Enclitics on Latin indeclinables

I maintain a Latin dictionary called velut. To decide what words to include and how to treat them, I often need to research some arcane facet of Latin grammar or phonology. This is an example.

In Latin, most nouns and adjectives decline, meaning their endings change according to their grammatical case and number (and gender for adjectives). However, a few do not, and are called indeclinable. Examples include fās and nefās (meaning “right” and “wrong”), names of letters of the Greek or Latin alphabets, and some obscure nouns borrowed from other languages.

An enclitic is a kind of affix that attaches to the end of a word. The three I’m most interested in are the ones that can be attached to almost any word in Latin. They are: -ne (introducing a question), -que (“and”), and -ve (“or”).

I wondered whether indeclinable words can be given those three enclitics. I could imagine using indeclinables with enclitics, but I could also imagine that the act of adding an enclitic might be seen as similar enough to declension that it would be forbidden for indeclinables.

An internet search for discussion of the topic yielded nothing.

There are 61 indeclinable non-proper nouns in velut. So I searched the Packard Humanities corpus for all 183 combinations of those 61 nouns with the three enclitics.

Some of the words are spelt the same as irrelevant words, and these must be discounted. Appearances of ove, for instance, are simply the ablative singular of ovis (“sheep”), not ō as the name of the letter O with an enclitic -ve. (It is also unrelated to ō, the interjection in Ō fortūna! or Ō tempora!)

Meanwhile, āque and dēque are indeed attested as encliticized forms, but (as encliticized forms) they are from the prepositions ā and dē, not indeclinable nouns for the letters A and D.

However, I did find the following attestations of words that cannot be anything other than an encliticization of a noun that I consider to be indeclinable.

Admittedly, velut currently has nihil, nihilum, and nīlum as the same lemma (headword), which makes nihil look like it’s not indeclinable. I’m going to separate the three lemmata in velut. The lemma nihil should have no forms except nihil, and exist in only the nominative and accusative singular (like fās and nefās). The lemmata nihilum and nīlum are regular neuter nouns with genitive in -ī and dative/ablative in -ō, albeit with no plural forms.

And labda is written as Greek in the De Litteris… source, making λάβδαve, but I’m still counting it.

I therefore include encliticized forms for all indeclinable nouns in velut. So (for example) ōve, āve, and dēve can be from names of letters, by analogy with Terentianus’ labdave.

For indeclinable proper nouns, I don’t know of any attestations with enclitics. There are nine indeclinable proper nouns in velut, but Packard Humanities had nothing when I added -ne/-que/-ve to them.

This is not strong evidence that indeclinable proper nouns cannot get enclitics. Most of my nine are biblical and a couple others are post-classical, but the corpus does not include many texts past 200 AD. (Indeed, Caere is the only indeclinable proper noun I know of from classical Latin.) I therefore would not expect to see many attestations with enclitics.

A general internet search revealed a mediaeval poem by Paulinus Aquileiensis that uses Iācōbque. I hadn’t heard of the poet, but it looks like good enough Latin.

Let’s see what encliticized attestations of indeclinable adjectives we can find, in the Packard Humanities corpus.

The total lack of -ne on the list is somewhat surprising. But, I suspect there are not as many direct yes‒no questions, proportionately, in surviving written Latin as there would have been in everyday speech. I don’t see why Latin speakers shouldn’t have been asking, Totne?, Septemne?, Centumne?, etc.

On the basis of the evidence in this article, I will include encliticized forms for all indeclinable adjectives in velut, alongside all indeclinable nouns (including indeclinable proper nouns).

Previous post: Launching a dev server in frameworks I’ve used

Next post: CSS flags