Wiktionary talk:About Latin

Definition from Wiktionary, the free dictionary
Jump to: navigation, search

Archives[edit]

Question on adverbs[edit]

Since {{la-adv}} seems to attract little attention, I'll repost my question here, hoping for an answer. What formatting is required to show that an adverb is missing the superlative form, whilst still showing it has a comparative and vice versa? An example of this would be ēlātē, which according to L&S and the OLD doesn't have the superlative ēlātissimē, and there is limited evidence on google books for any valuable citations for the superlative. Caladon 21:53, 29 March 2010 (UTC)

Use a dash (-} for parameter 4. Note that L&S don't actually say there is no superlative; they just don't document it with a quotation like they do the comparitive. There is an Italian descendant of the superlative (spelled identically), so the superlative is at least theoretically possible in Latin of some age. --EncycloPetey 22:44, 7 April 2010 (UTC)
Yes, I was reconsidering my wording afterwards. I have changed the entry for now until a citation is found for the superlative. The only thing, which is annoying, concerning the use of the dash, is that it links to -#Latin, which doesn't appear to be very useful to me. It may also be helpful in the future, if no dictionary explicitly says it is unattestable, to explain the fact that it is theoretically possible by use of some note or template, rather than either leave it showing that the superlative is not possible or that the superlative is possible but there's no citation provided for it. Caladon 23:03, 7 April 2010 (UTC)
True, but we have very little work being done on adverbs in any language. Also keep in mind that dictionaries don't attest every form of a verb's conjugation either. That's a project for a much, much later date. --EncycloPetey 23:12, 7 April 2010 (UTC)

Bad example[edit]

Icelandic gala is an inherited word, not borrowed from Latin. This has already been corrected in the word's entry, but here it is mentioned as a descendant of Latin gallus. This erroneous information must be removed. Unfortunately, this makes gallus less useful as an example entry. Could someone pick another example word with borrowed descendants and change the section? – Krun 16:50, 12 April 2010 (UTC)

I and J[edit]

Do we want to keep this distinction here? It's generally unsupported by editors (AFAICT). For example {{la-categoryTOC}} doesn't include a 'j', but there are Latin entries starting with a 'j'. I'd very much compare this to vp where v & u are now considered separate letters, but may not have been at a previous time. Nobody's saying that Latin doesn't use a 'J' - I've seen it myself, anyone can, the theory has been that they're simply variants of the same letter in Latin. Mglovesfun (talk) 16:48, 8 April 2011 (UTC)

  • Yes, "j" is just an "alternative-form-of" "i", and all our Latin terms containing a "j" should be of that format. So I don't suppose we need them separately indexed. SemperBlotto 16:55, 8 April 2011 (UTC) (p.s. My next Latin text to scour will be the w:Novum Organum, and it's got many "j" words (so I'll be adding them))

Latin -que compound words[edit]

Following up on the little discussion there was last year in RFV (when archived: Talk:fasque), I'd like to open a bigger discussion about our policy towards -que. To date, policy (reflected in the usage note on -que) has been not to include compound words like "populusque", "fasque", "sitisque"; I want to know if we should reconsider. An argument in favor of excluding the compounds is that they are SOP. An argument in favor of having them is that they're single words in post-Roman texts, i.e. texts that use spaces, and Latin isn't polysynthetic, so it makes as much sense to have them as to have the SOP English word chesspiece (or, arguably, unresolved or undress). A counter to that is that Latin was historically written without spaces, and "-que" was historically preserved in abbreviations like SPQR, making it hard to tell whether or not "SENATVSPOPVLVSQVEROMANVS" was originally three words or four. In case it wasn't clear, I personally am on the fence. If -que compounds are allowed, only attestable ones will be included. - -sche (discuss) 20:06, 7 February 2012 (UTC)

