Print

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

[Starver]

Well for building "irregular" verbs, tenses and genderisation etc. you could devine "Rules" this can to happen procedural. As example for rules: many Job-names/titles in German have the suffix "-in" to indicate that a female is meant.

Not sure if you're talking about irregularities there.  That's regular grammar to "add an '-in' to indicate feminine version", very much like the "-[e]ss" for English, I suppose.  (Mein Deutsche ist nicht sehr gut, verseihung.)  Author => Authoress.  Prince => Princess.  It gets mildly irregular with Actor => Actress (instead of "Actoress") and one could argue that Husband => Wife is at the very irregular end.  (Without looking up the etymology, I don't want to bring up the reverse relationship of Widow <= Widower.)

Ditto irregular (often 'borrowed') plurals.  Fox/Box => Foxes/Boxes, but Ox => Oxen.

It's often the most basic words that are irregular.  Local variants of "to be" (I am, you are, he is; ich bin, du bist (I think!); il est, nous somme, ils avais; etc) being the obvious examples in most (if not all?) natural and organic languages.  I can't remember if "to fire photon torpedos" is regular or irregular in Klingon, unfortuantely! 

And please don't think I'm saying it must have irregularities, I just thought it worth mentioning that irregularities are often a feature.  The most annoying one, for learners, though gender (including neuter and various other alternatives that feature in some scandiwegian languages, etc) and the French insistence on differing rules for even the regular versions of -er, -re and -ir verbs causes me most problems in languages other than my own (having learnt all of the usual exceptions, and at least some of the less usaul ones, long before I lost my infant brain's intrinsic ability to pick up such subtleties).

What I was going to add to my "five tins of catmeat" example was a suggestion of procedural mutation, as well, as it happens.  Like in Welsh.  (I won't go into it.  It's 'regular', but I never really got a hang of the rules behind it, without a look-up table.)

[Mel_Vixen]

Not sure if you're talking about irregularities there.

The example was on genderisation of words . [edit] Anyway with a good Database in the background you can compensate too iregularies if we put words in wordfields and categories by giving them tags for example.

A Machine-translation can be extremly accurate with artificial languages.

[Starver]

Not sure if you're talking about irregularities there.

The example was on genderisation of words .

Gotcha.  I thought you were conflating the two (what I'd call a "rules" process for genderisation, v.s. an ad-hoc "exceptions" table for the irregularities).

[edit] Anyway with a good Database in the background you can compensate too iregularies if we put words in wordfields and categories by giving them tags for example.

A Machine-translation can be extremly accurate with artificial languages.

Yep, a bit like my example markup differentiated "can" (object) and "can" (auxilliary verb) at source, pre-constrained and ready for purpose.  And also sub-subtleties to the latter.  Much better for the computer to have an idea what it's supposed to be saying rather than let us messy humans provide it random text to try and parse the meaning then churn out a suitable equivalent meaning.

_{An old hobby of mine: Fooling an Eliza program into making ridiculous responses by exploiting homonyms.}

[jaked122]

Edited:Removed stupid joke.
Let's see, Procedural Syntax Generation would be complex, Unless it was extremely small changes, such as agglutination(Probably not that small).
Generating Words procedurally is somewhat complicated unless you don't care if it's even possible to pronounce(and if you don't Bruteforce password generation techniques would apply easily here)

[MelloHero]

Concerning irregularity and such, I've always thought the concise nature of the Dwarf language in the game was rather like Classical Chinese, which I still have trouble believing could ever have been a natural language, due to the amount of intuition necessary to understand it. It seems dwarfy enough, a language made to be as time and space efficient as possible. Specifically the way that words don't inflect, and other little words are only necessary for translation. I'm thinking of one phrase that was mentioned here that was seven words in English and five in the original: "Clutter God the God of Godly Gods," or the more concise, better-sounding "Om Nom Nom Nom Nom" (plus, extra points for meme inclusion).
At any rate, what if we adopted a mode of thinking like that into the Dwarf language, at least? For example:

Spoiler (click to show/hide)

