Wiktionary talk:About Latin

• /Archive 1 (2006–09)
• /Archive 2 (2010–16)

Vowel length

First, if this has been discussed already somewhere else, I apologize.

The current standards state that macrons indicate long vowels, and forbid the use of the breve, saying that any vowel that is not long must be short. However, Latin was generally not written with indications which vowels are long and which are short. If none of the orthographic tools were ever used to indicate that a specific vowel is long, we can infer its length in several other ways:

• if the syllable is open (including cases of muta cum liquida), we can infer the length quite certainly in the majority of cases from the accent placement or the meter.
• if the syllable is closed, we cannot use the above methods. We can still use etymology and sound changes affecting vowel length that we know happened to determine the length. If we are lucky, a Roman grammarian will have commented on the length of the vowel.

The point is that it can happen that in some cases we simply cannot know whether it was short or long, beyond a random guess. If Wiktionary is to be exact, it needs a three-way distinction between 'long', 'short' and 'we don't know'.

Another problem is the distinction between syllable weight and vowel length. Generally, macrons are used to mark long as opposed to short vowels, whether in a closed or open syllable e.g. natūra and cōnsul. However, in cases where a double consonant is written as a single consonant, e.g. maior for [majjor], the preceding vowel is given a macron even if it is short, i.e. māior (see W. Sidney Allen, Vox Latina). In such cases the macron represents a heavy or long syllable, not vowel length. This results from an incomplete understanding of Latin pronunciation, orthography and/or phonology. Such notation is inconsistent, misleading and obscures the length of the vowel in case we do know its actual length (by one of the methods for closed syllables described above). While it is true that this practice is present in many dictionaries, it is based on wrong conclusions and should therefore be abandoned.

The guidelines should address these issues. 109.123.31.174 18:03, 14 February 2017 (UTC)

In some way some parts of your question were asked at Wiktionary:Information desk/2017/February#Latin vowel lengths (from February 2017 too), but this place is better anyway. In that discussion there is a focus on different vowel lengths mentioned in different dictionaries, but it also addresses New Latin terms with uncertain vowel lengths.

Sometimes when the length is unknown both macron and breve could be in use like "ā̆" - though it could also mean or be misunderstand as both long and short, so it also requires a note.

But there might be entries which simply use vowels without diacritics (simply a, e, i, o, u) but without indication of a certain vowel length. Like if one finds a Medieval or New Latin term and one doesn't know the length, one simply doesn't use any diacritics - which might incorrectly mean all vowels are short, when in fact they are unknown (at the moment or to the contributing editor).

-Slœtel (talk) 23:52, 18 February 2017 (UTC)

@109.123.31.174, Slœtel: Re “If Wiktionary is to be exact, it needs a three-way distinction between 'long' [¯], 'short' [˘] and 'we don't know' [ ].”, I think I agree with that. Lewis & Short, Gaffiot, and many others use both the macron and the breve, whereas the Oxford Latin Dictionary uses almost exclusively the macron (it has “sibī̆” and uses breves for i-initial words; the latter use is primarily a mark of vocality, however); I'd love to know the different rationales, especially the OLD’s. I definitely agree with your point about maior and its ilk; the macron should indicate long vowels, not heavy syllables. It might be worth bringing up these points in the Beer parlour, to attract a larger audience. Re combining-⏓, that should only be used for vowels that legitimately vary in length (sibī̆, for example). — I.S.M.E.T.A. 11:05, 3 March 2017 (UTC)

@Slœtel, I'm so meta even this acronym#NO BOLD: Actually, I would suggest using a special symbol for 'we don't know', while keeping unmarked (or optionally for emphasis, the breve) for 'short'. Three reasons. First, readers are used to the current convention, so they are used to unmarked vowels representing short vowels. Therefore, anyone that is unaware of the change in convention is likely to interpret the unmarked vowel incorrectly. Second, less words and less vowels need to be updated, making it easier to apply the convention to all the entries. Third, a new symbol will be unfamiliar to the reader, making them wonder what it means and finally leading them to a note saying that vowel length is not known. That is, it would make the vowels with unknown vowel length stand out from the rest, which would be a good thing. We would then get minimally different natūra, cōnsul, maior, sibī̆, Tra̚iānus (just as an example).

I wanted to bring this to attention, but I do not really have the time to follow this discussion. So if this needs to move on to the Beer parlour, I ask someone else to do what is necessary. Thank you! 109.123.31.174 16:13, 8 March 2017 (UTC)

@109.123.31.174: I take your points, but ˺ (U+031A COMBINING LEFT ANGLE ABOVE) is definitely not the right diacritic for this purpose — for one thing, it's barely distinguishable from the macron. The most intuitive mark for a vowel of ambiguous quantity would be a COMBINING QUESTION MARK ABOVE; however, such a mark appears not yet to have been encoded in Unicode, although Martin Schrage and Karl Pentzlin have proposed such an addition in these three contributions. Without a coherent proposal, I don't think this is worth taking to the Beer parlour; it needs more work before then. — I.S.M.E.T.A. 22:01, 3 April 2017 (UTC)

Vowel length of Latin first declension nouns borrowed from Greek and ending in a (ā or ă)

This (IMHO) is a too important matter so that it doesn't belong on a monthly WT:ID or WT:TR discussian page.
Quoting from: Friedrich Neue, Formenlehre der Lateinischen Sprache. Erster Theil, Stuttgart, 1866, p.53-56, §.18 (GB):

