Disputatio:Circulus Dupont

Page contents not supported in other languages.
E Vicipaedia

Stating the question[fontem recensere]

Credo rectius "Circulus Dupont" dici. Sed libentius "Dupont Circle" (ut in aliis Vicipaediis), nisi fontem Latinum habemus. Neander 00:57, 18 Aprilis 2011 (UTC)[reply]

I was taking my cue from the discussion about the order of the elements in the names of counties, which has led to Helveticus's creating hundreds of lemmata having the form seen in Clermont Comitatus (not Comitatus Clermont). See the general issue raised in Taberna at "Urbs or -opolis?" What's the basic pattern for these sorts of things? or is there one? IacobusAmor 02:46, 18 Aprilis 2011 (UTC)[reply]
My cue consists in the pattern exemplified by urbs Roma, flumen Tamesis, etc., the basic rule being this:
(1) [NP <proper_name> ]NP [NP [N <noun> ]N [ADJ <adjective_attribute> ]ADJ ]NP
I.e., if a proper name ("the antecedent") is modified by a noun phrase functioning as its apposition, and if the appositional noun phrase consists of a noun modified by an adjective attribute, the noun phrase is placed after the proper name: Roma urbs aeterna, Tamesis flumen magnum, cett.;
(2) [NP [N <noun> ]N ]NP [NP <proper_name> ]NP
i.e., if the apposition has no modifier in the same phrase, it precedes the proper name: urbs Roma, flumen Tamesis.
Because Latin is a free-word-order language, these rules have textual exceptions; yet, these are valid as syntactic generalisations and define how words are placed by default. Violations of this syntactic rule tend to be pragmatically motivated, although the motivation may not always be easily detected (and, of course, everybody has his euphonic preferences every now and then). Default word-order is generally favoured in titles. Therefore, the question is whether Clermont County and Dupont Circle are comparable to River Thames. If they are, as I first thought they are, Comitatus Clermont and Circulus Dupont look better. But if they are not, and I'm also reckoning with this possibility, I have no definite position on this matter. Neander 22:09, 20 Aprilis 2011 (UTC)[reply]
FYI: In the United States, Thames River is the default pattern. IacobusAmor 16:09, 21 Aprilis 2011 (UTC)[reply]
As you may have seen, Iacobus took this discussion on to the Taberna, where I made two points (that may well not get us much further!) One is that those names are not strictly comparable because if they stand alone they would not have the correct meaning (Tamesis and Roma would; Dupont and Clermont would not). The apposition is essential to establish what is being named. Romans rarely built names in this fashion: generally they used adjectives to achieve a comparable naming style (via Appia, Forum Iulium, Legio XX Valeria Victrix). However, my conclusion from this was that it is even more necessary, for the sake of the reader of normal text, that the general noun (e.g. circulus, comitatus) come first (.. English subjunctive being used there ...) so that the said reader is not disoriented by being unable to interpret the word Dupont or Clermont. Well, that makes the word order coincide with your case (2) anyway.
The other point I made was a purely practical one, which I preface by saying I don't (yet) agree that "Default word-order is generally favoured in titles". I'm not sure where that assertion comes from (but I would love to know if someone has published a grammatical study of Latin titles!) I think, for Wikipedia titles (not text), it is more helpful if the distinguishing name comes first, because it makes it quicker to find them. Hence
  1. in writing about English counties, I have moved the titles where necessary to get the distinctive name first (in this case normally an adjective, Somersetensis comitatus, but in the body of the text I have tended to write these names with the word comitatus first, because I agree with you that it's normal, and also (as I say above) so that the reader is not disoriented.
