PAGINA1    

Classical Latin

Before delving into the Italian language, I thought it proper to introduce it with a couple of notes on classical latin. These have nothing to do with conventional grammars - the rules of Latin diction and metre have been expounded ad nauseam by far more experienced authors are readily available. But I found that getting anything that explained the phonetics of Classical Latin was less easy to find even on paper. The one usually included in moderm grammars was designed for the convenience of the student of Latin in the middle ages - and it should only be regarded in this light.

This is also the pronunciation adopted by ecclesiastical/medieval latin,  and is modeled on the Italian phonetics. It is the pronunciation in use in the Vatican state, where Latin is still the official language. It is the one we hear at mass, in the rare hymns still sung in Latin. Many Latin teachers still use it in class, although a few corageous ones have popped up the questions recently and are teaching their pupils the actual pronunciation.

When faced with the task of explaining how Italian evolved from its parent tongue, even phonetically, I wondered what sense it would make to ignore the subject, especially since we are exploring the Italian language diachronically and we are to see how it evolved from its parent tongue. However, it is wise to keep in mind that this is a standard, and that even at Virgil's times there were vernacular variants in use by soldiers and the lower classes which allowed for a relative freedom within these rules, as happens for our modern standards today, and that even the great authors such as Cicero and especially Augusts, said to have a passion for vulgarisms, must have indulged in them wherever their official duties did not bound them to orthodoxy.

Classical pronunciation had been forgotten by the early middle ages: as vulgar latin arose, the word Caesar was pronounced as "sezàr" (Fr.) or chézar (It) instead of the original "kàesar", pisces was pronounced "piséys" (fr.) or "pisheys" (it) instead of original lat. "piskeys"; "qui" sounded like "key" in France and Spain instead of Lat. and it. "quee".  

In other words, the drive toward romance languages on the one hand and the destruction or loss of many documents during the years of political upheaval following the fall of the Roman Empire combined to disperse the pieces of the puzzle, which only the research of Renaissance philologists and modern comparative studies on romance languages in the 1800s could put together again. Comparative linguistics also allowed for the reconstruction of Protoindoeuropean, the mother of modern European languages, when no written trace of it remains and is not likely to emerge given the pre-historic age this language dates to (about 10,000 years A.D.)  

Giving a pronunciatio restituta of Classical Latin was, compared to that of our ancestral language, relatively easy, since scholars have not only based themselves on French, Italian, Romanian, Spanish, Portuguese etc but on dialects - which are known to be much more conservative because their use is local and sometimes spoken by isolated communities or ethnicities. As such they are under no pressure to change and evolve as, on the contrary, in the case of languages spoken by millions, exposed as they are to relentless political as well as technological innovation - we know how rapidly new terms are coined, at such step that not even a hardback dictionary can keep pace with it - especially since our standards have been floating across search engines, where neologisms spring up to life every day and more words die to make room for them.

Looking for lost sounds 

By comparing the phonetic traits of words we have come close to reproducing the original sounds of Classical Latin and have good clues to understanding that spoken in Republican times. Though only a tape recording could have rendered the exact intonation of the spoken word, it is safe to say that today's reconstructed pronunciation (pronunciatio restituta) would enable a hypothetical time traveler to communicate effectively with an ancient Roman. 

The technique of comparative linguistics is basically the same as that employed by forensic scientists to establish if two people are siblings: finding the genes shared by two or more individuals so that it is also possible to know the identity of the father or mother. Once applied to the offspring of latin, this system gave us surprising results. 

A telling example, though it also takes advantage of languages outside romance languages, is the word Caesar, when seen under different perspectives (its Latin spelling, "Caesar", "Cesare" in Italian, "kaiser" in German and czar in Russian)

Two modern languages out of three have a K sound; all of them bear the stress on the first syllable, and Kaiser has a diphthong that looks suspiciously Latin (ae>ai). The reconstructed pronunciation of Caesar must therefore be / 'kar-ey-sar/. Using the same technique we have learned that C is always velar, or hard /k/, even in words like Lat. scio 'I know', from which we have scientia, "knowledge") or scena, "scene". 

Similarly, G is never palatalised, so its sound is that of en. get. We also know that T is never palatalised in standard forms, so we should never say ratio with a 'sh' in Latin, but with a T similar to that of en. tin. In Classical Latin, as in Italian, no two vowels combine into one sound (as in 'deep'), and no one vowel can  be read as two (as in 'fate', each has its own separate sound which is the same in all position: what applies to æ (ar-ey) also applies to œ /or-ey/, au (cfr. 'ow' of owl), eu (approx. e of 'bed' + u of 'put'), etc. The value of the single vowels in each diphtongs is close to that of those found in today's Italian or Spanish. 

Q

QU- words like quod (that), quid (what?), qualis (which) never omit the u, as happens with Spanish or French. The correct sound is /KW/ as in en. queen or it. quello.