"Griechische Nomina auf ᾱ bewahren im Lateinischen zuweilen die Länge des Endvocals. Prisc 6, 2, 10, S. 681: Hoc etiam sciendum, quod omnis nominativus in a desinens corripitur, quamvis sit apud Graecos productus, ut Lydia, Syria, Phrygia, Italia, Hispania; unde accentus quoque cum tempore mutatur. In paucis tamen inveniuntur poetae Graecis servasse morem Graecum. Statius in IIII Thebaidos (V. 287): Non Tegéa, non ipsa deo vacat alite felix Cyllene. Idem in VI (V. 508): Te plangeret Argos, te Neméa, tibi Lerna comas Larissaque supplex poneret. Derselbe 7, 2, 5 S. 720: Apud Statium Nemeā, quia servavit a productam, accentum quoque Graecum servavit, id est paenultimam acutam, in V Thebaidos (V. 44): Nec faciis Neméā latas evolvere vires.
So ist Plaut. Epid. 4, 1, 10 Cantharā, Rud. 1, 4, 17 nach wahrscheinlicher Vermuthung Palaestrāne. Derselbe Mil. 4, 2, 73 hat Aetnā mons. [...] Aber Verg. Ge. 4, 343 hat Ephyrā nach dem Med., sonst Ephyre; und Aen. 3, 702 Gelā. Tibull. 2, 5, 67 Amaltheā, Prop. 3 (2), 14, 5 Electrā. Ovid. Amor. 3, 1, 7. 3, 9, 3. rem. am. 379. Her. 15, 7 elegiā, A. A. 1, 511 rem. am. 743 Phaedrā, Fast. 4, 177 Electrā, 4, 201 Rheā (als Name einer Göttin), 5, 115 Amaltheā, 6, 501 Leucotheā, Her. 14, 1. 53 Hypermnestrā. Colum. 10, 235 einarā. Sen. Herc. fur. 203 und wohl auch 1009 Megarā. Bei Stat. ist außer den von Prisc. angeführten Beispielen von Tegeā und Nemeā noch Theb. 1, 355. 2, 378. 4, 150 Nemeā, 2, 286 Pasitheā, 4, 45 und wahrscheinlich 7, 331 Mideā, 4, 224 Maleā, 12, 178. 255. 303. 373. 391. Argiā, Mart. 12, 31, 9 Nausicaā (nach dem Voss. c, in mehreren anderen wie Put. Voss. A B C, ist Nausica, in einigen Nasica), und 14, 187, 2 Glycerā (nach dem Thuan., in welchem nec licera ist, und anderen Büchern). Aber wie an die Stelle des Griech. η im Latein. oft ein kurzes a tritt, so ist auch das lange α [= ᾱ] des Griech. im Latein. an vielen Stellen verkürzt, selbst in den nämlichen Nomina, welche nach dem Obigen zuweilen das lange a [= ā] behalten. [... (examples with short a; like 3/4 of page)] und die adjectivisch gebildeten Namen von Frauen und von Ländern auf ia werden wie Latein. Nomina behandelt. [...]
Zu dem Nomin. auf ā gehört ein Accus. auf ān, Andromedan Ovid. Met. 4, 756, Argian Stat. Theb. 4, 91, Electran Ovid. Trist. 2, 395, Glaphyran und Glyceran Mart. 11, 20, 3. 11, 40, 1, Malean Lucan. 9, 36. Stat. Theb. 10, 537, Odyssean Auson. epitaph. 5, 2. In diesen Stellen ist am unstatthaft wegen des Vocals im Anfange des folgenden Wortes, hier und da aber ist die Var. -en vorhanden. Die Endung ān findet sich vor Consonanten in Andromedan Ovid. A. A. 1, 53. Met. 4, 670, in Electran Ovid. Fast. 4, 32. 174, in Megaran Mart. 11, 43, 6, in Phthian Ovid. Met. 13, 156; in einzelnen dieser Stellen ist als Var. theils -en theils -am bemerkt. Solche Nomina, welche im Griech. ein kurzes α haben, können auch im Latein. den Accus. auf kurzes an bilden. Prisc. 7, 2, 5 S. 730: Accusativum quoque Graecum in femininis invenimus a poetis proferri, sed raro. Statius in VII (V. 319): Namque ferunt raptam patriis Aeginan ab undis. So Camarinan Ovid. Fast. 4, 477, Cillan Met. 13, 174, Deidamian Stat. Achill. 2, 367, Iphigenian Ovid. Pont. 3, 2, 62, Maian Fast 4, 174, Orithyian Met. 6, 707, Ossan Prop. 2, 1, 19. Ovid. Fast. 1, 307. Sen. Herc. Oet. 1735. Aetna 49. [...]
Auch in Prosa stehen zuweilen Accus. auf an sowohl von Nomin. auf ā wie auf ă. Varro bei Non. S. 383 Odyssian (in denselben Worten bei Non. S. 480 wird Odyssiam gelesen), Cic. Divin 2, 27, 59. 2, 37, 79 Politian und tetrarchian, Liv. 45, 31, 14 Antissan, Plin. 35, 4, 10. 35, 10, 36, 101 Nemean und Nausicaan, Flor. 2, 8, 9 Euboean (doch im Nazar. Euboeam)."

That is:

Greek nouns in ᾱ sometimes retain the length of the last vowel in Latin. [Latin quotes]
So is Plaut. [..] Cantharā, Rud. [..] according to a likely conjecture Palaestrāne. [Other examples: Aetnā mons [...] Ephyrā (else Ephyre), Gelā, Amaltheā, Electrā, elegiā, Phaedrā, Electrā, Rheā (name of a goddess), Amaltheā, Leucotheā, Hypermnestrā, einarā, Megarā, Nemeā, Pasitheā, Mideā (likely), Maleā, Argiā, Nausicaā (depending on text/edition), Glycerā. Note: By context it's said that these should all be nominatives (not ablatives) in -ā.] But just like Greek η is replaced by a short a in Latin, the long α of Greek becomes short in many places in Latin, even in the nouns which according to the before sometimes retain the long a. [Examples of short a, and sometimes in words which can also have a long a according to the foregoing text, like Electră, elegiă, Phaedră, Rheă (name of a priestess), Pasitheă, Argiă, Cassandră, Clepatră, Ledă and more.], and in the adjectivaly formed names of women and countries in ia which are treated like Latin nouns." To the nominative ā belongs an accusative in ān, [examples]. In all these places am is inadmissible because of the vowel in the begining of the following word, but here and there is the variant -en. The ending ān occurs for consonants in [examples: Andromedan, Electan, Megaran, Phthian]; in some of these places is as variant partly -en partly -am a noticed. Such nouns which in Greek have a short α, can have in Latin the accusative in short an. [Latin quotes] [Examples: Camarinan (from Greek Καμάρινα (Kamárina) which here in WT has a short vowel ᾰ), Cillan, Deidamian, Iphigenian, Maian, Orithyian, Ossan, Aetna.] [...]
Even in prosa sometimes stand accusatives in an both from nouns in ā and in ă. [Examples: Odyssian (in another place Odyssiam), Politian and tetrarchian, Antissan, Namean and Nausicaan, Euboean (but somewhere Euboeam).]

