Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

oui!

17 (60.7%)

aucun!

11 (39.3%)

Total Members Voted: 28

[cryopyre]

I think the language needs circumfixes for item material.

And while I'm late to the thread, I hope you seriously consider the importance of dwarven culture on their language.

[Baffler]

Never too late, my friend! Come back anytime.

[Halfling]

Catten runs past Urist's cat.
Rendered as:

Catten rur ág Urist kun. (lit. Catten run over Urist cat.)
This leaves a lot more ambiguity. Now my question is, should prepositions be explicitly so with their own distinct form, as noted by either a special particle or a special construction? Or would it be better to stay with the English form?

Is the ambiguity because you don't know which is which? There seems to be no relationship noted between Urist and cat there. I sound like a broken record a lot, but "Catten rur ág Urist kun" would be "Catten run(s) over Urist the cat" by the vanilla naming rules.
Catten rur ág (Dodok) kun Urist -> Catten runs over (Dodok,) the cat of Urist

Your question was hard to understand. Are you asking whether prepositions should vary and come with extra particles or not depending on circumstances?

[Malleus Inferni]

In DF Dwarfish the rules you have are exactly:
x y, where x is a proper noun: "x the y" (as in Urist the cheesemaker)
x y, where x and y are common nouns: "x of y" (as in slayer of dragons)  <--- distinct from below with a space
xy, where x and y are nouns: x-y/xy (knife-murdering, dragonslaying)   <--- distinct from above without space
xy, where x is an adjective and y is a noun: "x(adj) y" (angry mechanism)
xzy, where x and z are adjectives and y is a noun: "x(adj) [and] z(adj) y" (angry, awe-inspiring mechanism)

Building a complete language around these particular constraints is a hairy proposition, but it sounds fun.

(Disclaimer: I am a linguist.)

The first two, I think, present the more troublesome ambiguity. If you see the sentence "X Y eats Z," and you don't know what part of speech "x" is supposed to be, you could read "X Y" as either "X the Y" or "X of Y." Or, in technical terms, you can't tell whether it's an appositive construction ("X, who is also a Y, eats Z") or a genitive construction ("X, who comes from/is owned by/does things to Y, eats Z"). Fortunately there's precedent for this kind of ambiguity. For example, Japanese uses the particle の to express both apposition and various genitive relationships, and you just have to figure out which is which from context. So that's not actually a big deal, at least for people who can learn to read "Urist of cheesemaker" as "Urist the cheesemaker."

The next problem we come to is the unmarked genitive case, which people have already mentioned here. That's very unusual cross-linguistically. In fact, if the nominative case (or whatever this language has instead of a nominative) is marked and the genitive is not, that's supposed to violate a universal and make it unlearnable by normal human beings. It's ambiguous to me whether this counts, since the two are just morphologically identical for no apparent reason (ie. Toady didn't feel like doing it another way at the time).

If you want to keep this constraint, and not introduce any noun classes where the default/subject case and the genitive are morphologically distinguishable from each other, I think this would have syntax implications. The head noun (including any incorporated modifiers, like compound prefixes) would have to be the first unbound morpheme in every noun phrase, forever. When constructing a sentence, the subject ought to be ordered relative to the verb in a very consistent way, such as immediately before or after it. I'm leaning towards "before," since noun phrases themselves are already stuck with right-branching. This also means person/number marking on the verb itself is totally redundant (the subject is RIGHT THERE, no matter what) and I'd go ahead and leave it out.

(There's more stuff about subordinate clause marking I can't quite work through right now. Maybe some other folks will have ideas there. I'll just offer that prepositions will work much better than postpositions or suffixes, for any kind of modifier.)

The last thing I want to note right now is the spacing issue. How do you pronounce a space? Written languages are encodings of their spoken counterparts, not languages in their own right. You learn to speak years before you learn to read, and when you do learn it, the written version must conform to the rules of the spoken one. At any rate, if word breaks are significant in writing, they must be significant in speech as well, and marked somehow. My preferred solution here is to stress the first syllable of each word consistently, ideally with a pitch accent. That means those accent marks represent different vowel sounds, not stress shifts. Also, individual words can be emphasized just by saying them louder, but there's still room for a system of particles you attach to a word to put emphasis on them, which would carry over to writing.

