Do you know conlang software tools?

[dɮ the phoneme]

I've recently moved over to doing everything in LaTeX (I use TeXShop on OS X), for both my grammars and lexicons. It's... a bit of a hassle, but boy do things come out looking pretty.

[Dē Graut Bʉr]

There's something to be said for pencil and paper in mapmaking though. And doing it well always takes time.

Which is precisely how I do it have have done ever since I was a little child.

[chris_notts]

I don't see how XML is much better. SQLite and Microsoft Access use a single file you can easily distribute. With MySQL, you can share the SQL dump.

I love sqlite. I haven't used it for a conlang dictionary, but I've used it a lot at work. I don't think I'd use any other database for a client-side application now... compared to a traditional database it's far easier to use, no admin, easy to backup, and unless you're working with many gigabytes of data it's actually faster due to a lack of network latency. Oh, and the library itself is so tiny that packaging it with your application makes more or less no difference to anyone. The only major limitation that forces me to sometimes use other solutions is the lack of support for concurrent write access.

Anyway, I always use LaTeX/XeTeX for writing grammers. I'm still struggling for a good dictionary solution... I've just started trying FieldWorks for the first time but I'm finding it a bit hard going. Too many steps seem to involve going off to a completely different part of the application and defining new categories just to get a simple dictionary entry to say what you want it to say.

[bbbosborne]

word 2007 all the way

[Sevly]

So in addition to ConWorkShop, which has been mentioned several times before and has some pretty cool community experiences and collaborative tools, I've recently discovered something called Lexiconga which seems like a not-bad solution for hosted (and thus cross-device) lexicon management. It has a pretty good mobile view, too. Now, I'm still not super impressed with some of the usability limitations of either so am using a custom solution myself, but they're decent options to check out.

[evmdbm]

There's something to be said for pencil and paper in mapmaking though. And doing it well always takes time.

Which is precisely how I do it have have done ever since I was a little child.

As I have always done it, but they don't scan well!

[storyteller232]

There's something to be said for pencil and paper in mapmaking though. And doing it well always takes time.

When you're as artistically bereft of talent as i am (I've seen videos of more elephants with more artistic talent than i have) there's something to be said for hiring someone to make a map...thankfully current novel is set in real world present day so no need for my own maps lol. However I agree, for my conlang i am doing much of it in a notepad but i am also tryimg to decide what software to digitize and organize it.

[Curlyjimsam]

The one big advantage of doing maps on the computer, I find, is being able to use layers to quickly customise the map for different purposes.

[zyxw59]

I like to sketch a map out on paper, and then scan it and trace it and edit it in GIMP. I'm much better at making freehand map-looking things with pencil and paper than with a mouse.

[Kuroda]

I keep a lot of notes and do a lot of doodling on paper, but for the 'master document' stuff -- lexicon and texts with interlinearization, at least -- I use SIL software: Toolbox or Language Explorer. Grammar descriptions proper, I just write out in plain old flat text.

Toolbox is probably my preference: it's simple, relatively easy to learn the wheels and gears under the hood, thus easy to customize, generally quite flexible, lets you do all kinds of crazy multi-lingual, multi-dialectal, multi-scriptal stuff (this is what you want to complete your dream illustrated pentaglot Georgian-Mongolian-Tangut-Middle Egyptian-Quenya encyclopedic dictionary), accommodates whatever level of attention to phonetics or etymology or morphology, or academic vs lay audience, that you want to have (or avoid); and once you're finished with the hard graft, it obviously can spit out well-formatted, pretty good-looking, print-ready (by modern standards) dictionaries in the push of a few buttons.

Aside from pure lexicon building/management, you can use it for comparative linguistics (word lists and cognate sets, though very little ability to work with sound change rules) and for working with or from texts. For the latter, it has a sturdy and sensible parser that lets you have some real fun; it cranks out interlinearized text fast, consistently, and just how you like it. It's as good as Language Explorer, IMO, despite being older, by a smaller developer team even when it still had one, and was never intended as a big broad software suite like...

...FieldWorks Language Explorer, which I've used for an actual salvage-linguistics dictionary-making project (once) and an actual corpus-based grammar + glossary project (once), but not so much for conlanging. It's kind of a mad Swiss Anthropologist knife, so that you can wander into a mysterious community wherever in the world and whip out this nifty big tool with ALL the tools folded into it, that will guide you in collecting, organizing, documenting, and analyzing data for just about any kind of linguistic/ethnography project.

