Making conlangs for practice.

[xBlackHeartx]

I've long suffered from not being able to finish a conlang. Yeah, I know conlangs can never be 'finished', but I rarely ever even start them! Even if I do actually start writing things down, my conlang always just boils down to a phoneme inventory (I'm primarily a phonologist) and some specualative notes on what grammatical features I want to include in it. Its very, very rare I even attempt to start on the vocab. Obviously, this doesn't make much of a conlang. Yeah, I know you could just re-use the same vocabulary over and over, but that may not work depending on the grammar of a conlang. For instance, a conlang with inflectional noun cases may require a limit on what kinds of endings nouns can take in their nominative/citation form.

Besides often suffering conflicting goals for my conlang(s), I'm also intimidated by the sheer amount of work involved. Like, where do you start on the vocabulary? You're going to need a word for everything eventually, so where do you start? I've tried to go with the most commonly used words, but doesn't generate a word list that lets you make sentences while you're developing the grammar.

One problem I have also is my perfectionism. I just can't stand the idea of starting a conlang until I know I can get it 100% right. Of course, this has caused me to produce NOTHING over the past 10 years other than a folder full of one-page files of phoneme inventories that never went anywhere.

I have come up with an idea to fix this though. All I've learned tells me that its pretty much impossible to not screw up with your first conlangs. So I thought, why not just make conlangs I never intend to use just to get used to making languages? Of course, I would completely abandon these conlangs and likely forget about them after I was done exploring them, but it could help me learn through experimentation rather than speculation. Besides, it would at least give me SOME accomplishments that I could look back on and actually feel like I've made progress.

However, I've never heard of someone making a conlang just to practice doing it. Yeah, people botch their first conlang all the time, but they didn't make it to 'practice', they did it because they just wanted to make it for the same reasons people typically make conlangs (most of the time for world-building, art, or just yet another auxlang attempt). Wikipedia has an article claiming that 'personal languages' are often made just to test out linguistic features and are by their nature short-lived, but of course wikipedia isn't the best source on most things.

Has anyone here actually attempted this, and if so, how did it go for you?

[linguoboy]

You're going to need a word for everything eventually, so where do you start?

Not necessarily. It depends on what you plan to use your language for. There could be huge lexical gaps in your language compared to English (or another real-world language) depending on the conculture where it is spoken. Why bother inventing domestic vocabulary if your conpeople don't live in houses?

If you focus on derivational processes then you'll have the tools you need to generate vocabulary as needed. That's a big challenge in itself, but it's more manageable than coming up with a couple thousand lexical roots right off the bat.