Disputatio Vicipaediae:De nominibus propriis

E Vicipaedia
Jump to navigation Jump to search

Loci[fontem recensere]

I think we can use latin suffixes if they easily attach. Almost any berg or burg could take burgum, as Myces mentioned in the article proper. Trouble of course arises in what's the bar for easy attachability.--Ioshus Rocchio 21:51, 19 Aprilis 2006 (UTC)

Actually berg and burg should be two different endings: berg means "mountain", and appears to correspond to -berga, while burg is "(hill) town/fort", and -burgum (but it seems they've been confused for some time). I was actually doing a little survey of name elements from lists of Latinized place names in England to how endings were normally Latinized; the few I found so far:
-field → -felda
-caster → -castra
-c(h)ester → -cestria
-don → -dunum
-stead → -steda
-stow → -stova
-hampton → -hantona
-ham → -hamum
-hampstead → -hamsteda
-ey → -ega
-more → -mora
-gate → -gata
-ing → -inga
-mer(e) → -mera
-mouth → -muthum
-b(o)urne → -burna
-well → -wella
-tree, -try → -tria
-dale → -dala
-combe → -cumba
-wood → -wuda
-ley → -lea
-leigh → -lega
Of course places in England of any antiquity will probably already have Latinized forms, but more recent places like much of America won't. —Myces Tiberinus 11:03, 20 Aprilis 2006 (UTC)
Of course...an iceberg isn't an ice town, clearly. My fault. This is a pretty sweet list off which to base things. -on seems to go to onia or inium, cf Londinium and Vasingtonia. Places like Astoria, Annapolis I think we can leave be.--Ioshus Rocchio 13:42, 20 Aprilis 2006 (UTC)
Hehe... well, the etymology factors into many of these, since the Latinizations are of earlier forms of the language, so I take the ending element, not just the last few letters per se. I'm not sure how exactly "London" segments, and my intuition is that the usual for -ton might be -tona. (I should mention the list I am beginning with is Disputatio:Index locorum in Regno Unito.) Of course Astoria, Annapolis, and other things that are clearly meant to be of Latin origin can probably remain the way they are (as would Londinium etc. which are too entrenched to change even if they did run afoul of whatever modern scheme we might use). —Myces Tiberinus 22:56, 20 Aprilis 2006 (UTC)

Latinisatio possibilis terminationes Slavicorum nominum urbium[fontem recensere]

-ov, -ów → -ovia (e.g. Pskov - Pscovia, Kraków - Cracovia)
-sk → -scum (e.g. Minsk - Minscum vel Minsca, Pinsk - Pinscum)
-grad → -gradum (e.g. Belgrad - Bellogradum, Belgradum vel Belgrada, Wischegrad - Vissegradum)
-gorod → -gardia aut -gradum (e.g. Novgorod - Novogardia sive Novogradum, Belgorod - Belogradum)
-slav(l)→ -slavia (e.g. Pereyaslav(l) - Pereaslavia, Yaroslavl - Iaroslavia)

Vide etiam: Index locorum in Russia. -- Alexander Gerascenco 02:23, 21 Aprilis 2006 (UTC)

What would be really good in both these cases is writing good, encyclopedic, well-referenced material on the usual Latinization of placenames, describing history, practice, recommendations, etc. under Latinizatio nominum locorum or some such. —Myces Tiberinus 16:48, 29 Aprilis 2006 (UTC)

A Hand-List of Latin Place Names with Their Modern Equivalents[fontem recensere]

The book A Hand-List of Latin Place Names with Their Modern Equivalents, by Edward W. Burke, S.J., contains a similar list on pp. 348-9. Of course this list is more concerned with the meanings of the suffixes than the vernacular equivalents—in fact, vernacular suffixes are rarely given at all. But perhaps it can still be of use to us, so here's a summary:

-berga ← -berg "mountain"
-brica ← "bridge" (also variant for -briga)
-briga ← (Iberian variant for -berga)
-briva ← "bridge, ford"
-burgus, -burgum ← -burg, -burgh, -bourg, -borg, -bork, -bury, etc. "fortified town, city"
-castria, -cestria, -cestra ← "Smith notes that in areas of England dominated by the early medieval Danes the suffix became Eng. -caster, while in areas under Anglo-Saxon rule it was transformed into -chester or -cester." Cool, eh? (and of course, it goes without saying that this is historically from castra)
-dūnum, dūn- ← "hill, hill town, fortified town" (he has a whole section devoted just to this suffix on pp. 349-355)
-dūrum, -dōrum, dūro-, dōro- ← "place on a stream"
-ētum ← "grove" (of course: this is a productive suffix in Classical Latin)
-furtum ← "bridge, ford" (Francofurtum)
-magus ← "field, market, town" (leading to my favorite: Blattomagus, which is Celtic for "flower-field", but Latin for "cockroach-wizard")
-mutha ← "river mouth"
-polis ← "city"
-stadium ← -stad, -stadt "city"

--Iustinus 16:25, 16 Septembris 2006 (UTC)

Arlington[fontem recensere]

Once that's been driving me nuts for some time is a decent Latinization of Arlington. I've found nothing on several forays. Sinister Petrus 21:57, 16 Maii 2006 (UTC)

Well...following the Latinization of Washington as Vasin/Vasingtonia...I think Arlin/Arlingtonia would be fine.--Ioshus Rocchio 04:35, 20 Maii 2006 (UTC)
I was going to suggest Arlintonia myself. I don't know why I didn't (who knows? maybe to avoid getting a response of "licet" or worse "vide catenam etiam"). Thanks for confirming my hunch. Sinister Petrus 02:33, 24 Maii 2006 (UTC)


Nomina Hominum[fontem recensere]

If someone's last name can be Latinized in a simple manner, it should be done. For surnames that are already in a plural form, the plural ending can be dropped and replaced with an "i". For example, Julia Roberts can be Latinized as "Julia Roberti" (Robertus is already the commonly accepted Latinization of Robert). Likewise, the surname Gonzalez can be Latinized as "Gonzali". Romance language surnames (Spanish, French, Italian, etc.) should also be Latinized, as these languages have originated from Latin and most of their words have Latin cognates. The surname Benitez can be Latinized as Benedicti, Romero as Romerius, Genovese as Genovicus, Pisano as Pisanus, etc. As for names using prefixes such as de or da, they can be dropped and replaced with a suffix inticating a belonging to. For example, Di Stefano can be translates as Stephanicus and DaSilva as Silvanus. If it is a Latinized last name in the singular form that ends in -us or another Latin ending, is thould take the feminine version when it is a female's surname. For example, the name Alexa DaSilva should be translated as Alexa Silvana. Likewise, Roberto DaSilva would become Robertus Silvanus.

The -s in Roberts and the -ez in Gonzalez are not plural markers. —Myces Tiberinus 11:50, 16 Septembris 2006 (UTC)
It's a sort of genitive, isn't it? That is, González = hijo de Gonzalo = filius Gonzali? So the effect might look the same: Roberti would be genitive singular instead of nominative plural. For Pérez, the Latin according to this proposal would be Petri, since it stands for something like filius Petri. (Btw, I didn't write the original paragraph here.) IacobusAmor 12:33, 16 Septembris 2006 (UTC)
It's not a sort of genitive it's just an old genitive. -es was the old genitive marker in english, hence -'s these days. You can see it still in german and scandinavian names, ie Leif Eiriksson. But i think I would agree with you that the genitive and plural looking the same in latin, we would be safe doing Gonzali or Roberti. The lowercase f abbreviation of filius in latin was a formality not a rule.--Ioshus (disp) 12:53, 16 Septembris 2006 (UTC)

First of all, the plural of nomen is nomina—to say nothing of populi, which Iacobus already corrected without comment. Second of all ... well, allow me to repeat some comments I posted here:

... The fact is ... that not Latinizing family names, for better or for worse, has been the default since, oh, somewhere in the early 19th century, I'd say. This practice is followed by pretty much all of the seriouse modern Latin "authorities" and publications. Some examples from the Periodica Latina I have handy:
  • "...praeclarum lyceum Mediolanense, Iosepho Parini dicatum, nonnulli discipuli inundaverunt." -p. 211 Horatii Antonii Bologna "Diarium Latinum: Lyceum Mideolanense inundatum est." Latinitas, An. 53, Lib. 2, 2005.
  • "Statua equestris Pauli Revere" -p. 576, Gaii Licoppe "De itinere Americano (III)", Vox Latina, Tomo 40, Fasc. 158, 2004
  • "Qui liber in tres est partes divisus : primam, quæ parentes tractat et liberos, a Christiano Læs compositam ; secundam, quæ in amores incumbit et mores venereos, ab Antonio Van Houdt tractatam ; tertiam autem a tribus libri auctoribus curatam, quæ de matrimonio agit." -p. 16, Volfgangi Jenniges Lovaniensis "De Novis Libris", Melissa 119, 2004
  • "Studiis Lindae Buck atque Richardi Axel ingeniosissimis..." p. 13, "3.1 Nasus humanus quantum valeat" Nuntius Leoniinus, Vol. 1 2004
  • "Bavari Owenum Hargreaves nequaquam volunt amittere." Ephemeris
Lest you think all these sources are merely influenced by each other, I could dig up older references, from the 1800s, if you like. But this will have to do for now. --Iustinus 21:41, 26 Augusti 2006 (UTC)

Ut mihi videtur, melius esset si cuique scribenti ratione sua uti sineretur. Interdum vertendum est praenomen tanum, tum cunctum nomen, alias neque praenomen neque nomen gentilium. Id e contextu pendet. Quid agitur in aliis linguis? Francogalli, exempli gratia, solent nomine peregrino adhibere, sicut "Tony Blair" (non Antoine Blair). Item Angli dicunt "Jacques Chirac" (non James Chirac). Mihi utique ridiculum videretur si Antoine Blair et James Chirac dicerentur. Ideoque cur invenitur Antonius Blair et Iacobus Chirac in ipsa nostra Vicipaedia? Non censeo hoc molestum esse, sed mea quidem sententia oportet istam regulam abrogare qua praenomina dicuntur vertenda esse, in vice autem nova lex molienda esse: "bona ratione tua adhibenda esse si velis scire num nomen vertendum sit." Danius 13:36, 17 Iulii 2009 (UTC)

Ridiculum non est, et id ridiculum videtur tibi modo quia linguam nostram cum regulis aliarum comparas. Sed lingua Latina proprios mores habet. Diu haec de primo nomine convertendo consuetudo est Latina et ubique, de qua te certiorem fieri potes legendo litteras Latinas hodiernas.
Proporro de opinata nominis quaerendi difficultate erras. Cum in arco quaerendi ad laevo latere nomen Tony Blair inpones, certus sum te paginam de Antonio Blair inventurum. --Rafaelgarcia 13:55, 17 Iulii 2009 (UTC)
Fateor linguam latinam suos mores habere, neque molestum esse vertere quaedam nomina; nihilo secius autumo nonnulla nomina non vertenda esse, secundum contextum. Numquid assentiris regulam "praenomina vertenda sunt" flectere ut quisque suo arbitrio uti possit? Danius 14:21, 17 Iulii 2009 (UTC)
Regulam censeo optimam dicentem nomina quae converti possunt convertenda. Ita Tony quidem convertendum est, quia millenia nos docent Tony valere idem ac Antonium in lingua nostra. Alia nomina tam facile Latine invenitur, sicut Britney Spears. (obiter, si scis quomodo recte Britney Latine converti possit, quaeso dicas...). Quid censes? --Ioscius (disp) 14:39, 17 Iulii 2009 (UTC)
Britney, secundum aequabilitates conversionis, iuste est Britneia. IacobusAmor 14:53, 17 Iulii 2009 (UTC)
Iosci: hoc non sufficit, quia ista regula tenemur omnia praenomina mutare dumtaxat praenomen latinum exstat. At tamen nonnumquam pristinum nomen conservandum est, mea quidem sententia. Si haec lingua vivens fieri volumus, necesse est quasdam consuetudines variare. Licet linguam latinam habere suas proprias mores, tamen sub recentius tempus opus est alias linguas imitari, in rebus hodiernis utique. Danius 15:06, 17 Iulii 2009 (UTC)
Non summus antecursores linguae latinae sed fautores assectatoresque eius. Nomen quoddam non converte, si de eo convertendo umquam dubites. Sed encyclopaediis opus est mores extra se sequi, nec quidem novos mores fingere. Nostra regula modo mores hodiernos linguae sequitur.--Rafaelgarcia 15:31, 17 Iulii 2009 (UTC)

Pocatello -> Pocatellum (Romanic names)[fontem recensere]

Why not change Romanic Pocatello into Pocatellum? --Alex1011 18:31, 16 Decembris 2006 (UTC)
Because the purpose of the rules is to avoid inventing names (which in Wikipedia terms amount to original research). Besides, it seems 'en:Pocatello' is a Shoshoni name, not a European one. Basically, the point is that Wikipedia, even in Latin, is a reference. We don't want people to come here and think "oh, Capita Animalium is Latin for Thierhaupten, I can use this on my T-shirts" when it's just something one of our users just came up with. (Pocatello may well ought to be Pocatello, -onis, or even Pocatellopolis.) If, though, a proposed name is something we can produce references for, then by all means, let us present it. —Myces Tiberinus 00:12, 17 Decembris 2006 (UTC)
See also Vicipaedia:Coining. --Rolandus 11:11, 17 Decembris 2006 (UTC)

Italian names Medici in particular (here is the right location for our discussion)[fontem recensere]

I ask me if is not better to let the italian name (cognome), because italians names are already relics of an old genitive (e.g Marchetti from "figlio di Marchetto" in Latin filius Marchetti), therefore it makes no sense to translate them in Latin. See Gerhard Rohlfs, Grammatica storica della Lingua italiana e dei suo dialetti, vol. "morfologia", nota n° 1 pag. 8: about the origin of the italian names "l'autore si pronuncia qui per un genitivo nato e consolidato nel linguaggio degli avvocati", the author believes that it is a type of genitive created in the language of lawyers". Should we move the pages Silvio e.g Berlusco to Berlusconi? Ciao--Massimo Macconi 11:01, 17 Decembris 2006 (UTC)

