-
Notifications
You must be signed in to change notification settings - Fork 1
Word Generator
The word generator in this app has two modes. The first is designed to be as straightforward as possible, while the second mode is more complex, but also much more powerful.
There are fields for what consonants and clusters can go word-initially, medially, finally, as well as vowels and polyphthongs are available, and what sounds should never appear next to one another. The Trial Words field will try to autocomplete a word as you type based on those simple rules, and the Generate Words button will make around 24 fresh words.
At the top of the word generator, there is a checkbox labelled Use Advanced Phonotactics. Checking it will switch the UI below it to an entirely different set-up (this does not delete any information you have in the default fields).
The first of these new fields is the Categories field. Here, you can define groups of sounds as you wish. The format is as follows in this example:
C:: m n p b t d k g f v l r th sh
V:: a e i o u ai au ei eu oi ou
Note that each token is space-separated, unlike categories in the pronunciation rules (which are comma-separated). You can have multiple-character length tokens here. It's up to you whether to use IPA or your romanization/other orthography here.
The next field is for defining, as you might guess, syllables. There are two ways you can do this. The first way is to simply write out all the allowed syllables line by line, such as
V
CV
ClV
CVC
ClVC
Note that non-category symbols used here are interpreted literally (the l's in the above example will actually just be l's).
The second way is to define categories of syllables: this will be useful for the next step, so we'll try that out.
S:: V CV ClV
1:: o Co a Ca
2:: ete Cete Clete
Words in your language usually aren't actually allowed to freely follow generic phonotactic rules. You usually have some sort of morphological rules as well, such as all verbs having specific endings, nouns genders being marked, et cetera. Structures allow you to fully control how syllables are put together to form words. Let's start with a set for nouns that all end in o or a.
S1 SS1 SSS1
That's it, for this simple example. All of the generated words will be 2 to 4 syllables and end in o or a. Let's now add a new set of structures by clicking the Add Set button and create a set for our verbs.
1S2 1SS2 1SSS2
In this hypothetical language, all verbs start with a gendered prefix and end with -ete. Now we can generate that.
Let's say we only want to have one of these sets active at a time. You can click on the checkboxes next to each set to toggle them, and when you click the Generate Words button, only the structures in enabled sets will be selected from. If you create too many sets, the - button next to each one will delete it. You can't delete your last set, but if all of your structure sets are empty or disabled, the word generator will just construct 1-4 syllable words composed of random syllable types from your syllables field.
If there are specific sounds that should never appear next to each other, this is the place to specify those.
ji wu mn ou ao
Again, space-separated tokens.
Because it's possible to accidentally (or perhaps intentionally) ban all of your syllables this way, the app will only try to come up with a word that doesn't have any illegals in it 20 times. If it fails, it will show [ ∅ ]. If you see this show up when you try to generate words, it means that it tried and failed to generate a legal word 20 times.