One pants? One scissors? Oh, no way! This is a follow-up to the @brianpck's question How do I specify how many “litterae” or “castra” there are?, and the accepted answer was to use distributive numerals with the pluralia tantum nouns, which are morphologically plural even when their denotate is a single object. The answser did not provide any exceptions for expressing the quantity of one such a thing, and the answerer even confirmed that a distributive form is used with a single castra or litterae in my follow-up comment.

I did not specifically learn Latin grammar by section and paragraph, and mainly developed my understanding from texts, using grammar books only for reference. The answer struck me as odd, as my internalized pattern (which often happens to be wrong because of my way of study) suggested using a gendered plural form of unum for one such thing.

An example that I immediately came accross was from (Caes. B.G., 1, 74): ut una castra iam facta ex binis viderentur: “[They mingled together so much] that two camps ostensibly made one”. A possible argument that castra could somehow be used as a normal countable noun of the feminine gender does not work here, because una facta castra (which might be considered as either NEUT-PL or FEM-SG from morphology) does not work syntactically, as the whole NP is the subject of the verb viderentur which is plural, which makes the *FEM-SG parse is impossible. Caesar here undoubtedly uses una with the neuter plural castra.

A&G $137.b. states [the bold highlight is mine]:

Distributives are used as follows:
b. Instead of cardinals, to express simple number, when a noun plural in form but usually singular in meaning is used in a plural sense: as, bīna castra, two camps

This conspicuously does not say anything about the singular entity. (To confuse matters further, they have to add that

(duo castra would mean two forts)

but this is not our gripe at the moment.) The PHI corpus, not being POS-tagged, does not help, as singuli is also a from of the ADJ singulum.

So what is the correct way to express one “litterae” or “castra” in Classical Latin?

  • Have you seen this question? It's certainly related, but I'm not sure if it's a duplicate.
    – Joonas Ilmavirta
    Nov 11, 2017 at 21:00
  • @JoonasIlmavirta: The very same Tom Cotton responded to my comment: “The appropriate distributive isn't unus, but singuli: singulae litterrae.” I perceive a contradiction here. Of course, in "one letter from each applicant" would use singulae, but I mean just the normal cardinal, non-distributive case “Alice wrote three letters, Bob one, and Charlie five.” Nov 11, 2017 at 21:06
  • @JoonasIlmavirta so strictly speaking yes, it looks like a duplicate (and I missed the question you are referring to), but I believe the question would certainly benefit from some additional clarification. Up to you, indeed. :) Nov 11, 2017 at 21:13
  • I'm still somewhat confused myself, too. Perhaps it's just that the distributive has more roles than usual with plurale tantum words. I hope to see more explanations on the matter. This question might benefit from clarifying what is confusing in relation to earlier questions, but I'm genuinely not sure. I just wanted to make sure you knew of the older question. (I'm not going to close this as duplicate unless you ask for it, but someone else might.)
    – Joonas Ilmavirta
    Nov 11, 2017 at 21:15
  • @joonasIlmavirta Did you notice that I expanded my original answer tokkm's question?
    – Tom Cotton
    Nov 11, 2017 at 21:30

1 Answer 1


It seems to me that you already have all the information you need for an answer. You just need to follow through the logic in the Allen & Greenough section that you quote.

That section says that distributives are used instead of cardinals when a noun that's plural in form but usually singular in meaning is used in a plural sense, and you want to express simple number for that noun. Therefore, 'two camps' is bina castra, as in Caesar BG 1.74.

However, you want to express simple number when the noun that's plural in form but usually singular in meaning is used in a singular sense. In this case, the condition in §137.b isn't met, and you therefore use a cardinal number. So 'one camp' is una castra, just as Caesar says.

Gildersleeve & Lodge §97, Remark 3 makes this point explicitly:

The Distributives are used with pluralia tantum: binae litterae, two epistles. But with these uni is used for one, trini for three; unae litterae, trinae litterae.

(If you want to say 'one camp apiece' or 'two camps apiece,' that falls under the rule in A&G §137.a instead.)

As to duo castra, OLD actually has separate entries for the neuter plural noun castra, 'military encampment,' and the neuter singular noun castrum, 'fortified post or settlement.' So if you see duo castra (using the cardinal number) instead of bina castra (the distributive number), you should be dealing with the word for 'fort' rather than the word for 'camp'; otherwise, the rule that A&G gives would be violated. (L&S just includes the plural castra in the entry for castrum; but if the A&G rule holds, the two meanings must have been felt to be sufficiently distinct so as to constitute separate words, in practical terms.)

Which is, of course, not to say that there aren't exceptions to the rule that A&G gives.

  • Thank you! I would disagree a little bit on the reading of A&G; when the book says nothing about the rule, I can assume a more general rule applies, but in this case it is not that obvious what is the more general rule. Oh, an is it too much to ask to elaborate a bit on the exceptions that you are mentioning in the last sentence? Nov 11, 2017 at 21:32
  • @kkm Please look at my expanded answer to your original question.
    – Tom Cotton
    Nov 11, 2017 at 21:34
  • 1
    @kkm, When A&G is silent on something, I usually take that as a sign that I should look in Gildersleeve & Lodge. And, in fact, G&L §97, Remark 3 makes explicit what is left implicit in the A&G rule: 'The Distributives are used with pluralia tantum: binae litterae, two epistles. But with these uni is used for one, trini for three; unae litterae, trinae litterae.' As to exceptions, I don't have anything specific in mind, but they may exist; confirming this would require a corpus search. In the meantime, I'm just leaving the possibility of such exceptions open.
    – cnread
    Nov 11, 2017 at 21:47
  • Thanks for the reference to G&L! I was sticking so much to my A&G (they always worked for me) that I dropped other great references from my radar! I'll try to work on exceptions. This is an interesting topic, especially if looked at diachronically (this sometimes gives an answer why they exist). Nov 11, 2017 at 21:52

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.