P.S I reproduce the page on wiki taberna

Translitteratio[fontem recensere]

We're going with ISO? I suppose it's best to have SOMETHING. But for the love of God, please do not render Arabic as t: that's just absurd! --Iustinus 17:43, 20 Ianuarii 2007 (UTC)

Also, might I suggest that ISO-transliterated names be put in italics? In general, when I give a forreign name with no Latin attestation, I prefer to do this, so that it's clear I'm merely citing it as a foreign word. --Iustinus 17:51, 20 Ianuarii 2007 (UTC)
Yes, it has been my policy to italicize any untranslated foreign word with a vaguely latin script (ie greek, or cyrillic, or scandinavian, or czech).--Ioshus (disp) 17:53, 20 Ianuarii 2007 (UTC)
We don't have to go with ISO. But it was the only neutral way I could think of for when attested forms don't exist. If someone can think of a better way... --Myces Tiberinus 18:01, 20 Ianuarii 2007 (UTC)
On a partially related note: we should probably start writing pages on historically attested systematic transliterations, e.g. the ancient system for Greek (obviously--and this one is very important, because a lot of people seem to prefer newfangled systems, even in inappropriate places), the (slightly less rigid) ancient system for Hebrew (and how this could theoretically be applied to Arabaic, though that might be a bad idea), the various attested systems for Japanese, and so on. --Iustinus 18:07, 20 Ianuarii 2007 (UTC)
I have noticed that for Russian, ISO 9 is actually quite different from the International scientific transliteration. E.g. it transliterates Я as 'â', not as 'ja'. I have never actually seen this used in practice, neither in English nor in German. I would propose that instead of ISO, we make the most widespread system, i.e. the Scientific transliteration, (sine fastigiis) our standard. I haven't checked what ISO says about Arabic and Chinese yet, but if it is different from the rules of the Deutsche Morgenländische Gesellschaft and Pinyin, respectively, I would suggest we also abandon ISO for these two.--Ceylon 07:29, 25 Martii 2008 (UTC)
For Chinese, I immediately agree. Pinyin is standard in China and among the great majority of people who write about China elsewhere.
I hadn't ever noticed the point you make about Russian. For Russian and Arabic, I would agree as a start that we should adopt a widespread system: since the ISO Russian system as you describe it is far from widespread, it looks as if we should stop recommending it. Andrew Dalby (disputatio) 10:04, 25 Martii 2008 (UTC)
On the Chinese front, no worries; Hanyu Pinyin is ISO 7098. As I mentioned earlier, the chief rationale behind proposing use of ISO standards was neutrality: they are made for international use and exist for most major writing systems (so the question of choosing standards from different bodies for different individual languages—which might also raise issues of NPOV—doesn't so much arise). Mind you, older versions of ISO 9 *do* use 'ja' for 'Я', though this is apparently one of the changes en.wiki describes as being done "in favour of unambiguous transliteration over phonemic representation", which is a reminder that what these are intended for is transliteration (which can be done more or less neutrally with reference to international standards), not transcription (which, us not having published guidelines for spelling things in Latin, is exactly the kind of unreferenced fictio we're trying to avoid with this page's guidelines). —Mucius Tever 01:38, 26 Martii 2008 (UTC)
But shouldn't we also be concerned about recognisability of names? None of the other wikipedias, as far as I can see, have adopted the ISO system for Russian. Plus if, as the rule says, we drop diacritics, we will not arrive at "unambiguous transliteration" anyway, since we would - according to ISO - write A for both А and Я etc. So I would still plead for the International Scientific Transliteration, unless we can find a genuinely Latin system of transliterating Russian. -- The same concern applies to Arabic, where ISO has again chosen to differ from the internationally accepted standard (Deutsche Morgenländische Gesellschaft), e.g. by going for ẗ (a t with a trema on top) to represent the ta marbuta (ة). This has been discussed in the case of Usama bin Ladin, with a consensus that the t in Usamat would be a nuisance.--Ceylon 07:08, 26 Martii 2008 (UTC)
1) The recognizability of names? If you're recognizing it, it's most likely not from any use in *Latin*. Are we to recognize the English form, the Italian form, the German form?
If it's transliterated in an internationally accepted scientific way (which is not always ISO), then I would argue it would be recognisable in all of these languages, including Latin.--Ceylon 07:05, 27 Martii 2008 (UTC)
2) The rule as first written said fortasse without diacritics, but I see diacriticophobia has been creeping. :(
3) I do wonder how internationally accepted DMG currently is. Deutsche Morgenländische Gesellschaft Romanization of Arabic doesn't Google well at all (the first page of results returns only one description of its use, in Persian, in a PDF alongside several others) and en.wiki advises ALA-LC. (It does seem, from what I can find, that DMG is very like ISO 233, though I can't find detailed description to properly compare). —Mucius Tever 01:19, 27 Martii 2008 (UTC)
Having studied Arabic for many years, in my experience yes, DMG (maybe not always called by that name, but that matters little) is indeed the standard accepted in international scientific literature; and no, as pointed out above, it is not entirely identical with ISO 233.--Ceylon 07:05, 27 Martii 2008 (UTC)
Mucius is right about it not googling well, however, here are two descriptions for the Arabic part: on Wikipedia and on the Göttingen University homepage. As far as I remember, it also applies to other oriental languages though (Persian, Amharic/Geez, Hebrew, maybe Syriac, Aramaic).--Ceylon 09:30, 27 Martii 2008 (UTC)
Italian Wikipedia uses ISO/R 9:1968. [1]. -- Alexander Gerascenco 11:49, 26 Martii 2008 (UTC)
Which is the old ISO version and identical with what I called the International Scientific Transliteration.--Ceylon 22:26, 26 Martii 2008 (UTC)
Yes, Scientific transliteration. —Mucius Tever 01:19, 27 Martii 2008 (UTC)
Yes, it has to be recognized that transliterations/transcriptions serve a range of purposes. One size doesn't fit all, which may explain why (for some languages) very few people anywhere have adopted the ISO standard! One purpose that standards-makers like to go for is "reversibility". This remains important for a gradually decreasing number of users (it's still essential for some of them), but the unicode standard (allowing the encoding of almost any script) has now rendered this an unneccessary constraint for most users. "Reversibility" doesn't matter in the least to people writing on a wikipedia platform, because we can cite names in any number of scripts. Nobody needs to look at our transliteration to see if the ta marbuta is there, or to see whether a Я is there, because they can look either in the original script we placed alongside, or in the Arabic or Russian equivalent of the article. The ISO standard, where it insists on reversibility, will often be too primitive for us, therefore: we want something reliable but more user-friendly. Our transliteration is principally there to help ordinary readers read the name -- people who don't know and don't even want to know what a ta marbuta is. To them, the t (even with two dots) is, as Ceylon says, a nuisance.
To me, the ISO rule is a stopgap. For each language where there is a range of choices, we need to arrive at a good choice as soon as we reasonably can, preferably with the help of Vicipaedians who are experts. For Russian, we have already User:Alexander Gerashchenko, and we need to hear from him. Andrew Dalby (disputatio) 08:49, 26 Martii 2008 (UTC)
[Later:] For Arabic (and Persian?) Ceylon may well be right that the DMG would be the best standard, but we have others who know Arabic and may have other ideas.
For Sanskrit and Indian languages there is also a generally accepted international standard, that of an Orientalist Congress about 1890 (I don't remember the details). It's easy to apply, because Indian scripts are so scientific anyway. I have already used it in articles about Sanskrit literature. [I have the details now. The system I mean, the only one worth choosing for Indian languages, is called "IAST", is described on the page en:Devanagari transliteration, and is practically identical with the Calcutta National Library transliteration to which that page also links.] Andrew Dalby (disputatio) 08:57, 26 Martii 2008 (UTC)
IAST according to en.wiki does not differ substantially from ISO 15919. —Mucius Tever 01:19, 27 Martii 2008 (UTC)
Yes, agreed, they are practically identical.
The same system can be used to transliterate Tibetan (this is regularly done) and Burmese (and some scholars do this too). In these two cases, however, although it works for the scripts, it's utterly different from modern pronunciation. In an encyclopedia you would want to give a pronounceable transcription alongside. Andrew Dalby (disputatio) 09:19, 27 Martii 2008 (UTC)
It would be feasible to deduct a chart of (albeit medieval) Latin transliterations for foreign alphabets from how Latin writers (primarily of geographical and ethnographic literature) used to call persons and places in Russian, Arabic, etc. speaking regions. But of course, we would (especially in Asia) end up with more than one variant for some letters, since transliteration was not systematic (even today it is not fully so!).
The alternative is to adopt widely accepted transliteration standards such as ISO or others for particular languages; but this is not devoid of problems either, especially if we drop the diacritics which many of these standards rely on very heavily to distinguish phonemes.
Maybe we should therefore give a thought to drawing up our own transliteration charts for important languages, which are based on the accepted standard, but adapted to Latin pronunciation (e.g.: Chekhov [= Чехов = Čehov ISO9 of 1995 = Čechov ISO9 of 1968] would become Cehov according to our present rule, but might actually be more properly pronounced in Latin when transliterated Tsecov or Tsechov). Probably these modifications would only affect very few letters in each case. I know fingere is anathema, but one would probably even be able to back up such amendments of the standards with attestations if one looks hard enough. In this case, we could continue the discussion on the pages for the specific languages and put a link here.--Ceylon 22:26, 26 Martii 2008 (UTC)
Of course, in the pronunciation of many people, 'Ce-' would have a closer pronunciation to the original than 'Tse-'. But that's a separate matter, of course. —Mucius Tever 01:19, 27 Martii 2008 (UTC)
Is that our present rule? Following the reconstructed pronunciation, I'd pronounce Latin Cehov as [kehov], but isn't the "v" voiceless in the original, hence [f]? So not Cehov, but Kehof? ¶ What was the original of Vercingetorix? If the last morpheme there, rix, is related to modern German reich, and if the old Romans heard that "ch" as their "x," then the closest approximation to Чехов might start with an "x" (Xehov). Or is there an /s/ in there, so that rix is more like reichs? ¶ The larger point, whether the Vercingetorix speculation is apt or not, is that even when we have no ancient Latinizations for specific words, we probably do have ancient Latinizations for specific sounds, and there's no reason we can't freely apply them to new words, unimagined by the ancients. IacobusAmor 23:39, 26 Martii 2008 (UTC)
Well, there's one reason, and that's that an encyclopedia shouldn't make things up for itself, but refer to external, respected sources for its facts and rules. —Mucius Tever 01:19, 27 Martii 2008 (UTC)
That misses what I'm suggesting. We definitely do have "external, respected sources" for how the old Romans transliterated specific sounds in other languages. We wouldn't be "making anything up": with regard to those sounds, we'd be maintaining their methods. IacobusAmor 03:16, 27 Martii 2008 (UTC)
I think what we find is that such sources don't take us far enough, because (a) no Roman ever did a full phonemic inventory or transcription of a foreign language [except, you might just argue, Greek]; (b) Romans were not even consistent with one another about the way they transcribed single foreign phonemes or sounds [not even for Greek]. If we try to concoct a system from such chance transcriptions, and apply it to other languages, we may (1) make no sense to anyone, and (2) fail to skip over Mucius Tever's rope.
The Romans were not unique in their phonemic ineptitude. If you look at old English spellings of Indian words, any time before the mid 19th century, you will often find them hard to attach to any real phonology. If you were building a transliteration system, you wouldn't want to start there. Andrew Dalby (disputatio) 09:08, 27 Martii 2008 (UTC)
Probably not apt, Iacobe, bad luck. He was a Gaul, not a German, and the online Celtic dictionary in discussing this -rix form here does not seem to reconstruct a [χ] in it. Andrew Dalby (disputatio) 00:23, 27 Martii 2008 (UTC)
I suspected as much, which is why I thought it might be "related to" reich—which it is, according to Watkins's Dict. of IE Roots. However, if it's purely Celtic, where does the [s] implicit in the Latin "x" come from? Watkins says the Celtic was *rīg- 'king' (with no [-s]), whereas Germanic was *rīks 'king', which would explain the [s] of -rix. In any event though, the remarks on [χ] aren't apt, since that was apparently a development later than JC's time. IacobusAmor 00:55, 27 Martii 2008 (UTC)
My suggestions: either they matched it up with rex in their own language; or one of the cases in the Celtic declension of this word contained a sibilant; or a bit of both. Andrew Dalby (disputatio) 09:08, 27 Martii 2008 (UTC)

Rules for Converting Polynesian Terms to Latin[fontem recensere]

I here propose rules for Latinizing Polynesian terms. My purpose is to establish Latin forms that will be as regular as possible, so that indigenous spellings of terms can be reconstructed from the Latin as readily as possible. I base these rules on analogies with attested forms (not all of which I cite below); however, for all but a few Polynesian placenames & personal names, no attestions exist—and most that do exist have been contrived by scholars unfamiliar with Polynesian linguistics, and are therefore not necessarily the aptest forms for Polynesian languages. ¶ It's conceivable that, for some Polynesian terms, nineteenth-century Latin forms exist in writings by Roman Catholic missionaries—but it's unlikely: I've seen some such manuscripts, and most (for reasons we don't need to get into here) are in French. One manuscript I examined carefully had a few Latin terms (e.g., Status Animarum for a census of parishioners), but these weren't for indigenous concepts. To make matters more curious, nineteenth-century prelates' Latin titles referred to places in the Old World, not locally relevant ones; for example, a Roman Catholic bishop active in Samoa was formally known as the Bishop of Tipasa, an abandoned diocese in Africa, which the prelate had never visited.

