The trouble with nominatives

Quite early on in the writing of my book, Italy Before Rome: A Sourcebook (which now has a cover image!), I decided that I wanted to avoid Romanising or Hellenising the names of people, places and gods if I could help it. This seemed like a simple enough decision – if someone called himself Lúvkis Veleíis, why would I change it to Lucius Velius in my translation? Habitually Romanising all the names always seemed to me to erase some of the variety of names and naming practices across Italy, and to give a false impression of how ‘Roman’ people in the third and second centuries really were.

This all seemed straighforward enough. But I hadn’t counted on a particular problem – not all words are attested in the nominative in Oscan.

For people’s names, this is not too problematic. For almost all names, we can guess the correct nominative ending from the case ending given (usually genitive) because most names follow the same pattern. In Oscan, for example, men’s names generally end in -is or -iís. A few irregular praenomina are a bit more difficult, but usually they are common enough to have been attested in the nominative somewhere.

However, the names of deities are much more complicated. I’ll show you what I mean with one example: the hero and/or deity known as Herakles (in Greek) and Hercules (in Latin). He was a very popular deity all over Italy, and he is mentioned fairly often in inscriptions; he also appears on a lot of pots made or purchased in Italy (as in the picture on this page).

In Oscan, this name is attested in the dative as hereklúí (meaning ‘for Herakles’) and herekleís (meaning ‘of Herakles, belonging to Herakles’). But it is not attested in the nominative anywhere. This is mostly an accident of what kind of writing survives: we have dedications to Herakles, but nothing that labels him. (The situation is different in Etruscan, where characters are often labelled on mirrors and pots depicting various myths – so we know that in Etruscan he was called hercle.)

Attic black-figure amphora, found in Vulci, depicting Heracles slaying the Stymphalian birds with sling, c. 540 BCE. British Museum.

So I had to make a choice. If I was translating an inscription that included one of these words, should I Hellenise it to Herakles? Or Romanise it to Hercules? Or should I make up an Oscan nominative?

Oscan is a particularly annoying language to try to make up nominatives for, because of the way it adds vowels (anaptyxis) and drops vowels (vowel sycope) in various positions in the word. This means, for example, that the word ‘garden’ appears as húrtúm (pronounced ‘hortom’) in the accusative and húrtúí (pronounced ‘hortoi’) in the dative, but as húrz (pronounced ‘horts’) in the nominative, because the vowel before the final -s is dropped.

By the same principle, if we were to reconstruct the nominative of Oscan Herakles it should come out something like *herekls. But that final consonant cluster doesn’t work in Oscan either. I consulted my colleague Nick Zair, and he suggested something like Herekel or Herekell as a plausible Oscan nominative for Herakles. But then that creates another problem. Does Herekell look recognisable to a reader as a version of Herakles? Or is it making the translation overly difficult to follow?

At least with famous deities like Herakles, I could compromise slightly and use a better-known version, on the basis that that would be more helpful to a reader. This is what I decided to do in the end, using the Greek Herakles. But what should I do with local deities without a well-known equivalent, like the dative vezkeí in the Agnone tablet? Making up a nominative Vezks seems like the only way forward, but it feels a bit strange to make up a nominative that isn’t attested anywhere.

In the end, the names in my book have ended up as a bit of a mixture of recognisable forms, attested forms, and reconstructed forms, depending on the situation. It’s probably not the most elegant or consistent solution, but sometimes a compromise is needed.

Newly available cover image for the new book.

4 thoughts on “The trouble with nominatives

Add yours

  1. I’ve been wondering the same thing for so long! Syncope of short vowels before final -s seems to make everything messy. Looking at other Italic languages might be inspiring. Nominatives in -er and -el seem straightforward given plenty of examples such as famel or ager. For stems ending in a dental stop we have a few examples, mers in Umbrian for older *medos and húrz. For stems ending in nasal we have meitims in South Picene if I remember correctly and the humuns gives us an idea. For stems ending in a guttural stop, -ks seems a reasonable outcome, and in the case of vezkeí, we might suppose an anaptyctic vowel so that we’d have a nominative *veseks. Let’s not forget also that vezkeí might have suffered syncope in the oblique cases, perhaps like in numnéis. One might also wonder, what would a cognate to latin magnus look like in Oscan? In my opinion, seeing how consonant clusters were avoided with anaptyxis, as in aragetud instead of *argetud and anafríss instead of *anfríss, I dare to suppose something like *magans 🤔 I’ll keep on thinking about this 🤪


  2. A potentially interesting parallel, about which I know all too little, would be the early transliteration of Akkadian/Sumerian names and toponyms. Gilgamesh, for example was originally translated as Izdubar iirc.

    Obviously these issues are less severe with alphabets derived from Phoencian/Greek/Etruscan, but Motylos’ comment on Hittite brought this example to mind.


  3. This is such an interesting example, thank you. Syllabic scripts certainly add another layer of difficulty to this. We also have the problem of choosing whether to write the nominative or the root in Oscan, particularly for a-stem nouns, where the Oscan nominative is -ú but all the other cases have an -a-. Obviously an -a nominative ending signals a feminine noun to lots of readers, but it wouldn’t be quite right!


  4. It was a similar but less difficult problem with rendering Hittite and contemporary Anatolian names. He first difficulty was that they were written in syllabic cuneiform, so that a name would be in transcription written as ᴵḪu-u-tu-pí-ia-an-za-aš ku-iš DUMU ᴵZi-da-a would be transcribed and translated as ‘Ḫūtupiyanzaš who (was) the son [DUMU a Sumerian word written habitually written down for the Hittite word, as yet unknown] of Zidā(š)‘. The superscript (ᴵ) is a determinative for a male person, and the doubled vowels (ū, ā) might indicate stress or a long vowel sound. But often, as in this example, the nominative form of the name is not extant. So, whereas early Hittite scholars transcribed the names as Ḫūtupiyanzaš and Zidāš, modern practice is to just give the stem as the acknowledged name: Ḫūtupiyanza and Zidā/a.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create a website or blog at

Up ↑

%d bloggers like this: