Conlangerie.
Dec. 19th, 2004 09:55 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
pauraque_bkCouple days ago I was talking to ![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
wootsauce about the conlanging process, and realized I haven't really written anything about what I do when I'm working on langs.
This week I happened to start a lang from scratch, because I wanted to conlang, but needed a break from all the Amíu stuff I've been doing. This language doesn't have a name yet, but it belongs to the Amíu-speakers' southern neighbors. The terrain they inhabit resembles British Columbia.
This is nothing like a proper grammar -- it's just an example of what my preliminary notes on a conlang look like, with commentary on my initial thought processes. I prefer to conlang on paper. It's always a pain to computerize everything when I'm ready.
I wanted a lang with some stop consonants for a change, since Amíu has none. But I didn't want to move too far away from the "catlike" phonology I've got going on this continent, so I started working with voiced stops, nasals, and approximants. I also had a random notion that this lang would have lots of circumflexes ^ in its roman orthography. Again, perhaps because Amíu doesn't have any, and I like them.
When I started, I was just playing with sounds and wordforms. I had no idea what these words would mean, if anything. As I went on, I started a tentative phonetic chart. I decided the second series of voiced stops would be aspirated.
Then I got distracted by a shiny idea -- what if they wrote with colored bits of braided string? I went and got some string and played with it, and wrote down some color possibilities. This was a very effective time-waster, and it took me a while to get back to the language.
I started to think about the vowels. How many did I want? Would lax vowels be allophonic? Did I want tones? Tone contours? Did I want the acute accent to invade my circumflextastic paradise?
I took a moment to note to myself that the Amíu word for urine (of a person) should have the determinative 'sea', and to remind myself of some other Amíu terms I was going to need. The eagle-eyed reader will notice that the Amíu is backwards. I actually find it very difficult to write right-to-left, so a lot of my Amíu notes are mirrored.
I wrote down a couple of Japanese words at random, thinking that I would try to mimic that language's pitch-accent system. But that idea was scrapped, and I decided I wanted stress to indicate something of importance in the verbal morphology, and that the circumflex would mark that stress. I started trying to figure out what happens morphologically with single-syllable verbs. Meaning still hadn't entered into things yet; dlaw and dlawâ were nothing more than "verb of inflection #1" and "verb of inflection #2".
Down at the bottom of the first page, I noted something from the book I'm reading on the Classic Mayan language, where the ergative case marks both subjects of transitive verbs and possessors of nouns. I wanted to try this.
Every time I've tried to do an ergative lang in the past, it ended up active (split-S), or accusative with heavy elision. Now I was determined to really look at ergativity, and do it right. I fiddled around with some case forms for the first person, and for the first time settled on some actual meanings for the words. I also decided on a stative/inceptive distinction in the verbs; I've done that before, but it was very much fun. Again, again!
With web pages open in front of me, I tried to figure out how the reflexive, antipassive, and possessive would work. I got confused. "Does possession require a genitive case?" I asked myself. I still do not know.
Aha, serious note-taking. Accusative langs group together the Agent of transitives and the Sole argument of intransitives, calling them nominative. Ergative langs group together the Sole argument of intransitives and the Object of transitives, calling them absolutive. I knew this.
The rest is stuff I did not know. While the accusative passive promotes Objects to Sole arguments (nominative) and shifts Agents to an oblique case, the ergative antipassive "promotes" Agents to Sole arguments (absolutive) and shifts Objects to an oblique case. Okay. But what is the antipassive for?
It turns out that only absolutives can head relative clauses and participate in coordinate clauses. A very good reason to be able to turn A's to S'es. Eureka! Syntactic ergativity!!
Now I could make my lang truly ergative. In an accusative lang, "I struck the bear and died" means that I died. The same construction in my lang would mean that the bear died. To get the "I died" meaning in the coordinate clause, I use the antipassive voice! Whee!
Next step was to add a tier to my verbal system. Instead of a stative/inceptive distinction, I'm going for a three-way: inceptive/stative/cessative. "Get", "have", and "lose" are easily derived from the same root.
I would also need a transitivizer. I made it a causative, and took great pleasure in deriving "kill", "ressurrect", and "make sure (it) stays dead". I also prodded at a reflexive, making "commit suicide".
So, after a few days, some scribbling, a lot of reading, and a lot of frayed pieces of string, I have a rough phonology, four or five meaningful roots, and a smidgen of morphology. This may seem like minimal output, but the output isn't what primarily pleases me. It really is the process.