I. Declinability. Most Polynesian nouns are indeclinable in Polynesian, but we can reject the idea that Latin should treat them as indeclinables. Attestations of the declining of Polynesian-Latin nouns exist, and adjectival forms attested for centuries in scientific taxonomy show derivations from quasi-Latin stems, implying that the nouns decline. In all but the simplest syntax (grammar like "A est B"), retaining the indeclinability of Polynesian nouns would confuse most readers. ¶ The changes in spelling required by these rules alter Polynesian terms much less than Polynesian languages do when they convert non-Polynesian terms to Polynesian. For example, in Samoan, the nineteenth-century missionaries Drummond, Heath, Hardie, MacDonald, and Sunderland became Talamoni, Ite, Ale, Makona, and Sitanile, respectively: the original names are not reliably reconstructible from their Polynesian forms.

II. Phonetics. Most letters are retained as in their original orthographies.

A. Vowels are five, pronounced similarly to their Latin analogues.
1. Polynesian a = Latin a ; likewise, e = e, i = i, o = o, u = u.
2. As in Latin, Polynesian languages distinguish between long & short vowels, but long vowels are not to be marked; they're almost never marked in Polynesian customary orthographies.
B. Consonantal graphemes equivalent to Latin ones are b, d, f, h, l, m, n, p, r, s, t.
C. Problematic letters.
b: used in Fijian for /mb/; hence Latin mb.
c: used in Fijian for /þ/; hence Latin th.
d: used in Fijian for /nd/; hence Latin nd.
g: when representing the phonemic /g/ = Latin g.
g: when representing the phonemic /ŋ/ = Latin ng. Example: Samoan Toga = Latin Tonga.
k: always pronounced /k/; hence Latin c.
ng: when representing the phonemic /ŋ/ = Latin ng. Example: Tongan Tonga = Latin Tonga.
q: when representing (in Fijian only) /ŋg/ = Latin gg (as per W. Sidney Allen in Vox Latina). Example: Fijian yaqona = Latin iaggona 'kava'.
v: sometimes /v/ and often bilabial, /β/, but best retained in Latin as v.
w: sometimes /v/ and often bilabial, /β/, attested as Latin w in the species epithet pelewensis (referring to Palau), but probably best in Latin as v.
y: used in one or two languages to represent a sound like the Latin consonantal i ; hence Latin i.
ʔ: glottal stop, often representing Proto-Polynesian /k/, and sometimes the minimal point of difference between modern Polynesian pairs of terms, but seldom written by native speakers, and always to be ignored in Latin. Example: Hawaiian Hawai'i, Latin Havaii.
D. Unused consonants of the modern Roman alphabet: j, x. Some old Tongan texts use j for the pronunciation of /ti/ as /tsi/, but this cluster in Latin readily becomes si.
E. Note:—To Polynesianists, converting /k/ to /c/ results in spellings that look really, really odd ; but K, though a genuine Roman letter, doesn't seem to have been ordinarily productive for more than two millennia now, and I'm not sure we have a warrant to revive it, so I've left it dead.

III. Gender & termination. All Polynesian nouns—yes, without exception: all, all, all—end in a vowel. This fact makes their conversion to Latin relatively simple & regular, though eliminating or altering their final vowels often results in truncating Polynesian roots.

A. Gender.
1. Placenames. The default gender is feminine in the first declension and neuter in all other declensions. Semantics—if the term is marked as to sex in the original culture—may overrule the default.
2. Personal names. The gender matches the sex of the referent.
B. Termination.
1. Ending in A. These are invariably in the first declension. Semantic reference marks a few as masculine; for example, the Samoan mountain Vaea is mythologically the body of a man, Vaea, and so Vaea is masculine. Example: Tikopian Faea = Latin Faea, -ae, f. Also, a widely encountered nonproper noun: Hawaiian hula = Latin hula, -ae, f. ¶ Exception: The proper name A is possible (but I haven't found an example); if it exists, it has to be indeclinable.
2. Ending in E. These retain their original spelling in the nominative and then decline in the third declension, with terminations added to whatever stem results from the deletion of the e. Since the first word of an article is typically the titular term in the nominative (not the ablative, also ending in e), this seeming irregularity will help naive readers grasp the indigenous spelling. Example: Tane, Tanis, a Polynesian god. ¶ Exception: The proper name E is possible (but I haven't found an example); if it exists, it has to be indeclinable.
3. Ending in I. These are plural nouns of the second declension, much like Latin castra, -orum 'camp' and not unlike the first-declension placenames Athenae, -arum 'Athens' and Tres Tabernae, Trium Tabernarum 'Three Taverns'. Examples: Havaii, -orum 'The Hawaiian Islands', Viti, -orum 'the Fijian Islands, the Feejees'. Also: Tahitian Ari'oi = Latin Arioi, -orum, a performing-arts society instituted by the god Oro. ¶ The plural attestations that I've seen are to placenames, not personal names, and so it might be better to handle personal names ending in i differently, perhaps like those ending in e. Example: Hawaiian Kaha'i = Latin ?Cahai, Cahais. An alternative might be to retain the full original term as a stem and add -um for placenames, -us for male persons, and -a for female persons—but that process may be too complicated to keep straight. This class of nouns remains problematic. ¶ Exception: the proper name I exists; it has to be indeclinable.
4. Ending in O. On the basis of the attestation of "Cairo" = Latin Cairum and "Chicago" = Latin Sicagum, such nouns are invariably in the second declension. Example: Samoan Falealupo = Latin Falealupum, -i, a village; Hawaiian Lono = Latin Lonus, -i, a god. ¶ Exception: The proper name O is possible (but I haven't found an example); if it exists, it has to be indeclinable.
5. Ending in U. These are invariably in the fourth declension. Example: Hawaiian Kumoku = Latin Cumocu, -us, n., a place on Lanai. ¶ Exception: The proper name U is possible (if it exists, it has to be indeclinable. (And note the non-Polynesian name U Thant.)

IV. The beginnings of articles. Where necessary, as has become customary in Vicipaedia, the first phrase of an article indicates declension, gender, and indigenous spelling:

Falealupum (-i, n. ; Samoane: Falealupo). . . .
Lonus (-i, m. ; Havaiane: Lono). . . .

Because of absolute regularities in the proposed principles of conversion, however, this bother is unnecessary for nouns indigenously ending in -a and -u :

OK: Apia, maxima urbs et caput Samoae. . . .
Not necessary: Apia (-ae, f. ; Samoane: Apia) maxima urbs et caput Samoae. . . .

V. Redirects. Where the Latin nominative has changed the indigenous spelling, a separate page with a redirect may be wanted as an aid to readers; for example, the term Nadi may want a redirect to the article Nandi.

VI. Extending to other languages. Most of these principles can be extended to the non-Polynesian languages of the rest of the insular Pacific, especially Melanesia & Micronesia, but consonantal terminations occur there, and more examples & possibilities need to be considered before the rules can be made more general. IacobusAmor 22:32, 20 Ianuarii 2007 (UTC)

Responses[fontem recensere]

Wow, this is quite a piece of work, and overall quite usable. I would make the following comments:

  • What should be the name of this system? It would be useful to say, e.g. "Egger gives X in contrary to the Jacobean Y" or wharver we're going to call this.
  • II.C. Glottal stop might in some cases be represented by an optional <h>: compare Israhel, and perhaps even native words like mihi, nihil. (Of course in the latter class the <h> tends to turn into [k] in medieval pronunciation, which of course is à propos to the Polynesian glottal stop ;) )
  • II.E. in fact k isn't ENTIRELY unused in Neo-Latin terms, but it definitely looks ugly to me. The problem is that the Italic pronunciation for Latin is still EXTREMELY common, which will result in a lot of unwanted c > ch changes. I'm not sure what we should do about this though. The usual solutions to this problem are:
    1. The Germanic solution: before a front vowel [k] is written with <k>. This has the obvious problem that k was virtually never used in Classical Latin at all, and it is quite jarring to see. Another problem with this system is what to do with [g] (Germans pronounce it [g] no matter what).
    2. The Western Romance solution: before a front vowel [k] is written <qu> and [g] <gu>. Problem: this causes havoc for speakers who use ANY of the major pronunciations still in use, turning velars into labiovelars.
    3. The Italic solution: before a front vowel [k] is written <ch>. The main problem with this is that, I assume, Polynesian /k/ is not aspirated. But if there is any leeway there, this might be the best way to go. Of course, another problem with this system is that [g] becomes <gh>, which is even less Latin than <k>!
    • By the way, it might be hypocritical to get too worried about this, when we already have Italians going around pronouncing the most important Latinized Polynesian word there is as VEE-chee ;)
  • III.A. don't forget that while the names of countries and regions can be any gender, Latin regularly considers mountains and rivers masculine. There may be other geographical gender rules that could be of concern as well.
  • III.B.2. nominative -e for non neuters is more than a little odd.
  • III.B.4 actually, I think Cairus is more common, not that that's relevant. Another good example is Tochium (variously spelled): in the Jesuit system, Japanese nouns in -o or -ou regularly become second declention (of course, so did -u and sometimes -i).
  • III.B.5 the problem is that neuter fourth declention nouns are extremely rare even in native words. Furthermore, in Classical Latin all proper nouns fall under the first three declentions (if you can find an exception, let me know. I am not including the Archaic Latinization of Greek names into the fourth declention: that pretty much disappeared early on).
  • I would like to make allowances for previously attested Latizations to overrule (at least optionally) the systematic one. You know I'm all about attestations.

--Iustinus 23:10, 20 Ianuarii 2007 (UTC)

  • I worry that this is the same color as "original research" (i.e. inventing/creating things for the encyclopedia rather than collecting them from external sources). The primary purpose of this very page (VP:TNP) is to recommend to people that most of the time it's better not to invent Latin names if there aren't any. As for the argument that indeclinables are grammatically opaque, this is solvable by using ordinary nouns (e.g. cum systemate wiki) or pronouns when the word's grammatical role is not immediately obvious.
  • The only reference I see to using gg for /ŋɡ/ in Allen is the Grecizing suggestion of the pre-classical poet Accius (who also advocated things like spelling long vowels with double letters) — I'm not sure this is commendable.
  • I agree about nominative non-neuter -e. Especially since the Latins didn't have any trouble with nominative -es, having it in three entirely separate declensions (Anchises, -ae; feles, -is; dies, -ei...).
  • C is fine, I'm sure. If the Italians and their ilk don't care that the Romans themselves pronounced /k/ where they have a softer sound, why would the /k/ of other peoples be any more worth preserving? ;p (This is a similar but somewhat more flippant restatement of the principle I'm using in the general guidelines I am putting together [for non-wiki use, that is].) And of course if they really want to nothing keeps them from pronouncing the /k/, as the English have recently learned to do with the Celtae. —Myces Tiberinus 01:43, 21 Ianuarii 2007 (UTC)
Thanks, Myces; I'll try to get to your comments later. (Yes, I've read the bit about "original research." Having published my own dead-tree encyclopedia—hey, it weighs more than six pounds, and in the U.K. can cost as much as a hundred seventy-five—I've had the experience of thinking practically about what an encyclopedia is and can be.) For now:
Thanks, Iustine, for your attentive responses. How should we proceed? I intended the original to be a "working document": should I amend it as we go? Eventually, the idea would be to move a more final version over to the regular page, but it needs more tweaking before that happens. Yes, I'm all about attestations too—except remember that Egger & Company didn't have a clue as to how their versions would look to Polynesian-speakers, so when they themselves are the only attested evidence, one rightly worries. Responses:
I don't give two hoots what you call this "system." (And it's not finished yet.) Does it need a name?
II.C. Since indigenous writers almost always leave a sign for the glottal stop out (indeed, only in the last few decades have linguists reliably established where all instances of it should go), there's ample attestation for ignoring it. Also, we'd get into trouble by marking it as "h" in languages that already have an "h." Hawaiian ho'ohu'ihu'i, for example, would become hohohuhihuhi, and that looks really odd; a form like hoohuihui should be much more easily comprehensible to Polynesians.
II.E. Hmm. I hadn't thought about those unwanted c > ch changes. I'd say anything the orthography can do to head them off would be welcome! One shudders at the possibility of "VEE-chee" from our vici in reflection of Hawaiian wiki ! If only we (or rather, you all, before my time) had gone for Vikipaedia, I'd say make them all k, every last one of them. Adding an h sometimes (but not all times) would complicate the system, and I'd hate to advocate it. ¶ Now that I think of it, I inadvertently today created the article Makemake with a k. That's the unchanged Rapanui spelling. Does it really look so bad? If we stick with c, it should become Macemace, a form that looks strange to people familiar with the concept. ¶ Incidentally, if you compare this article with its analogue at en:, you'll see that the Latin one is better. It has the advantage of not saying en:'s last sentence, for which there's no substantial ethnographic weight; indeed, the evidence (for which Métraux cites many sources) is heavily against it.
III.A. I know about those masculine mountains, but until anno 79, the two most famous mountains were feminine: Aetna & Olympia. And so are the biggest mountains, the Alpes. The Polynesian word for 'mountain'—mauna and maunga and mauga— itself looks feminine. Here's a fair sampling of Samoan mountains, peaks, hills, and such. Does anybody see a gender-related pattern? Incidentally, this is a good sample to challenge any system of translinguification. See how you'd Latinize each of them and what problems their shapes pose (the g here is /ŋ/ = Latin ng):
Afolau, Afutina, ‘Alao (mythologically masculine), ‘Alava, Anaota, Fa‘ani, Fao, Fiso, Fito, Fuiavea, Lalomauga, Lanomoa, Lanutata, Lanuto‘o, Lataiuta, Latauta, Lauti, Leafafa‘alava, Lepu‘e, Mafane, Malata, Mata‘aga, Mānu, Matafao, Matavanu, Maugaafi, Maugaiolo, Maugaloa, Maugamua, Maugasā, Maugasilisili, Misimala, Mulimauga, Olemaga, ‘Olomaine, ‘Olomoana, ‘Olosa‘a, ‘Olotele, Palapala, Penafu, Pioa, Piua, Piumafua, Pu‘e, Puga, Siga‘ele, Silisili, Siope, Si‘usi‘uga, Tafatafao, Tafua, Talito‘elau, Taumata, Te‘elagi, Tofua, To‘iavea, Tumu, Va‘aifetū, Vaea (mythologically masculine), Vailele.
(Surely you mean Olympus ?) —Myces Tiberinus 11:16, 21 Ianuarii 2007 (UTC)
Ah, of course. I was thinking of Olympia, -ae, f., defined in my dictionary as 'Olympia; a sacred region in Elis Pisatis, with an olive-wood, where the Olympian games were held; here, too, were the famous temple and statue of Jupiter Olympius'. Maybe that comes under the rubric of "it's countryside, hence feminine." But gender is linguistically unmarked in Polynesia (pronouns, for example, don't make he-she-it distinctions), and imposing ancient Roman ideas about it on Modern Worldwide Latin (tm) could be an exercise in unnecessary tinkering, not to mention cultural arrogance. ¶ As for your earlier comments, I like the reminder about nominatives in -es and shall post an updated set of rules using it. Discussion of "original research" belongs in its own thread; if one already exists in Vicipaedia, maybe we should go there. IacobusAmor 14:26, 21 Ianuarii 2007 (UTC)
III.B.2. I know, I know: nonneuters ending in e look odd, but they may look acceptable from the Pacific. Check out Makemake again. If you try to wedge him into the o-declension, you obliterate the reduplication if you trim the final vowel with Makemakus, -i : accordingly, you'd want to keep it, and the result would have to be Makemakeus, -i. Reduplications are semantically essential and should be kept in any system of translinguification. Polynesian and other Austronesian languages abound in reduplicated words, to an extent that native Latin (and Greek, for that matter) speakers might find surprising. And partial & full reduplication can convey meaning, as in Samoan nofo 'sit (sing.)', nonofo 'sit (pl.)', nofonofo 'sit a while, lounge around'. Latin has the partial kind in the perfect stem of verbs, but it's rare: in Polynesian languages, both kinds of reduplication occur—in verbs, nouns, adverbs, and adjectives, and often!
III.B.4. Yes, putting o-ending nouns into the "o-declension" (the second) is something of a no-brainer. It would be a pity, though, if this process were to obliterate reduplications (see the comment on III.B.2.).
III.B.5. This whole gender-of-placenames thing is a problem—perhaps in any non-Indo-European language. In Latin, mountains & rivers tend to be masculine, and cities & countries tend to be feminine (with plants & trees). But maybe that's just an Italic or Mediterranean bias, which needs to bend a little when it meets the rest of the world. Latin doesn't belong only to Europeans anymore. The goal of translinguification, I think, is simplicity & transparency. The ability to reconstruct original forms from Latin declined forms would be a bonus. The alternative is universally to make Polynesian nouns indeclinable, but that's inadvisable in view of declined forms already attested and out there. IacobusAmor 02:44, 21 Ianuarii 2007 (UTC)