As with German "Nouncompounds" and Hebrew words with clitics, I think we should have these because an English speaker (English Wiktionary's audience) will not know where to break up the word so will look up the whole thing.​—msh210 (talk) 20:35, 7 February 2012 (UTC)
If we do decide to have entries for Latin -que words, how will they be defined? — Raifʻhār Doremítzwr ~ (U · T · C) ~ 20:56, 7 February 2012 (UTC)
I suppose something like the way fasque (noun) is now.​—msh210 (talk) 21:15, 7 February 2012 (UTC)
That shouldn't be acceptable, in my opinion. See my proposed way of presenting these que terms: I propose using only {{head|la}} (to prevent these entries from being categorised), using the Phrase header (because, if que is indeed a clitic, that makes all these que terms not words, but rather phrases; this means that que terms which are attested in multiple parts of speech can be presented more compactly under only one POS header), and restricting these terms' definitions to just [X] + que (I assume that a very simple template can be used to ensure consistent presentation). I think that would strike the best balance between user-friendliness, accurate information, and the elimination of user-maintained datedata redundancy. What do you think? — Raifʻhār Doremítzwr ~ (U · T · C) ~ 23:36, 7 February 2012 (UTC)
(I don't know what you mean by "date redundancy". And I believe you're correct that "a very simple template can be used to ensure consistent presentation". Now to what I think are your main points.) There are several differences between the entry you link to (which I'll call "yours") and the one I do ("mine"). AFAICT they are: (1) Mine uses a Noun header; yours a Phrase. Phrase sounds more reasonable to me, but I don't know Latin, and I think this (like POS decisions in general) should be left up to those who do. (2) Mine includes a gloss ("and religion; and divine law"); yours does not. IMO yours is better in this regard, as fas (for example) may have too many meanings to list in a short definition line. The only exception would be if (e.g.) fasque can, for whatever reason, be used for only some of fas's meanings — but even then, I might prefer a usage note at fasque (and maybe one at fas also) than a gloss on the definition line. (3) Mine uses a longer explanation of what fasque is, on the definition line, than yours: mine says "form of fas with the conjuction que (and), suffixed to it", while yours says "fas + que". IMO a verbal explanation is better than a plus sign — but I don't know that "mine"'s explanation is the best possible, or even the correct, explanation. Any other differences? I note that both versions categorize as a noun. I'm not sure that's best — perhaps as a phrase or nothing is better — but would prefer to leave that decision, too, to the Latin editors, I think. I commented here merely for incusibility; it was not my intent to get involved in a discussion on formatting.​—msh210 (talk) 00:40, 8 February 2012 (UTC)
Yeah, having entries and formatting them like {{&lit}} entries (templated information, rather than full definitions like this) seems like a good solution. The template text (I would prefer template text to a mere plus sign) can even explain the "-que" element, so users only have to click on the initial element to find out what it means. I'm blundering my way through a possible template as we speak. - -sche (discuss) 00:48, 8 February 2012 (UTC)
By "date redundancy", I meant "data redundancy" (which I've now corrected); i.e., redundant information in entries that are properly dependent for most of their meaning on some lemma of sorts. I agree that, on reflexion, the template we use should explain what the clitic que is doing in the entries that use that template (let's call it {{que-def}} or something), rather than presenting the information as I first suggested (i.e., [X] + que). My version categorised the entry as a noun only because Category:Latin nouns was included at the bottom of the entry; had I noticed that, I would have removed it, and the entry would have remained uncategorised, as it ought to be. — Raifʻhār Doremítzwr ~ (U · T · C) ~ 01:02, 8 February 2012 (UTC)
I think we should not have these, for the same reasons that we do not include English possessives or other words with enclitic particles. And almost any word in Latin can have this suffix form added. --EncycloPetey 04:33, 8 February 2012 (UTC)
FWIW, I agree with EP; however, from reading the RFV discussion about fasque, I realise that some sort of compromise (as msh210, -sche, and I are working out above) is more likely to win general support. — Raifʻhār Doremítzwr ~ (U · T · C) ~ 17:24, 8 February 2012 (UTC)
Oh, I am willing to concede that we may end up with a few case-by-case exceptions, but we should be explicit about that if we come to such a decision. --EncycloPetey 03:59, 9 February 2012 (UTC)
I am ambivalent on this topic. I am a staunch supporter of "all words in all languages" and these certainly look like words to me - a concatenation of alphabetic characters with no spaces. Perhaps we should allow them if they are actually accompanied by three citations (not just three citations being possible), but state somewhere that we have better things to do with our time. SemperBlotto 08:31, 8 February 2012 (UTC)
p.s. And the same goes for -ne and a few others that I can't remember on the spur of the moment. SemperBlotto 08:32, 8 February 2012 (UTC)
I support Semperblotto's suggestion (I would adopt such a rule in some other cases, e.g. numbers, but this is another issue). Lmaltier 21:44, 9 February 2012 (UTC)
I can't support the idea that every "concatenation of alphabetic characters with no spaces" should be included, because ancient manuscripts often included no spaces between words. The oldest Biblical texts in Koine Greek were written without any spaces, so that the entire text of the Gospel according to Matthew could be included as a single "word" under this criterion. --EncycloPetey 03:26, 10 February 2012 (UTC)
I agree: we should not include all of Matthew as an entry. But no one will look it up: they'll see the text and realize it's not one word. On the other hand, anglophones (English Wiktionary's audience) who see Latin text with spaces between words and come across fasque in it will look that up.​—msh210 (talk) 17:28, 10 February 2012 (UTC)
How do you pronounce fasque anyway? Is it different to "fas que"? I'm no Latin expert, but seems to me that if fasque, or other such words, are written without a space in text that otherwise has spaces then they should be included. I think if we can handle the 100 inflections of each Latin verb, then we can handle a few unusual noun endings when they appear in text. (-ve might be another that hasn't been mentioned above) Pengo 04:39, 10 February 2012 (UTC)
Que doesn't make much of a difference to the pronunciation of fas in fasque; however, in many cases, it does affect stress significantly, as in the case of populus, pronounced /ˈpo.pu.lus/, which becomes populusque, pronounced /po.puˈlus.kʷe/. — Raifʻhār Doremítzwr ~ (U · T · C) ~ 14:54, 10 February 2012 (UTC)