Exotic letters


In Latin, Y and K were only used for exotic names. The letters J and W did not exist, their semivocalic sounds / j / and / w / are always written as I and V respectively.

The H

H is always aspirated as in English or German in Classical latin: hora, heros, historia, although we know that non-standard, informal or vernacular use often weakened it considerably as in some English dialects.
Though such aspiration is weaker in between syllables (as in nihil) it must never be omitted.

K, X, Y are mainly of Greek origin, the Y retaining the value of Gk. y (similar to the u of fr. une, lune). Actually in preclassical times there was no distinction between C and G sounds and C to indicate both was used for a much longer time. The absence of semivocalic J, however, does not imply the absence of the / j / sound.

The letter for this sound was the I, which was pronounced differently according to its position in the sentence: in fact the 'i' of Lat. iustum is that of you PH is of Greek origins, but do not be tempted to read it the way it is usually pronounced today and that some careless teacher has taught you at school: just say the P and then the H, because PH was no F at those times.

This sound is at least 10,000 years old and comes from the Proto-indoeuropean language, the ancestral language coming from central and western asia of which abundant traces remain in Sanscrit and Persian. V and U Never, ever read V as in novels ! the Romans did not know the this modern sound. The Roman character either stands for U or W. It sounds /W/ (as in war) at the beginning of a word but not between consonants, where it is /oo/ (short, as in put). Compare the first with the second V of vervm /weroom/.

V or U ?

Before an I or any other vowel, however, V sounded as W: qui /kwey/, quaedam /kwar-ey-dam/, cuidem /kwi-dem/. There would be no way to render the pronunciation of en. viable using the Latin alphabet. Any Roman would mistakenly have pronounced it as ' wiable' as still happens in a number of dialects in central and southern Italy (Southern Latium, many parts of Calabria and Sicily...) where speakers tend to say woglio, wero for it. voglio (I want), vero (true) etc. because the absence of V was inherited by their own phonetic system.

The graphic symbol U is therefore a graphic convention, introduced much later to enable readers to tell the W/U sounds in vulgar Latin and in romance languages from the modern 'v' sound. Once this convention was widely recognised, it was also used for transcribing works of classical authors, hence the mistake in pronouncing the V in classical texts.

V

Our reading of V (as in 'live' or 'vast') is a modern outcome. In  Latin, the V sign defined either en. W (as in 'woman') or U (as in 'put') according to its place in the sentence. So, a Roman scribe unaware of english would spell 'woman' and 'put' as voman and pvt.
 
GL 

Because of the neighboring L, hard G has changed too in Italian and many romance languages in certain positions, especially when followed by a palatal consonant or vowel (see G and C). The Italians read GL as Sp. LL of caballo when GL is followed by an I (for eg. in figlio). Otherwise, the Italian pronunciation is like that of classical Latin: G+L (two distinct consonants) in all positions: glis, gloria, etc. The latter is also the pronunciation of classical Latin in all positions.  One Italian exception is geroglifico (hyeroglyph).

GN

Caesar may have pronounced GN in en. 'gnome' and 'magnificent' corectly.  In latin 1. you must omit the G at the beginning of a GN- word. 2. Inside a -GN- word, however Ceasar might hesitate between two options: GN of magnum as in en. 'ma-G-N-ificent' or maNNum (geminated, as in Italian), perhaps with a preference for the NN, easier to pronounce. 

The Italian language doubles the value of all its consonants, but only when it is reflected in the spelling (gatto, pacco etc). Interestingly, the pronunciation of Sardinian words in GN is ideantical to that of classical latin. Twin consonants are quite frequent in Italian as a result of an assimilation between neighboring consonants as the one already used in Lat. magnum: Lat. victoria > vittoria, actum > atto and so on.

In fact, most of the changes which take effect during the middle ages are sometimes seen in nonstandard variants (and errors) by uneducated speakers in early classical times, but it takes centuries before they actually upset the equilibrium of Latin sound system.  During the empire, the stabilizing factors are the schools and the power (and prestige) of Rome. But when the schools are destroyed by wars and Rome loses its grip on its provinces, unallowed variants emerge rather abruptly and standard Latin is eventually forgotten. 

PS group

Even here, it is important to look at where the PS is placed, whether at the beginning or not. Native English speakers will find it natural to use the PS group correctly, since P is never sounded at the beginniong of any word. En. psychology, psalm use the same rule as in Lat. psaltria or Psȳchē. In all other positions both P and S are heard. 

PH and TH

Words in consonant +H are usually Greek loans. Although we tend to read them as 'F' and en. 'Th' the sounds are not merged in classical Latin. P does not change, and the H  that floows is aspirated. All other -H combinations follow this rule.