Rules for Converting Polynesian Terms to Latin, update 1 (revised)[fontem recensere]

Here's an update, with rationales removed so the rules stand bare. The purpose remains simplicity, transparency, and potential reconstruction of the original.

I. Phonetics. Most letters are retained as in their original orthographies.
A. Vowels are five, pronounced approximately as in reconstructed Latin of the Golden Age (as summarized by W. Sidney Allen).
1. Polynesian a = Latin a, Polynesian e = Latin e, Polynesian i = Latin i, Polynesian o = Latin o, and Polynesian u = Latin u.
2. Polynesian long vowels are not marked, except in a record of an original form (as in the article Nafanua).
B. Consonants equivalent or quasi-equivalent to Latin ones are b, d, f, h, k, l, m, n, p, r, s, t, v. Likewise, the quasi-consonantal u (e.g., Samoan uati 'watch') is retained as "u."
C. Problematic consonants.
b: used in Fijian for /mb/; hence = Latin mb.
c: used in Fijian for /þ/; hence = Latin th.
d: used in Fijian for /nd/; hence = Latin nd.
g: when representing the phonemic /g/ = Latin g.
g: when representing the phonemic /ŋ/ = Latin ng.
k: always pronounced /k/; hence (N.B.) = Latin k.
ng: when representing the phonemic /ŋ/ = Latin ng.
q: when representing (in Fijian only) /ŋg/ = Latin gg.
w: except where attested (as in species epithets) = Latin v.
y: when used like the Latin consonantal i = Latin i.
ʔ: glottal stop (often orthographically ) = nil (ignored in Latin).
D. Generally unused consonants of the modern Roman alphabet: j, x, z (but they remain available if necessary).
II. Declinability. Polynesian nouns, though indeclinable in Polynesian, are declined in Latin. Exception: one-letter proper names (A, E, I, O, U) are indeclinable.
III. Gender & declension. The gender & declension of Polynesia nouns in Latin depends in part on their final vowel.
A. Gender.
1. Places & objects. The default gender is feminine in the first declension and neuter elsewhere unless otherwise indicated. Semantics may overrule the default.
2. Persons. The gender matches the sex of the referent.
B. Termination. Ideally unchanged in the nominative.
1. Ending in A. Declension 1.
Apia, maxima urbs et caput Samoae. . . .
Nafanua (Samoane: Nāfanua), dea. . . .
2. Ending in E. Declension 3, declined like mixed i-stems (e.g., Latin nubes) with an irregular nominative.
Tane (Tahitiane: Tāne ; Havaiane: Kāne), deus. . . .
3. Ending in I. <---Still problematic: bear with me.
a. Proper names. Declension 2, plural. (Except where convention adds -um to the nominative; e.g., Havaiium.)
Havaii (Havaiane: Hawai‘i), insulae (sunt, et civitas CFA). . . .
Havaiium (Havaiane: Hawai‘i), insula (est). . . .
b. Common nouns. Add -um. Declension 2. Genitive singular, -i.
tikium, (-i, n. ; Tahitiane: tiki), statua. . . .
4. Ending in O. Declension 3. Genitive singular: -onis.
Falealupo, vicus. . .
Lono, deus. . . . ~ Lono (-i, m. ; Tahitiane: Rongo), deus. . . .
5. Ending in U. Declension 4. Genitive singular: -us.
Tau (Samoane: Ta‘ū), vicus in Manuis insulis. . . .
IV. The beginnings of articles. Because of regularities in conversion, indications of declension & gender are often unnecessary, but they may be given. Indications of macrons & glottal stops in the original should be given. There's a distinction between Havaii, -orum (the archipelago & state) and Havaiium, -i (the island).

Fanum Sancti X[fontem recensere]

Among modern Latinists, the usual method of rendering cities whose names mean "St. X" (where St. i sthe local word for "saint" and X is some anthroponym) seems to be X-polis. Yes, this is stupid, and yes it is much uglier than Fanum Sancti X, but it is pretty ubiquitous. Egger is firmly commited to that method as well (though when I find that reverse index I should check if he lists any in the Fanum form). It seems to me that we are going to, at the very least, surprise a lot of people if we don't follow that naming scheme.

Sometimes I could swear that if the town of Ἀθῆναι did not already have a well known Latin name, Latin speakers would insist on calling it Minervopolis. --Iustinus 09:48, 23 Ianuarii 2007 (UTC)

Google gives 100 "Fanum Sancti" (exact word order). --Alex1011 13:35, 23 Ianuarii 2007 (UTC)
Aren't we accepting Sanctiacobi ? And Angelopolis must be an exception, because there isn't any "St. Angel." IacobusAmor 13:44, 23 Ianuarii 2007 (UTC)
Googling is a good idea, but 100 hits isn't really all that much. NOte how many hits there are for Jacobopolis. There are even more for Paulopolis, but I suspect something else is going on there. --Iustinus 18:42, 23 Ianuarii 2007 (UTC)
OH, and are we accepting Sanctiacobi? Sounds pretty awful to me. --Iustinus 18:46, 23 Ianuarii 2007 (UTC)
We have a substantial article with that title! IacobusAmor 19:51, 23 Ianuarii 2007 (UTC)
Well, yes, but we have lots of substantial articles with bad titles ;) --Iustinus 19:53, 23 Ianuarii 2007 (UTC)
You can add another six google hits for "Fanum Sanctae". --Alex1011 22:50, 23 Ianuarii 2007 (UTC)

Naming policy question at Disputatio:Octavius Mirbeau[fontem recensere]

Those of you who are interested in questions of naming policy, please come contribute to the discussion at Disputatio:Octavius Mirbeau. Most of it is in French, but if you don't know French there shoudl be enough Latin and English for you to figure it out. The basic issue is this: many authors have adjectival forms based on hypothetical Latinizations of their names, e.g. Shaw > Shavian, Giraudoux > Giralducian. Should these count as attestations of a Latin name in themselves, or not? --Iustinus 18:33, 27 Ianuarii 2007 (UTC)

Recent naming policy issues[fontem recensere]

Latinate names[fontem recensere]

Many European names, especially in Finland, Scandinavia, and sometimes the other Germanic areas are already Latinate in form. Should we treat these as declinable?

  • My opinion: For names that end in -us, definitely. I know I saw at least one locus where Svante Arrhenius declined his name. For other names, it is less clear cut. I'm told, for instance, that there are still some Germans with the surname Pistor. My feeling for non -us names is that we should decline them if they are obviously Latin, but not if it's uncertain. Note that these comments apply especially to names after 1800 or so. Before that, obviously if a name looks Latin, it probably is. --Iustinus 05:48, 29 Ianuarii 2007 (UTC)

Are names inheritable[fontem recensere]

If we know the Latin name of one family member, can we assume it for others?

  • My opinion: definitely in the case of noble families, e.g. Medices, Stuartus. Less certainly otherwise: note the usual example of Ioannes Dominicus Cassinus vs. Iacobus Cassini. It might perhaps be appropriate in some cases if the two lived close to the same time, and not around the 18-19th century when the naming practice was really changing. In general it does seem dangerous, and will definitely require a footnote if we allow it. --Iustinus 05:48, 29 Ianuarii 2007 (UTC)

Patronymics[fontem recensere]

Should patronymics be translated, or treated as surnames?

  • My opinon: when patronymics are true patronymics, they should definitely be translated, e.g. Leivus Erici filius. When they are simply inherited surnames, they should be treated as such: not translated without an attestation, e.g. bin Ladin, Johnson etc. --Iustinus 05:48, 29 Ianuarii 2007 (UTC)
Agreed on this. In the case of Sanchez - which is why I'm here in the first place - following the logic of the proposed rule, this would appear to be as Iustinus says "inherited". If his father's name was Sancius (or whatever the Spanish is, Santo?) then sure, 'Sancii filius. Nisi, let's go with Sanchez.--Ioshus (disp) 20:40, 16 Februarii 2007 (UTC)
The Spanish is Sancho. -ez is a residual pre-indoeuropean suffix (probably Basque, but nobody really knows) of the genitive, which was assimilated under the Visigothic kingdom as son: e.g. Sanchez=son of Sancho. In the middle ages, the patronimic changed every generation, and surnames did not get fixed until the XV c. Hence, in the cas of our discussion the surname should remain Sánchez, whereas in the Reges Aragoniae, for example, patronimics should be kept. Is that all right then? --Xaverius 00:51, 17 Februarii 2007 (UTC)
Just a note for future reference: the page in which "Sanchez" was under discussion was Disputatio:Gonzalus Sánchez de Lozada.
As for the suffix -ez, I have usually heard it attributed to a generalization either of the Latin 3rd declention -is genitive ending, or of the Gothic o>a-stem and i-stem genitive endings, likewise -is. --Iustinus 01:58, 17 Februarii 2007 (UTC)
According to this journal, Russian patronymics are translated into Latin and placed after surnames (which are preceded by first names). E.g., "Epistulae selectae Alexandri Schwartz Nicolai f. atque Theodori Korsch Eugenii f. ad Sergium Sobolewski Ioannis f. nunc primum in lucem editae". -- Alexander Gerascenco 07:58, 29 Maii 2008 (UTC)

Adjectives[fontem recensere]

Do vernacular adjectival forms that presuppose a hypothetical Latin form count as an attestation?

  • Oy. Can of worms. See disputatio:Octavius Mirbeau for details. I do note though that there definitely seem to be degrees of reliance on Latin: some of these adjectives cannot possibly be derived without Latin in mind, whereas others are less obvious. --Iustinus

Similarly: if we have a solid attestation of a true Latin adjective, should we be allowed to guess the nominal form?

  • This is tricky. My opinion would be that while it is on firmer ground than vernacular adjectives it should be done only as a last resort, with a footnote. --Iustinus 17:06, 2 Februarii 2007 (UTC)

Strictness[fontem recensere]

How strictly do we need to enforce this policy?

  • The current debate at disputatio:Octavius Mirbeau is making me wonder how much energy we really want to put into this, especially in borderline cases. Perhaps in some cases we could just put up a template saying something like Controversum est de huius commentationis titulo. Vide disputationem (which would be useful in other cases too), or maybe Titulus huius commentationes cum praeceptis VP:TNP non conformat. Si attestationem affere potes, quaesumus ut eam addes (which is more specific, and open ended), and leave it at that. On the other hand, not doing anything at all in such cases, just invites new users to do the same. --Iustinus 05:48, 29 Ianuarii 2007 (UTC)
  • Another point, which I quote from Andrew Dalby, in the aforementioned debate: "I think Vicipaedia in real borderline cases should tilt approximately 1 degree in favour of the opinions of local experts (who write it, after all)." --Iustinus 17:55, 29 Ianuarii 2007 (UTC)

Re: † Some Latinization patterns exist though, such as ... (e.g. XYZ)... use these?[fontem recensere]

(From the Project page) † Some Latinization patterns exist though, such as (name => name-polis) and some endings are consistently Latinized (e.g. -burgum, -vicum, -felda, -dunum...)... use these?