Urist McDwarf, a digger of mines acting as mayor, made his own pick and killed an entire elf army with it.
Ｕｒｉｓｔ　ＭｃＤｗａｒｆ、爲市長之挖礦者、作其銶、以殺全木妖軍。
Urist McDwarf, being-mayor-(adj.) dig-mine-er, make his pick, by kill whole elf army.
The literal, awkward translation, for the CJK-illiterate, is intended to show the modular nature of the language. Longer, appositive phrases ("acting as mayor") can have 之 (ordinarily a possessive) added afterward for clarity, but can be omitted. "Digger of mines" is "digs mines" followed by a character for "person," like the suffix -er. 以 means "by way of," and here the object of "it" (which can also be written with 之) has been omitted since it can be inferred.

I guess this description could have sold it a little better, but if anyone is interested, do say so and I'll say more and try to make it less TLDR.

[Rowanas]

Sounds good, but I think all that would be needed is to add in conjuctions, and we're set. I see no need for structure rewrites, or even a language that is generated more than once. We have a dwarven language now, why not just create a fuller dictionary of dwarven words that becomes the new permanent dwarven dictionary. Of course, some words won't exist, and other words that we don't have, will. That's not too much of an issue to fix, so why not feed a gobbledooker (a translator set to generate translations for a language that does not exist, in this case dwarven, as was suggested by the OP) the English dictionary and be done with it? We can add in other language words if they are necessary (do we need 16 words for love? I'm sure we need more than 1) and that'll work beautifully.

Any reason why not?

[bjlong]

One way to give a language flavor, I've generally found, is to make and remove degrees of implications. For a basic example, in English you have an article system that boils down to indefinite-definite, or "Unspecified x" and "Previously specified x." You might make an article system like "Unspecified x," "Previously specified x," "X to be specified later" "X that I am now specifying." A little more complex, naturally, but it could (possibly) work.

It would help keep off the English-analogue that seems to be emerging. Simply feeding a RNG a list of words to create would work better if we get a few key differences to make the language a bit more complex.

[LordZorintrhox]

For the two problems, we could have a RAW that defines some words so that  Urist, Armok, Ber, Olan, Kogan and Usan are all still in the game.

I have good luck with dwarves named Mebzuth, myself.  I don't know why.

As a student of linguistics, I support this thread.  However, language is more complex than anyone who hasn't taken a couple syntax classes likely knows.  Leaving English grammar in place is the least of all evils, I think.  Even something as simple as going from SVO to SOV requires remodeling at deep structure level.  Adding in procedural generation of syntax rules would be...intense.  I wouldn't be at all disappointed or surprised if a robust language system proved to sit outside the bounds of the coveted goal of Modeling Everything.  I'm not a programmer, though, and I'm only in my second year of being a Ling major, so maybe there's someone on the forum who knows some more on the subject than I do.

Words, though, okay.  Phonology is easy (at least that's what the grades on my academic transcript tell me).

Yeah, y'all should heed this.  Language is way more complicated than you know.  Even irregularities are, by their very nature, strange.  This will rear its ugly head if any attempt is made to randomly generate a unique language in the truest sense of the word.

You might get far, but the most difficult thing of a language to simulate is the...I don't know if their is a term for it, but the feel of a language.  The subtle interplay of vocabulary, connotation, definition, and the speaker's cognition that leads to how the language views and expresses reality is complex and very often describable.  For instance, Spanish has a theme in its structure concerning the permanent/concrete/explicit and the impermanent/mutable/vague.  There are two verbs for "to be," one for intrinsic, immutable things (ser) and one for general, relative, temporary things (estar).  Furthermore, there are the indicative and subjunctive moods for verbs, which map similarly.  Relative location words fall into this as well: you can say "here," "right there," both relating to things more or less in the immediate vicinity, and "over there," for things way the hell over there.

This thematic distinction in the language colors how you speak it and how you convey information, and it colors the vocabulary.  How do you simulate such things randomly?