  2. in discussing the American counties (of which, as we know, there are vastly more) I have urged that this same word order be adopted in titles, specifically because, if we do this, a searcher will not have to type out the word comitatus in full before getting anywhere. I think that's a legitimate use of the free word order that Latin allows us!
I admit that the easier-searching issue could also be dealt with by redirects. And no doubt Robert's bot could make all those redirects in a trice. Andrew Dalby (disputatio) 09:10, 21 Aprilis 2011 (UTC)[reply]
Neander's argument is helpful, and so maybe we should change Dupont Circulus to Circulus Dupont, not least because the latter form was my instinctive one, before I tried to accommodate the reasoning exhibited in disputatione de comitatibus. If the argument has general appeal, certain long-established (in Vicipaedia) titles may need to be changed. These include Argaeus mons (Luna), Arcansianum Flumen, Augia insula, Coloratum Flumen, Dunestorum castrum, Pikes Mons, Platte Flumen, Portunia insula, Quercuum Paeninsula, Sancti Laurentii fluvius, Sicagoensis Fluvius, and Turris Insula. Note, however, that the order seen in Crocodilorum Insula (Taivania) is attested in an impeccable source. ¶ Yet then again, see here. IacobusAmor 14:22, 21 Aprilis 2011 (UTC)[reply]
Look at how many Insulas come last in that book... Sanctae Crucis Insula, Sacti Jacobi Insula, Vincentii Insula, Trinitatis Insula. But then we get Aestuarium Reginae Charlottae, so now there's a big "Huh?" above my head. And Chapter XVII (Geographical Names) of that book is pretty damned awesome. --Robert.Baruch 14:40, 21 Aprilis 2011 (UTC)[reply]
As regards extraterrestrial geographical features, Latin names in the word order Specific name + descriptor are official; thus names such as Argaeus mons (Luna) and Ishtar terra had better stay the way they are, I think. Andrew Dalby (disputatio) 14:35, 21 Aprilis 2011 (UTC)[reply]
If it's official, then of course it should stand! IacobusAmor 14:47, 21 Aprilis 2011 (UTC)[reply]
Actually Argaeus mons (Luna) was my mistake (see below) but Ishtar terra is OK. Andrew Dalby (disputatio) 14:04, 22 Aprilis 2011 (UTC)[reply]
As regards terrestrial geography, what do other encyclopedias do? I suspect they practically always begin with the specific name. Iohannes Iacobus Hofmannus, Lexicon universale (1698) ~ would offer many examples. At present I am suspicious of Neander's crucial claim "Default word-order is generally favoured in titles" :) Andrew Dalby (disputatio) 14:35, 21 Aprilis 2011 (UTC)[reply]
Lexicon Universale has Insula Barbara, Insula Bocardi, Insula Franciae, Insula Jordani, Insula Sacra. So somewhere along the way someone moved Insula. --Robert.Baruch 14:56, 21 Aprilis 2011 (UTC)[reply]
Well, but in all those examples, the second member appears to be an adjective or a genitive noun: in no instance is it unequivocally an appositive. If the item here in question were Circulus Dupontianus, how many of us would be wanting to change the lemma to Dupontianus Circulus ? IacobusAmor 15:09, 21 Aprilis 2011 (UTC)[reply]
Sorry, I don't understand [Robert's] last sentence. Maybe we should be writing Latin ... :) Do those appear to be the majority of islands, or special cases? I suspect the latter. "Insula Franciae", of course, is not an island. Andrew Dalby (disputatio) 15:05, 21 Aprilis 2011 (UTC)[reply]
I'm looking at Isidorus Hispalensis, Etymologiae -- the nearest thing to an encyclopedia in more-or-less classical Latin. There's a mass of geographical names in book XIII. He uses the freedom of Latin word order, when he's giving a list, to put the descriptor first and then the whole list of specific names ("sinus dicuntur maiores recessus maris, ut in mari Magno Ionius, in Oceano Caspius, Indicus, Persicus, Arabicus"); but where he's not giving a list he seems nearly always to adopt the form Specific name + descriptor ("Myrteum mare dictum a Myrtili lapsu ... Aegyptius autem pelagus Asiae datur ... sciendum Ionium sinum esse inmensum"). I don't think we've yet measured the complexity of the "rules" that underlie the "freedom" of word order in this field. Andrew Dalby (disputatio) 15:21, 21 Aprilis 2011 (UTC)[reply]