Not exactly very specific... As I may decide to make pages for different parts of the city of Leeds, should I apply these endings and latinise them etc. (I have already made Horsforth → Horsfordia), or should I just treat them as indeclinable? (in which case I would have to use something like vicus, regio, urbis pars, vicinum) --Harrissimo 22:18, 16 Augusti 2007 (UTC)

De nomine, or, Whether to transfer forenames to Latin[fontem recensere]

[Copied across from Disputatio usoris:IacobusAmor)

I didn't really know where I should write this, so here goes. I honestly think that "re-naming" everybody into Latin is HIGHLY superfluous and, even, stupid. Sure Latin 'has' an equivalent of the name "Edward" but that doesn't mean we change that some guy's name into "Edvardus" or what ever. There is an Icelandic version of the name Edward as well (Játvarður)- but nobody on the Icelandic Wikipedia would ever change the name "Edward Island" into Játvarður Island like you are doing here. And in most cases you are simply making up names. To people who's speak Latin natively that would sound weird; i.e. changing the spelling of a name to adapt it to the declension of said language. I don't know if this post has been coherent, but I hope you get my point. I just think it's stupid to call George Bush 'Georgius Bush'. Why not simply call him George? --BiT 11:53, 19 Augusti 2007 (UTC)

Its often necessary in latin to decline a name because of the grammar. For example especially to form dative, genitive and ablative. For this reason there is a long tradition in latin lit. of translating names. In middle ages people would translate their own names. Thus we have Isaacus Newtonus, Leonhardus Eulerus, etc...--Rafaelgarcia 12:00, 19 Augusti 2007 (UTC)
Well, Icelandic has declension just the same, and yet we have no need to alter the names of people. I can't shake the feeling that this is just "we can and we know how to, therefore we do". --BiT 12:17, 19 Augusti 2007 (UTC)
Out of the interwiki links to Georgius W. Bush, 4 or 5 (who use the latin alphabet) translate his name to various forms. A very interesting one, which seems to have a very good policy, is Lithuanian, which also has grammatical declension. 1st line: "George Walker Bush (Džordžas Volkeris Bušas, g. 1946 m. liepos 6 d. New Haven". It later refers to him in the Lithuanian form in the lower paragraphs. That may be a solution which could please both parties? --Harrissimo 12:21, 19 Augusti 2007 (UTC)
Vicipaedia has thousands of articles that do it the other way: the lemma is the Latin form, and alternate forms (including those in the original languages) go in parentheses. IacobusAmor 12:34, 19 Augusti 2007 (UTC)
I think stupid is the wrong word. It is commonly done even in today's English in some cases (like en:George Papandreou and en:John the Baptist); it was done more commonly in the past, and Latin is among the languages in which it was done regularly. People did have the habit of having a Latin name, or a name in another language, by their own choice. In some countries people still do this. At most it's old-fashioned, perhaps, rather than stupid. And, yes, Vicipaedia is old-fashioned!
I was uncomfortable with the practice when I first came here. It does, however, help the writing of articles (as Rafael says) because you need to decline words. In fact, if you look through a Russian article such as Буш, Джордж Уокер (that's George W. Bush) you will find that the English personal names in the text of that article are declined, just as they would be in Latin. I think, for Latin as for other flexional languages, such adjustments makes sense. Andrew Dalby (disputatio) 12:28, 19 Augusti 2007 (UTC)
But for example the word anime, at first I thought that it should be declined in Latin, but then I found out that I had been doing that in Icelandic, and there was always something wrong about it. And then I was like.. It hit me, you're supposed to say "Anime þáttur" = "Anime episode", so you use the words Anime as an indecl. adjective, and you actually decline the word episode. Same with manga, calling it a "manga novel" or "~ book" which sounds much better than trying to force a declension on a word (won't work on names though). But I digress, when I found that out, I also found something else. Manga could very well be an Icelandic word. A lot of Icelandic words end in -a, and they decline normally, but manga didn't feel right simply because it's not part of the language, and maybe that's how people who's native tongue is Latin (I always think about how the Romans would think the Latin sounded) would find it. (I ramble) but I still find it a pretty good perspective, we might add '-us' to some names, and '-us' is one of the most common endings in Latin, but it could still sounds wrong to a Roman. Like my name; Baldurus. That just sounds..lame! x( And in conclusion, Vicipaedia is (in my mind) supposed to be a new breath of life to Latin, like I feel that "Iaponia" is such an obsolete spelling.. oh but never mind me --BiT 12:37, 19 Augusti 2007 (UTC)
Someone (I can't remember whether it was me or another) decided that mangaca was too good a word to lose, so now we have a Categoria:Mangacae. You decline it like nauta of course. Andrew Dalby (disputatio) 12:44, 19 Augusti 2007 (UTC)
What was wrong with mangista? Cf. organum 'genus polyphonicum &c' and organista 'qui composuit organum'. IacobusAmor 13:52, 19 Augusti 2007 (UTC)
What was wrong with it? Its lack of ubiquity. I saw mangaca used in some article or other; I never saw mangista. Pure chance, I guess. Andrew Dalby (disputatio) 14:05, 19 Augusti 2007 (UTC)
Everybody recognizes -ista as a productive suffix, referring to 'one that does, performs, is associated with' (English & German: -ist; Spanish: -ista; French: -iste; all from Latin -ista, from Greek istes); but those who recognize -ca as a similar suffix must surely be few & far between. IacobusAmor 14:48, 19 Augusti 2007 (UTC)
I simply never saw the word till now. It's ideal, Iacobe! Feel free to change the category if you want to. It only has about seven members. Andrew Dalby (disputatio) 18:03, 19 Augusti 2007 (UTC)
Mangaca.. wouldn't you rather use "manga scriptor"... ^^ I also know you've gotten enough of my rants, but many of the words you've got now for stuff sound like the words that the "Icelandic Official Wordsmith Organization" comes up with. Long, compound words which people laugh at. "Sjálfrennireið" (Auto-sliding-machine) was once the word for car, now we use "bíll". "Rafreiknir" (electronic-calculator) or "rafheili" (electronic-brain.. >.< stupid!) used to be the "official" words for computer. Now we use "tölva". ^.^' I guess what I'm trying to say, is that I'm not trying to bug you, but rather to think "if Latin were used by millions of people every single day, would people really use these words which might make sense (cf. electronic-calculator) but might sound stupid". --BiT 12:53, 19 Augusti 2007 (UTC)
That's it, exactly. They would. How words are converted for use in other languages is very interesting -- and more variable than you think, BiT. Icelandic is a bit special because for hundreds of years it hasn't abutted on the territories of any other languages. For Latin it was totally different. 63 (or more) languages were spoken on the territory of the Roman Empire, and Latin had to adopt names (and did adopt common nouns too) from most if not all of them. Many of the place names that we see in our articles on Spain, France, northern Italy, Austria, etc., were adapted into Latin from pre-Latin languages, Celtic, Iberian, etc. In they went, and they nearly always got shoe-horned into one declension or another. That's one way in which Latin grew. Andrew Dalby (disputatio) 13:01, 19 Augusti 2007 (UTC)
But how can you know if a word would sound weird to a native speaker? (I'm not talking about words like aghippppppppzblaghus) --BiT 13:09, 19 Augusti 2007 (UTC)
Personally, I don't care about how a word would sound. When I am writing about Finland and come across words like 'Järvenpää', I just decline it using the Declension table. I once changed Anjalankoski to Anialancoscia, but as many pointed out, this is just wrong. If we were Nuntii latini, or an oral latin group then maybe we should care, but Vicipaedia is just for reference and reading, as far as I see it. --Harrissimo 13:17, 19 Augusti 2007 (UTC)
I guess I just think way to much about how Latin would be, were it my native language (well, I guess first of all I'd know it better ;]). --BiT 13:20, 19 Augusti 2007 (UTC)
You and Vercingetorix! IacobusAmor 13:53, 19 Augusti 2007 (UTC)
I must add that when I look again at our headings Britannia Spears and Coemgenus Federline, I think BiT could be right after all. Andrew Dalby (disputatio) 14:44, 20 Augusti 2007 (UTC)

Britannia isn't so bad is it? And if you look at the etymology of Kevin, it's just like most other names (i.e. just adding -us from the original). I did suggest above (referring to Lithuanian Wikipedia) "1st line: "George Walker Bush (Džordžas Volkeris Bušas, g. 1946 m. liepos 6 d. New Haven". It later refers to him in the Lithuanian form in the lower paragraphs. That may be a solution which could please both parties?" Which got sort of stampeded over. --Harrissimo 16:48, 20 Augusti 2007 (UTC)

What we do here, of latinizing only first names, is pretty standard among modern latin writers and is already a compromise. At the other extreme, in the past I've seen some people tend to get carried away here translating George Bush as Georgius Vepris and Albert Einstein as Albertus Monopetrus. I don't think Bush or Einstein would be pleased with those choices!--Rafaelgarcia 17:07, 20 Augusti 2007 (UTC)
As to Britannia, well (1) who says that Britney was a misspelling or an intentional variant of Brittany? (2) who says that Brittany means "from Britain"? (3) who says that "from Britain" translates into Latin as "Britannia"? (4) who says that "Britannia" is an acceptable Latin forename? (Hint: the answer to (2) is en:wikipedia, but no source is given. The source would no doubt be one of the books about your kids' names, and those books are not exactly sources of truth.) Andrew Dalby (disputatio) 17:46, 20 Augusti 2007 (UTC)
I can't really answer all of your questions individually, but [2] says that all of the Britney, Brittany etc. are interchangeable under latin Britannia, however it has no sources and most of its recommended links are Baby name sites. [3], which isn't a Baby name site says that they are all forms of Brittany (as in the North-West part of France) which is in Latin Britannia Minor, hardly a suitable first name. Maybe the writer came to the conclusion of Britannia by shortening it. --Harrissimo 18:09, 20 Augusti 2007 (UTC)

To return to BiT's original point, I must admit I rather sympathise with it. Andrew is right, of course, when writing: "People did have the habit of having a Latin name, or a name in another language, by their own choice" (emphasis Neandri). Those days even the surnames tended to be latinised. The situation seems different, when it comes to our times. For example, it would look quite quaint to latinise the Nightwish people, say, Erno Matti Juhani Vuorinen as Ernestus Matthias Ioannes Vuorinen (who wants to be called Emppu Vuorinen), or Marko Tapani Hietala as Marcus Stephanus Hietala (who calls himself Marco Hietala). Somehow this just isn't at home in the world of rock and its derivatives, imo. I'm almost sure these guys would like more their names latinised plene (Ernestus Matthias Ioannes Montanus, Marcus Stephanus Harenarius). I don't know but in any case Latinate and Romance names somehow just don't mesh with Finnish names. "Giscard Ryhänen" would make an ordinary Finn smile (at least behind the back...). But I won't press on this, because I know my irresponsibility in matters administrative... :-) Neander 19:54, 20 Augusti 2007 (UTC)

To state another point, Neander and BiT, I have always thought that's one of the quirks that makes Vicipaedia unique to other wikipedias. We're one of the few extinct language wikipedias, and certainly the best. We're allowed to be quirky and old fashioned, right? Do we really want to be the same as everyone else? --Ioscius (disp) 21:03, 20 Augusti 2007 (UTC)
I totally agree. I think researching place and proper names and translating them (in the latter case) is one of the more fun things about Vicipaedia. I don't want us to 'ascend' into the Über-P.C. world of English Wikipedia. --Harrissimo 21:37, 20 Augusti 2007 (UTC)
Hear! hear! IacobusAmor 21:43, 20 Augusti 2007 (UTC)
Ok ok, I've no problem with looking at the Sache from this angle, too. Mr.Zelig alias Neander 22:43, 20 Augusti 2007 (UTC)
Yea, but that kind of brings me to what I remarked about before; "we can and we know how to, therefore we do", which (as remarked above, by pointing out Georgius Vepris and Albertus Monopetrus) can end ludicrously. Translating place names by calque is in my idea very stupid. If you haven't gotten tired of my "Icelandic" references here's one more; "New York" was at some point translated as "Nýja Jórvík", a phrase which no one uses, and just sounds confounding and I can't help getting the feeling, that when people do this is Latin it's simply a way to say "I can translate wordz! I sum l4t1n hax0rr!". In my opinion translating words using their "etymology" and then translating then calque is a very dead way to create neologisms. Like Einstein, someone looks at it and thinks "hey, I know what that means in German- it means "one stone", better get my English-Latin dictionary and loan-translate the hell out of it!" --BiT 16:22, 21 Augusti 2007 (UTC)
From my point of view it's a good thing to translate the first names, names and the names of towns and nations which already have a Latin name. If we do not do it, it will remain very few latin words in our pages because of theirs shortness and there is also (as other have already said) the problem of declension. It would instead be very strange to translate modern names which doesn't have a known Latin traslation, e.g Silvius Berlusco was very amusing but it doesn't seem serious. Silvius Berlusconi it's the best solution.--Massimo Macconi 16:39, 21 Augusti 2007 (UTC)
Although like I've said that it is possible to keep borrowed proper nouns indeclined, sometimes Icelandic declines them so that they do not change in form in the nominative case, but have minor changes such as an "-s" in the gen. and "-i" in the dat. Is that not possible in Latin? --BiT 16:47, 21 Augusti 2007 (UTC)
Maybe, but it won't always sound like Latin, because many non-Latin names (e.g., George, Ralph, Trish) end in sounds that are highly unlikely in Latin nominatives. It works for Polynesian names, however, because all of them end in -A, -E, -I, -O, or -U. ¶ New York is an unuseful example for your argument because there's a standard, well-attested medieval Latin term for York, and adding Novum to it is something of a no-brainer. IacobusAmor 17:51, 21 Augusti 2007 (UTC)
That is to suggest George, Georgis? New York, Newis Yorkis? You can't be serious? Georgius is fine. I do agree with you, Baldur, that Monopetrus is excessive, and silly. But I disagree that Albertus is.
To remark on Massimo's point about Berlusco, I only chose Rocchius because I found Antonius Rocchius and Joseph Rocchius.--Ioscius (disp) 17:15, 21 Augusti 2007 (UTC)
By no means did I intend for that declension for EVERY Latin load word! I have nothing against Latin names given to countries and cities, but for certain names it does feel weird. --BiT 17:35, 21 Augusti 2007 (UTC)

Frankly I'm astonished that this would even be controversial. Latinizing forenames has always been the standard when writing in Latin. The exact same sources I cited above to show that surnames are usually not Latinized in the modern context, also show that given names usually are. ---Iustinus 19:26, 21 Augusti 2007 (UTC)

Bradford[fontem recensere]

If the surname Bradford (which was originally found in Yorkshire as the town is today and obviousbly has the same etymology as the town) is latinised as De Vado Lato[4][5]. Can the town be Vadum Latum? Harrissimo.