macrons: only guesses?[edit]

Am I right in thinking macrons are always based on guessing? Scientific guessing I'm sure, but guessing nonetheless. Pronunciations by the same token, are always based on guessing too. Am I right? Mglovesfun (talk) 14:31, 16 August 2012 (UTC)

  • Same as etymology !
  • Actually, I think that you can study poetry and figure out if a vowel is longer or shorter than normal. But different people put macrons in different places - and I can't really summon up much enthusiasm for them at all. I wish we didn't show them - it would make all the templates a lot easier to use. SemperBlotto (talk) 14:45, 16 August 2012 (UTC)
    • Regarding "I think that you can study poetry and figure out if a vowel is longer or shorter than normal" that is indeed what my 'scientific guessing' referred to. Mglovesfun (talk) 14:48, 16 August 2012 (UTC)
      • There are some inscriptions that use an apex mark to indicate a long vowel, and some Latin grammarians (i.e. native speakers) say which vowels are long. Otherwise, vowel length can be determined by historical linguistics (both working up to Latin from Proto-Indo-European and working forward from Latin to the Romance languages as well as to Latin loanwords in Germanic and Celtic languages) and, as you mentioned, by scansion in poetic texts. So some knowledge of vowel length comes first-hand from native speakers, and the other is based on scientific evidence, which is rather more than "guessing". —Angr 18:31, 16 August 2012 (UTC)
        • I suppose hypothesis is a better word than guess. But it's very, very far from being fact. Mglovesfun (talk) 22:58, 27 December 2012 (UTC)
          • Our knowledge of what vowels are long in Latin is as close to being fact as our knowledge of the pronunciations of the individual letters. Like all widely accepted scientific theories, it's testable and is not contradicted by any known facts. —Angr 23:02, 27 December 2012 (UTC)
            • Which makes it a theory, which is stronger than a hypothesis. There is no such thing as a fact in science; everything is falsifiable. —CodeCat 23:03, 27 December 2012 (UTC)
              • Testable? How? Mglovesfun (talk) 23:05, 27 December 2012 (UTC)
                • I hate scansion with a vengeance, but I will admit it's pretty damn useful. Latin poetry is extremely regulated, so we can use it to figure out vowels that must be naturally long. —Μετάknowledgediscuss/deeds 23:07, 27 December 2012 (UTC)
                • In science, testing a theory means trying to find evidence that proves it false. As long as there is no such evidence, and the theory correctly accounts for the evidence that is present, it is considered valid. —CodeCat 23:09, 27 December 2012 (UTC)
                  • No, no, no, you're really wrong there. What you're saying is that if something is not testable, it should be assumed that it is valid. Mglovesfun (talk) 23:12, 27 December 2012 (UTC)
                    • But the vowel length of Latin words is testable, because it is possible to find evidence that proves a hypothesis false. To give an example: assume two possible hypotheses, one that says "cantare" has a short vowel, and another that says it is "cantāre" instead. There will probably be evidence that is compatible with either hypothesis, but sooner or later you will come upon evidence that is not compatible with one of them, and therefore that hypothesis cannot be true. Therefore, the other must be true. That's how science works. —CodeCat 23:19, 27 December 2012 (UTC)
                • (ec) Against the scansion of poetry, the statements of the grammarians, the inscriptions that use the apex, and the developments of the sounds in the modern Romance languages and Welsh (which, unlike modern Romance languages, kept Latin short ă and long ā distinct). —Angr 23:10, 27 December 2012 (UTC)

macron redirects[edit]

The current policy reads:

...within the text of the page, macrons should be used wherever appropriate. For example:
==Latin==
===Adjective===
'''līber''' {{m}}, '''[[libera|lībera]]''' {{f}}, '''[[liberum|līberum]]''' {{n}}

This is needless and moronic make-work. The text should read

'''līber''' {{m}}, '''[[lībera]]''' {{f}}, '''[[līberum]]''' {{n}}

If the definitions are (per consensus) located at macronless namespaces, there should be redirects to point to the proper location from the macron-employing variant. There is no reason to require multiple piped forms of the same term in every instance where Latin appears on a page. However, attempts to create such redirects – necessary, as mentioned above, in order to use many templates – have been deleted, apparently following a separate policy that we must avoid redirects at all costs.

