Stuck in a rut

[vegfarandi]

I've been revising and reworking Imutan for the past year or two and for some reason it's turning out to be harder than any previous language to work out. I keep discovering new systemic ways to "make it cooler" which require going back and redoing previous sections. And after I've redone one section, I have to go through everything else to make sure the cascading impact of those little changes is captured throughout. And several times I've gotten so tired of revising, I decided to start the grammar from scratch, transposing and rewriting things in order from the existing documents but I've ended up with 10 sources of various levels of accuracy to how I currently envision the gestalt in my head.

I'm basically just looking for some commiseration, and wondering if any of you have fallen into similar cycles with a particular language, and if so, if you've found more productive ways to move them forward. My goal is to get Imutan into publishable state this year, but looking at my messy grammar file now, and anemic lexicon after the fourth or fifth relex I'm worried that won't be doable.

In the end, I'm super excited about where it's headed. It feels like an amalgamation of all the cool features I want in it and I can really envision the final product, but for some reason having a very hard time articulating it.

EDIT: Specifically, I've figured out ways that the proto-language probably worked and created a vague working sense of it. Would it perhaps be helpful to simply go back, sketch out the proto-language in more detail, and approach Imutan as a derivation as opposed to a priori? The thing that scares me about that is that I've never really been successful at making a posteriori language. I find working through sound changes extremely tedious.

[Salmoneus]

I've been revising and reworking Imutan for the past year or two and for some reason it's turning out to be harder than any previous language to work out. I keep discovering new systemic ways to "make it cooler" which require going back and redoing previous sections. And after I've redone one section, I have to go through everything else to make sure the cascading impact of those little changes is captured throughout. And several times I've gotten so tired of revising, I decided to start the grammar from scratch, transposing and rewriting things in order from the existing documents but I've ended up with 10 sources of various levels of accuracy to how I currently envision the gestalt in my head.

I'm basically just looking for some commiseration, and wondering if any of you have fallen into similar cycles with a particular language, and if so, if you've found more productive ways to move them forward. My goal is to get Imutan into publishable state this year, but looking at my messy grammar file now, and anemic lexicon after the fourth or fifth relex I'm worried that won't be doable.

Yes, welcome to conlanging.

Well, specifically I think this is where a lot of us end up after a while. The more substantial your project, the more impossible it ever is to actualy finish. I find ideas for one's conlang tend to be of two types: exciting new ones, which, being new, require changing everything else, or once-exciting but now-old ones, which you still approve of, but because they're old they're no longer exciting, so it feels tedious actually explaining them for the nth time...

I think the only real solutions are a) discipline, really forcing yourself to put everything on paper as it is and leave it alone (it helps if it's a minor project you don't care about too much) or b) learning to enjoy the process, rather than the result...

EDIT: Specifically, I've figured out ways that the proto-language probably worked and created a vague working sense of it. Would it perhaps be helpful to simply go back, sketch out the proto-language in more detail, and approach Imutan as a derivation as opposed to a priori? The thing that scares me about that is that I've never really been successful at making a posteriori language. I find working through sound changes extremely tedious.

I don't know. I do think that working out more about the proto-language can often be helpful, particularly in dealing with exceptions and complexities. But that doesn't necessarily mean you have to sit down and write an entire description of the proto-language and a perfectly detailed description of every soundchange, particularly if you don't intend to make any related languages.

[zompist]

I suspect this syndrome is common-- Tolkien was a notorious sufferer, and Jeff Burke who many may remember from this board.

I don't have any good advice, but maybe some perspective: there's no need to make a conlang perfect, because it won't look perfect anyway once you've done several of them. Or to put it more positively: it's great that new ideas that you love keep coming. But you don't have to put them all into the language you're working on. If they don't fit, they can go in the next one. Or throw them in as dialect or register variants.

There are certainly projects that I'm stuck on; I usually just let them be stuck and work on something else.

[Zaarin]

I have a triconsonantal conlang that is also the main focus of its particular project; I go back to the drawing board every time I start to understand Semitic a little better (something that was significantly helped by my a posteriori Punic conlang). At the moment I'm pretty happy with how it's developing...but we'll see how long that satisfaction lasts.

As an aside, I once had a project in mind that was intended to have ~400+ languages. Even I eventually acknowledged that was insane.

[Pabappa]

Are you changing things because you've made mistakes, or because you've thought of better ideas? If the former, yes, conlanging is a lot of work, but you'll get there, because eventually all the mistakes will be fixed. If the latter..... having more than one conlang might help. After all, what if two of the features you like happen to contradict each other? I think Ive gotten lucky with Poswa in that I happened to combine many ideas into a seamless whole, and none of the features contradict each other. However, Ive had to leave some things out ... in some cases because they dont fit with the rest, and in other cases because I just didnt think things through well enough early on but it's too late to redo everything. For example, Poswa's /tʲ nʲ/ > /f v/ sound change is based on a misreading of Russian's sound changes that I came up with about twelve years ago, but I can't change it because it's crucial to the grammar. Ive just had to resign myself to the fact that I cant make a perfect conlang, even with more than a decade of effort, and so I've decided to live with that flaw. (Poswa's phonology is highly unrealistic from a typical human language standpoint, but I've always prided myself on using naturalistic sound changes.) I've found other glaring mistakes, but in most cases I've fixed them, because no positive features of the language depended on them. For example, I forgot to shift /hu/ > /fu/ in some words, and just figured I'd explain it away as analogy since some of their inflected forms would contain /hau/ instead. But I have since changed all of them, at least the ones I found. Its a lot of work but I have a lot of time.