Is Aestuarium Reginae Charlottae a special case? Yes, because it's Queen Charlotte's Sound, not Queen Charlotte Sound. So perhaps our rule should be: placenames with genitive and adjectives are ordered as Type + Gen/Adj. (e.g. White Island -> Insula Alba, Cat's Island -> Insula Felis, Queen Charlotte's Sound -> Aestuarium Reginae Charlottae) while others are ordred as Name + Type (e.g. White Island -> White Insula (where White is the name of a person), Zoloft Island -> Zoloft Insula, Queen Charlotte Sound -> Regina Charlotta Aestuarium). Excepting where a name is already attested? --Robert.Baruch 15:26, 21 Aprilis 2011 (UTC)[reply]
A problem with this analysis is that the names of geographical features have changed over the centuries. What's nowadays Hudson Bay used to be—and not so long ago!—Hudson's Bay. (Similarly in another field, Down's Syndrome has become Down syndrome.). So perhaps it doesn't matter, and Circulus Dupont doesn't involve an appositive, as Dupont here should be understood to be an indeclinable adjective, and therefore Circulus Dupont is the default word-order after all. IacobusAmor 16:09, 21 Aprilis 2011 (UTC)[reply]
Hudson Bay -> Sinus Hudsonius AND Hudsonius Sinus, in contemporaneous references (but one Sinus Hudsoniensis). But always EX Sinu Hudsonio (sometimes ex Sinu Hudsonia), and never ex Hudsonio Sinu. --Robert.Baruch 16:47, 21 Aprilis 2011 (UTC)[reply]
Agree. Circulus Dupont. --Robert.Baruch 17:01, 21 Aprilis 2011 (UTC)[reply]
The "excepting where a name is already attested" rule would allow for all the island names in Botanical Latin which violate the general rule I posited, so Sanctae Crucis Insula, not Insula Sanctae Crucis.
Oh, and what fun: there are many "Insula Sanctae Crucis" when after ex or in or ab. Searching for both word orders in Google is an interesting exercise. --Robert.Baruch 16:01, 21 Aprilis 2011 (UTC)[reply]
Vide etiam ordinem verborum in locutione "Collegium Sanctae Crucis" et in nominibus permultarum universitatum! IacobusAmor 16:37, 21 Aprilis 2011 (UTC)[reply]
I'm afraid that I can't buy the arguments for searching. I don't think the title should be affected by the limitations of Wikipedia's search mechanism. But I am certainly being swayed towards the Comitatus X side, excepting already established word order. --Robert.Baruch 14:36, 21 Aprilis 2011 (UTC)[reply]
This is only one consideration among several, but I disagree with you strongly that it should be discounted. Yes, to us it happens to be a question of "Wikipedia's search mechanism", because Vicipaedia is what we are writing; well, in writing an encyclopedia it's important to be right, but being right is useless if people don't find it or don't read it. So, in principle, when writing an encyclopedia, we have to think about how people will get to our pages. Now, for a long time it was true that no computerised information source was as quick to search as an alphabetical reference book. What finally changed that (I think) was en:word completion. It's such a crucial feature of data entry and searching that it makes sense to work with it (just as, with printed reference books and in indexes, it makes sense to put the specific word ahead of the general word because finding entries is quicker that way). Andrew Dalby (disputatio) 11:44, 22 Aprilis 2011 (UTC)[reply]
Would this be an appropriate use for redirects? So that the article is Circulus Dupont, but there is a redirect at Dupont Circulus to game the search mechanism? --Robert.Baruch 14:01, 22 Aprilis 2011 (UTC)[reply]
Oh, yes, that would work for me. Andrew Dalby (disputatio) 14:07, 22 Aprilis 2011 (UTC)[reply]