These three policies in tandem – employing macrons throughout the mainspace, avoiding macrons throughout the namespace, and avoiding redirects – are a disservice to the users and the editors. The proper solution isn't to remove the macrons throughout, but to use redirects where needed.

Are there any other talk pages necessary to get this conversation started? LlywelynII (talk) 22:08, 27 December 2012 (UTC)

Redirects from titles with macrons are okay in cases where Latin is the only language with a word so spelled, but if some other language whose writing system uses macrons (like Latvian or Hawaiian or Romaji Japanese) has a homograph, then we can't make it a redirect. —Angr 22:40, 27 December 2012 (UTC)
Is there consensus for allowing redirects at all, even if Latin is the only language? WT:REDIR says that redirects should not be used to link from forms without diacritics to forms with diacritics. Presumably the reverse is true too. —CodeCat 22:45, 27 December 2012 (UTC)
I will post there, but I don't see why the reverse would be true. The rationale for that policy has no bearing (in fact, completely fails to apply – the search does not automatically go to the non-accented version). The current Latin policy is creating needless busy work writing words twice (once as the macronic running text, once as the macronless link) and breaking some templates (which require the macronic form to display correctly but don't create the proper link). LlywelynII (talk) 22:53, 27 December 2012 (UTC)
(ec) Again, I think that rule is in place because of the possibility of a redirect getting in the way of a different language. For a relatively long Latin word with a distinctly Latin ending (e.g. conversābiminī redirecting to conversabimini) I think it's unproblematic because there's virtually no chance of a different language having its own conversābiminī. But I'd never support a redirect from amō to amo. I think we should be allowed to use common sense in our interpretation of the rules. —Angr 22:54, 27 December 2012 (UTC)
If amō has no homographs and is currently a redlink, why wouldn't you support a redirect?
Because there are dozens- if not hundreds- of underrepresented languages that use macrons, and we don't know enough about all of them to say for sure which spellings for simple VCV-structured words with macrons are never going to have an entry. Chuck Entz (talk) 18:34, 28 December 2012 (UTC)

LlywelynII (talk) 01:13, 28 December 2012 (UTC)