Nova Regula[fontem recensere]

I just posted a new rule but it's probably wise that I should write here just to make sure everybody agrees. What I said was:
In Biology, Botany, Names of Dioceses and Zoology, many places have adjectival forms of their names (often ending in -ensis). Use these but before the type of settlement. For example

  • Malhamensis Vicus (for Malham, a village)
  • Tristanenses Insulae (For Tristan da Cunha, a group of islands)
  • Ebebiyinense oppidum (For Ebebiyín, a town) [6]
Consentitisne? Harrissimo 20:40, 26 Novembris 2007 (UTC).

Ex pagina hunc removi[fontem recensere]

We might need separate rules for how to best Latinize, e.g. what to do if patronymics are invoved, or if multiple Latinizations exist like Iohannes/Ioannes, or with the many forms of Gulielmus...

  • Another thing that needs to be determined is what to do with infixes like von, de, ben, etc. For instance, for ben, ought we just supply the genitive of the father's name and remove the infix altogether? This could get bumpy with names that are not easily latinizable. For de as well, use de plus ablative or remove the infix and use genitive of place? Same problem as above arises...
  • Some attested Renaissance composers (from facsimiles in Willi Apel, Notation of Polyphonic Music, 1953), omitting numerous obscure theorists:
Guillaume du Fay = Guillermus Dufay (pp. 102-103) [Dufay liked to write the FA part of his name in musical notation as the B-flat just below middle C = fa.]
Josquin [des Prez] = Josquinus (p. 154)
Francesco Landini = Magister Franciscus Caecus Horganista [sic] de Florentia (pp. 390-391)
Johannes de Muris [sic] (p. 395)
Johannes Ockegem [sic] (p. 167)
Philippe de Vitry = Magister Phillipoti de Vitriaco (p. 395)IacobusAmor 02:30, 22 Augusti 2006 (UTC)

Loco Sancti etiam Divi invenitur: Divi Georgii pro Sankt Georgen (St. George), vide Instrumentum Pacis Osnabrugensis

  • Some Latinization patterns exist though, such as (name => name-polis) and some endings are consistently Latinized (e.g. -burgum, -vicum, -felda, -dunum...)... use these?
  • The question arose today about Sancta Lucia (civitas). I don't think it works to change this to Fanum Sanctae Luciae -- if the name refers to something as big as an island or an independent state, can it really be described as a Fanum? See this discussion. Andrew Dalby (disputatio) 12:39, 28 Februarii 2008 (UTC)

Re. acronyms: (But how language-independent is language-independent? A majority of the languages of Europe, gauged by Wikipedia interlanguage links?)

For ancient Greek we should probably transliterate as the Romans did, but not so for modern Greek names. The pronounciation of Greek and Latin has changed since Roman times so it no longer makes sense to transliterate Greek as it was 2000 years ago. Michael Dukakis's surname, rendered in Greek as Δουκακης, would have been pronounced as Du-kak-ays in Roman times, thus giving the transliteration of "Ducaces" perfect sense at that time. However, that name is now pronounced as Du-kak-is and, while if we use the classical Latin pronounciation "Ducaces" it would not be pronounced too differently from the modern Greek, if we use the medieval/church pronounciations that some prefer it would be pronounced as Du-ka-chays and if we would use the Americanized pronounciation it would be Du-kay-sees, which are way off the actual pronounciation of the name. "Dukakis", however, using all commonly used pronounciations of Latin, would be pronounced the same way as the modern Greek. We should use the ISO rules when transliterating modern Greek except for when the use of the Roman method of transliteration would produce a pronounciation identical to the modern Greek. For example we should use "farina phyllo" rather than "farina philo". -KedemusKedemus 04:41, 19 Octobris 2007 (UTC)

Re: "The pronounciation of Greek and Latin has changed since Roman times."—Those of us who aren't Roman Catholic prelates and their partisans use the reconstructed classical pronunciation. I myself would do as most people around the world have always done, and proceed not from the spelling, but from Michael Dukakis's own pronunciation, which in this case yields Latin Dukakis (or conceivably Ducacis, though K is a productive letter in my dialect of Latin, so I'd prefer K, not least because the folks at the Vatican would have to pronounce Ducacis as /dukačis/). Vicipaedia's policy on modern surnames—in general, don't change their spelling—would yield the same result: Dukakis. IacobusAmor 11:17, 19 Octobris 2007 (UTC)
For Greek of all periods Vicipaedia actually does use the Roman system: you can already find on Vicipaedia hundreds of examples of modern Greek people and places so transcribed. So, if Dukakis was a Greek politician, living in Greece and habitually speaking and writing in Greek, he would write his surname Δουκάκης and his name here would be Ducaces. But in fact he is an American politician: he usually writes his name in the Roman alphabet, and so do most of the people who address him and write about him. Therefore the question of transliteration doesn't arise. Following our usual rule, we use his own form of his surname, Dukakis. Andrew Dalby (disputatio) 11:40, 19 Octobris 2007 (UTC)

De translitteratione consonantium [tʃ] (e pagina IacobiAmor depromptum)[fontem recensere]

Hi! There is a lot of page you can rename...Michael Gorbachev ---> Michael Gorbasev, Maria Sharapova ---> Maria Sarapova, Petrus Tchaikovski ---> Petrus Saikovski... However, I think it's not a good way to transliterate!! :) Btw, can you link me a page explaining how to transliterate from Russian, ne erram again?? Good work and sorry for my errors (and my English)! Osk 10:13, 18 Maii 2008 (UTC)

It's an experiment. An alternative, and perhaps a preferable one, is <Siefsienco>. According to the models you cite ("Gorbachev" et al.), you should originally have transliterated the name as <Shevshenko>. Using <sc> is horrible here because it makes everybody—except Italians & certain Roman Catholics—say [sk] where you want them to say [ʃ]. ¶ The models you cite may be taken from the traditional English method of transliteration, and there's no good reason that should prevail. Let's consider them tentative. ¶ For an example of [tʃ], see Socolata; you'll note that the German source wants to render the first phoneme like the German <sch> or the English & Romance <ch>; but again, there's no good reason why those systems should prevail. IacobusAmor 12:40, 18 Maii 2008 (UTC)
It's worth noting, maybe, that there has been discussion of transliterating names from Cyrillic script on the page Disputatio Vicipaediae:De nominibus propriis. Andrew Dalby (disputatio) 20:24, 18 Maii 2008 (UTC)
Maybe somebody should prepare a table of transliteration from Cyrillic. ¶ In which we might have Sieftsienco or Sieftsienko (Anglice: Shevchenko, not an uncommon Slavic name), Gorbatsiov (Gorbachev), Siarapofa (Sharapova), Tsiaikofskii (Chaikovskii, Tschaikowskii, Tchaikovski), etc. IacobusAmor 20:39, 18 Maii 2008 (UTC)
The Regula is Ševčenko (or Sevcenko), Gorbačev (or Gorbacev), Šarapova (or Sarapova), Čajkovskij (or Cajkovskij) etc., in accordance with ISO. Until there is consensus on another rule, this is the one everybody should stick to.--Ceylon 21:00, 18 Maii 2008 (UTC)
I agree with Ceylon. ISO is used also in Italian wiki... (ISO <Gorbačëv>) Osk 21:57, 18 Maii 2008 (UTC)
<Ševčenko> is OK if we allow diacritics, but why do we allow diacritics? <Sevcenko> horribly mispronounces as [seukenko], so that's out. And <Gorbacev> mispronounces as [gorbakeu]. So Osk tells us that Gorby is ISO <Gorbačëv>, and Ceylon tells us he's ISO <Gorbačev>: which is he? <Čajkovskij> & <Cajkovskij> look strange, at least to English-speakers, in view of en:Tchaikovsky. IacobusAmor 23:07, 18 Maii 2008 (UTC)
Unfortunately, so many of the sounds are missing in Latin, period. Unless you keep the diacritics or otherwise indicate it's supposed to be pronounced as foreign, they're all going to be horrible mispronunciations. It's probably better to leave things systematic with the rules, whatever they are, unless you have some kind of evidence that [seu'ken.ko] is somehow more horrible than, say, the pentasyllable [si.eut.si'en.ko], which the classical pronunciation would make of Sievtsienco ([j] is restricted in its environment—the Romans did not have it after consonants, except sometimes across morpheme barriers, as in abieci; normally it must be [i]). On a side note, if C is was actually that horrible a substitute for [tʃ] in Latin letters, it's a wonder so many languages thought to equate them, from English to Romanian; even Slavic languages like Russian have interchange from к to ч (плакать/плачу, etc.), so many people are used to relating /k/ to /tʃ/ (more, I'd wager, than relate /tsj/ to /tʃ/—at least in Europe). —Mucius Tever 02:44, 19 Maii 2008 (UTC)
Re: "[j] is restricted in its environment—the Romans did not have it after consonants, except sometimes across morpheme barriers, as in abieci; normally it must be [i]."—It wouldn't be surprising if people in the midst of everyday life (as opposed to poets, reaching for metrical precision) pronounced, say, <ciere> as [kje:re], rather than [kĭe:re]. ¶ As for the conversion of the Cyrillic <ч>, that's the sound of [tʃ], and in at least one common case, Vicipaedia renders its Spanish & English analogue by an "s" (see Socolata). So there you have it: phonetic [tʃ] = Cyrillic <ч> = Latin "s." I'd still argue for <si> though, because fresh in one's ears is the Polynesian method; in Samoan, for example, English <choke> becomes <sioka>, and (for a voiced example) English <Joe> becomes <Siō>. IacobusAmor 03:20, 19 Maii 2008 (UTC)
And here's another problem with /ʃ/, the same #$%#$% phoneme: see our lemma for Aung San Suu Cii. Birmanice, the name is rightly AungSanSuuKyi1.png), which English renders as Aung San Suu Kyi; however, that last bit is pronounced, not like [kji], but rather like [ʃi:], for which Cii (Latine = [ki:] is (I assert) worse than Si or Sii would be. The original contributor spelled it Tsiis, but noster Andrew changed it to Cii, and so here we are with another mess. Perhaps "Kyi" was formerly Birmanice pronounced [kji], and that's what Andrew was approximating? IacobusAmor 03:54, 19 Maii 2008 (UTC)
"I will not be blamed!" (quote from some novel or other). The Burmese pronunciation is [tʃi:] rather than [ʃi:]; not that that's much help. With Burmese, as with Tibetan, pronunciation is a long way from spelling. The c (if I can think back to that decision through the mists of time) was intended to suggest the widespread post-classical Latin pronunciation [tʃ]; the letter c is sometimes handy for such a purpose, allowing us to reserve k and s for their well-known and unambiguous uses.
But there are no perfect answers. A year ago, when Aung San Suu Kyi was first named here, I was feeling my way with Vicipaedia and foreign scripts. I am now more confident than before that we should adhere as closely as we can to some independent standard. And for Cyrillic, luckily, that's not too problematic. Andrew Dalby (disputatio) 09:01, 19 Maii 2008 (UTC)
Since I like to be accurate (or pedantic), let me add that the Burmese spelling of that last syllable, if transliterated IPA fashion, would come out as /kjañ/. Since the history of Burmese spelling goes back via medieval Mon to late-antique Pali, with compromises and clever adjustments at each stage, it would be hard to be certain whether in Burmese the kj or the were ever pronounced anything like the way they look. Andrew Dalby (disputatio) 09:11, 19 Maii 2008 (UTC)
Re: "The Burmese pronunciation is [tʃi:] rather than [ʃi:]."—I've marched with hundreds of Burmese chanting her name, and they all say [ʃi:], not [tʃi:]. Maybe this variance reflects the difference between the literary language and the colloquial one—in which case Vicipaedia should probably recognize both variants. ¶ Re: " The c . . . was intended to suggest the widespread post-classical Latin pronunciation [tʃ]."—That's the kind of ambiguity I think we should avoid! In the classical pronunciation, c is always and ever [k], and that's how I'll pronounce it if that's how you write it. Avoiding that ambiguity is why we might prefer s there, and k in words with a genuine [k]-sound, like Tokio (compare Tocio), for which we shouldn't be surprised if attested sources of Tochio or Tochium can be found: our medieval & renaissance forebears were reluctant to use k, but they knew that ci would mislead their readers, so they inserted an otherwise gratuitous h. Likewise we have Pechinum for what might better have been spelled Pekinum (or Pekino or Pekin). If, as we're told upthread, we can freely deploy things so startling as modern ISO diacritics, we can surely deploy k & y, letters that have more than 2000 years of attestation in Latin. IacobusAmor 11:43, 19 Maii 2008 (UTC)
As to Aung San Suu Kyi, it seems that while en:wiki and my teachers are in agreement with me on the pronunciation, you have hundreds of native speakers in agreement with you; we must diagnose free variation in current Burmese, I guess ...
And I notice this on the en:wiki disputatio page: "Listening to a Burmese pronouncing her name, the final part sounds like IPA:[tkʰì] rather than IPA:[tʃì]." Further free variation??
It's possible—nay, even probable—that the pronunciation of a shouted or chanted word (as at a political rally) will differ from the pronunciation of a quietly & carefully spoken word. IacobusAmor 13:36, 19 Maii 2008 (UTC)
On the spelling, though, in my pedantic paragraph, I should have said krañ not kjañ as the literal equivalent of the Burmese spelling. Sorry about that.
As to c, we may possibly never agree, Iacobe! But it is worth reflecting why early Latin had three letters, c, k and q, for what is widely taken to be the same phoneme. My answer, for what it's worth, is that they represent three allophones, the q being the furthest back, and the c being the furthest forward -- i.e. the most palatalized of the three. So I would agree with you that "in the classical pronunciation c is" always and ever /k/, but not that it was always and ever [k]. Andrew Dalby (disputatio) 12:12, 19 Maii 2008 (UTC)
Maybe. English has at least two of those: the /k/ of keep (in the front of the mouth) and the /k/ of cool (in the back). ¶ Also, note that Latin c in castra moved forward in the mouth to become English ch in -chester. IacobusAmor 16:29, 19 Maii 2008 (UTC)

