Glosa Synonym Criteria (obsolete page)


I worked on development of the Glosa dictionaries from 1997 through 2000 but no longer maintain them. I've kept my stuff here for historical and reference purposes. The latest and greatest dictionaries, currently maintained by Marcel Springer, can be found at the Glosa Website. See my Glosa Page for Glosa info and links.

. . . . . . . . .

Suggested criteria for the selection of the "preferred" synonyms:
  1. If the synonym spawns a Generic, make it the preferred synonym!

  2. The synonym which is used most often in common compounds should be the preferred synonym. The existence of the compound setimana (seven mornings = a week) argues for seti and mana rather than septi or matina. On the other hand, we also find septi-di!

  3. In male/female pairs, or any group of corresponding words created by means of generics, a generic combined with a neutral term should take precedence over a unique term. Thus, fe-sponsa is preferred over uxori.

  4. The preferred synonym should not look or sound too much like some other unrelated word. This will probably be the trickiest criterion to apply, because the Glosa dictionary already contains many words which differ only in a final vowel.

  5. For groups of logically related words, consider how the words might be used. For example, directions are often abbreviated - so Nord, Est, Sud, Uest are preferable to Boreo, Orienta, Austra, Ocide (which, however, make for wonderful literary synonyms, in my opinion!)

  6. For groups of logically related words, the preferred words should all come from the same language, for predictability.

  7. Originally, Glosa used the CVCV pattern for its words. As Robin Gaskell pointed out, some of the later additions - more often the Greek ones - are a bit harder to articulate. Consonant clusters are bad news, and should be avoided. I too believe that aqa is easier to pronounce than hidra.

1998-03-19 gene krea, 1998-03-30 gene muta.