Andrew is quite right in calling into question my statement "Default word-order is generally favoured in titles". This is what I thought off the cuff without realising that not even titles are exempt from pragmatics: while I'm still clinging to the rules (or syntactic generalisations) stated above, I'm ready to admit that Andrew's practical principle of having the distinguishing name first in titles implies a legitimate use of the flexible word-order of Latin for a pragmatic purpose ("quicker to find"). § As to Argaeus mons, I'm not quite sure what makes it official; but certainly it is in line with the pragmatic principle of easy detectability. Note, however, that Plinius (nat. 6.8) writes sub monte Argaeo (which I take to be the default order), whereas Ammianus (20.9.1) has urbem ... sub Argaei montis pedibus sitam, which is due to the well-known transformation (in the propagation of which Iacobus has been very helpful) that moves a complex NP (N+Adj, e.g. mons Argaeus) modifier in front of its head-word (here: pedibus), at one fell swoop reversing the NP constituent order; i.e., sub pedibus [montis Argaei] (expectable in school texts) => sub [Argaei montis] pedibus. Here, the default order is overridden by another syntactic rule. Neander 22:50, 21 Aprilis 2011 (UTC)[reply]

Well, touché, I didn't cite any source for my assertion about "official" extraterrestrial names. Who knows, an extraterrestrial authority might turn up to dispute the legitimacy of any terrestrial body's claims here: but until that happens, I do claim the authority of the IAU. Databases of names that they have approved are normally in Latin and normally use this [specific name + descriptor] word order. For lists of many of them see the subcategories of en:Category:Surface feature nomenclature of solar system bodies (I note that that category page includes two very handy external links).
[Added later:] The big exception, however, is the Moon (see e.g. en:Category:Maria on the Moon -- nice ambiguity there) and I haven't found any explanation of the inconsistency. Many features on the Moon were named before the IAU got involved. It may be that the word order applied by the IAU on all other planets and satellites was adopted for the very reason I have urged above -- it makes for quicker finding in gazetteers: if there was another reason, I can't guess what it was. (Because the Moon is an exception, my earlier move of Argaeus (Luna) to Argaeus mons (Luna) was wrong, I now realise: it should have been Mons Argaeus (Luna). Mea culpa.)
I didn't mean my assertion to apply retrospectively to the terrestrial Argaeus mons, which must be subject to any general rule that we evolve. And I noticed yesterday evening, when just setting out to the first barbecue of what may be a long summer, that Pliny's usage, in his lists of geographical names, differs strongly from that of Isidore. Whether this is because Pliny is listing them in a geographical order (and therefore wants to emphasise that the next thing in the list is a flumen and not a portus) while Isidore is classifying them (and so has no need for that emphasis), or whether the reason is quite different, I haven't worked out. Andrew Dalby (disputatio) 08:47, 22 Aprilis 2011 (UTC)[reply]

Attempting a summary[fontem recensere]

While I don't feel too competent to take a firm stand on US place names, let me say by way of summary that there are three basic construction types involved here:

Type Unmarked order Examples Marked order Examples
(1) N + ProperN urbs Roma; urbs Veii; mons Argaeus ProperN + N Roma urbs; Veii urbs; Argaeus mons
(2) N + ProperNAdj via Appia; comitatus Somersetensis ProperNAdj + N Appia via; Somersetensis comitatus
(3) N + PropNGenDefinitivus urbs Patavii; urbs Pisauri PropNGenDefinitivus + N Patavii urbs; Pisauri urbs

Given the fact that Latin is a free-word-order language — which means that all of those ordering types listed can be expected to occur in running texts — I feel the necessity to distinguish, for descriptive purposes, between unmarked (default, neutral, natural, pure syntactic) order and marked (contrastive, pragmatic) order. This is a typologically-motivated theoretical distinction "emanating" from the fact that Latin is a SOV language. That this distinction is "psychologically real" as well is evidenced by the fact that, whereas via Appia is generally regarded as the normal construction, Appia via is in fact textually more frequent in Cicero. § Type (3) is scarcely met with in Caesar or Cicero (though the latter has Averni lacus); but still it may be considered a post-classical feature. § As to Dupont Circle, I now see various possibilities: Circulus Dupont (the unmarked case of type (1)), Dupont being an indeclinable noun); Dupont circulus (the marked case of type (1)); Circulus Dupontianus, as suggested by Iacobus above [notice that Circulus Dupont, Dupont circulus, circulus Dupontianus are equally unattested, so I don't see why Dupontianus were less admissible], and even Dupontianus circulus (the marked case of type (2)). § These musings of mine are supposed to offer a bit of basic research and groundwork for further implementations, whether they'll be done in Robert's or in Andrew's way. Neander 01:59, 23 Aprilis 2011 (UTC)[reply]

I find that extremely helpful. (Very interesting about the frequency of Appia via: I wouldn't have expected that.) You convince me about the unmarked order, and I must agree that it is preferable from the Latinitas point of view (even if the necessary pragmatics of titles has not yet been written) to have our titles in this natural, unmarked order if there is no exceptional reason for doing it the other way (as e.g. with official topographical names on Venus and Mars!) Robert and I already agreed above that redirects from the marked order (e.g. Somersetensis comitatus > comitatus Somersetensis) would solve the practical problem I've raised. Andrew Dalby (disputatio) 08:37, 23 Aprilis 2011 (UTC)[reply]
That's a most helpful summary! and since we all seem to be in agreement, Ima change the title to the unmarked (default, neutral, natural, pure syntactic) order, which would have been my choice in the first place, had the discussion of the counties not loomed so large. IacobusAmor 12:03, 23 Aprilis 2011 (UTC)[reply]
Excellent summary! So now I think that when I generate all those US county pages, I should stick with the unmarked title (e.g. Comitatus Cecil (Terra Mariae)) and add redirects to help the searcher (e.g. Cecil Comitatus (Terra Mariae)). I've just been waiting until May, when the Census Bureau releases the more detailed demographics. Then I can scrape and beef up the articles a bit. --Robert.Baruch 00:44, 25 Aprilis 2011 (UTC)[reply]

