Appendix talk:Latin cardinal numerals

"When is used to form compound numerals, such as    ("twenty-one"), the masculine nominative singular is used, and does not inflect."

The same for "duo", "tres". Is this statement correct? Google Books gives many instances for "viginti unius anni", "viginti uno anno", "viginti una" and "una et viginti", but only one for "viginti unus anni" (perhaps a typo) and none for "viginti unus anno". Cicero also used "unius et viginti":. A comprehensive Guide to Wheelock's Latin seems to have mistaken (and also mistakenly named "viginti unus" etc. ordinals instead of cardinals). "Twenty-one horses" is certainly "viginti unus equi", but this word combination should decline: "viginti unius equorum", "viginti uni equis", etc. Burzuchius (talk) 19:44, 22 June 2013 (UTC)


 * If you can provide fuller citations, with sources identified, I might be able to address the issue. Finding the sequence of elements "viginti uno anno" does not necessarily indicate that the writer was talking of "twenty-one years".  It also does not indicate whether the source was a Classical text, Medieval text, Renaissance text, Modern text, or what.  That said, Latin did change enormously over time, and many writers of later periods did not follow the "rules" of Classical writers.  Additionally, the Appendix as it currently stands had its information assembled largely from several published texts and textbooks on Latin, but I am certainly willing to believe that information they included was either incomplete or incorrect.  Numerals are one part of speech that is typically neglected or is badly covered both in textbooks and by grammarians, and not just in the case of Latin either. --EncycloPetey (talk) 22:16, 24 June 2013 (UTC)


 * Sorry, I have mistaken. The combination "viginti unus anni" certainly exists (as a nominative). I meant only one instance for "viginti unus annorum". According to Google Books, "viginti unus" (when the meaning is clearly "twenty-one" from the context) may be followed by either singular or plural forms of nouns ("unus et viginti" by plural only), and in any case "unus" should be declined. This grammar clearly has examples with the declined forms ("viginti tribus", "tribus et viginti") from Classical Latin. This text is Classical and has a lot of compound numbers, and every word in them that can be declined is declined. Burzuchius (talk) 14:05, 25 June 2013 (UTC)


 * OK, I want to clarify a couple of things before responding to the content. (1) You cannot say "According to Google Books..." because Google Books does not make any claims one way or the other. It searches for strings of text in a collection of files, and returns are inexact. (2) Citing from a grammar can be iffy, but at least in this situation there are clearly marked citations that allow me to trace back to the original source for inspection. (3) I can't verify the Frontinus text.  It's posted on a website and has page numbers, but the site doesn't tell which text or whose edition it came from.  So, I can't tell for certain it's been accurately copied or the nature of the source text.  I know who Frontinus was, but don't know which of his works this is supposed to be part of, and so can't examine people's translations of the work either.  That kind of information is of utmost importance when trying to give support to assertions. (4) It's not helpful to say "Here's a link to an 83 page text, go hunt for stuff".  I or any other reader would need more precise information than that.


 * So, with regard to content: tribus et viginti does appear in Suetonius' "Julius Caesar", section 28. Because the Fontinus text can't be tracked back to a source, that's all I've gotten out of your latest post.  Unfortunately, a single instance like this isn't enough to build a general pattern from, as almost any rule will be found to have exceptions somewhere.  What we need are a series of clear citations demonstrating usage from the original material.  Take a look at the entry for  (you might have to click "Show quotations").  Here are a pair of clear citations demonstrating usage of a Latin word.  That's the kind of data needed in this situation, as that's what we use on Wiktionary to support all our assertions about grammar, usage, and definitions, and ideally we'd want three instances of each form, as that's what Wiktionary looks for as a matter of policy.  (See WT:CFI.)  We wouldn't have to do this for everything if we can demonstrate the pattern for a few situations, but we do need a number of independent examples of the same phenomenon to gauge that, indeed, there is something going on here that isn't an isolated phenomenon.  --EncycloPetey (talk) 21:07, 25 June 2013 (UTC)


 * I know this is an old discussion but I just stumbled upon this. I myself have had to wrestle with this topic since I began teaching Latin to my wife and 6 year old daughter about 5 months ago using a conversational approach. We are trying to carry on all daily conversation in Latin, which really forces me as the teacher to dig into the grammar books for things I never really considered before, this being one of them. I have come to the conclusion that the numbers 1-3 in compounds should decline. I noticed that many textbooks avoid this topic, but in 'A Primer of Ecclesiastical Latin' by Collins when he introduces cardinal numerals he lists "viginti unus, -a, -um" (p 305). It seems that he believes it should decline. Now, of course, this is a text for Ecclesiastical Latin, but this is how I'm using the numbers in daily conversation. I think that it should be pointed out that Greek has a similar phenomenon with compound numerals and in the Greek compounds the declinable part declines. The fact that '21' could be said "viginti unus", "viginti et unus", or "unus et viginti" demonstrates that to the native Latin speaker these numbers were really felt as compounds, i.e. made up of two parts. Therefore, it seems that the "unus" in "viginti et unus" is really just plain old "unus" and must decline as "unus" usually does. I can't imagine that the phrase "viginti et unus feminas vidi" would feel natural for a native Latin speaker; "viginti" is unmarked for gender or case, but "unus" screams out nominative and masculine. Especially if phrased thus "unus et viginti feminas vidi", this just seems ungrammatical. The one citation of "tribus et viginti" seems sufficient to show that native speakers felt the parts as separate.