I guess that should imply the following:

• Vowel length in Latin terms coming from Greek depends on the Greek, that is when it's -ᾱν in Greek, it's always -ān and not -an in Latin and when it's -αν in Greek it's always -an and not -ān in Latin. (In theory less educated Romans and pupils sometimes could have used -an for Greek -ᾱν and -ān for Greek -αν, but this would have been an error.)
• If one assumes that a Latin -an (be it -ăn or -ān) is -ān, the nominative should also be -ā. Though it could also have an alternative nominative in -a and accusative in -am. (In practice it could be that sometimes only a accusative in -ān and a nominative in -a are attested, but that's just like with many other forms, especially vocatives, which aren't attested.)

PS: Dictionaries often do not mention a long -ā just like they sometimes don't mention an accusative in -an/-ān or some proper nouns at all; while sometimes they have a form with -ē besides -a or forms with additional vowels instead of -ā like "ĕlĕgīa (ĕlĕgēa, in Ov. ĕlĕgēĭă) [...] form elegeia, Ov. Am. 3, 1, 7; 3, 9, 3; id. R. Am. 379." in L&S. But Georges at least has "Nomin. Phaedrā gemessen bei Ov. art. am. 1, 511." which would mean that the nominative can have a long -ā. So just like it's done in other cases, it should be justified and make sense to assume vowel lengths like in Greek. -Slœtel (talk) 23:52, 18 February 2017 (UTC), 01:03, 19 February 2017 (UTC)

@Slœtel: What about Latin borrowings ending in -ia from Greek feminine abstract nouns ending in -ῐ́ᾱ (-íā)? There are a lot of them. — I.S.M.E.T.A. 11:09, 3 March 2017 (UTC)

@I'm so meta even this acronym: The text above should say that an educated Roman could have used it in a Greek way with -iā in the nominative and -iān in the accusative. When Latinising it just like changing Greek -os into -us, -on into -um, it would become -ia and -iam. But if you are asking for examples, I'm don't know if there are any.

• -os and -us are different in writing and one can differ between them, but -ā and -a would look the same in prose. There are ways to mark long vowels, but they are optional.
• Depening on edition one can find elegeia/Elegeia/Elegeïa or Elegia from ἐλεγεία (elegeía) in Ovid, which should be ĕlĕgēĭă (L&S) or ĕlĕgīā (Friedrich Neue above has ā, and ει becoming ī is common). IMHO ĕlĕgīā makes more sense than ĕlĕgēĭă with a change from ε (e) to ē. But that should be the feminine -είᾱ (-eíā) from -εῖος (-eîos) and not -ῐ́ᾱ (-íā).
• It should be very likely that one can find an accusative in -ian/-iān from a Greek word ending with -ῐ́ᾱ, at least in New Latin. It should be -iān and when the nominative is nearby it should be -iā. But I doubt that there is a clear proof for the vowel length. Ways to proof a long vowel are finding a text which marks long vowels or finding a good poem; ways to assume but not to proof a long vowel are finding amnestia near to amnestian.
  • In case of New Latin one could have doubts regarding the vowel lengths. -as/-ās often is short in biblical proper nouns like Lucas in modern European languages and it's short in Church Latin according to {{la-IPA|Lūcās|eccl=yes}}. Just like that -as/-ās could have been short sometimes in New Latin, and so -iān could have become -ian. But as New Latin went back to the roots, -ās would be the proper form.
  • Valerius Maximus (1st century A.D./C.E.) has "quam Athenienses amnestian uovant" (The Latin Library). Compared with ᾰ̓μνηστῐ́ᾱ (amnēstíā) it should be amnēstiān and then the nominative should also be amnēstiā like in Greek. But I don't know if one can clearly attest this nominative. This book at GB has amnestia near to amnestian, so one can at least assume that it is amnēstiā.

-Slœtel (talk) 19:08, 8 March 2017 (UTC)

@Slœtel: I'm happy to accept the validity of the inference of a nominative -iā from an attested accusative -iān (with or without direct evidence of vowel length). I think, however, that it's a bit of a stretch to infer from the attestation of some -iān forms that Greek-type declension and vowel lengths should be rolled out as standard to the entire class of nouns ending in -ia that derive from an etymological -ῐ́ᾱ. — I.S.M.E.T.A. 22:35, 3 April 2017 (UTC)

@I'm so meta even this acronym: As far as I know the accusative in -ān (or more specific -iān) with nominative -a/ā (or -ia/-iā) is rare anyway. Terms where I found a mentioning of an accusative -iān in a dictionary where aetiologiān and sphaeromachiān (both in Sen. Ep.), analogiān (Quintilianus), polītīān and tetrarchia (Cicero, both in the same text), theologiān (Augustinus). While Lewis & Short, Gaffiot and online Pons do not mention this accusative, Georges has it although often without any specific cite or reference and OLD has it with a citaton. For many other terms both Georges and OLD don't give an accusative with -īān, e.g. not for acyrologia, aeschrologia, alogia, Amantia, amphibologia (or amphibolia), anōmalia, aperantologia, apologia, astrologia, carphologia, cōrycomachia, crīthologia, dēmarchia, dilogia, ecclēsia, ēthologia, etymologia, eulogia, geneālogia, genethlialogia, hysterologia, Lamia, macrologia, monarchia, monomachia, nauarchia (or nāvarchia), naumachia, palillogia, perimachia, perissologia, philologia, philosophia, physiologia, pithanologia, prosōdia, tautologia, toparchia, trāchīa, tropologia (some of these words aren't in the OLD). And even if an accusative in -ān (or -iān) is attested, it could be rare and an exception anyway. Personally, I would usually assume nominative -a (or -ia) with accusative -am (or -iam), unless an accusative in -ān (or -iān) or somehow a nominative in -ā (or -iā) is attested. If an accusative in -ān (or -iān) is attested, there could be a usage note regarding the nominative. E.g. as Augustinus has theologian multiple times, his theologia could have -ā too, but one can't see the vowel length.
Just be the way, similary, I would usually assume that terms from Greek ending in -eus (-εύς -eús) have a diphthong, although it can have two syllables which likely is rare too (Phalereus is attested as Phălērĕŭs).
If New Latin is concerned, -ān might be less rarer than in ancient Latin, but that's New Latin anyway. -Slœtel (talk) 20:02, 3 May 2017 (UTC)