How to transcribe in Latin the sound given as IPA [tʃ]?

c—Ceylon (secundum ISO et Regula)
c—Andrew Dalby ("to suggest the widespread post-classical Latin pronunciation," sed solum snte e et i?)--Ceylon 06:28, 22 Maii 2008 (UTC)--Ceylon 06:28, 22 Maii 2008 (UTC)
č—Ceylon (secundum ISO et Regula) + Osk
ch—Vicipaedia (s.v. Chilia, Massachusetta)
s ante i—Vicipaedia (s.v. Sina)
s ante a, e, o, u—Vicipaedia (s.v. Socolata)
si ante a, e, o, u—?
tch—Vicipaedia (s.v. Petrus Tchaikovski)
ts ante i—Vicipaedia (s.v. Domus Tsingiana et Pronuntiatio Ecclesiastica: "ci pronuntiatur tsi") + IacobusAmor Ecclesia Catholica (Qingdao = Tsingtao: http://www.catholic-hierarchy.org/diocese/dtsng.html)
tshEcclesia Catholica (http://www.catholic-hierarchy.org/diocese/dtshu.html)
tsi ante a, e, o, u—IacobusAmor
tz—Vicipaedia (s.v. Tzadia, Tzekia)

I'd prefer to keep s(i) for [ʃ], as in Sicagum, and according to the practice of the Ecclesia Catholica (e.g., Sinyang = Xinyang: --Ceylon 06:28, 22 Maii 2008 (UTC)http://www.catholic-hierarchy.org/diocese/dsiny.html). IacobusAmor 16:29, 19 Maii 2008 (UTC)

I only come into this because you remembered my spelling of Aung San Suu Cii, a year ago, and I tried to recall my reasons for it! My current opinion, as I hinted above, has a different starting-point. I think we should find a good international standard for transliterating each language written in a non-Latin script, and adhere to it. Andrew Dalby (disputatio) 17:15, 19 Maii 2008 (UTC)
There's the complication that we should cite attested forms when they exist— but of course we don't have to cite them as the primary spellings if we develop rules whose consistent transliterations are at variance with them. IacobusAmor 17:32, 19 Maii 2008 (UTC)
Yes, I'm with you on that. Andrew Dalby (disputatio) 17:35, 19 Maii 2008 (UTC)
As for the use of ISO forms, I don't see why they couldn't profitably appear in parentheses, while the lemma remains as Ciceronian & nonnewfangled as possible, like this:
Demetrius Demetri filius Siostakovits (Russice: Дмитрий Дмитриевич Шостакович, Dmitrij Dmitrievič Šostakovič). IacobusAmor 17:40, 19 Maii 2008 (UTC)
One reason against doing that is that it gives the newfangled "nonnewfangled" form a sort of unmerited validity compared to the ISO form—the invented form gets the dignity of the lemma, while the standard but non-Latin form is getting the parenthetical aside. —Mucius Tever 21:11, 19 Maii 2008 (UTC)
The nonnewfangled form would indeed get that dignity, but whether it would be merited or not is what we're discussing. IacobusAmor 22:03, 19 Maii 2008 (UTC)
Except to further the cause of the Encyclopedia That Adjusts Inconvenient Facts To Suit Itself, I'd really have to say no. Foreign words are foreign, and it's not our place to nativize them; let them look foreign (hopefully putting aside the mysterious allergy to diacritics while we're at it), or let them be nativized elsewhere. —Mucius Tever 00:39, 20 Maii 2008 (UTC)
As if Vicipaedia didn't nativize foreign words left & right, almost everywhere. Perhaps you're arguing that, for example, our lemma Vladimirus Eliae filius Lenin should be changed to Владимир Ильич Ленин? After all, "foreign words are foreign." IacobusAmor 02:24, 20 Maii 2008 (UTC)
It's fine to respell words in characters recognizable to those who use the language, and this is what the rule saying to use the ISO transliteration is about; where we cross the line is when we try to invent new forms corresponding to how we think the Romans (or, for whatever reason, the Polynesians) would have spelled the words if they had borrowed them from an illiterate culture. As for the rule nativizing forenames, I have been thinking lately that it is a little bit broad in its application; but, as my understanding is that it's not a custom original to Vicipaedia, I can't pretend to hinder it here—I would, though, probably invent a slightly different rule, if pressed, for non-Vicipaedia use. —20:50, 20 Maii 2008 (UTC)
To whoever wrote that: as with all living languages, "the Polynesians" borrowed many of their words from other languages without the benefit of literacy: they heard people speaking, and took the spoken words into their own languages. English check became Samoan siaki; chalk became sioka, as did choke; champion became siamupini; cold chisel became kolosisi: so we know that English [ʧ] became Samoan [s]. Similarly, English shilling became sēleni and cashier became kēsia: so we know that spoken English [ʃ] = Samoan [s]. With voiced consonants, English giraffe became serafa; jack became siaki (and Jack became Siaki); germ became siama; jam became siamu; jug became sioki; Joe became Siō; cabbage became kapisi; college became kolisi: so we know that spoken English [ʤ], no matter how it was spelled ("ge," "gi," "j"), became Samoan [s], with a tendency to be plain [s] before high vowels, and [si] before mid & low vowels and when the original [ʤ] was at the end of a word (before mid & low vowels, [si] usually becomes [sj]). That's the way living languages borrow: from the sounds of the words as spoken: they alter foreign words to fit their own phonology. It's a matter of personal regret that I seem to be the only person here treating Latin as a living language, rather than a dead one. ¶ Examples from all the world's languages could probably be adduced in evidence. A language at hand, via the internet, is Japanese, and you see how it adapts foreign words to its own phonology: alpha = aruhua, acre = eekaa, cycle = saikuru, kilometer = kiromeetoru, franc = huran, bolt = borutu, volt = borutu. The living Latin language will do the same. ¶ That's not to say that literary borrowing doesn't also occur in living languages, but it's a different process, involving artificial construction or even the deliberate revival of dead forms. Missionaries forced English jubilee to become Samoan iupeli instead of its naturally expected form, *siupeli; and by going back to the original Hebrew, they forced cherub to become kerupi instead of *selupi; likewise chariot became kariota instead of *selioti. Thanks to fastuous meddling, modern Samoan takes a nonnegligible number of words directly from ancient Hebrew and Koiné Greek (e.g., words for the concepts of "concubine," "eagle," "fox," "gold, "mustard," "snow," "wolf"). The natural process of borrowing is usually less obscurantist. IacobusAmor 13:25, 22 Maii 2008 (UTC)
I wrote that; hit an extra tilde and thus my name didn't make it through. Anyway. There is more than one way that living languages borrow words—by sound, or by spelling. When borrowing by sound, some concessions may be made; it doesn't necessarily have to be an exact copy of the foreign word, and it doesn't necessarily have to be wedged into the borrower's phonology; for example, older Russian borrowed /f/ but not /θ/ in Greek words, though they originally had neither. When borrowing by spelling, concessions can be made as well, such as transliteration, or the shedding of diacritics. The rules for what kind of borrowing happens can be unpredictable: Japanese borrows from English by sound and ignores the spelling, but when it borrows Chinese words it takes the spelling and uses its own pronunciations. English tends to borrow both foreign sounds and spellings (so a French name, for example, will keep its silent final letter, and unusual sounds like /x/ or initial /ts/ may be retained among those who know how to pronounce them). A language that prefers to write phonetically might respell everything (like your Samoan example), or it might only respell some common nouns and leave others, and proper names, visually recognizable (as in Spanish or Italian, say). A borrowed spelling does not necessarily indicate an affected foreign pronunciation. At any rate, what the Vicipaedia should do should be based on what the Romans did, not arbitrarily what the Japanese or the English or the Samoans or the Spanish do. Pre-classically, of course, the Romans are known to have taken the Samoan solution; in those times the Greek φ, θ, χ, υ, and ζ made it through in borrowings as their nearest equivalents, p, t, c, u, and ss. But from the second century B.C. onward, including the classical period we hold up as a model to imitate, they'd already discarded that practice and taken to borrowing spellings: z was borrowed wholesale, sound and spelling; ph, th, ch, and y were written even though they were not, by many people, pronounced any different from p (later f), t, c, and u (later i). If the Romans were not squeamish about borrowing even entire letters from foreign alphabets to represent that language's foreign sounds, even if those sounds went unused, why should the much more moderate route of plain transliteration be rejected as too extreme? If the Romans had no qualms about inventing digraphs to represent alien phonetic units, why should that privilege be denied to those who borrowed the Romans' letters to write sounds not alien to them? —Mucius Tever 03:27, 23 Maii 2008 (UTC)
I think la.wiki should admit foreign names, so I agree with Mucius Tever (imvho latinising proper names is awful). Latins used Latin alphabet: that's why I think we should use the Latin alphabet (with diacritics) transliterating Russian (or Arabic, etc.) names. Osk 15:43, 20 Maii 2008 (UTC)
Re: "latinising proper names is awful":—Latin has Latinized proper names since the recorded beginnings of the language. All languages adapt foreign words to their own phonologies. IacobusAmor 13:25, 22 Maii 2008 (UTC)
Incidentally, I was thinking today that socolata is a bad example for /tʃ/, without evidence that it is supposed to be an /s/ representing /tʃ/—it could just as easily be for the original Nahuatl x /ʃ/, which apparently the common /tʃ/ is an over-Hispanicization of; come to that, it may even come from a different indigenous language that did render it by [s], if there were any such. —Mucius Tever 20:50, 20 Maii 2008 (UTC)
A small point off topic (sorry)- I've seen ch as a way of writing [tʃ], but then, how would [x] be written?--Xaverius 14:46, 21 Maii 2008 (UTC)
What do you mean by [x]? IacobusAmor 03:39, 24 Maii 2008 (UTC)
The [tʃ] (and all other phonemes) in proper names taken from languages which do not use the Latin alphabet should be rendered in different ways according to which transcription is standard for the language they belong to. This is what English does when writing Deng Xiaoping (where x is not 'ks') --Ceylon 06:28, 22 Maii 2008 (UTC)
In writing the name as Deng Xiaoping, we're ignoring diacritics that might indicate the tones, essential features of the words in the original language. This practice is common in English, where tonal marks are ordinarily omitted from names taken from Chinese, Navajo, Yoruba, and other tonal languages, and where the French café can be acceptably written as cafe, façade as facade, naïve as naive, résumé as resume, and rôle as role. Writing-systems can get along without diacritics. IacobusAmor 13:25, 22 Maii 2008 (UTC)
or Kyat (pronounced 'chat') or al-Qaeda (where q is not 'kw') - using standard transcriptions like these is becoming more common even with Russian names such as Čechov / Chekhov. Latin allows foreign last names unaltered, as in Vinstonis Churchill (not Siursill), which is no more or less problematic than allowing Čajkovskij (which, however strange for the English eye, is the standard transcription used by scholars worldwide). --Ceylon 06:28, 22 Maii 2008 (UTC)
That's a problematic use of the notion of standard. If English, the world's universal language (aside from Latin, of course!), were taken for a standard, then Tchaikovski might be regarded as the standard transcription. IacobusAmor 13:25, 22 Maii 2008 (UTC)
Indeed—and the choice of ISO for transliteration was motivated not merely to have a standard (standards are many), but to have a relatively neutral standard not tied to some third language, and a standard that addressed many scripts (rather than consulting first one entity then another for rules when encountering different languages). —Mucius Tever 03:27, 23 Maii 2008 (UTC)
A good point, but reality is going to be messier. We've already seen that, for [ʧ], we've got attested Latin examples of "ch," "s," "tch," "ts," "tsh," and possibly "tz." Of these, the most frequently attested in the names of Roman Catholic dioceses may be "ts." IacobusAmor 03:39, 24 Maii 2008 (UTC)
Indeed—and the choice of ISO for transliteration was motivated not merely to have a transliteration (transliterations are many), but to have a relatively neutral standard offered specifically for international use (rather than consulting first one entity then another for forms when encountering different words, and producing forms de novo contra NOR when none can be found). Incidentally, I was thinking today that Sinae is a bad example for /tʃ/, without evidence that it is supposed to be a Roman /s/ representing an immediate foreign /tʃ/; disregarding the fact that the root Sin- is often considered to be from a separate source than the root Chin-, the word first came through Greek and before that, according to my dictionary, through Arabic, and it was the Arabs, or (if you reject that hypothesis) at least the Greeks, who turned the original sound ([tʃ]-like or not) into /s/, not the Romans, so I don't think it's a good idea to mark it as a Latin practice; they were just borrowing the Greek word. (Similarly, many Biblical names were Graecized before they were Latinized.) —Mucius Tever 06:53, 24 Maii 2008 (UTC)
If there were a genuinely Latin system of transcribing certain scripts (as English has for Russian, thus arriving at Tchaikovsky), this should be adopted. But I doubt such a system can be reconstructed for even the most common languages such as Russian, Chinese, or Arabic (much less for Samoan) - there will never be consensus on how to write dozens of strange phonemes, and even if there were one, it would have to be learned and followed by all Vicipaedia authors. This is why standard transcriptions are useful - because even if you are unfamiliar with say the Burmese system of using 'ky' for modern [tʃ], it is easy enough to check how the name is written in English. (Shouldn't this discussion be copied to Disputatio Vicipaediae:De nominibus propriis? --Ceylon 06:28, 22 Maii 2008 (UTC)
Re: "even if you are unfamiliar with say the Burmese system of using 'ky' for modern [tʃ], it is easy enough to check how the name is written in English":—It may be "easy enough," but people won't do it. They'll know that the Classical Latin consonant "k" is pronounced [k], and the Classical Latin vowel "y" is pronounced like the French "u," and so the Latinized Burmese name "Ky" is not going to be pronounced [ʧi] (except perhaps by people already familiar with Burmese culture). The Burmese name Ky and the Vietnamese name Ky, if they're written identically in Latin, are going to be pronounced identically. IacobusAmor 03:39, 24 Maii 2008 (UTC)
I was rather thinking about contributors to Vicipaedia, who can easily check in other wikipedias how to transliterate foreign names according to a widely accepted standard (rather than having to memorise concordances of peculiar Latin transliterations), than about readers. As far as readers are concerned, by your logic, they would be mispronouncing Kyat and Deng Xiaoping and al-Qaeda in the context of English or any other European language as well. In German, for instance, phoneme-grapheme equivalence is usually quite straightforward (as in Latin), but when you see a foreign name or loanword, everybody knows that they have to pronounce it according to the rules of its original language.--Ceylon 08:36, 24 Maii 2008 (UTC)
"Everybody"? You must never have heard Americans pronouncing the American names of non-American places: Beaufort (/ˈboʊfɚt/ "BO-furt"), North Carolina, but Beaufort (/ˈbjuːfɚt/ "BYEW-furt", South Carolina; Berlin ("BURR-l'n"), Ohio; Havana ("huh-VANN-uh") Arkansas, Florida, Illinois, North Dakota, and West Virginia; Montpelier (/mɒntˈpiːljɚ/ "mont-PEEL-y'r"), Vermont; New Madrid ("new MAD-rid"), Missouri; Versailles ("ver-SALES"), Illinois, Indiana, Kentucky, Missouri, Ohio, and Pennsylvania. People usually pronounce foreign spellings according to the phonology of their own language. ¶ Re: "by your logic, they would be mispronouncing Kyat and Deng Xiaoping and al-Qaeda in the context of English or any other European language as well.":—Thanks for pointing this out, because that's exactly what they do. For many Americans, al-Qaeda is "ahl-KAY-duh" (with the syllable "KAY" ruining all four of its original phonemes); for many Americans, especially military personnel who've served in the Middle East, Iraq is "eye-RACK." The man whose name used to be spelled Mao Tse-Tung was in America universally pronounced "mousey TONGUE," and of course Americans (and probably most Europeans) still pronounce Deng Xiaoping and the names of other famouse Chinese leaders without any sense of the original tones whatsoever. IacobusAmor 11:54, 24 Maii 2008 (UTC)
No doubt you are right, but this supports my analogy: If it is okay to use transliterations prone to mispronunciation in English, why is it not okay in Latin? (Since surely mispronunciations they are - whoever (Wade-Giles?) chose to transliterate Mao Tse-Tung (Pinyin: Zedong) this way, did not intend the last syllable to be pronounced /tang/ (but neither did he count on anyone getting the tone right).--Ceylon 08:08, 25 Maii 2008 (UTC)