Which gives it two big drawbacks: 1) if you know exactly what you're doing and have your project defined already, big swathes of the package will just be irrelevant dead weight to you, and what you do want to use won't be set up quite the way you'd like (and FLEx is more resistant to customization than Toolbox); 2) if you don't have your project clearly defined at the start, or even if you do but are newer to fieldwork methodology... it's a lot to take on and start trying to boss around. Very database, so categories!

(Need to cross-reference the birthplace and age of your informants against their use of evidential markers in any clause with past-tense reference (narrator's voice only, not in quoted speech) in their folktales (just folktales, no other kind of material)? Ones narrated at the marketplace, not in homes or in private (unless after 12:00 noon on Sundays); and leave out stories involving any kind of birds if they're referred to with a collective plural form, please? Hey, buddy, chillax: FLEx is TOTALLY set up for that already!)

Lexique, I thought, was developed to be a presentation & distribution front-end for SIL-formatted dictionary projects. I.e., you'd do the lexicographic labor with Toolbox or FLEx, then import the data file to Lexique to create a user-friendly interface, with handy index tabs, drawings and photos, audio and video files, hyperlinks everywhere, and all that -- and which could be put online or distributed electronically more easily. I've never done that personally, but I've seen a couple really nice uses of Lexique in that front-end, electronic tool role though. (And using Lexique to build a dictionary from scratch seems like it would be all PITA and no plusses.)

There's a ton of newer software to do these kind of things, but ISTM that the more polished and specialized and practical and superpowers-having they are for doing real-world language documentation (with, like, actual speech communities and end users and all that), the more awkward or unhelpful they are for conlangers.

So I guess I'm for Toolbox because it's a real step up from Word or Excel, without going down the rabbit hole. (The other rabbit hole, I mean. The one that doesn't lead to the con(ey)lang burrow. Hraka!)

[Qwynegold]

I use LexiquePro for dictionaries. The bad thing about it is that you end up redoing a lot of stuff because you change your mind about how information should be organized. But I guess you'd have that same problem with any software. Here's a guide on how you're "supposed" to make your dictionary in LexiquePro: http://elibrary.bsu.az/books_400/N_297.pdf. It's a long read, but it gives you a better understanding what you're supposed to use different things for.

[Kuroda]

I use LexiquePro for dictionaries. The bad thing about it is that you end up redoing a lot of stuff because you change your mind about how information should be organized. But I guess you'd have that same problem with any software.

On their website, SIL should still have some stand-alone tools to help reformat, reconfigure, and enforce consistency on databases using their dictionary format (Lexique). Haven't looked recently though so I don't know specifically what's available. They would almost certainly be faster and easier than trying to do that stuff within LexiquePro.

The two programs I mentioned, Toolbox and Language Explorer, have never given me those kinds of problems, FTR -- and believe me, I change tracks while deep into something a lot ;( Both make it pretty fast and easy to restructure your databases. And the latter has some bulk edit tools that have saved many dozens of hours of slogging if I were using Lexique or Excel.

(FTR, not affiliated with SIL the organization in any way, or the developers of those softwares. Not a shill, honest!)

[Alvin Littell]

To generate a random root I use a python program:

Code: Select all

import random

def rc():
  return random.choice("bccdffghhjkklllmmmnnnppqqrrrssttvvwxxyz")

def rv():
  return random.choice("aeiou")

if 3 == random.randint(1, 3):
  print(''.join( [rc(), rv(), rc(), rv(), rc()] ))
else:
  print(''.join( [rc(), rv(), rc()] ))

It says that I need one root of either CVCVC or CVC, and CVC is preferred twice over CVCVC. Also among consonants voiceless fricatives are mentioned 2 times in the list for they appear more often. And sonorants are mentioned 3 times, and they appear even more often. This is easily configurable to make more types of roots e.g. CCVC, to play with phoneme frequency, to forbid phonemes in specific positions etc etc up to your imagination. And to make 521 roots at once, I use the following

Code: Select all

import random

def rc():
  return random.choice("bccdffghhjkklllmmmnnnppqqrrrssttvvwxxyz")

def rv():
  return random.choice("aeiou")

for i in range(1, 521):
  if 3 == random.randint(1, 3):
    print(''.join( [rc(), rv(), rc(), rv(), rc()] ))
  else:
    print(''.join( [rc(), rv(), rc()] ))

[xxx]

First time, in the 80', I wrote acceptable phonemes on pieces of paper and put them in a box... one tirage, one root...
But I rapidly prefered planification to random...