My auxlang

[Qwynegold]

I had a thread about this in the old ZBB. It's basically an international auxiliary language, but there are some ideas I want to incorporate in it that maybe aren't that good for an IAL, so maybe one could consider it to be a bit of an engelang. In this project I have like three main goals:

1. The conlang should be based on a large variety of natlangs.
2. It should be as easy as possible for as many people as possible to learn.
3. There should be as little ambiguity as possible in the language.

Naturally these goals clash against one another at times, especially goals 2 and 3. In these cases I tend to choose ease of learning over disambiguity.

Right now I have a decision to make that I just can't make up my mind about, so I thought about asking here if anyone has any opinions. What should I do about compounds? There are three options for how to handle compounds orthographically:

1. Spell compounds together, e.g. teapot.
2. Put a dash in compounds, e.g. dependent-marking.
3. Spell it as separate words, e.g. water buffalo.

Option 2 would be best for ambiguity's sake. Some compounds could possibly consist of two very short words, so a dash helps you see that it's a compound and not a single word. The dash also helps you identifying the morpheme boundary.
However, option 3 might be best for ease of writing. At least in Sweden and Finland, one of the absolutely most common spelling mistakes seems to be writing compounds as separate words. But how severe would the ambiguity this creates be? In Swedish there are a lot of compounds that change their meaning if you separate them, e.g:

femhundralappar = 500 kr bills
fem hundralappar = five 100 kr bills
femhundra lappar = 500 notes

vår kassapersonal = our cashiers
vår kassa personal = our crappy staff

djupfryst kycklinglever = deep-frozen chicken liver
djupfryst kyckling lever = deep-frozen chicken is alive

vårlök = spring onion
vår lök = our onion

grönsak = vegetable
grön sak = green thing

pumpagubbe = pumpkin man (perhaps a jack-o'-lantern)
pumpa gubbe = to f**k an old man

But maybe there's just something special about Swedish that causes this to be so common. Still, I think there are some types of compounds that may cause problems. For example ADJ+N compounds; how do you differentiate this from just a regular adjective modifying a regular noun? There may also be cases where an adjective is modifying a compound and you don't know if it's modifying only the first word or the whole thing. Compounds involving a verb and a noun could also be potentially confusing, because you don't know if the noun is a verb argument.

One could have a rule where the non-final words in a compound are modified somehow to show that they're part of a compound, but that would be just as hard to learn to do as it is to learn to write compounds with a dash inbetween.

So I'm just asking which method you think is the best, or if you have any other input.

[mèþru]

So a more diverse Lojban?

[Qwynegold]

Nah, I wouldn't say it's anywhere as logical as Lojban.

I have been thinking about how to handle place names, and have come up with some kind of idea. I haven't really thought this idea all the way through, but I thought I'd share it here and see if anyone can find flaws in it. All place names get assimilated from their local language into this conlang, e.g. Budapest > Putapeste, Qingdao > Cigtau [tʃiŋtau]. But words in the toponym that relate it to some other place get translated. I have no lexicon yet, so in the following examples underlined words should get translated into the conlang:

West Virginia > West-Weciniye, because the word west relates it to Virginia, which it is west of
New York > New-Yok, because the word new relates it to York, England
Great Britain > Big-Periten, because the word great relates it to "Little Britain" (Brittany in France)
Stoke-on-Trent > Estouk-on-Terente, because the word on relates the city to the river Trent

Place names should also carry suffixes that identify what type of place it is. Cities should have a city suffix, countries a country suffix, rivers a river suffix, etc. If the toponym already contains a morpheme with this meaning, it gets dropped, e.g. Germany (Deutschland) > Toicestan (not Toiclantestan).

[Salmoneus]

Nah, I wouldn't say it's anywhere as logical as Lojban.