Comparatives and superlatives

@I'm so meta even this acronym, JohnC5, kc kennylau, CodeCat, EncycloPetey, Wikitiki89, Slœtel, I've been meaning to bring this up for a while now, but rather unusually on here, comparatives and superlatives of Latin adjectives (e.g. altus) are generally neither indicated in the headword line nor given in the declension table itself, but simply listed below the table. Is this something we want to keep, or should it be changed? In any event, it should be written in to the relevant module (I already asked Metaknowledge, and he thinks it should be changed). Esszet (talk) 01:42, 8 April 2017 (UTC)

Comparatives and superlatives were groups that I didn't have a clear idea how I thought we should present them. On the one hand, they are "forms" of the base adjective, but in another sense, they have their own "lemma" and declension, and can carry meanings that are not present in the standard adjective. Since I was wavering about how to tackle this, I never set up anything in the templates, nor generated more than a couple of test cases on Wiktionary. --EncycloPetey (talk) 01:48, 8 April 2017 (UTC)

I'd like to see them in inflection tables, but their full inflection doesn't have to be there. In other words, the inflection table of altus should list altior and altissimus, but things like altiōrem and altissimō wouldn't be at altus but only in the inflection tables of altior and altissimus themselves. —Aɴɢʀ (talk) 07:31, 8 April 2017 (UTC)

I'd prefer to see them in the headword line, but yeah, not the full inflection, just the two lemmas, as in the Polish word wysoki. Esszet (talk) 11:19, 8 April 2017 (UTC)

{{la-adj}} already has parameters for them, so we might as well use them. —CodeCat 12:10, 8 April 2017 (UTC)

@Esszet: As CodeCat intimated, many (if not all) of the adjective headword templates have |comp= and |sup= parameters, which can be used to specify an adjective's comparative and superlative degrees (as I've just done for altus). Pace Aɴɢʀ, I'd rather not see the comparative and superlative degrees in the declension tables, if for no other reason than the fact that some adjectives are comparable and others are not will result either in incorrect information being presented or in aesthetic inconsistency in presentation; however, if you still want to add the comparative and superlative degrees to the declension tables, consider the way ObsequiousNewt did that for Ancient Greek adjectives with {{grc-adecl}}. — I.S.M.E.T.A. 00:33, 10 April 2017 (UTC)

@JohnC5, kc kennylau, Wikitiki89, Slœtel Any input? Esszet (talk) 19:58, 15 April 2017 (UTC)

So, having been the one to set up the |comp= and |sup= parameters in mod:la-headword, I'd prefer them in the headword. Indeed we could add acceleration if anyone uses such things. My main issue is the categories like cat:Latin comparative adjectives. The truth is that Latin comparative and superlative adjectives and adverbs do not function grammatically differently from normal adjectives and adverbs. On the other hand, I do like categorizing such things. I'd prefer Latin comparative adjectives to be both placed under the heading of cat:Latin adjectives and cat:Latin comparative adjectives as opposed to just the latter. —JohnC5 23:56, 15 April 2017 (UTC)

@JohnC5: I myself would like to keep the positive-degree adjectives separate. How about having a Category:Latin positive adjectives or Category:Latin absolute adjectives of them, and then having it as a subcategory of Category:Latin adjectives alongside Category:Latin comparative adjectives and Category:Latin superlative adjectives? — I.S.M.E.T.A. 00:34, 16 April 2017 (UTC)

@JohnC5? Subcategories seem to be the way to go here. Esszet (talk) 15:42, 26 April 2017 (UTC)

@Esszet: That could be fine. The current templates {{la-adj-comparative}} and {{la-adj-superlative}} are wildly insufficient. They should display exactly like the {{la-adj-3rd-2E}} and {{la-adj-1&2}} respectively with the added notation that they are the comparative/superlative degree of some positive adjective. —JohnC5 15:51, 26 April 2017 (UTC)

That kind of information goes in the definitions, e.g. {{comparative of}}. —CodeCat 16:47, 26 April 2017 (UTC)

True, then where should we add the Category:Latin comparative adjectives category? —JohnC5 16:59, 26 April 2017 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ Never mind, they already are subcategories of Category:Latin adjectives. Esszet (talk) 22:48, 26 April 2017 (UTC)

We can still use {{la-adj-comparative}}, and the comparative adjectives category should be added by that. I'm just saying that we shouldn't actually list "comparative of X" in the headword line. —CodeCat 23:14, 26 April 2017 (UTC)

I made some changes to {{la-adj-comparative}} and {{la-adj-superlative}}. —JohnC5 01:45, 27 April 2017 (UTC)

Did you mean something else? Those templates haven't been edited since November. Esszet (talk) 14:59, 27 April 2017 (UTC)

@Esszet: Well, technically, I edited mod:la-headword (poorly) so that the usage of {{la-adj-comparative}} and {{la-adj-superlative}} would remain the same, but the output would change. —JohnC5 15:34, 27 April 2017 (UTC)

Sorry, I didn't check the actual code. Would you mind getting rid of the "comparative/superlative of" headers? I also think that information would be better off in definitions. Otherwise, once someone gets a bot ready, I can put auto-generation in the module. Esszet (talk) 22:34, 27 April 2017 (UTC)

@Esszet: So, I can easily remove the "comparative/superlative of" text, though the text of the second positional parameter will need to be removed from entries. What do you mean by "put auto-generation in the module"? —JohnC5 04:38, 28 April 2017 (UTC)

@Esszet: I've made the changes. —JohnC5 04:47, 28 April 2017 (UTC)

Make it generate them automatically and create a |nc= parameter for adjectives that aren't comparable. Esszet (talk) 14:41, 28 April 2017 (UTC)

Actually, since there are a lot of adjectives that are comparable but don't have the comparative and superlative listed at all (e.g. gramineus), the best thing to do would be to create the |nc= parameter now, put it in where necessary, and do the bot run afterwards. That way, we won't have a lot of adjectives that are comparable listed as being not. Esszet (talk) 14:56, 28 April 2017 (UTC)

