Robert Couchman discusses evidence for the pronunciation of Latin.

Evidence for the Pronunciation of Latin


As an ancient language and one no longer spoken today (except by the Catholic church, who have their own pronunciation separate from the Classical Latin spoken by the Romans), you might expect that we don’t know how it was pronounced. However, we actually have a lot of evidence from various sources (explicit statements from Greek and Roman grammarians, poetry and the modern Romance languages among others).

I’m just going to talk about a few letters that I think are interesting or particularly different from their modern pronunciations, not the entire language. The term “Classical Latin” generally refers to what was spoken from roughly 100 BC to 150 AD so that’s the time period I’ll be talking about.

But first, a word about the Latin used by the Church – the pronunciation of this was standardised by the king of what would become France, Charlemagne, in roughly 800 AD, but not long after that people just started pronouncing Latin words like they would in their native language. SInce about 1850, Italian-like pronunciation has been most commonly used. So for example “ci” is said as in “cheese and “gi” as in “jeans”.


The letter C was always pronounced as K. In 95 AD, the author Quintilian wrote the following:

“As for K my view is that it should not be used at all except in such words as may be indicated by the letter standing alone as an abbreviation. I mention the fact because some hold that K should be used whenever the next letter is an A, despite the existence of the letter C which maintains its force in conjunction with all the vowels.”

After the Romans had fully developed their alphabet, K was replaced by C in almost all words, except a few like kalends “the first day of a month”. Presumably these were the words Quintilian thought that K should be an abbreviation for. The latter part of the extract heavily implies that the rule that K should be used before A was arbitrary and that C and K were pronounced the same.

There’s also some evidence from Greek. In the Greek alphabet, there’s a letter for K, kappa -Κκ- and one for S, sigma -Σσ (ς at the end of a word). In ancient texts the names Caesar and Cicero are spelt Καίσαρ (Kaísar) and Κικέρωνος (Kikérōnos), which also show that C was pronounced as K.


V was pronounced as W. But before going into that, I should talk about the letters V and U themselves. In Classical Latin U didn’t exist and V could be used as either a vowel (long or short U) or as a consonant. It was only much later on, in the 16th Century, when U started to be used a vowel and V as a consonant. As late as that V was being used as a vowel in English texts. In Latin texts printed today, U is generally used anyway to make reading easier though (so servvs “slave” is usually spelt servus”. Anyway, back to pronunciation: the Roman politician Nigidius Figulus (98-45 BC) once remarked that when saying tu “you (singular)” and vos “you (plural)”, the speaker’s lips point towards the person they’re addressing. If vos was pronounced with a modern English V, this wouldn’t be true therefore it was pronounced as W.

Again, there’s some evidence from Greek. In Greek there was neither a V nor a W sound, but the name Valerius was transcribed as Ουαλεριος (Oualerios). This still doesn’t really get there but when ου was said very quickly it sounded vaguely like our W. From Greek we can tell when V changed to be pronounced more like it is today, as Ουαλεριος became Βαλεριος “Balerios”. It looks like the change started in the mid-1st Century AD but by 400 some people were still using the W pronunciation.

In a minority of Latin texts printed today, you’ll see some words with the letter J in them, for example janua “door” and cujus “whose”. The Romans didn’t actually have J – they only had I. J is used to show when I is used as a consonant as opposed to an ee sound as in qui “who”. It seems that the most important source we have for this is poetry, as by analysing the rhythm we can see that I sometimes has to be a consonant to fit the poem’s metre. Some evidence for what this consonant actually was comes from looking at modern Romance languages. In Spanish, J is pronounced as a throaty sound similar to ch in “Bach” and “loch”, and in French and Portuguese it’s the same sound as the S in English measure. In Italian there is no J apart from in words taken from other languages. Very rarely there is a consonantal I, for example in Savoia. Sometimes consonantal I in Latin has changed to gi (pronounced jee) in modern Italian, for example Iulius becoming Giulio. Through analysing how languages normally change and taking other factors into account, we can conclude that the consonant was probably similar to our Y sound.


There’s a lot of evidence to say that when a vowel was followed by a nasal consonant (N or M), the consonant wasn’t pronounced but did make the preceding vowel nasal. From analysing rhythms in poetry we know that when a word ended with a vowel and the next word started with one, the vowel at the end of the first word was ignored. So in in the line subligat atque ensem collo suspendit eburnum from Aeneid XI, atque ensem would be pronounced atqu’ensem. We also know that this rule also applies to words ending in a vowel and M, so in the line in et maestum Iliades crinem de more solutae, also from Aeneid XI, maestum Iliades would be maest’Iliades. If the M was pronounced there would be no reason to take away (elide) the -um. As well as this, in graffiti from Pompeii this M is dropped entirely, for example ancillam was once spelt as ancilla. We know that the M made a difference because it changed the grammatical case of a noun (eg. ancilla is the subject of a sentence whereas ancillam is the object). There’s evidence from N as well – for example in early texts consul was sometimes written as cosul. We can see this happening in modern French as well – in hon hon hon the N nasalises the short O sound and in interdit there’s a nasal vowel sounding more like our short A than anything else at the beginning. This evidence from French shows that the vowel being nasalised is the most likely change to happen to it.

Of course, we can never be absolutely certain how Classical Latin was pronounced since it’s been about 1500 years since it was spoken. This is all just our best guesses at what it sounded like.



Vox Latina by W. Sidney Allen via and Wikipedia

Quintilian’s Institutio Oratoria via

Evidence for the Pronunciation of Latin by Ralph L. Ward, accessed from JSTOR