Nowadays, most Latinists learn the "reconstructed classical" pronunciation: an attempt at reconstructing the way Cicero, Caesar, or Vergil might have spoken in formal settings.

However, it seems fairly clear that this wasn't how an everyday Roman would have spoken in the forum. For example, Catullus 84 wouldn't make sense if /h/ was still pronounced in everyday vulgar speech. And later sound changes in Romance, like the loss of vowel length, probably didn't appear out of nowhere at the end of the Empire.

So: what do we know about how an everyday pleb would have spoken on the streets of Rome, somewhere around the first century? This question is unanswered, which I take to mean we don't know enough to make a full recording of it, but scholars certainly must know something about this topic.


It turns out, we know quite a bit about this!

There are three main sources for Vulgar Latin pronunciations: Classical texts imitating (or mocking or correcting) Vulgar speech, graffiti from actual plebs, and reconstruction from the Romance languages.

For the first of those, we have bits by Petronius and Catullus (the Cena Trimalchionis and Carmen 84), as well as the third-century Appendix Probī. And for the second, we have well-preserved writings from first-century Pompeii. These give hints at how Vulgar Latin was pronounced (by showing how it was spelled), and Romance reconstructions can fill in some of the gaps.

Here's a short summary of what I've found:

  • H disappeared completely: harenaarena "sand"
  • Qv merged into c before front vowels: coqvenscocens
  • Unstressed vowels in medial syllables (not the start or end of a word) disappear: spéculumspéclum "mirror"
  • Unstressed e and i turn into j before vowels: víneavínja "vineyard"
  • V started being pronounced /v/ instead of /w/; later, b between vowels merged with it. (They both probably started as [β], but there's no way to know when that changed.)
  • Nasals at the end of words and before s became nasalized vowels, then disappeared: consulcõsulcosul "consul"
  • Consonants at the ends of words disappeared in some regions: graffiti in Pompeii has valeatvalja "may he flourish", but modern French still shows remnants of final ts, so this wasn't universal
  • Vowel length was slowly turning into vowel quality: ī, ū, and o remained intact, while the pairs ō/u, ē/i/oe, ā/a, and e/ae merged into something like ō ē a e
  • Diminutives were all over the place: faxfacla "torch", neptisnepticla "granddaughter", ānusanucla "old woman"
  • Regular first/second-declension forms spread to other words: glis/glirisglirus/glirī "dormouse", pauper mulierpaupera mulier "poor woman"
  • "Qv merged into c before front vowels: coqvens → cocens" But aren't there other words that show different reflexes of c and qu before front vowels? E.g. the question/relative words (French qui and que are not pronounced like ci and ce), Spanish quedar, and French coi. Given that coquo has back rounded vowels after the qu in the present tense in the first-person singular and third-person plural, and that the qu was always delabialized to c in the supine and perfect stems because of the following consonant, paradigmatic leveling seems like a potential explanation of cocens. – Asteroides Mar 22 '20 at 7:08

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.