 * I have done some research on perseus and it seems undeniable that the components of compound numerals behave just as they do when uncompounded, i.e. they decline. I did a search for 'viginti' and it retrieved some 600 results. I worked my way through the first 300 results and this is what I found: First, authors generally prefer to use round numbers just as we do today; '20' or '25' are found much more frequently than '27' or '23'. Second, in every instance in which the numbers '21' '22' or '23' are used, they are declined according to gender and case. 'Unus' 'duo' and 'tres' appear only where they would be expected, i.e. when they agree in case and gender with the modified substantive. 'Unus' is always used in the singular, which was something that I wondered about before, since 'unus' does have lesser used plural forms, e.g. 'unae scopae' -> 'one broom' (fpl). The singular usage makes sense though because it seems that 'unus' is treated as truly independent from 'viginti'. Here is the list: http://perseus.uchicago.edu/cgi-bin/philologic/showrest_?conc.6.1.18671.0.603.LatinAugust2012 The relevant instances are: 2,5 (Ammianus Marcellinus) 29, 30 (Aulus Gellius) 86, 87 (Julius Caesar) 96, 115, 117 (Livy) 129 (Seneca) 135, 138 (Cicero) 209, 210, 224, 225, 226 (Pliny the Elder) 247, 256 (Suetonius) 272, 273, 277 (Tacitus). The latest of these authors is Marcellinus who was born 325/330. Only one or two of these instances occurred with 'milia'. I believe that the idea that 'unus' would be indeclinable in 'unus et viginti' is really an intrusion of English thinking upon Latin. In English, twenty one is really felt as a single number, not as a compound. You will never hear an English speaker on the street say twenty and one, or one and twenty. The 'one' is bound to the 'twenty' as one word. Contrast this with the Latin examples which freely move the components around. To the native Latin speaker it seems that 'viginti unus' wasn't a 'compound numeral', rather it was really two different numerals, i.e. 'viginti' and then add on to that 'unus'. '21' is conceptually '20 + 1' is what I'm suggesting. The linguistic evidence seems to support this. 'Undeviginti' by contrast is really a true compound, it is indeed one inseparable word and the 'unus' has reduced to it's stem 'un'.Rysh (talk) 21:33, 6 August 2014 (UTC)

Vocative / Genitive of duo
Here unus does have a vocative, while duo and tres don't.

(1.) What's with the vocative of duo and tres? Is it the same as the nominative and thus omitted, or does something like "O you two men, please help me!" not exist?

In several grammar books a vocative isn't mentioned isn't mentioned either, but some (e.g.: Arnold's First Latin Book, 1857; An epitome of grammar: or, a short introduction to the latin tongue, 1733; Aldus Pius Manutius' Grammaticarum instituionum libri IV, 1575) mention a vocative of duo and tres which equals the nominative.

(2.) In said grammar books (in case of Arnold's only for the singular) the vocative of unus is like this: and not which seems to be correct (-us ~> -e in voc.)

(3.) Doesn't duo also have the genitive plural "duum" (this was mentioned in a grammar book) resp. "duūm" (so in Lewis & Short @ perseus.tufts.edu) resp. "duûm" (with signum contractionis, so it's in Georges' dictionary and older books in Latin), though maybe just m.&n. and not f. (so in grammar books in Latin, though the dictionaries don't say so, and though one grammar book in English even said it is used for all 3 genders)? PS: duum does exist and says it's just m.&n.

-80.133.114.159 02:54, 30 May 2015 (UTC)

Other numerals
As Latin has other numerals which are missing at en.wt, I'd like to leave this here: Sources & Notes: Various grammar books and dictionaries, especially [www.zeno.org/Georges-1913 Georges' dictionary], [en.wikipedia.org/wiki/Karl_Gottlob_Zumpt Zumpt's Latin grammar] & [www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0059%3Aalphabetic+letter%3DA Lewis&Short]. Zumpt also gives other additive numbers instead of substractive numbers in case of 18 & 19 &c., but not sure whether or not that's a grammarian's invention or did/does exist. Also: There were/are forms like "XIIII" instead of "XIV" which are missing here. Inflection notes like "(-a, -um)" of course could be replaced (e.g. by a star * which at the bottom gets explained as either "indeclinable" or "declinable", while the entries of each number should contain the declension). -93.196.241.177 12:51 & 14:14, 31 May 2015 (UTC) & 14:16, 1 June 2015 (UTC)
 * [*...]: uncommon, but mentioned in several grammar books. (Note: some more -plex & -plus numerals are questionable, but were used in (New) Latin, which can easily be verified through a google books search.)
 * substantival numerals: ūnio (-ōnis, f.), bīnio (-ōnis, m.), ternio (-ōnis, m.), quaternio (-ōnis, m.), quīnio (-ōnis, m.), sēnio (-ōnis, m.), _, _, [*novēnio (-ōnis, m.)], [*dēnio (-ōnis, m.)]