In response to Angr's example above where a macronic form would refer both to Latin and (e.g.) Hawaiian, I don't see the problem. The templates and uses would link to the page's Latin definition, which would list it as a variant form of the main entry.
The alternative there would be not listing the Latin at all which is far less helpful. LlywelynII (talk) 22:53, 27 December 2012 (UTC)
LlywelynII, could you please avoid inflammatory words like 'moronic'. Mglovesfun (talk) 22:50, 27 December 2012 (UTC)
I see your point w/r/t civility, but I'm not using inflammatory rhetoric w/r/t any other well-meaning editor but w/r/t a combination of three well-intentioned policies which (taken together) do in fact have a moronic effect. LlywelynII (talk) 22:53, 27 December 2012 (UTC)
w/r/t? Mglovesfun (talk) 22:55, 27 December 2012 (UTC)
Sorry. "With regard to". Have been reading too much DFW lately. LlywelynII (talk) 22:56, 27 December 2012 (UTC)
We do have w/r/t. —Angr 22:57, 27 December 2012 (UTC)
I'm sure this is a SNOOT haven. =D LlywelynII (talk) 22:59, 27 December 2012 (UTC)
SNOOT? Anyway, I doubt that people are searching for terms with macra, because almost no texts except dictionaries, grammars, and texts re-edited for beginners even use them. We couldn't ever write templates assuming there to be redirects, because I think many of us would oppose their creation en masse. Oh, and @Angr: isn't it a bit difficult to draw the line between amō and conversābiminī (I don't think I've ever even used the latter...)?—Μετάknowledgediscuss/deeds 23:03, 27 December 2012 (UTC)
It's difficult to define the line between them, but it's not difficult to use common sense at the extremes of the spectrum. —Angr 23:04, 27 December 2012 (UTC)
If being correct involves more work, I'd rather do the extra work and be correct. Is this a pro-laziness proposal? Even if it were, it would have the opposite effect; making a lot more work. So it would mean a lot of work to be less accurate, an all-round loss. Mglovesfun (talk) 23:07, 27 December 2012 (UTC)
Lines like "needless and moronic make-work" suggest that Ll. thinks of this as essentially "pro-laziness", although with better connotations. —Μετάknowledgediscuss/deeds 23:09, 27 December 2012 (UTC)
I agree that we should do more work (have the macronic forms in the running text) and be correct (have the main entries at the macronless forms). Making those work together is precisely what we're discussing. LlywelynII (talk) 23:51, 27 December 2012 (UTC)
Consider also this situation. It is a given that not all forms with macrons will exist only in Latin, since some other languages also use them and can (and probably do) happen to have the same word. That implies that any link to a form with macrons could lead to a page that isn't a Latin entry, but rather a "macron spelling of" entry with yet another link. Such a situation is obviously not desirable. But if we allow links to macron spellings, then we cannot prevent at least a few links to such entries. —CodeCat 23:13, 27 December 2012 (UTC)
It's more desirable than the macronic entry not having any Latin or link to the Latin entry (e.g., amō not showing any Latin use whatever). Certainly another possibility would be to change the current Latin policy to make the macronic form the main entry for every single Latin word, but that would (a) involve much more work, (b) be rather anachronistic, and (c) be less helpful to the end users who will mostly type without the macrons. LlywelynII (talk) 23:51, 27 December 2012 (UTC)
Oh, and Llywelynll, I think you're wrong about the templates being broken, because I've never seen it happen. Can you find even a single example of this? —Μετάknowledgediscuss/deeds 23:19, 27 December 2012 (UTC)
Just want to point out that when Scribunto is deployed, we will be able to make templates that automatically remove the macrons in the link, effectively solving this problem. --WikiTiki89 23:26, 27 December 2012 (UTC)
So... for how many years have we been saying stuff like that? I can't wait, but personally it still doesn't seem anywhere near imminent. —Μετάknowledgediscuss/deeds 23:28, 27 December 2012 (UTC)
Any template where the running text display is the same as the generated link. LlywelynII (talk) 23:51, 27 December 2012 (UTC)
I don't think any exist. I have asked you to find just one example, and you can't. —Μετάknowledgediscuss/deeds 23:55, 27 December 2012 (UTC)
I mostly edit things that I notice are missing, rather than scripting bots or spending hours dumping dictionaries into approved templates. I'm sure there are (as I said, one of the other, more experienced editors complained about it in the discussion directly above this one) and that some of the other editors here can point them out to you. {continued below}
The thing is, you made this claim, so I'm asking for some evidence. Now you tell me you have no evidence. That's OK, but I just wanted to make sure. —Μετάknowledgediscuss/deeds 00:23, 28 December 2012 (UTC)
I think what he means is not exactly "broken" templates, but that you can't just do {{l|la|amō}}, and instead you have to duplicate the word as {{l|la|amo|amō}}. --WikiTiki89 00:26, 28 December 2012 (UTC)
I gave it in the paragraph immediately below, but you posted here and rewrote my comment to make it seem like it was a complete and non-responsive entry. I didn't unsign anything. Cease rewriting and editing my comments and respond to the points at hand. LlywelynII (talk) 00:30, 28 December 2012 (UTC)
It would be nice if you would stop accusing me of ridiculous crimes like "rewriting" your comments. As WikiTiki says, that's not "broken" (i.e. it works). (And unsigned just means that you forgot to leave your signature there.) —Μετάknowledgediscuss/deeds 00:33, 28 December 2012 (UTC)
I didn't forget: the bottom of my post was signed. You precisely rewrote my comment to make it seem like the midpoint was the end of my post and I had not answered your question, both of which were untrue. LlywelynII (talk) 00:40, 28 December 2012 (UTC)
You are repeating false statements, and this annoys me. I'm not intending to respond to any more comments here that are not specifically about the macra in question. —Μετάknowledgediscuss/deeds 00:45, 28 December 2012 (UTC)
I'm fine with getting back to the topic, except the complaint is not false and you can review the history to see that. You created and then restored the misleading "signature" on my "post".
[My apologies to WikiTiki, who takes exception to my having moved his post to where it should have been had your edit been less deceptively placed. I wanted to keep the macron discussion honest and separate from this aside about you and your deceptive editing of my posts, but if I'd seen he was moving his comment – rather than you – I wouldn't've returned it a second time. I have no idea why he misleadingly added a timestamp to what was never my full post.] LlywelynII (talk) 01:10, 28 December 2012 (UTC)
In any case, very common templates such as {{term}} do create exactly such links; correcting the display requires filling in a second field with the same word as before. LlywelynII (talk) 00:19, 28 December 2012 (UTC)

Incidentally, I did a summary and clarification of my main points here at the Beer Parlor. Meta redirected the conversation there to here, but the summary is there if anyone was confused about how these three policies work together to create a bad effect for the users and editors. LlywelynII (talk) 23:53, 27 December 2012 (UTC)

Emend.: The policies work together in a way that you and you alone feel is bad. You have not demonstrated any factual problem, but merely that your opinion about the issue differs from that of the community. --EncycloPetey (talk) 19:15, 28 December 2012 (UTC)

Old Latin inscriptions[edit]

TK archive icon.svg

The following discussion has been moved from User talk:-sche.