Again, these are thoughts specifically on the kind of language the existing in-game constraints would imply, and "throw it away and make a new language" is still a viable alternative.

[Halfling]

In DF Dwarfish the rules you have are exactly:
x y, where x is a proper noun: "x the y" (as in Urist the cheesemaker)
x y, where x and y are common nouns: "x of y" (as in slayer of dragons)  <--- distinct from below with a space
xy, where x and y are nouns: x-y/xy (knife-murdering, dragonslaying)   <--- distinct from above without space
xy, where x is an adjective and y is a noun: "x(adj) y" (angry mechanism)
xzy, where x and z are adjectives and y is a noun: "x(adj) [and] z(adj) y" (angry, awe-inspiring mechanism)

The next problem we come to is the unmarked genitive case, which people have already mentioned here. That's very unusual cross-linguistically. In fact, if the nominative case (or whatever this language has instead of a nominative) is marked and the genitive is not, that's supposed to violate a universal and make it unlearnable by normal human beings. It's ambiguous to me whether this counts, since the two are just morphologically identical for no apparent reason (ie. Toady didn't feel like doing it another way at the time).

Is it really that unusual? What are these universals? That's definitely an interesting concept and one anyone designing a language should be aware of if it's based on science.
I'll admit I don't really understand the difference between a genitive and a possessive cross-linguistically. Actually I may have been speaking incorrectly previously because of this. But there's this on languages that work similarly, regarding not marking the possessive:
http://web.archive.org/web/20040224062437/http://linguistlist.org/issues/12/12-1577.html
The DF language could also optionally signify possession with a particle like classical chinese, mandarin or vietnamese ("Urist own cat"). That would not break the syntax.

Throwing this argument away might be sane too if it's just unattractive.

What would your suggestion look like in a sample sentence?

[Malleus Inferni]

Is it really that unusual? What are these universals? That's definitely an interesting concept and one anyone designing a language should be aware of if it's based on science.

A "universal" is an unbreakable crosslinguistic rule. The basic idea is that we can conceive of languages with certain properties, but we would never encounter a real language like that for some reason. This could be because thinking that way is unrealistically difficult, or because the parts of the brain that learn language are totally unable to express certain ideas.

It's a topic of ongoing research in linguistics. Progress is very slow, because experiments have to be devised very carefully, and I have only a vague memory of one example of an experiment on a hypothesized universal. Something to do with number agreement, IIRC.

The universal I was referring to had to do with markedness hierarchies in general. I can't find any good material on that stuff right now, but I'll ask around.

I'll admit I don't really understand the difference between a genitive and a possessive cross-linguistically.

"Possessive" refers to a semantic relationship between the owner and that which they own. "Genitive" refers to a morphosyntactic feature (ie. something about the shapes or relative order of words) which may be used to signal a number of semantic elements, depending on the language. Usually there are several of these, and the one that you call "genitive" is the main way of expressing possessive constructions, but that's not always the case.

When I interpret "of Y" as a translation of a genitive, I'm noting first of all that it doesn't necessarily indicate a possessive relationship; for example, the dragons do not own their slayer in the same way that I own my arm or my sword. More broadly, the translation is consistent with the way genitive phrases of other languages are rendered in English.

Actually I may have been speaking incorrectly previously because of this. But there's this on languages that work similarly, regarding not marking the possessive:
http://web.archive.org/web/20040224062437/http://linguistlist.org/issues/12/12-1577.html
The DF language could also optionally signify possession with a particle like classical chinese, mandarin or vietnamese ("Urist own cat"). That would not break the syntax.

This would be a great way to incorporate the "different registers" idea. At least, I'd be more easily sold on this than on the proposals with presence/absence of inflection.

What would your suggestion look like in a sample sentence?

Just as a random example? Let's say an uninflected verb is present tense, there's no person agreement, and the particle "åz" marks the direct object. So "Urist eats [a/the] cat" -> "Deb Urist åz kun."

It occurs to me you could also use a topic/comment construction, where the topic is unmarked/unsuffixed but also isn't explicitly the subject of the sentence.