I have no clue if this is interesting to anyone but me. Oh well.
wootsauce about the conlanging process, and realized I haven't really written anything about what I do when I'm working on langs.This week I happened to start a lang from scratch, because I wanted to conlang, but needed a break from all the Amíu stuff I've been doing. This language doesn't have a name yet, but it belongs to the Amíu-speakers' southern neighbors. The terrain they inhabit resembles British Columbia.
This is nothing like a proper grammar -- it's just an example of what my preliminary notes on a conlang look like, with commentary on my initial thought processes. I prefer to conlang on paper. It's always a pain to computerize everything when I'm ready.
I wanted a lang with some stop consonants for a change, since Amíu has none. But I didn't want to move too far away from the "catlike" phonology I've got going on this continent, so I started working with voiced stops, nasals, and approximants. I also had a random notion that this lang would have lots of circumflexes ^ in its roman orthography. Again, perhaps because Amíu doesn't have any, and I like them.
When I started, I was just playing with sounds and wordforms. I had no idea what these words would mean, if anything. As I went on, I started a tentative phonetic chart. I decided the second series of voiced stops would be aspirated.
Then I got distracted by a shiny idea -- what if they wrote with colored bits of braided string? I went and got some string and played with it, and wrote down some color possibilities. This was a very effective time-waster, and it took me a while to get back to the language.
I started to think about the vowels. How many did I want? Would lax vowels be allophonic? Did I want tones? Tone contours? Did I want the acute accent to invade my circumflextastic paradise?
I took a moment to note to myself that the Amíu word for urine (of a person) should have the determinative 'sea', and to remind myself of some other Amíu terms I was going to need. The eagle-eyed reader will notice that the Amíu is backwards. I actually find it very difficult to write right-to-left, so a lot of my Amíu notes are mirrored.
I wrote down a couple of Japanese words at random, thinking that I would try to mimic that language's pitch-accent system. But that idea was scrapped, and I decided I wanted stress to indicate something of importance in the verbal morphology, and that the circumflex would mark that stress. I started trying to figure out what happens morphologically with single-syllable verbs. Meaning still hadn't entered into things yet; dlaw and dlawâ were nothing more than "verb of inflection #1" and "verb of inflection #2".
Down at the bottom of the first page, I noted something from the book I'm reading on the Classic Mayan language, where the ergative case marks both subjects of transitive verbs and possessors of nouns. I wanted to try this.
Every time I've tried to do an ergative lang in the past, it ended up active (split-S), or accusative with heavy elision. Now I was determined to really look at ergativity, and do it right. I fiddled around with some case forms for the first person, and for the first time settled on some actual meanings for the words. I also decided on a stative/inceptive distinction in the verbs; I've done that before, but it was very much fun. Again, again!
With web pages open in front of me, I tried to figure out how the reflexive, antipassive, and possessive would work. I got confused. "Does possession require a genitive case?" I asked myself. I still do not know.
Aha, serious note-taking. Accusative langs group together the Agent of transitives and the Sole argument of intransitives, calling them nominative. Ergative langs group together the Sole argument of intransitives and the Object of transitives, calling them absolutive. I knew this.
The rest is stuff I did not know. While the accusative passive promotes Objects to Sole arguments (nominative) and shifts Agents to an oblique case, the ergative antipassive "promotes" Agents to Sole arguments (absolutive) and shifts Objects to an oblique case. Okay. But what is the antipassive for?
It turns out that only absolutives can head relative clauses and participate in coordinate clauses. A very good reason to be able to turn A's to S'es. Eureka! Syntactic ergativity!!
Now I could make my lang truly ergative. In an accusative lang, "I struck the bear and died" means that I died. The same construction in my lang would mean that the bear died. To get the "I died" meaning in the coordinate clause, I use the antipassive voice! Whee!
Next step was to add a tier to my verbal system. Instead of a stative/inceptive distinction, I'm going for a three-way: inceptive/stative/cessative. "Get", "have", and "lose" are easily derived from the same root.
I would also need a transitivizer. I made it a causative, and took great pleasure in deriving "kill", "ressurrect", and "make sure (it) stays dead". I also prodded at a reflexive, making "commit suicide".
So, after a few days, some scribbling, a lot of reading, and a lot of frayed pieces of string, I have a rough phonology, four or five meaningful roots, and a smidgen of morphology. This may seem like minimal output, but the output isn't what primarily pleases me. It really is the process.
I have no clue if this is interesting to anyone but me. Oh well.
