When creating words I have often run into the problem that too many words sound similar even when I have allowed many different phonemes. Are there methods to avoid too much similarity between words, while making words similar enough to feel part on one language?

  • 4
    Natlangs all have lots of similar words, not to mention homophones and polysemy. Can you explain more why you think you've actually got a problem? – curiousdannii Feb 17 '18 at 6:41

A modern approach to this problem is to generate the words automatically according to some formula describing allowed words. You can filter out words that are too close to already generated ones, but usually the randomness used by a computer program will be sufficient.

There are well-known examples of Conlangs with computer generated vocabulary. Loglan and Lojban used some vocabulary lists from natural languages as input for their vocabulary generating algorithm.

Not the answer you're looking for? Browse other questions tagged or ask your own question.