[Snaake]

I hate to keep pushing this back, and at the risk of sounding like a broken record, but schoolwork hit like a ton of bricks this weekend, so I don't expect to have my proposal up for another few days, though I will probably continue posting here. C'est la vie.

Now then, I am somewhat stuck on a certain point: prepositions. Not as words, but in their usage. I know that in English we have prepositional phrases, which can't be broken up and while they technically CAN end a sentence, in almost all cases it doesn't really work. I'm not really concerned with coming up with words like above, below, under, etc. at this juncture but I am having some difficulty with form. Consider this simple English...

Catten runs over the hill.
This would be rendered currently as:

Catten rur ág inen (lit. Catten run over hill)
This works fine, but in a more complex sentance:

Catten runs past Urist's cat.
Rendered as:

Catten rur ág Urist kun. (lit. Catten run over Urist cat.)
This leaves a lot more ambiguity. Now my question is, should prepositions be explicitly so with their own distinct form, as noted by either a special particle or a special construction? Or would it be better to stay with the English form?

I asked earlier if dwarven mapped one-to-one, in the very next example you personally said it did, and yet know you go against that statement? language_DWARF.txt does have [T_WORD:OVER:ag] and [T_WORD:PAST:geth] (no idea if this is an adjective or a noun though, or if the game even distinguishes between them, no indications towards that in the RAWs at least...), though, so you were right the first time. The example quoted above would be many(in English)-to-one, and I suspect most/all real languages' vocabularies' relations to each are technically many-to-many at least in some cases due to slight differences in the meanings, associations and connotations of words.

I do get your point about "Urist cat" being ambiguous though. Urist's cat, or a cat called Urist? The answer that pops into my layman's mind would be that in these sort of cases, the word order would matter (ie. owner/name are defined to be on different sides of the noun), or maybe the owner would be expressed as an adjective tacked onto the noun. Not sure how easily this can be generalized. The particle option is relatively simple and easy to remember, in my opinion. "Own" translates to "sanreb" though, which is maybe a tad too long. Maybe "greed" = "dal" for extra dwarfiness? ;P Or just invent something new.

...The next problem we come to is the unmarked genitive case, which people have already mentioned here. That's very unusual cross-linguistically. In fact, if the nominative case (or whatever this language has instead of a nominative) is marked and the genitive is not, that's supposed to violate a universal and make it unlearnable by normal human beings.
...

They're dwarves. Rather strangely-thinking ones, at that, if dwarven AI and the fanon around their thinking is anything to go by

[Loam]

I believe the "past" in the raws is temporal, not locative: "past, present, and future." You can look it up in language_words.txt to see how it's used (sort of).

Regarding the in-game Dwarvish: we only see it used for names. So, as I saw it, it might use a system peculiar to proper nouns, without regard to actual spoken Dwarvish; more malleable in meaning, and therefore in construction.

For example, the name Alexander "means" Defender of Man; but no one ever called Alexander the Great "Defender of Man the Great." Just so, no dwarf is named "dagger," although many are named Urist; and no dwarven civilization was ever called "The Oaken Tomes," though it might be called Kinmelbil. So if someone said "I took the oaken tomes," he might not say "I took kinmelbil."

Does that make sense? Sorry if it's less than clear: it's a difficult distinction to make.

[Snaake]

I believe the "past" in the raws is temporal, not locative: "past, present, and future." You can look it up in language_words.txt to see how it's used (sort of).

I get this point, and did consider that the "past" listed is probably more time-related, but was unaware of the usage being specified in language_words.txt. "Over" also can be used in multiple senses, too eg. "overworld" (as opposed underworld), something being over, or running over (all related, but still). These usages are etymologically related, I guess, but eg. the meaning of "someone ran past me" is also related to the temporal usage, even if you're expressing location.

Regarding the in-game Dwarvish: we only see it used for names. So, as I saw it, it might use a system peculiar to proper nouns, without regard to actual spoken Dwarvish; more malleable in meaning, and therefore in construction.

Possible. But if so, then that removes much of what there is to work from regarding constructing the language, and although I'm unlikely to do much in this thread than offer a few comments, I'd much prefer something more closely related to the in-game stuff, rather than just throwing that away and starting from scratch.

