As I've mentioned before, Shabkiuza has a high degree of regularity, but I don't want it to be completely regular. This dichotomy is emerging, at the moment, in my attempts to build a vocabulary.
I downloaded a list of the word base of Simple English, and I have been assigning words to it. I have also found a list of all English language words, which includes not just root words but every plural form, every -ing and -ed on the end of the verb, as a separate entry.
When I enter a gloss in the simple list, it appears automatically in the full list. I can then visit it, fill in the glosses for the morphological forms. But that doesn't address the other associated words. For example, -ness and -ive and -tion and -ful and -ly endings. These are all clearly based on the root word, and part of natural language is adapting my root words into these other parts of speech.
I have two different mechanism I use, depending on whether my starting point is a noun/adjective or a verb. If I'm starting with a noun, I create a verb form that is clearly based on the same root as that noun, but without a specific in-language mechanism for its creation. For example, the word for "awake" (adj.) is briatvo. The verb form I created for "to wake" is bria, which fits my self-imposed phonotactics for verbs. There isn't a productive suffix in the language -tvo, although I might create a few more adjectives with that ending in a similar semantic space. But the two words are clearly connected.
If I start with a verb, then there are already mechanisms built into the language to produce nouns and adjectives. The verbal noun, which acts like a gerund, involves the suffix -kko. Therefore, I get briakko for "waking", gammakko for "cutting", etc. If I want to further derive a noun from the verbal noun, I remove the -ko ending. For example, gammak is "a cut". In general, this form is used to indicating the result of the action of a noun, rather than the action itself.
Another form is the verbal adjective, which is like a past participle. It is an adjective indicating the state of having the action performed. So for example, the adjective "cut" would be gammad, the root verb plus the -d ending.
What about more complicated suffixes? Here are a few examples
cuttable: amu talgammad' "able passive+cut+verbal-adj"
wakeful: fozh briatvo "full-of awake"
soldier (literally, "cutter", one who cuts): gammaza "cut+person-who"
At the moment, I have defined 286 "simple" words, but thanks to these word generation processes, both productive and non-productive (in in-language terms), my total word count is up to 535, and growing.