This discussion is no longer live and is left here as an archive. Please do not modify this conversation, though feel free to discuss its conclusions.


Inscriptions and whatnot

I'd like to add some Old Latin terms used in inscriptions, but I'm not really sure what format to use and how to do within the frame of Latin. For example, the Garigliano Bowl uses the word SOKIOIS, the ablative plural of *SOKIOS. Should I show this as {{Old Latin|lang=la}} {{obsolete form of|sociis|sociīs|lang=la}} or as an inflected form? What if the lemma isn't citable? Also, what about different readings of a mangled inscription? What if the same author reads it two different ways? What if the writing system used is not the Latin alphabet? TIA —Μετάknowledgediscuss/deeds 17:57, 17 March 2013 (UTC)

Also, should the pagename be sokiois or SOKIOIS? Only the latter is attested, of course, but WT:ALA would advise using the former. —Μετάknowledgediscuss/deeds 20:33, 17 March 2013 (UTC)
First, we need to answer the question: is Old Latin a form of Latin ({{la}}) or a separate language (in need of its own code)? What do you think? What do academic authorities think? That influences what to do with SOKIOIS et al:
No matter what we consider Old Latin, if a form is attested (e.g., if the ablative plural of *SOKIOS, SOKIOIS, is what's attested), that form should have an entry. But if Old Latin is a form of Latin, I see no reason not to follow WT:ALA in normalising its all-caps inscriptions the way we normalise newer Latin's all-caps inscriptions — whereas, if Old Latin is a separate language, there'd be a strong case for keeping it in the all-caps form it's attested in.
If OL is a form of Latin, sokiois could be listed as an obsolete ablative plural form of socius (the way hath is an archaic third-person singular of have), or, if we're confident the nominative would have looked like sokios, sokiois could be listed as the ablative plural of sokios, and sokios could be created as an obsolete/OL form of socius (read on for justification and precedent for that).
If OL is an independent language, and we're reasonably confident the nominative would have looked like SOKIOS, it should definitely have an entry (IMO), the justification being that the word is attested (in inflected forms), and a word's definition has to be stored somewhere (in one of the word's forms), and Wiktionary has decided to make the nominative singular that somewhere. Precedent for this is ample: I know we do it for Gothic, Old Norse, and sometimes even German, and I expect we already do it for Latin and Greek. - -sche (discuss) 22:03, 17 March 2013 (UTC)
If there are multiple readings of an inscription, I think we should have entries for all of the readings; that's how I've handled Pictish and Khazar.
If an inscription isn't in the Latin alphabet, I would enter it (in whatever alphabet it's in) all the same. I thought some other old languages already had entries in multiple script; I can't find any examples, but I know that e.g. Old Norse is attested in both the Runic and the Latin scripts, so it's only a matter of time before someone adds Runic Old Norse... - -sche (discuss) 22:28, 17 March 2013 (UTC)
That's a valid question. Wiktionary considers Old Latin to be Latin. The main problem is that there's some rather tame "Old Latin", which can be found in ample literature, and often survived as archaisms into Classical Latin. By tame "Old Latin" I mean spellings like quom and second declension genitive plural endings in -um instead of -orum. This is fully comprehensible if it is read by somebody as fluent in Classical Latin as I am.
There almost certainly is another language, closely related to Latin, which I might call legitimate Old Latin, like the Carmen Saliare, which was difficult even for educated Classical Romans to read, and which I have been reading up on lately. Forms like FHEFHAKED (from the Praeneste Fibula) are nothing short of impossible for someone like me to read without specific study of Old Latin, and that's actually a transliteration. That brings me to another point, which is that truly Old Latin is often not in the Latin alphabet, but instead in the Etruscan alphabet. The corpus is tiny, because only inscriptions remain (manuscripts had their spellings and other features largely normalised). The real question is whether it's worth creating a language just for these inscriptions, when it can't really be named in a non-confusing way (I would prefer 'Primitive Latin', but I don't think that such a term is used academically) and is definitely mappable to Latin. —Μετάknowledgediscuss/deeds 01:11, 18 March 2013 (UTC)
Post scriptum: LinguistList separates Old, Classical, and Vulgar Latin. We essentially already separate Vulgar by ignoring it when attested and adding it to the Appendix namespace when not attested. —Μετάknowledgediscuss/deeds 01:30, 18 March 2013 (UTC)
Hello? —Μετάknowledgediscuss/deeds 16:53, 24 March 2013 (UTC)
Oh, I didn't realise I was supposed to reply to this; I'm sorry. I don't see a problem with mapping Old Latin to Latin. Considering FHEFHAKED and fecit to be the same language seems no harder than considering Middle English, in all its variation (might, mikte, misten, mauht...), to be a language. As you suggest, it doesn't seem worth it to create a new code/L2 for such a small corpus of terms that can be mapped to {{la}}. - -sche (discuss) 22:54, 24 March 2013 (UTC)
Sorry about that; I guess I was just relying on you to make a decision one way or another. {{la}} it is, then. —Μετάknowledgediscuss/deeds 23:06, 24 March 2013 (UTC)
On one hand I see value in treating Old Latin as a distinct language, because it simplifies the treatment of inflection tables. One of the main changes from Old to Classical was the reduction of vowels and diphthongs, so that makes our current set of tables pretty much useless for those old forms. We could, of course, add them to the normal Latin tables, but those would not be helpful for the majority of users and just clutter up the table. So in a sense, we'd consider Old Latin (as a language) to be a "dumping ground for archaic forms that are not found in classical texts", which keeps the classical Latin entries tidier. On the other hand, Old Latin only really ends a hundred years or so before the time where a lot of the classical literature was written. So Old Latin and Classical Latin really compare more to Middle and Early Modern English; from Plautus to Cicero is more or less the same as from Chaucer to Shakespeare, after all. —CodeCat 18:54, 26 March 2013 (UTC)
I wasn't really thinking of making Old Latin inflection tables. The language was far too irregular and unpredictable, even. The corpus is quite limited, too, so I fear we'd have to rely on guesswork or PIE far too much. Keeping Classical Latin tidy is important to me, but I think that we can preserve that with Old Latin forms marked with {{Old Latin}}. I really just want to get inscriptional forms recorded here. As a side note, in my schooling, Shakespeare was taught from the original, but Chaucer from a translation. —Μετάknowledgediscuss/deeds 20:06, 26 March 2013 (UTC)
Even if some endings or some words are sufficiently well attested to merit being in tables, I don't think "considering them Latin" and "keeping them out of the Classical entries' tables" are contradictory. When the Classical Latin spelling and the Old Latin spelling of the citation form of the word differ, the tables have no reason to be mixed: sociis can be in a table at socius while sokiois can be in a table at sokios. - -sche (discuss) 15:10, 27 March 2013 (UTC)