I have been thinking about how to handle place names, and have come up with some kind of idea. I haven't really thought this idea all the way through, but I thought I'd share it here and see if anyone can find flaws in it. All place names get assimilated from their local language into this conlang, e.g. Budapest > Putapeste, Qingdao > Cigtau [tʃiŋtau]. But words in the toponym that relate it to some other place get translated. I have no lexicon yet, so in the following examples underlined words should get translated into the conlang:

West Virginia > West-Weciniye, because the word west relates it to Virginia, which it is west of
New York > New-Yok, because the word new relates it to York, England
Great Britain > Big-Periten, because the word great relates it to "Little Britain" (Brittany in France)
Stoke-on-Trent > Estouk-on-Terente, because the word on relates the city to the river Trent

How do you find 'relate', given that so many placenames are relational in meaning?

What names does your language us for:
Stow-on-the-Wold
Stow cum Quy
Stony Stratford
Ashby-de-la-Zouche
Upperthong
Little Wenlock
Much Wenlock
Westmorland
Kirkby Lonsdale
Middleton
Upton Snodsbury
Nomansland
Nether Wallop
Westward Ho!
West Mudford, Up Mudford, and Mudford Sock
Little Marsh and Marsh Gibbon
Hole of Horcum
Pratt's Bottom
King's Lynn
Royal Tunbridge Wells
Knowle St. Giles
Cricket Malherbie
Camden Town
Kentish Town
Ruyton-XI-Towns
Brafield-on-the-Green
Bridge of Weir
Mull of Oa, and Kyles of Bute
Faringdon Without
Pontarddulais
etc?

Germany (Deutschland) > Toicestan

Ironic. I'd almost suspect you included that as a joke...

[zompist]

That's a pretty quirky criterion... would you translate New York but not Newfoundland, since it doesn't relate to Old Foundland? You're going to have Chinese provinces named East Guang and West Guang, while others remain opaque. Do you translate Tokyo as East Capital, though there's no West Capital? (Or if you consider Kyoto as the west capital, should Tokyo be East Kyoto?) Should Jiangxi be West River, or West Jiang? The same character is used in Jiangsu, which however is named from two cities, Jiangning and Suzhou.

This isn't to say your idea is outlandish; European languages used to do this, though not consistently... e.g French has New York but Nouveau-Mexique.

Also, if you're going to distort words to the extent of Periten and Weciniye, that's... a pretty Volapükian approach to auxlangs. The modern tendency in English has been to use native transliterations rather than the old adaptations, and I'm not sure why an auxlang would want to use the old method.

[Qwynegold]

How do you find 'relate', given that so many placenames are relational in meaning?

What names does your language us for: [...]

Thanks, this gave me lot to think about. I didn't attempt to translate the names that I couldn't find etymologies for (tbh I didn't look very hard).

Stow-on-the-Wold > Estou-on-(the-)Wolte (the conlang might not have articles)
Stow cum Quy > Estou-with-Kuwai
Upperthong > Upper-Fog
Little Wenlock > Little-Wenlok
Much Wenlock > Big-Wenlok
Nomansland > Noumanselante, because there's no Everymansland or anything. The -stan suffix would not be used for land if it's not a country or region.
Nether Wallop > Down-Walop, I'm not exactly sure how nether would be translated
Westward Ho! > Westward-Hey
Marsh Gibbon > Maskipon
Little Marsh, this is a good question. Should it be Small-Mas, even though that doesn't quite go together with Maskipon? Or should it just be Litelmas?
Hole of Horcum > Houlowehokum, unless there is a place called just Horcum
Pratt's Bottom > Peratsepotom, because Pratt is a surname, so it doesn't relate to another place
King's Lynn > Kigselin, the king is a person so it doesn't matter. The etymology of Lynn is apparently unclear, so I'll have a rule that such things will not be translated.
Royal Tunbridge Wells > Royal-Tamberic-Spring, the word royal here is a specific title given to the place, so it might get translated. Wells should be translated, because it distinguishes this place from the town of Tonbridge.
Camden Town > Kanten-town, all towns, cities etc. will have a suffix identifying what it is. (Here I have not added them to the other places names.) So the suffix "eats up" the word town.
Kentish Town > Kentis-Town; unrelated to Kent. The etymology is also unclear.
Ruyton-XI-Towns, this one is hard. If just Ruyton is a separate place, then some part of Ruyton-XI-Towns needs to be translated. The suffix I use for e.g. town might be different than the independent word for town itself, in which case it probably should not pluralize. I really don't know how to deal with this one.
Bridge of Weir > Pericowiye, (don't know the original pronunciation). Weir is apparently not a place, so the word of doesn't get translated either.
Mull of Oa > Ou-GEN-Promontory; I don't know how exactly I will translated mull as this language might not make that many distinctions