The problems get worse the more natural features to language you add.  For instance, say the humans are speaking some inflecting language, kinda like Spanish or Latin (that is, words change depending on their grammatical purpose in a thought).  Now you add a high rate of change to their language.  Simulating the change is pretty easy; sound changes and contractions are always regular and follow patterns that a computer can easily understand.  But that is the problem: the sounds changes are regular, so if a specific sound change will screw up a whole category of the possessive form of words so they look like there are plural and part of a prepositional phrase, well too bad.  How will the computer deal with this?  What humans do with it isn't even predictable; anything could happen.  Hell, English has collapsed Dative and Accusative into one case for Christ's sake.

And then there is just the weird shit out there.  Basque adds a word to every sentence composed of a particle for each word in the sentence that indicate the grammatical purpose and gender of that word.  That's be like "Red cars drive faster" becoming "Red cars drive faster nominative_feminine_adjective-nominative_feminine_noun-present_thirdPerson_verb-superlative," though in far fewer sounds.  No other language does that, so how does one simulate strange-ass things like that?  Should one even bother?

There are so many ways to build a language that it would be an epic undertaking in itself just to come up with a passable place holder.  Prepositions can't even be taken for granted: Latin uses two cases that more or less randomly associate with the prepositions, English just groups them with the noun phrases they modify, and Chinese probably dispenses with prepositions altogether.  Polish has an extra half-dozen cases just for prepositional statements alone.

If you want to get fancy, insert the noun inside the noun it is within for "on," or say the word "cat" a few tones below the word "table" which the cat is underneath.

For a game that simulates the tensile strength of materials, attempting to randomly generate language can't ignore these things; it would be against the spirit of the random content.

With that kind of possible complexity, it would be easier to just keep a static conlang for each race, or dub over English.  If you're going to dub over English, why bother with randomly each time?

[bjlong]

Static conlangs are good. I'd reccomend that we do that, but keep it just enough like English for people to translate easily. I'm sorry if that didn't come across in my post.

[Rowanas]

Sounds perfect. go for it.

[Fieari]

The most important thing a random language needs, I think, is "word color".  By this I mean a system of having words with similar meaning sound similar.  This isn't hideously complex.

We already have the beginnings of the system in our sphere and symbol word mechanics.  When generating a word, I'm sure we'd use something like the S V C system.  What you'd do is have a RAW file for each language that would specify the SVC for each sphere/symbol for that language.  As an extra benefit, you could also have the RAW file overwrite the random generation for specific words you want to use.




For those of you who don't know, the SVC system of word generation works like this:

A (S)yllable consists of (V)owels and (C)onsonants. To make a word, you randomly arrange Vowels and Consonants in particular word orders.  So if your Vowels are "A E O" and your Consonants are "B N T", the word BAB could be formed with the pattern CVC, which could also produce BAN, NET, TOB, but never NEE, since that'd be CVV.

It gets more interesting when you specify different sets of vowels and consonants to be allowed at different parts of a word.  So you could have C1 have "B, N, T" but C2 only has "M, S".  A word would then be something like C1 V V C2, and create words like BEOM and NEAS.

And of course, these groups can also have multiple letters in them, so C3 could contain "gn, hm, ph" and that sort of thing.  And then you can get even more complicated by adding S in.  S would consist of C1VC2, and then the word creation list could consist of "SVVS".  So they stack.

[Mechanoid]

If you want to get fancy, insert the noun inside the noun it is within for "on," or say the word "cat" a few tones below the word "table" which the cat is underneath.

Guess what happens when the dwarf gets a injury related to his voice-making body parts and "breaks" his language. Social out-cast, and serious stories.

[Dr. Melon]

I think the OP idea would work.
However, I suggest this modification:
Any word that is not defined in the language raws is generated. That way, you keep Urist, but all the words that aren't written in by hand into the language file are generated. So the word "casino" is not in the language file, ergo it gets generated.

[PTTG??]

I think the OP idea would work.
However, I suggest this modification:
Any word that is not defined in the language raws is generated. That way, you keep Urist, but all the words that aren't written in by hand into the language file are generated. So the word "casino" is not in the language file, ergo it gets generated.

That's excellent.

[Rowanas]

Hmm. that way, when you find a word you really like, you can keep it forever.