/au.sum/ > /au.zum/ > /au.sum/[edit]

Discussion moved from User talk:Chuck Entz.

This is in reference a revert (diff) of changing the pronunciation in ausum to show intervocalic voicing of the s.

I thought intervocalic single-s in Latin was voiced as /z/ before it became /r/ or /ɾ/ in later Classical Latin. Is /au.zum/ not the older form of /au.rum/ before rhotacization? —This unsigned comment was added by 76.243.38.4 (talk) at 12:22, 15 February 2014 (UTC).

I could be wrong, but I believe that would be Old Latin, not Classical Latin. In cases such a honos/honor where the change was later, the intervocalic sound change happened first, then was later extended to the rest of the paradigm by analogy. Also, many sources attribute later ausum to the Sabines, who retained intervocalic s. I really don't know for sure, but I felt it was better not to change it without evidence or references. I wouldn't object to presenting both, but completely eliminating the usual interpretation without evidence didn't seem right. Maybe User:Fsojic might have more information. Chuck Entz (talk) 16:40, 15 February 2014 (UTC)
The consonant was certainly voiced in Old Latin, and was probably voiced since Proto-Italic times (it's the only way to account for changes like -sr- > -br-). But I don't know if the voicing was phonemic, it probably wasn't. Other Italic languages have similar allophony for "f" as well, which was voiceless initially but voiced medially. Latin had no medial "f" so it didn't occur there. —CodeCat 19:38, 15 February 2014 (UTC)
And although there isn't really any evidence for it, it's highly likely that /h/ was voiced word-medially in Latin as well. —CodeCat 21:04, 17 February 2014 (UTC)
In Phonetics and Philology: Sound Change in Italic Jane Stuart-Smith page 52 is written the following paragraph:
In Latin intervocalic *-s- became /r/ by the middle of the fourth century BC; for example, aurora 'dawn'<*ausōs, cf. Skt ușas- (Leumann 1977:178-80). We may assume that *-s- had become voiced [z] before this date (represented by S). We shall see that the intervocalic voicing of *-s- to [z] probably dates back to common Italic (cf. rhotacism in Umbrian; Meiser 1986: 38).—This unsigned comment was added by 76.243.38.4 (talk) at 20:53, 17 February 2014 (UTC).