Specifically, I've figured out ways that the proto-language probably worked and created a vague working sense of it. Would it perhaps be helpful to simply go back, sketch out the proto-language in more detail, and approach Imutan as a derivation as opposed to a priori? The thing that scares me about that is that I've never really been successful at making a posteriori language. I find working through sound changes extremely tedious.

I'd say yes, but a lot of words can get lost through sound changes, so it's OK to let Imutan be the dominant language and therefore seed the proto-language with words that survive only as pieces of Imutan words, rather than having a 1:1 correspondence all the way through. THe way you say "simply go back" makes it sound like it would take relatively little effort, in which case I say sure, go for it. But if it means a sixth run through your dictionary uprooting everything all over again it might not be worth it.

As an aside, I once had a project in mind that was intended to have ~400+ languages. Even I eventually acknowledged that was insane.

I have about 90 conlangs right now, but more than 80 of them are just intended as naming languages. They have sound changes, but I make the words up on the fly by looking at the shared proto-language. Not bad since placenames tend to resist replacement more than normal vocabulary words do.

[Ælfwine]

To be honest I am the same way, I'll fixate on certain things for weeks before dumping the project or moving on to something new and fresh. Recently I've switched to focusing on one or two main projects and an additional few side projects to tinker with when I run out of steam.

[gach]

I don't have any good advice, but maybe some perspective: there's no need to make a conlang perfect, because it won't look perfect anyway once you've done several of them.

I find it useful to think that after a while no natlang will look perfect either. They have grammatical processes that are less elegant than what you thought when you first learned about them and you keep finding poorly explained word forms that are probably due to internal loans or incomplete sound changes but look just like errors in running the sound changes for a conlang. One of my own hurdles is getting invested in exploring certain typologies and not daring to set things in stone before reading just a bit more literature on the subject.

Having several less important side projects is good advise. The downside is that if you are too worried in putting half cooked ideas into your main conlang, you may soon find doing all sorts of interesting things with the side projects and not working at all with the main conlang that you hold the dearest to you.

My best advice is to try to learn being less perfectionist. Perfectly capturing a nebulous ideal of a conlang in written form is as impossible as producing flawless art.

[akam chinjir]

One of my own hurdles is getting invested in exploring certain typologies and not daring to set things in stone before reading just a bit more literature on the subject.

Having several less important side projects is good advise. The downside is that if you are too worried in putting half cooked ideas into your main conlang, you may soon find doing all sorts of interesting things with the side projects and not working at all with the main conlang that you hold the dearest to you.

My best advice is to try to learn being less perfectionist. Perfectly capturing a nebulous ideal of a conlang in written form is as impossible as producing flawless art.

The scratchpad format (at least as I've interpreted it) has done wonders for my conlanging productivity precisely because of how it's helped me deal with this sort of issue. A post is enough of a thing that I'll want to think it through a fair bit and ring some changes, but it's also obviously nonfinal so I don't feel like I need to be completely on top of things before putting it up. And having gotten a bunch of stuff done, I often end up getting a lot more out of subsequent reading, because it's relevant to things I've started working through in the language.

(Before I discovered this format, my conlangs mostly died in morphophonological swamps, because I didn't know enough---either about the languages themselves or about morphophonology in general---to write a proper phonology chapter, but apparently I'm incapable of doing proper iterative development in the privacy of my own notes, and without a phonology chapter more or less done I was never able to get to serious work on other aspects of the languages.)

[Raphael]

I've been revising and reworking Imutan for the past year or two and for some reason it's turning out to be harder than any previous language to work out. I keep discovering new systemic ways to "make it cooler" which require going back and redoing previous sections. And after I've redone one section, I have to go through everything else to make sure the cascading impact of those little changes is captured throughout. And several times I've gotten so tired of revising, I decided to start the grammar from scratch, transposing and rewriting things in order from the existing documents but I've ended up with 10 sources of various levels of accuracy to how I currently envision the gestalt in my head.

I'm basically just looking for some commiseration, and wondering if any of you have fallen into similar cycles with a particular language, and if so, if you've found more productive ways to move them forward. My goal is to get Imutan into publishable state this year, but looking at my messy grammar file now, and anemic lexicon after the fourth or fifth relex I'm worried that won't be doable.

If it's any consolation, if you have an actual language worked out, you've already accomplished a lot more than me. I've basically run into a wall the moment I tried to start working on sound changes, and I'm not even trying to get actual languages - just naming languages. Good luck!

[So Haleza Grise]

I spend a lot of time here! But I don't really have any good advice than "persist." Sometimes I think it's time away from a grammar that makes me want to throw it out/revise anything/start from scratch. But if I work on it regularly, I come up with ideas that I don't commit to paper right away but that I can add in, giving me an incentive to keep coming back to make (even small) additions. And when I'm making an addition I can usually find something to clean up (e.g. an out of date example) that makes it look better, so it feels like progress.

[HourouMusuko]

Yes, unfortunately, I tend to be both A) indecisive and B) a perfectionist, so I'm constantly rethinking decisions I've made about Lihmelinyan, and wondering whether I should make significant changes. Just now I'm considering overhauling the verb system and I've already changed it several times in the past year. I still don't know if I want /b/ as a phoneme because I seem to never use it outside of noun case endings. Sometimes I do think I could benefit from creating a second conlang, as that would give me the chance to try out of some of these "options", but going back to square one doesn't sound fun given how long I've been working on Lihmelinyan and how much effort I've put into it. So I understand the "rut" feeling.