@Esszet: Ok, you were saying autogenerate from the positive templates to the comparative and superlative. I was confused because I had already added autogeneration of the neuter comparative in {{la-adj-comparative}} and of the feminine and neuter superlatives in {{la-adj-superlative}}, so I didn't know which autogeneration you meant. You're plan sounds good. Please tell me anything I can do to help. We also have a lot of comparative and superlative entries that are not so-labelled that will need to be converted. Also, should we make {{la-adv-comparative}} and {{la-adv-superlative}} now? —JohnC5 15:22, 28 April 2017 (UTC)

I guess, if for no other reason than to put them in categories. I just looked at Category:Latin adjectives, and since there are so god damn many, I don't think we're going to be able to put |nc= in manually. Of course, I also don't see a way for a bot to be able to detect grammatical uncomparability, so it looks as though we have a problem on our hands. Any ideas? Esszet (talk) 20:47, 28 April 2017 (UTC)

Yeah, I was curious how you intended to do that. Maybe @kc kennylau, CodeCat have some ideas. —JohnC5 03:46, 29 April 2017 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ Doesn't look like it, I guess just move existing things to the headword line for now and we'll add auto-generation later on. Esszet (talk) 14:49, 3 May 2017 (UTC)

@Esszet: I guess the comparative and superlative fit into the header like as {{la-adj-1&2|altus|alta|altum|comp=altior|sup=altissimus}} now in altus.
This short reply is put long below, but it's commented out as not to bother anyone.

As for the uncomparable/comparable distinction and categorisation, I see some problems. Often one can find a positive and then often one doesn't know whether or not it's comparable, so one can't say whether it's comparable or uncomparable. So one always should have to add something like "|nc=1" (not comparable = true, that is it's uncomparable) manually.

-Slœtel (talk) 20:02, 3 May 2017 (UTC), replaced an it at 20:26, 3 May 2017 (UTC)

What do you mean? Can you provide examples? Esszet (talk) 20:05, 3 May 2017 (UTC)

According to Wiktionary, English is comparable, while Chinese is not, and dead is sometimes comparable. So how should one know whether or not danicus/Danicus (= Danish) or mortuus (= dead) is comparable? One could search for all possible comparated forms and look into dictionaries, but that's much work. Hence Latin adjectives without having the parameters |comp= or |sup= in it, shouldn't be listed as uncomparable. Instead one should always have to add something like |nc=1, if one knows that an adjective is uncomparable. -Slœtel (talk) 20:26, 3 May 2017 (UTC)

I've edited the entry for Chinese so it no longer says it is not comparable. I found plenty of instances of "more Chinese" and "most Chinese" (with the adjective, not the noun) on Google Books: phrases like "Taiwan is more Chinese than China" and "that was the most Chinese thing I've ever done". —Aɴɢʀ (talk) 20:53, 3 May 2017 (UTC)

Ok, and thank you for mentiong it. Then maybe one could use another English adjective or a non-English adjective as example, but maybe one can find comparated forms of it too. Alternatively one could give the argumentation of some dictionaries, grammars or some people. canoo.net for example explains: "The so-called "absolute" adjectives cannot be compared." but with "Nevertheless, these adjectives are occasionally compared [...] This type of comparison is frequent in colloquial language and in advertising". While for English it should be relatively easy to find comparated forms of such adjectives (e.g. "most optimal" for optimal) and while it's sometimes possible for German to find colloquial comparated form, it could be different in Latin. Anyway, I don't think it would be a good idea to add adjectives without the parameters |comp= or |sup= in it into a category like Category:Latin uncomparable adjectives. -Slœtel (talk) 21:52, 3 May 2017 (UTC)

I guess where there is ambiguity, it should just be put in a footnote (if anywhere). I can't imagine things like 'more dead' and 'most dead' would need to be in the headword line if they're going to be included at all. Esszet (talk) 20:47, 7 May 2017 (UTC)

Since it appears that Slœtel goes long periods of time without contributing, does my idea sound good to everyone? Esszet (talk) 10:58, 15 May 2017 (UTC)

@JohnC5? Esszet (talk) 23:32, 21 May 2017 (UTC)

@Esszet: Quickly outline what you've finally landed on as a solution? I find this back an forth a bit hard to follow. —JohnC5 00:32, 22 May 2017 (UTC)

We really haven't reached a solution overall, the consensus at this point is just to move existing things to the headword line. Esszet (talk) 10:32, 23 May 2017 (UTC)

@Esszet: Which "existing things"? I'm probably fine with whatever changes, but I'm still not sure exactly what's going where. Anything is better than the current state of confusion. —JohnC5 14:28, 23 May 2017 (UTC)

If they're already there, just move them to the headword line if they aren't there already. Otherwise, we're going to have to figure out a way to make a bot detect grammatical uncomparability (if that's possible at all). Esszet (talk) 14:55, 23 May 2017 (UTC)

Cool, go for it, I say. —JohnC5 15:13, 23 May 2017 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ I pinged you because I have no idea how to work with bots. Esszet (talk) 17:10, 23 May 2017 (UTC)

I can sometimes do AWB stuff, but no bots. —JohnC5 17:37, 23 May 2017 (UTC)

Then @I'm so meta even this acronym, CodeCat, EncycloPetey, Angr? Anyone here know how to work with bots? Esszet (talk) 20:34, 23 May 2017 (UTC)

It's not just knowing how to work with them- you have to have your bot approved by a vote (see WT:BOT). The routine procedure is to make a request at the Grease Pit. If you ask nicely and the bot owners aren't all too busy, someone will usually help you. Chuck Entz (talk) 02:15, 24 May 2017 (UTC)

Alright, thanks, I'll do that now. Esszet (talk) 23:33, 24 May 2017 (UTC)

@Chuck Entz Sorry this is so late, but my request simply wasn't taken up…should I just ask again? Esszet (talk) 22:05, 3 November 2017 (UTC)

Appendices in need of cleanup

I recently discovered Appendix:Mutations of the letter and sound a in the Latin language, linked at a. It is very badly formatted and needs cleanup. The Latin words should be italicized and language-tagged, and probably linked; it should also be more consistently marked with macrons and breves. Some of the Greek was messed up. I began doing both these things, but there's more to do.

I am not sure about the capitalized words (representing the letterforms used in inscriptions); they should perhaps be converted to lowercase.