It seems to me there are three issues:

  1. Is this referring to Old Latin, Sabinian, or some combination of the two?
  2. Based on the above, was this voiced in the Old Latin, but not in the Sabinian, or in all cases?
  3. If it was voiced, but strictly due to a predictable rule, should z be used in a phonemic representation (I assume from the "/ /" that's what we have here).

In any case, I'm a little out of my depth, so I'll let others decide. Chuck Entz (talk) 22:15, 17 February 2014 (UTC)

It's both. The change [s] > [z] medially is Common Italic, and is parallelled by [ɸ] > [β], [θ] > [ð] and [x] > [ɣ]. Most of those sounds don't survive as such in Latin, but the voicing distinction is clear because it affects the outcome. Aside from [z] > [r], there's also [β] > [b], [ð] > [d] and sometimes also [ɣ] > [g], whereas [ɸ] and [θ] both develop into [f]. w:Proto-Italic has more about this. So [z] was the pronunciation of these occurrences in all Italic languages, including Old Latin and Sabine. The voicing was definitely allophonic to a large degree. In the Osco-Umbrian branch of Italic, [ɸ] and [β] (or [f] and [v], it's hard to tell) are both spelled as "f", just like [s] and [z] are both represented as "s" in those languages as well as Old Latin. So using /s/ for the phoneme would be correct, but it would be phonetically misleading, and it's only the [z] allophone that becomes [r]. —CodeCat 22:34, 17 February 2014 (UTC)

comparative & superlative adjective lemma dilemma[edit]

Do the root forms of an adjective's comparative and superlative inflections get lemma status?

for example:
virīlis, virīlior, virīlissimus

Shall these all be included in both "Latin adjectives" and "Latin lemmas"? Naturally, I'll eventually wonder this about adverbs, also. Charts of the appropriate linkage to appurtenant cumulative files for each POS would help tighten up these conventions; maybe someone can direct me to these. I'm new around here.

We actually have two different categories for comparatives and superlatives, depending on lemma status. Treated as a non-lemma, they go in "adjective comparative forms", while treated as a lemma, they go in "comparative adjectives". I don't know which one is used for Latin though. —CodeCat 19:55, 6 August 2014 (UTC)
Thanks CodeCat, very helpful. So, I guess I should have asked: do we consider the Latin comparative & superlative root (nom. masc. singular) forms to be lemmas, or just yet two more of the dozens of forms which result from Latin inflection. I'm a beginning Latin student, not any sort of lexicographer (other than the Wiki sort). Based on the templates that I've encountered, it appears that there might be two schools of thought. --Doebee (talk) 07:20, 7 August 2014 (UTC)
All right, I've convinced myself that each part of speech should only have one lemma, ergo: root comparatives & superlatives are just inflected forms, not lemmas. I bet I'll even stumble across a Latin comparative or superlative that has no positive root form, and thus might have to be considered a lemma. Ah, how about bonus -> melior -> optimus ! ...Doebee 17:57, 7 August 2014 (UTC)
There is already a headword template for Latin comparatives and superlatives: {{la-adj-comparative}} and {{la-adj-superlative}}. —CodeCat 19:22, 7 August 2014 (UTC)
Thanks CodeCat, got that. How can I determine which accumulation files a given template contributes too? Example: la-adj-3rd-2e contributes to "Latin Lemmas", and to "Latin Adjectives", but I can only discover that by executing code. Some of the templates that I have come across, apparently aren't documented (case in point: "Template:la-adj-comparative"). Anyway, thank you for directing me to sandbox, but I don't think the sandbox will help me understand where the templates are putting new entries... The take away is, thank you for keeping me from harming anything; I'm working on it. --Doebee 19:59, 7 August 2014 (UTC)
The categories (that's the official term) that a template adds a page to can change depending on the parameters given to the template, as well as the page the template is placed on. So there is no easy way to determine it other than actually adding it to a page and clicking "preview". You could also try Special:ExpandTemplates, which lets you see exactly what text is added by a template on a given page, including the categories. —CodeCat 20:23, 7 August 2014 (UTC)