SC group

Do not be tempted to palatalize words like scena or scientia the way you do with 'scene' or 'science'!  Do not take anything for granted: SK is always the correct pronunciation. As we said, palatilisation is typical of romance languages, has passed from French into English but it does not belong in Classical Latin.  The Italians love sciare ('shee-uh-rey'), but what would they say if the snow is scarsa ('skar-sa), that there little of that?

C and G

The sound shift of romance languages has occured because E and I tend to change velar or 'hard' into palatal consonants . In earlier times Lat. C and G would be always be velar but had actually a unique K sound in early Classical Latin (the fact that GN or CN abbreviation for 'Cnaeus' on tombstones is used interchangeably seems to confirm the hypothesis). Later, G and C became two distinct, but still velar, consonants as in en. cat and get. The sound of C and G would not be affected by  -e and -i (Sardinian reflects classical latin in this respect: cĕntum sounds 'kentum', gĕlum 'ghelum').

There were no such Cs or Gs as can be found in it.  cinema ('chinema') or genio (read as 'genius') today. Similarly, T and D are not palatilised before an I as in Ecclesiastical Latin: in words like pronunciatio use the dental T as in Sp./ it. toro, it is not alveolar T as in England).

N and M

Words or syllables ending in -m and -n nasalised the preceding phonemes, as in referendum. However, changes were only nasalised, and vowels did not shift in tone.

If -M words are followed by a syllable or another word whose consonant has a different point of articulation the M may sound like an N: Cum grano salis > 'cungranosàlis' and cum libro > 'cunlìbro'. Just to be on the safe side, remember: that the stable consonant clusters are N + K/G/T/D/F/V) and M +P/B/.  M + G, K etc are always at odds. 

M will tend to sound like the N of junk, while your N +B will take the sound of lamb. Liquids Ls did not differ much from today's sound, though some variants might occur whether the following vowel was front or back.

R

More remarkably, the R might be uvular (as in French or German), alveolar (closer to the Italian) or even a trill like in Spanish. It was burred and never ever 'rounded' as that of British or U.S. English. These sounds have not changed much in Italian.

B & D variants

In some cases, B and D sounded as P and T. This happened by assimilation of B and D (voiced stops) to some following voiceless consonants, as T or S:  BS of obsoletus sounded like 'opsoletus', 'adpatrus' like 'atpatrus' (and then appatrus).

Different words may also influence one another because of their proximity, as in ad patrem > 'appàtrem', sub die > 'suddìe'.

The vowels

Phonological and therefore semantic difference was based on a short/long vowel contrast: mălum, (it.male, "evil, bad") vs. mālum, (it. mela, "apple"); lătus (it. lato, "line") vs. lātus (it. esteso, "wide") and so on. Like Greek, to a good extent, English and many more Germanic languages. Latin was based on quantity rather than quality, because it used a flexive system (its six cases were differentiated by vowel length). Romā (long A) means 'in Rome', but Romă (short A) is vocative (Oh, Rome!).

Therefore, the round arch signifies that a vowel is short and open, while the long line stands for long and closed. To get an idea compare en. ship (with an ĭ ) and sheep (with an ī ). The Italian has scrapped the flexive system since it was no longer needed when the cases were replaced by preposition + noun: a Roma, per Roma, con Roma do not need short or long As because the case 'markers' are now expressed by prepositions. 

The long/short vowel contrast implies that the syntactic and semantic function of a given word is not given by the its place in the sentence. Eg. Caesar Romā est (Ceasar is in Rome), Caesar est Romā  are equivalent, and in more complex sentences words occurs in a virtually infinite combination, allowing for a great stylistic freedom.

AE - OE -AU and other diphthongs


Diphthongs are classified as long vowels but  are made up in fact of distinct sounds: AE = a (uh) + e ( of 'bed'), OE = or + e (same 'e').  The same goes for AU (ar + 'u' of should, put), EU (approx. e of bed + u of put. This applies to all other diththongs whose vowels are close in tone to Italian or Spanish ones. Merged diphthongs is post-classical and belongs to the middle ages, when it is not difficult to find mispellings where one graphic symbol is used instead of two.

Where the stress falls

But, where does the stress fall? If the penultimate syllable contains a long vowel, the stress falls there: 'cŭ - pī - do (I desire), but if the penultimate syllable contains a short one, then the stress falls on the antepenultimate: cŭ - pĭ -dō ('greedy', dative of cŭ - pĭ -dŭs). In any case, the stress can never fall before the antepenultimate or on the last syllable.

It falls on the penultimate if the syllable is followed by two or more consonants (in this case we say it is long by position), as in vĭ-gĭn-tĭ ('twenty'). The case where the stress seemingly falls on the last (illŭc, ĭllĭc, adhŭc) is only apparent since the c is a 'leftover' of an old syllable (arch. il-l-ce, illūcĕ), so the same old rule applies here too.

NEXT>>