I think there may be other appendices like this, but as this one was not categorized, the others may also not be. This page was created by User:Liliana-60, who is now blocked from all Wikimedia sites. — Eru·tuon 02:58, 23 April 2017 (UTC)

The text seems to be: “A”, in Charlton T. Lewis and Charles Short (1879) A Latin Dictionary, Oxford: Clarendon Press - and L&S often doesn't mark the vowel length properly and uses capitals for inscriptions. 84.161.19.68 17:55, 26 April 2017 (UTC)

Wiktionary:Votes/2017-05/Clarifying LDL status of extinct languages

I have created this vote, which mainly impinges on Modern Latin. Although it does not change existing practice, please take a look if you're interested and vote when it starts in a week. —Μετάknowledge^{discuss/deeds} 02:42, 3 May 2017 (UTC)

If it impinges on Modern Latin, then it could change existing practice, and then there might be entries or information which would have to be deleted. -Slœtel (talk) 20:02, 3 May 2017 (UTC)

Page is incomplete

There should be a paragraph about Old Latin, with a reference to these discussions:

• Wiktionary:Beer_parlour/2014/February#A_proposal_to_treat_Old_Latin_as_a_separate_language_from_the_other_chronolects_of_Latin
• Wiktionary_talk:About_Latin/Archive_2#Old_Latin_inscriptions

as well as about New Latin, Contemporary Latin, etc. Don't know where to add that. --Barytonesis (talk) 19:29, 31 May 2017 (UTC)

Old Latin is considered its own language on Wiktionary, it has its own code itc-ola. So it should also have its own page Wiktionary:About Old Latin. Of course, that page could just redirect to this one to keep it all together. —CodeCat 14:17, 6 July 2017 (UTC)

Substantive adjectives/participles

I noticed that an IP and User:Robbie SWE were getting into an edit war at pigmentarius over whether {{lb|la|substantive}} in an adjective sense should be {{lb|la|masculine substantive}}. I "solved" this by splitting the substantive sense off into a noun POS. Then I looked at the edit history and realized that the adjective-with-substantive label setup was intentionally done that way by a veteran Latin editor and that this was standard practice. Given that I'm a completely self-taught la-1, I figured I should bring it up here before doing anything further.

The only discussion I could find in the archives here is from a decade ago, so it's possible that revisiting this might be beneficial. The problem, as I see it, is that substantive adjectives have a specific gender and function as nouns, but aren't categorized as nouns or by their gender. They also don't have their own headword line, so this kind of information has to be tacked on with the equivalent of paper clips and duct tape. Yes, I realize that these are also still adjectives, but this arrangement seems a bit ad hoc and clumsy. Chuck Entz (talk) 08:00, 6 July 2017 (UTC)

All quiet on the reversion front, I gave up on finding a reasonable solution so the edit war is averted. My reason for reverting this anon's contribution in the first place was that I haven't seen any other entry where {{lb|la|masculine substantive}} was used. With that said, I'm glad you brought it up so we can come to a feasible solution within the community. --Robbie SWE (talk) 08:58, 6 July 2017 (UTC)

As I understand it, substantive is synonymous with the modern definition of noun, so a "substantive adjective" should always be placed under a Noun header. In the past (starting in Latin and Greek), noun (or the equivalent) was a cover term for what we call nouns, adjectives, and most pronouns and determiners, because these were things that all had case forms. The historical term for what we call noun would have been substantive noun, I think, and adjectives were called adjective noun or (with French word order) noun adjective. Later, those were shortened to noun and adjective.

If there are any so-called adjectives of fixed gender with a meaning that is not clearly derived from the ordinary meaning of the adjective, they are probably nouns and should be categorized as such. However, if the meaning is just "someone or something that can be described with the adjective", it's probably not worth having a separate POS section, because you can noun any Latin or Greek adjective in that way, and so that would give several noun entries for almost every Latin adjective. Sort of like how you can add the to many English adjectives and create a plural noun. (However, we do have an entry for the plural noun poor.)

So I think you did the right thing here. L&S and LSJ typically treat substantivized adjectives under the adjectives that they derive from, but that is inconsistent with Wiktionary's practice, in which we always have separate POS headers for different parts of speech. (LSJ sometimes defines derived words under the lemmas from which they derive, which is even more inconsistent with Wiktionary's practices.) — Eru·tuon 09:27, 6 July 2017 (UTC)

In Latin gender is important. That is, gender has to be mentioned, and if it's just a label, then the gender has to be somehow in the label, be it {{lb|la|masculine substantive}}, {{lb|la|masculine|_|substantive}}, {{lb|la|substantive of masculine gender}} or whatever. Thus the reverts by the admins were no improvements.   Of course, alternatively the substantive sense could also be put in an own POS section.   In any case, the declension has to be considered too. For third declension substantivisations something like {{la-decl-3rd|NOM|STEM}} could be wrong as substantivisations could preseve the adjectival or i-stem declension. E.g. in in New Latin one can commonly find the ablative in plurali (short for in numero plurali or in plurali numero) instead of in plurale. For second declension substantivisations the only question is "What's the correct vocative singular of the masculine substantivisation?" as all other forms are the same. The new entry with a new separate POS section gives a probably unattested vocative which could doubtful - while the old entry with the label did somewhat conceal the correct declension of the substantivisation and while also in some way implying it's -ie which is probably unattested and doubtful too. Maybe properly the table in the separate POS section should have "|voc_sg=?" or something else in it.
(Note 1: mānuārius could imply that the vocative of masculine substantivisations is -ī, but that's just one and maybe a not so good example. Note 2: ēgregius, genius - and possibly vulturius/volturius - which are other words ending in -ius and have an attested masculine vocative singular don't proof anything for substantivisations like pigmentārius. Note 3: Old grammars sometimes give a rule like "proper nouns have vocative -ī while common nouns have -ie" or "proper nouns have -ī while common nouns except some exceptions (like filius, genius) have -ie", but the first is wrong and the second probably without attestion.) -80.133.117.77 13:31, 9 July 2017 (UTC)

Fourth principal part

Wiktionary always lists the supine as the fourth principal part of a Latin verb. But my Latin language textbook (Ecce Romani), a commonly used study program for the Latin language, lists the passive perfect participle as the fourth principal part instead. For example, we have "dico, dicere, dixi, dictum" here on Wiktionary, but "dico, dicere, dixi, dictus" in my textbook.