I don’t know if this is a voting matter, but if it is I would like to express my support for the point of view defended by Mucius Tever and Ceylon: in the absence of a well-founded Latinised form of a name, we should conform to the ISO-norm. Moreover, if this produces a form that doesn't invite (by analogy) a certain declension, the form should be treated as indeclinable. Declension may then be provided by the addition of generic nouns ('de praeside Bush', 'in urbe Nouakchott', etc.) I also consent with Ceylon's suggestion that this discussion be moved to Disputatio Vicipaediae:De nominibus propriis. --Fabullus 11:52, 25 Maii 2008 (UTC)

Rule 3 in Libri et arte facta[fontem recensere]

I rewrote this rule I wrote a long time ago because it was written poorly and left quite a bit of things in a loophole status (there is really no reason why a written work's title should be treated any differently from an oral work's title).

Previous discussions:

Mucius Tever 17:25, 14 Iunii 2008 (UTC)

Nexus[fontem recensere]

Addidi nexum ad paginam inscriptam Index_nominum_Latine_redditorum. Nam illae tabulae maximo usui sunt eis, qui opem quaerunt in nominibus propriis vertendis. --Iovis Fulmen 17:43, 27 Iulii 2008 (UTC)

de translitteratione nominum Russicorum[fontem recensere]

Vide Disputatio:Antonius Chekhov#De orthographia nominis. --UV 14:57, 3 Maii 2009 (UTC)

Following this discussion, there are three questions:
  • Shall we change the norm, so that j and diacritics would be kept in ISO transliterations, just like it is kept in German or French names?
  • For Cyrillic, should the transliteration system be based on ISO 9:1995, ISO/R 9:1968 sub-standard 1 or something else? ISO 1995 sticks on 1 Cyrillic letter = 1 Latin letter, while ISO 1968 breaks this principle, but looks more natural (from my POV at least). There are also other standards that look even more natural, such as the System of Academy of Sciences of the Soviet Union, but this one is for Russian only — maybe we should be more universalist and have one system for all Cyrillic languages.
    • Probably, there should also be an exception for non-Latin-alphabet names originating from Latin-alphabet ones — such as Otto Schmidt, a Russian with German surname.
  • Whatever the norm is, shall it be enforced? Currently, it would require dozens or maybe hundreds of articles to be moved. --Gabriel Svoboda 06:40, 6 Maii 2009 (UTC)
In my point of view, ISO/R 9:1968 (or, possibly, the System of Academy of Sciences of the Soviet Union) should be used for translitterating the Russian surnames (but not French, German, Polish, etc., surnames possessed by Russian or Soviet citizens - in that case the original spelling should be preferred) that don't have translitterations attested in Latin texts outside Wikipedia. -- Alexander Gerascenco 09:57, 6 Maii 2009 (UTC)

De nominibus compositis Anglicis[fontem recensere]

Haec removi quia, nisi fallor, non iam disputavimus et usque adhuc hanc regulam non accepimus:

  1. Ex
    -ton (Campton)
    -ham (Birmingham)
    -bury/burg/bourg/borough (Ludwigsburg)

    fiat
    -dunum (Campdunum)
    -haemum (Birminghaemum)
    -burgum (Ludovicoburgum). Brennus Regan: De toponymis Anglicis Latine vertendis, quae in "-ton", "-ham" aut "-bury" cadunt. VOX LATINA. Commentarii periodici favore et subsidio Studiorum Universitatis Saravicae comparati. ISSN 0172-5300. Saraviponti. 45/2009/176.

Id fortasse faciunt alii quidam; sed in Anglia "-ton" saepius Latine -tonia sonat (cf. Tantonia), "-ham" -hamia (cf. Birminghamia), "-bury" -buria (cf. Malmesburia). Ac si fontem non reperimus, non vertimus. An recte dixi? Andrew Dalby (disputatio) 15:50, 23 Aprilis 2011 (UTC)

More Latin suffixes[fontem recensere]

There is a nice list of placename suffix conversions in Botanical Latin (Stearn, 4th ed., 2008), but lest I be accused of copyright infringement, I shall not reproduce it here. Nevertheless, he says that the list is based on one given by Saalfeld (1885). I tracked that reference down to "Deutsch-lateinisches handbüchlein der eigennamen aus der alten mittleren und neuen geographie zunächst für den schulgebrach zusammengestellt von Dr. G.A. Saalfeld." whose text sadly cannot be found in Google Books. But luckily for us, a copy exists in the Internet Archive. I have painstakingly copied that list here, in the hopes it is of some use.

The rest of the book is something like 600 pages of names with Latinized names. The list is, however, apparently highly Euro-centric. It does not contain typical American names, for example. --Robert.Baruch 02:24, 2 Maii 2011 (UTC)

-ach, -ack → -acum, -achium
-ad → -adum
-agne → -ania
-ailles → -alia
-ain, -aine → -ania, -anium
-al → -alium, -alia
-am → -amum
-an (ain) → -anum or -anium (for place names), -ania (for country names)
-ant, -anz → -antia
-ar → -aria
-at → -atum
-atsch, -atz → -atium
-au → -avia, -oa, -augia (for place names), -ovia (for country names)
-aux → -atium
-berg → -berga
-borg → -burgum
-born → -borna
-burg → -burgum
-c → -iacum
-cester, -chester → -cestria
-dorf → -dorpium
-e → -a
-e or -é → -as, -aeum
-ec, -eck → -ecca, -eca
-eglia → -elia
-ei, -ey → -eia (place), -ia (country)
-eil, -eille → -elia
-ein → -inum
-em → -emum
-en → -a, -ena, -um, -ium, also -ia for country names
-ence, -enz → -entia
-ent → -entium
-er → -era
-euil → -olium
-feld → -felda, -feldia
-fels → -felsa
-ford, -fort → -fordia
-furt → -furtum
-gard, -gart, -garden, -gorod, -gord → -gardia
-gen → -ga (place), -gia (country)
-hafen, -haven → -havia
-hausen, -husen, -hus → -husa, -husium
-heim → -hemium
-hofen → -hofa, -hovia, -hovium
-holm → -holmia
-holz → -holtia
-horst → -horstium
-hut → -hutum
-ich, -ick, -ig, -ik → -icum
-ie → -ia
-igno → -inium
-im → -imum, -ima
-in → -inum, -inium
-itsch, -itz → -icium, -itium, -icia
-kirchen → -querca, -kerka
-land → -landia
-leben → -leba, -lebia
-mold → -moldia
-mond, -mont → -montium
-mouth → -muthum
-münde → -munda
-n → -iacum
-o (in Latin names) → -um
-oglio → -oleum
-ogne → -onia
-oise → -osia
-oping, -öping → -opia
-oux → -osum
-ow → -ovia, -ovium
-pel, -poli → -polis
-que → -ca
-r → -iacum
-sand → -sanda
-stadt, -städt → -stadium
-stein → -steinium, -stenium
-stock → -stochium
-t → -iacum
-thal → -thalia, -dalia
-us → -usium
-wegen → -vegia
-werth → -verda
-y → -ium, -iacum (especially after c, n, r, t)
-zell → -cella
-zza → -ssa

Compositions are often formed by the genitive and dative, and sometimes for reasons of euphony a vowel is inserted: e.g. Marienberg → Mariaeburga, Augustusberg → Augustoburgum.

The adjective derived from these names usually gets the ending -ensis, -ense. But the following cases are considered:

1. Names with -ena only get -sis appended, as Gubena, Gubensis.

2. With -a, the adjective sometimes gets -ensis, and sometimes -anus: Iena, Ienensis, but Gotha, Gothanus.

3. With -burgum, the adjectives sometimes gets -ensis, and sometimes -icus: Marburgum, Marburgensis, but Hamburgum, Hamburgicus.

4. With -husa, use -anus: Nordhusa, Nordhusanus.

5: With -husium, -dorpium, and -stadium, use -inus.

"Si nomen Latinum nondum exstat usitatum, non vertatur, eis tamen exceptis quae in sermone suo vulgari constant e verbis cottidianis atque intelligibilibus."[fontem recensere]

This rule has been called into question. Discuss! --Robert.Baruch 15:52, 2 Iunii 2011 (UTC)

Lutetia (Texas) and the like[fontem recensere]

I used this as an example (see the above link) and I copy the relevant comments here, adding a response:

... I think when we make a name such as ... Lutetia (Texas) (as we might do) we are misunderstanding the function of place names and the nature of translation. It is true that other writers of modern Latin sometimes make such names; in which case, our rules allow us to follow them (though we also mention the everyday name). But we shouldn't be the inventors ... Andrew Dalby (disputatio) 09:14, 1 Iunii 2011 (UTC)
If (as historians assume) Paris, Texas, was named after Paris, France (i.e., Lutetia), what's wrong with Lutetia (Texia)? IacobusAmor 12:09, 1 Iunii 2011 (UTC)
It doesn't work that way. Why fr:London (Ontario) and it:Paris (Texas), not *Londres (Ontario) and *Parigi (Texas)? Why Noviodunum = fr:Noyon, but also Noviodunum = fr:Jublains? Because Londres is the French name for a particular city in England, not for everything that happens to be called "London" in some other language. Because Parigi is the Italian name for a particular city in France, not for everything that happens to be called "Paris" in some other language. Because the Latin name Noviodunum belongs to several cities, and the names in other languages of each city are different. If we have a Noviodunum for which we cannot be sure of a French name (and indeed we have), will we call it Noyon in French because one of the others is so called? Of course not. Noyon is the proper name in French of a certain city, not of all the cities whose proper names were "Noviodunum" in some other language.
Now it's possible, if (as you say historians assume) the place in Texas is named after the big Paris, that some local Latinist has already called it by one of the three usual Latin names of the big Paris (classical and modern "Lutetia", late antique "Parisii", or medieval "Parisius"). If so, and we find that source reliable, we can follow. Until someone has done that, none of those three is a name for the small Paris. Its name is Paris. Our job is to describe things as they are.
That's my view, anyway :) Andrew Dalby (disputatio) 18:41, 2 Iunii 2011 (UTC)

De formula Convertimus[fontem recensere]

The section Nomina geographica, number 4 implies that "College Park" should be left untranslated, but it currently is: Saeptum Collegii (Terra Mariae) (and see the discussion there). I don't really have an opinion here, but it would be great to get more clarity in general on when it's OK use Formula:Convertimus.

"Fanum Sancti Ioannis" vs. "Sanctus Ioannes"[fontem recensere]

Oportetne regula de locis ex sanctis nominatis in parte Vicipaedia:De nominibus propriis#Nomina geographica mutari? Vide, exempli gratia, quod usor Theobaldus contra dixit. Lesgles (disputatio) 17:25, 18 Augusti 2014 (UTC)

Huic regulae (si regula sit) nunquam obtempero. Possumus in textu dicere "ad fanum/ad urbem S. Germani in Laya". Haud necesse est "Fanum" vel "Urbs" in titulo inserere nisi fontes adsunt. Andrew Dalby (disputatio) 17:56, 18 Augusti 2014 (UTC)
Propter hanc rationem, nemine contradicente, textum paginae mense Decembri 2014 mutavi, verbis "... si fons Latinus non reperiri potest ..." additis, nominibus alternativis "Fanum/Urbs" suggestis. Non est cur verbum "fanum" praeferre debeamus; non est cur talem consuetudinem contra fontes fideles tenere liceat. An recte feci? Andrew Dalby (disputatio) 12:31, 25 Martii 2015 (UTC)