Did the Romans have a theory for the origin of their language? I assume there were several ideas, and it would be great to see a summary of them. No need to go very deep on any individual theory; a reference and a brief description of each would do. For example, did some grammarians or other scholars think that Latin descends from Greek, or that Latin was given to the Romans by the gods, or that there was a global language of some kind which later split to several different ones, or something else?

All Roman thoughts on the matter are welcome as answers, no matter how short or simple. I don't know what kinds of theories they had or if any single one was more prominent than others, so I can't make the question much more specific.

  • 1
    Great question! Lucretius has an elaborate account of how man may have acquired the skill of language naturally, somewhere in the second half of DRN V, probably inspired by Epicurus as usual. In general, the there was debate among the Greeks (and hence the Romans) as to whether language was 'invented' nomôi or phusei (by convention or naturally). As to Latin specifically, I am unaware of any specific classification or account of the position of the Latin language in the development and branching of historical phonology.
    – Cerberus
    Sep 17, 2017 at 22:54
  • 5
    Related: linguistics.stackexchange.com/questions/6675/…
    – TKR
    Sep 18, 2017 at 3:52
  • Well, Stevens 2006/7 seems to be the most comprehensive account on this - thanks to Nick Nicholas for drawing our attention to this paper.
    – Alex B.
    Nov 24, 2018 at 13:55
  • 2
    @AlexB. It would be good if someone could post a summary, even a short one, of that as an answer. The answers on Linguistics do offer useful insight, but the question has a different focus.
    – Joonas Ilmavirta
    Nov 24, 2018 at 19:46
  • 1
    I've added the three sources I could find (thanks Cerberus for the mention of Lucretius!) as separate answers, so they can be upvoted or downvoted separately. If you have other sources, please do feel free to do the same!
    – Draconis
    Apr 26, 2019 at 1:59

4 Answers 4


For one last answer, it seems another author has quoted a source no less Roman than Cato the Elder himself (though Cato's original doesn't survive):

ὁ Ῥωμύλος, ἢ οἱ κατὰ αὐτόν, δείκνυται κατ' ἑκεῖνο καιροῦ τὴν Ἑλλάδα φωνήν, τὴν Αἱολίδα λέγω…Εὐάνδρου καὶ τῶν ἄλλων Ἀρκάδων εἰς Ἰταλίαν ἐλθόντων ποτὲ καὶ τὴν Αἱολίδα τοῖς βαρβάροις ἐνσπειράντων φωνήν.

At that moment Romulus, or his retinue, showed off the Greek language—I mean, the Aeolian dialect…because of Evander and the other Arcadians, who went to Italy at one point, and scattered the seeds of Aeolian among the barbarians there.

(Translation mine.)

I haven't been able to locate the original text, unfortunately, but this seems like a pretty good indication that the quintessential Roman thought that his language did indeed descend from Aeolian.

(Source: Benjamin Stevens, Aeolism: Latin as a Dialect of Greek, 2006)


Dionysius of Halicarnassus, in his Ῥωμαϊκὴ Ἀρχαιολογία (Roman Antiquities) 1.90, explains that Latin was actually a dialect of Greek, corrupted by contact with European barbarians:

Ῥωμαῖοι δὲ φωνὴν μὲν οὔτ᾿ ἄκρως βάρβαρον οὔτ᾿ ἀπηρτισμένως Ἑλλάδα φθέγγονται, μικτὴν δέ τινα ἐξ ἀμφοῖν, ἧς ἐστιν ἡ πλείων Αἰολίς, τοῦτο μόνον ἀπολαύσαντες ἐκ τῶν πολλῶν ἐπιμιξιῶν, τὸ μὴ πᾶσι τοῖς φθόγγοις ὀρθοεπεῖν…

The language the Romans speak isn't totally barbarous, but also isn't completely Greek. No, it's a mixture made with something of each; the majority of it is Aeolian, and the only "benefit" they get from being so promiscuous with other [languages] is that they don't pronounce all of their sounds right any more…

(Translation mine.)

Jeffrey Henderson, in his notes on LCL 319, says that there were probably three reasons for this:

  • The Greek dialects diverged a lot, so it's reasonable that Latin could be a very very divergent dialect
  • Latin had the letter F, derived from the Greek waw/digamma Ϝ, which eventually died out in all Greek dialects except Aeolian (though the Aeolians pronounced it /w/ and the Romans pronounced it /f/)
  • Aeolian and Latin both kept long /aː/ in words like μᾱτηρ/māter, while Attic and Ionian shifted it to long /ɛː/ (μητηρ)

I'll add one more to this list:

  • A big part of Dionysius's goal in writing the RA was to convince the newly-conquered Greeks that Rome was properly civilized, so drawing a contrast between the "Greek" Latin and the "non-Greek" barbarian languages would help with that

For yet another account, more pure-Roman than Dionysius of Halicarnassus, Quintilian (1.6) seems to agree that Latin is close to Aeolian:

Etymologia, quae verborum originem inquirit, a Cicerone dicta est notatio, quia nomen eius apud Aristotelen invenitur "σύμβολον", quod est "nota". […] Continet autem in se multam eruditionem, sive ex Graecis orta tractemus, quae sunt plurima praecipueque Aeolica ratione, cui est sermo noster simillimus, declinata, sive…

Etymology, which investigates the origin of words, Cicero calls "notation", because the word "symbol" was found in Aristotle, which means "note". […] See, it involves quite a lot of learning, whether we're facing down [words] from Greek—which are all over the place, especially the ones derived from the Aeolian style, which is most similar to our own language, or…

Quintilian also mentions the Aeolian letter digamma (Ϝ, pronounced /w/) several times, for example in 1.4:

…in his "servus" et "vulgus" Aeolicum digammon desideratur…

…in these words, "servus" and "vulgus", we long for the Aeolian digamma…

And 1.7:

…neutro sane modo vox quam sentimus efficitur, nec inutiliter Claudius Aeolicam illam ad hos usus litteram adiecerat.

Neither version [seruos or seruus, for modern servus] properly makes the sound that we hear. Claudius did a good deed when he adopted the Aeolian letter for these purposes.

(All translations mine.)

But in all these other instances, he's complaining that Latin doesn't have digamma for /w/: he's definitely not conflating it with Latin F, and might not have seen the connection there at all. There were other correspondences between the Latin and Greek alphabets that were seen as pure coincidence, like Latin P /p/ with Greek Ρ /r/, and F/Ϝ might have been just another one of those.


Lucretius argued in Dē Rerum Nātūrā book 5 (lines 1041-1045) that language was innate; he had no explanation for why Greeks and Romans spoke differently.

proinde putare aliquem tum nomina distribuisse
rebus et inde homines didicisse vocabula prima,
desiperest. nam cur hic posset cuncta notare
vocibus et varios sonitus emittere linguae,
tempore eodem alii facere id non quisse putentur?

So to think that one human first gave out names to things, and that humans then learned their first words from them, is downright idiotic. Because, why should this one hypothetical person be able to name everything with words, and make the various sounds of the language, when you think other people at the same time couldn't do the same thing?

(Translation mine.)

He goes on to explain that different words are equivalent to the different noises animals make by instinct.

  • By the way, if anyone knows where I can find an electronic copy of DRN with line numbers, it would be much appreciated! I'm using TheLatinLibrary which doesn't use any finer division than books.
    – Draconis
    Apr 26, 2019 at 1:04
  • How's this?
    – Joonas Ilmavirta
    Apr 26, 2019 at 1:17
  • @JoonasIlmavirta Perfect! Thank you!
    – Draconis
    Apr 26, 2019 at 1:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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