slashes, symbols are the closest approximation to how a native speaker would pronounce them; if in brackets, it is an exact representation. In Norwegian dialects, there are six to eight different ways to pronounce the letter R (excluding the retroflex flap), if you include unvoiced pronunciation, so /r/ could in Norwegian be: [ɾ, ɾ̥; r, r̥; χ, ʁ; ʒ, ʃ].¹ All these will be showed as /r/ in phonemic writing. ¹ The pairs are standard (east, Trøndelag and north west (and some more)), standard emphatic, south and south west, and many areas of Northern Norway. – Canned Man Jan 20 at 22:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.