Circulus, for the record[fontem recensere]

That's the obvious cognate of English circle, but is it the best? Conceivably pertinent is an English synonym used only (?) in New England: rotary, which might imply Latin rotariu(m,s) ; but we might want to save that word—or maybe better rotariolu(m,s)—for Wheelie. Another pertinent term is Roundabout, used mainly (?) in England and by engineers, but roundabouts aren't exactly the same thing as traffic circles, so these concepts will want separate articles and therefore separate names (something from rotund- for 'roundabout'?). A worthy encyclopedia will eventually have hundreds of articles on named examples of these and related entities. Even now, in the infancy of the project, en:Category:Squares, plazas, and circles in Washington, D.C. shows that Wikipedia already has twenty-nine pages on just the ones in Washington, D.C., of which twenty are circles. IacobusAmor 12:03, 23 Aprilis 2011 (UTC)[reply]

Oh, the trials and tribulations of being at the forefront of modern Latin. ;) I find it exciting, but then, I'm strange. --Robert.Baruch 00:45, 25 Aprilis 2011 (UTC)[reply]

Traffic circle[fontem recensere]

Is anything better than orbis commeatús ? IacobusAmor 12:03, 23 Aprilis 2011 (UTC)[reply]

Circuitus? --Robert.Baruch 01:28, 25 Aprilis 2011 (UTC)[reply]

Xystus ? + Default word-order same as with circulus above? IacobusAmor 13:21, 23 Aprilis 2011 (UTC)[reply]

Maybe this is one of those places where we say that Latin doesn't have different words for Avenue and Street? Just Via? --Robert.Baruch 01:30, 25 Aprilis 2011 (UTC)[reply]
Hmm, but (they tell me) US cities make such heavy and consistent use of these two terms that it would be good to allow for both if we could. I was told at school that Latin cardo (maximus) would be the north-south main road in a planned city, and decumanus (...) would be the east-west main road. Could one of those words serve for US "avenue"? Andrew Dalby (disputatio) 08:44, 25 Aprilis 2011 (UTC)[reply]

Addresses[fontem recensere]

E.g.: "339 10th Street" = 339 Via 10 (339 Via Decem)? 339 Via 10a (339 Via Decima)? 339 Viá 10á (339 Viá Decimá)? 339 Viae 10ae (339 Viae Decimae)? or what? IacobusAmor 13:21, 23 Aprilis 2011 (UTC)[reply]

Well, if we ignore the sense of a mailing address (in which case you better use the English!), I'd say 339 10th Street is short for "House number 339 on 10th Street" = Domus cui numerus est 339 in Viā 10ā. "House number 339" could also be treated using an ordinal (although I admit 339th house isn't quite the same): Domus 339a in Viā 10ā. So, for short, I think either 339 Viā 10ā or 339a Viā 10ā will work. I prefer the former, though. --Robert.Baruch 01:38, 25 Aprilis 2011 (UTC)[reply]
Harrius Potter (Needham) renders "No. 4 Privet Drive" as aedes Gestationis Ligustrorum numero quattuor signatis. Just another data point, where the street becomes genitive and the number becomes ablative (but still cardinal). --Robert.Baruch 01:42, 25 Aprilis 2011 (UTC)[reply]
The numero is ablative, not the quattuor (if I am parsing correctly). House 4 "by number". That would well suit usage in a lot of US cities (I think) in which it wouldn't mean "the 339th house" but "the house to which we have given the number 339 based on the block it's in and its distance from the crossroads". Or am I wrong?
As to European habits, our house is no. 8 [there's the same Latin ablative in "no."], but it isn't the 8th house. Nos 2 and 4 don't exist, and probably never will. Andrew Dalby (disputatio) 08:52, 25 Aprilis 2011 (UTC)[reply]
My house number is 76, and there isn't even room for 76 houses on my street, let alone five, but that's pretty typical: you take the length of the road from intersection to intersection, divide by 100, and wherever your house is, you get a proportional number. § But anyway, I thought the parse was "the house of the Drive of Privets signified by the number four", so that "four" is in apposition to "number", and thus is ablative (even though quattuor is indeclinable). --Robert.Baruch 13:03, 25 Aprilis 2011 (UTC)[reply]

Quadrants in addresses[fontem recensere]