The problem here is that I would need to check the etymology of each place name I translate. However, I'm not planning on adding anything but the capital and maybe one or two other major cities of each country in the dictionary. Or maybe I should make things simpler for me and translate only the most obvious things, like words like new, cardinal directions, upper, lower, etc.

Germany (Deutschland) > Toicestan

Ironic. I'd almost suspect you included that as a joke...

I like the -stan suffix so I want to include it. All countries would be stans.

[Qwynegold]

That's a pretty quirky criterion... would you translate New York but not Newfoundland, since it doesn't relate to Old Foundland? You're going to have Chinese provinces named East Guang and West Guang, while others remain opaque. Do you translate Tokyo as East Capital, though there's no West Capital? (Or if you consider Kyoto as the west capital, should Tokyo be East Kyoto?) Should Jiangxi be West River, or West Jiang? The same character is used in Jiangsu, which however is named from two cities, Jiangning and Suzhou.

This isn't to say your idea is outlandish; European languages used to do this, though not consistently... e.g French has New York but Nouveau-Mexique.

Also, if you're going to distort words to the extent of Periten and Weciniye, that's... a pretty Volapükian approach to auxlangs. The modern tendency in English has been to use native transliterations rather than the old adaptations, and I'm not sure why an auxlang would want to use the old method.

Yeah, Newfoundland would just be Niyufauntelante with an island suffix. Chinese place names are going to be problematic. For example: Beijing > North-Capital, North-Cig or Peicig? I should probably not translate the entire name, because then there's like nothing left. But then it's like "why not translate Jing if you're translating Bei?" So Peicig seems like the least bad option. Tokyo would not be translated. Most place names are translatable, but one must not translate too much. So I have to set some kind of reasonable limit.

My reason for assimilating place names rather than borrowing them unchanged is that otherwise different speakers may pronounce them in such various ways that they can't understand each other, and because the pronunciation is not clear from spelling. If I assimilate the names, then everyone (of the speakers) will pronounce them in a way that can be understood by everyone, everyone will be able to read them, and everyone will be able to spell them.

[missals]

For compounds - you could use some kind of interfix or special particle. Something genitive-like. It could just be, say, o. Or lo. I guess the form would depend on the typical shape of root words. If the roots are more CVC-like, then it might look like kat + gom = katogom. Or kat-o-gom. If the roots are more CVCV, then maybe something like lata + misa = latalomisa or lata-lo-misa. I guess it's a bit cumbersome, but it would at least yell out "This is a compound!!"

Not all languages have compound words, anyways - Spanish doesn't have much in the way of noun-noun compounds, and many languages don't allow them at all. And compounds are a big source of ambiguity.

Though I guess that route is difficult, since an auxlang will have to calque or otherwise provide translations for a lot of compound and compound-like scientific and technical vocabulary common to many major world languages.

Anyways, this is interesting. What you've suggested so far looks a lot like how I'd design an auxlang, especially with the phonological incorporation of placenames. How on earth would one pronounce things like Archangelsk and Baghdad if they're just dropped into the text of an auxlang with a much simpler phonological system? Are people to code-switch into their native language every time they say a place name?