For example, the name Alexander "means" Defender of Man; but no one ever called Alexander the Great "Defender of Man the Great." Just so, no dwarf is named "dagger," although many are named Urist; and no dwarven civilization was ever called "The Oaken Tomes," though it might be called Kinmelbil. So if someone said "I took the oaken tomes," he might not say "I took kinmelbil."

Does that make sense? Sorry if it's less than clear: it's a difficult distinction to make.

It does make sense, although (again with the nitpicking!) I'm uncertain how good the real-life examples are, since in western languages at least nowadays most people have forgotten the original meanings of names (eg. Chinese is different, of course) like Alexander, Daniel ("God is my judge" in Hebrew), etc. And it's probably impossible to say how large a proportion of peoples' names, at least, originally had meanings like this.

[Baffler]

@snaake
I realize the contradiction, and I don't like that it ended up that way. I just could not for the life of me think of a better example. My understanding of linguistics is elementary at best, and it definitely shows at times.

Regarding the in-game Dwarvish: we only see it used for names. So, as I saw it, it might use a system peculiar to proper nouns, without regard to actual spoken Dwarvish; more malleable in meaning, and therefore in construction.

this is an extremely attractive to think of it, actually. I can't think of any mechanical issues off the top of my head with it anyway.

[Loam]

I'm uncertain how good the real-life examples are

True: but they are just examples. The only other ones I've got are from Tolkien - e.g. "Feanor" = "Spirit of Fire," but of course he wasn't referred to as "Spirit of Fire." This might be a better example of what I mean, since the Elves hadn't (presumably) forgotten the etymology of their names as we have.

It would be nice to have a language that could be applied to the in-game names, but I am not sure it can be achieved without being very limiting. I'd like a language that could maintain clarity without such a hard and fast structure.

this is an extremely attractive to think of it, actually. I can't think of any mechanical issues off the top of my head with it anyway.

All I can think of (and I'm probably just shooting myself in the foot here) is titles, like "Cerol Ochrewinds the Deity of Raptors." Again, these could be thought of as "names," but as English uses them they are simply appositives: again, like Alexander the Great. If they aren't considered part of the name, one must assume they use at least a little of the Dwarven structure.

[chevil]

I just had a idea that dwarvish should have strict and clear rules so that there is a chance that toady integrates it into DF.

[Orange Wizard]

I'm posting to show my support! I definitely want a fleshed-out Armokian Dwarvish. I've seen threads on it come and go, many times. With any luck, this one will create something useable.

Sadly I'm not likely to get involved. My knowledge of linguistics is limited to an absolutely flawless understanding of English, about nine words of French, and half that of Latin.

I just had a idea that dwarvish should have strict and clear rules so that there is a chance that toady integrates it into DF.

I wholeheartedly agree, but being a constructed language this shouldn't be a problem.

[Loam]

(hrk... grrrm... hnngh... I'm not dead yet! ... urk...)

Sorry for the necro, but I got around to working on this (somewhat) again. I just spent the better part of an afternoon making an alphabet, and a computer font to go along with it (the font is pretty crude, though, but it's functional).

Spoiler (click to show/hide)

It's a bit off, since there are some differences between the Roman and Dwarvish alphabets. Dwarvish doesn't have J, Q, W, X, or Y, and only has H in the letters Th and Sh (which, along with Ng, are missing from the font, although I did come up with symbols for them).

Just for fun, I also made "names" for the letters, taken from Dwarvish words:

Spoiler (click to show/hide)

(this follows the Dwarvish alphabet sequence)
Arel, "water"
Ser, "head"
Rith, "bell"
Lun, "cloud"
Tun, "door"
Nil, "hammer"
Muz, "fungus"
Etur, "boulder"
Kil, "wheat"
Gusil, "copper"
Dák, "tree"
THash, "mead"
Ber, "earth"
Isin, "grass"
Otin, "wine"
SHis, "fang"
Zon, "helm"
NGárak, "beetle"
Vim, "tear"
Uben, "soap"
Cog, "boot"
Feb, "arrow"

[Kaelem Gaen]

Looks nice,  but I hope I didn't catch on to this to late. 


I also wish I had some suggestions to offer.  Hopefully this won't die off like many of the other's I've seen.   I did start watching the wiki on wikidot.