Vasingtonia and probably many other cities have quadrants (NE, SE, SW, NW), which must be specified in addresses; likewise, streets in many cities have east & west parts, or north & south parts, which similarly must be specified (so "339 Pecan Street" would be wrong for "339 E. Pecan Street"). How best to do that in Latin? IacobusAmor 13:21, 23 Aprilis 2011 (UTC)[reply]

Well, we have:
English Latin (direction) Latin (wind) Latin (wind alternate)
North Septentriones Aquilo Boreas
South Meridies Auster
East Oriens Eurus
West Occidens Zephyrus Favonius

We could use B-A-E-Z, or A-M-O-Z, or S-M-O-Z, or whatever. I kind of like B-M-E-O. So your quadrants would be BE, ME, BO and MO, and your address would be 339 Via Pecan E. We could also have two systems, BMEO and B-M-Or-Oc. --Robert.Baruch 01:21, 25 Aprilis 2011 (UTC)[reply]

So then should Laogai Museum (though that's how the name above the front door reads) be changed to Museum Laogai? And thus all other museums? Neander? IacobusAmor 17:01, 23 Aprilis 2011 (UTC)[reply]

Personally I am all for it. I'm still puzzling over all the X Insula that Linnaeus used, though. --Robert.Baruch 01:44, 25 Aprilis 2011 (UTC)[reply]
I think we already do most museums that way, though there may well be exceptions. We already do languages that way ("lingua Latina" etc.), and nearly all universities ... As to why Isidore of Seville, Linnaeus and the IAU prefer it the other way, we may never know, but Latin word order is free, so they aren't wrong. Andrew Dalby (disputatio) 08:31, 25 Aprilis 2011 (UTC)[reply]
Maybe free, but not random! The "Latin Word Order" book makes a good case well backed up with data. --Robert.Baruch 13:04, 25 Aprilis 2011 (UTC)[reply]
This is just academic :) I'm not disputing that we have made a right decision for us, but I would back Isidore and Linnaeus, even against the "Latin Word Order" book, to have made a decision that was right for them. Even the IAU nearly always gets its Latin right for it. [I think in fact I'm merely agreeing with your "not random".] Andrew Dalby (disputatio) 13:14, 25 Aprilis 2011 (UTC)[reply]

Iterum atque iterum: ordo verborum in lemmatibus[fontem recensere]

And yet after all that about the best order of words in lemmata, we find in VP:TNP:

Quod ad Biologiam, Botanicam, Nomina Dioecesium et Zoologicam attinet, multis nominibus formae adiectivales (saepe in -ensis exientes) accedunt. Quas pone ante verbum quod locum designat.
Exempli gratia
Malhamensis Vicus (pro Malham, vicus)
Tristanenses Insulae (pro Tristan da Cunha, collectio insularum)
Ebebiyinense oppidum (pro Ebebiyín, oppidum)

"Quas pone ante verbum quod locum designat"?! Oh well. IacobusAmor 00:15, 2 Maii 2011 (UTC)[reply]

Sed solum cum nominibus formam adiectivalem habentis. Ergo, Malhamensis Vicus quia est verbum "Malhamensis, -is, -e" et Circulus Dupont, quia non est verbum "*Dupontiensis, -is, -e". Non tamen diligo quia mihi constantem non videtur. Et unde venit hanc legem? --Robert.Baruch 01:03, 2 Maii 2011 (UTC)[reply]

More words on this. An analysis of Livy done by Dr. H. A. Sanders in the journal Transactions and Proceedings of the American Philological Association, v. 32 (Iul 1901): "With names of towns which were not well known, Livy quite regularly appends or less often prefixes the words urbs, oppidum, colonia, vicus, caput, castrum, portus, emporium, etc.; but in the case of Rome he places urbs first, using urbs Roma, urbs Romana, or rarely urbs alone, but never Roma urbs." So I guess that's another point in favor of putting the geographical feature type after the name. --Robert.Baruch 02:36, 2 Maii 2011 (UTC)[reply]

And we also have Fischer, Latin Grammar: Etymology and an Introduction to Syntax (1876), who says: "§108 3. Latin nouns in apposition generally are placed after their governing nouns, while in English they often precede. as: Cleopatra regina, queen Cleopatra; Scamander fluvius, the river Scamander." (he also says that English "city of..." gets converted to apposition) Yet another vote for after. --Robert.Baruch 02:42, 2 Maii 2011 (UTC)[reply]