Which type of fourth principal part is more commonly used? I'm guessing the supine, but my Latin teacher has seldom encountered usage of the supine as the fourth principal part. I'm from the U.S; does it depend on geographical region or something? Please help me out here, and thanks in advance. Johnny Shiz (talk) 21:22, 20 March 2019 (UTC)

@Johnny Shiz: I grew up in the U.S. too, but my Latin textbooks in junior high and high school always used the supine as the fourth principal part. I think the reasoning is that intransitive verbs don't necessarily have past passive participles, but all verbs have a supine. —Mahāgaja · talk 07:22, 19 April 2019 (UTC)

@Mahagaja: Makes sense. It should also be mentioned that for intransitive verbs, the future active participle is used instead; hence, "ambulo, ambulare, ambulavi, ambulaturus." Johnny Shiz (talk) 20:42, 22 April 2019 (UTC)

Latin verbs in etymologies of other languages

@Rua: this section could be made more general. Often I see etymology sections of modern Romance-language nouns that take the modern form back to the Latin accusative instead of the lemma nominative (e.g. "from Latin patrem" instead of "from Latin pater"). So we do need to say that the "lemma from lemma" principle holds for nouns too. That said, however, I feel like there could be a few exceptions. For example, I'd be in favor of explicitly pointing out that French rien comes from the accusative rem, if only because monosyllabic accusative singulars are so rare in Latin, and this is one of the few cases where the nasal at the end of the Latin word actually survives (in the form of a nasalized vowel) in the modern language. Also, in cases like Portuguese Deus, Spanish Marcos, and French Jacques, it could be worth pointing out that the modern lemma form exceptionally comes from the Latin nominative, not the accusative as usual. —Mahāgaja · talk 07:20, 19 April 2019 (UTC)

Yes, we can note the exceptions when applicable. —Rua (mew) 11:18, 19 April 2019 (UTC)

• What about providing non-lemma forms but linking to the lemma forms? E. g. {{der|en|la|exigo|exigere}}. Ain92 (talk) 12:26, 7 May 2020 (UTC)

  I dislike that because it seems to violate both the spirit of the "lemma from lemma" principle and the principle of least astonishment (if I click on a link that says "exigere" I don't expect to be taken to "exigo"). —Mahāgaja · talk 12:30, 7 May 2020 (UTC)

  @Mahagaja: Neither do I discern a lemma from lemma principle nore is it astonishing to click on a citation form and land on another citation form (Latin verbs have four citation forms one seeks anyway, Arabic ones two …). Fay Freak (talk) 12:48, 7 May 2020 (UTC)

Allowing macrons in quotations

I consider the rules stated on this page quite antiquated. In my opinion, an ideal Wiktionary Latin entry would have fully macronised quotations, at least for non-Medieval works. A prohibition on macrons seems very unproductive; while discussions on their usefulness have been had before, I propose that the use of macrons be made optional. This is a dictionary, after all—a lexicographic resource, and it were ideal to offer readers the maximum amount of help in an entry. ―Biolongvistul (talk) 10:57, 21 April 2020 (UTC)

@Biolongvistul It is indeed. Now we add stress marks to Russian texts which didn’t have it (so dictionary users can partake of native Russian speakers knowledge of the complicated patterns) and we vocalize Arabic quotations. And texts from antiquity are always manipulated for modern readership anyway. Fay Freak (talk) 12:53, 7 May 2020 (UTC)

Laxening other orthographic rules for quotations

Several years ago other medieval/renaissance/early modern diacritics were prohibited in quotations, but in 2015 creating such forms were allowed, how are they supposed to be illustrated? Demanding orthographically normalized quotations doesn't make any sense there, and indeed @Desaccointier have been breaking these rules in cõfundantur etc. Ain92 (talk) 14:18, 7 May 2020 (UTC)

Admittedly, I had not entirely familiarised myself with the regulations when I created that entry, as well as others of words scribally abbreviated; it was a mistake on my part to neglect to do so, instead merely following the language-wide guidelines set out in WT:QUOTE. I will try to fix the quotations in these in accordance with current rules best as can be done, but I very much agree with you in that not allowing the rendering of quotes with scribal abbreviations makes little sense when it seemingly makes impossible the unique examplification in citation of many apparently permissible entries' terms. —⁠Desacc̱oinṯier 14:56, 7 May 2020 (UTC)

Of course if your point is to show abbreviations then it is totally cromulent to go the whole hog and transcribe the text diplomatically; the policy page is an incomplete think tank. But the main issue is that Unicode is not quite apt for computerizing abbreviations. And for the page itself, nobody will ever search such Unicode encodings online therefore. Fay Freak (talk)

Vowel length on entries prefixed with in-

As of late I've noticed a number of pages which all have the same error, Latin words prefixed with in- have their initial vowel lengthened to īn- as I've seen them I've been correcting them to in- after verifying them against Lewis and Short's and the Oxford Latin Dictionary. I don't know if this is a user who is accidently contributing this error or a bot. Most recent page I corrected was insurgo. While I can spot them instantly, just knowing they look off, any inexperienced Latinist could easily be mislead here. The most recent edit correcting the page is from my IP, I usually look up entries in incognito mode since I'm checking 100s of words a night for my projects. What's the best course of action here?

Found the culprit "update macrons in la-adj-form: if before two cons, per Bennett corrected by Allen/Michelson; update macrons in inflection of: if before two cons, per Bennett corrected by Allen/Michelson" the long vowel before two consonants doesn't apply in these situations, since in- is a prefix, so and there are many other exceptions as well, just had to fix infestō from īnfēstō, both Lewis and Short and the Oxford Latin Dictionary confirm this is incorrect. Can these changes be mass reversed? — This unsigned comment was added by ColinJParry (talk • contribs) at 05:06, 21 April 2021‎ (UTC).

@ColinJParry: who says the prefix in- doesn't have a lengthened vowel before s and f? You can't check with L&S or the OLD because neither of them marks long vowels before two consonants, such as īnsula and mēnsa, which both dictionaries write without a macron. Langenscheidt, OTOH, does write macrons on īnsula and mēnsa – but also on īnsurgō and īnfestō (in the latter case adding "(ē?)" as the editors are uncertain whether the e was short or long in this word). —Mahāgaja · talk 06:46, 21 April 2021 (UTC)

You have miscorrected these entries. The prefix in- did have a long vowel when attached to a base starting in f- or s-; see this quote from Cicero, which is mentioned in a number of grammars or pronunciation guides such as W. Sidney Allen's Vox Latina: https://www.google.com/books/edition/The_Latin_Language/UHFiAAAAMAAJ?hl=en&gbpv=1&dq=%22litterae+sunt+quae+in+sapiente+atque+felice%22&pg=PA136&printsec=frontcover --Urszag (talk) 07:21, 21 April 2021 (UTC)

Yes, yes, I see that now. Oxford has proven themselves to be lazy enough to misguide me. Thank you for the update.

@ColinJParry: Welcome to the website. I don't think it's fair to blame the OLD when they say in the preface that "Normally only long vowels in a metrically indeterminate position are marked". I'm afraid you were simply too rash and didn't take the time to determine the cause of the disagreement. It's all the more evident because the first thing you corrected was fēlīx, when the vowel length is clearly visible from oblique forms; that this word is marked short in the OLD regardless should have been a smoking gun. I think it's important to recognise this so that the next time this happens, you make sure there's no mistake on your part before you suspect it on the part of other editors. I've reverted your edits; please let me know if some edits were made from a different IP so that I could revert them too. Btw, please don't forget to sign your posts by typing 4 tildes (~), cheers. Brutal Russian (talk) 22:51, 21 April 2021 (UTC)

I think it's pretty fair to blame the institution which worked on a book for over 50 years to do the whole "You shoulda known" thing, I find it lazy to purposefully leave out information and leave it to the purchaser to know the difference. I really wasn't trying to act rashly, checking two separate dictionaries before I modified anything. The fact that multiple dictionaries purposely leave out information is all the more alarming. I thought I could rely on such texts, but they're proving more and more unreliable as time goes on. I think macrons are more important than ever with schools often neglecting them entirely. Which is part of the project that I'm currently working on. Most edits I've ever are made from this IP, which is my place of work. ColinJParry (talk) 23:32, 21 April 2021 (UTC)

@ColinJParry: Don't you entertain the possibility that they did it very very consciously? I mean, it's an academic dictionary. Look at the list of editors. What might the reasons be? I think it's evident that they chose to include only those macrons that had been established with certainty, with the aim at making the dictionary error-free and future-proof.¹ One has to only take a look at L&S to see the effects of the opposite approach; even on this website these are far from all being sorted out. That said, Gaffiot 2016 and LaNe still include them. These are the two most reliable dictionaries in that regard, by the way, short of consulting De Vaan 2008. You can find both on Logeion.

1: Excluding them from endings like -īx is probably a mechanical consequence of this; even so it's visible from the oblique stem, which is given immediately after.

p.s.: Yay for anyone who supports educating people on the correct pronunciation of Latin, and macrons as a virtually inescapable aid in its acquisition. Brutal Russian (talk) 23:51, 21 April 2021 (UTC)

@Brutal Russian: Of course they did it "on purpose" I'm still greatly annoyed with it, I think that line of thinking, the "they should know" as a bit of typical British academic snobbery to it. But now I'm seeing this indignātiō should it not be indīgnātiō then? If so, this is why I'm so incredibly frustrated at this point. Because I've now got 3 incorrect dictionaries to work with. I'll look at those you pointed out, but there is a serious lack of macronized materials, I'm actually trying to produce Latin texts with macrons, and I'm coming to fear that nothing is reliable at this point.

@ColinJParry: I wanted to retort that I was reliable, but then remembered that my opinion is subject to change =D Complete agreement on all the lengths is impossible for a language that's no longer directly attestable (no speakers), and the indirect attestations can and are interpreted in various ways. That apparent, epigraphically- and one-grammarian-attested lengthening before gn, for instance, has to my knowledge left absolutely no trace in Romance. It might have existed in some specific sociolect that's left no trace; I think we've been leaving it out, sometimes even in those cases which might have been original (cōgnōscō, īgnōscō). If I find more theoretical substantiation for this, I might change those. Neither Gaffiot 2016 nor LaNe marks them as long, however. Here's a webpage you can sometimes turn to for evidence, but not necessarily for conclusions (which your bot culprit was implementing). As for the OLD, again, I don't believe they were thinking of anything like "they should know". I believe they were thinking "we don't know and there's no Brutal Russian to help us out" xP Brutal Russian (talk) 00:35, 22 April 2021 (UTC)

@Brutal Russian: Well, I appreciate the confidence and resources, which I'll use when I can. I was confused about the -gnus -a -um, and had to reread the paragraph a couple times. And while I understand your forgiveness of the producers of the OLD, I have read every entry, and there are 469 (out of 39,589) entries that are left untranslated and more that are given an incomplete or lackluster translation, such as "fellator" simply being "A type of sexual pervert" or similar. So, to get it straight, you view Gaffiot 2016 and LaNe as the most reliable, I noticed Lewis did have the macrons for the word in question, but was the only one that did.

@ColinJParry: > I have read every entry [...] (out of 39,589) O___O *jaw hits floor*

Yes, I noticed that some of these words' meanings were left to the reader's imagination - but all the untranslated words I've seen were defined in the Latin quotation. Lewis is the Elementary Lewis and Short, either from 1890 or corrected a couple of decades later. It reflects the same then-current discoveries and views as found in the scholarly works from that era quoted by the Alatius page (Bennet, Marx). In essence, it included all the newly-postulated longs from recent literature, and in some cases ended up overshooting as compared with the current consensus. The OLD was apparently being conceived shortly after that, while the debate on several of those issues continued for decades after, not to mention the ones that still can't be decided. With enough information and with the L&S being clearly outdated in that regard already by the start of its writing, the OLD editors' decision is understandable. —The two dictionaries I recommend are from the last decade; if in doubt, go with LaNe. Brutal Russian (talk) 00:03, 23 April 2021 (UTC)

@Brutal Russian: I may know someone who, while reading all 39,589 Lexical Entries in the OLD, typed them up in an excel spreadsheet because he was tired of there being no real quickly searchable English-Latin dictionary available. Who, also may or may not have written a Vulgar Latin Dictionary of all the filthiest words in Latin, with gems such as oppedo -ere (+ dat.) to fart in the face of... ColinJParry (talk) 01:56, 23 April 2021 (UTC)