Now there are plenty of KOhA sumti to go around. In fact, if you've run out of words by getting to ko'u, you can start over again with fo'a, fo'e ... fo'u. There is a problem, though: you have to remember (a) which sumti was assigned to which KOhA word, and (b) to assign the sumti in the first place. There's nothing to say that this will not become commonplace in future Lojban usage. Right now, however, there is a feeling that this is a little too calculated to work spontaneously. And Lojban cannot readily use the little hints natural languages pepper their grammar with (like gender and number), to keep track of who is who.

As a result, yet another strategy has been introduced to refer back to sumti. This strategy dates back from 'Institute' Loglan, before Lojban arose in its modern form. (Yes, Lojban has a history and a prehistory. No, we don't really have the time to go into them here.) The strategy involves acronyms. Simply put, if you see a Lojban letter being used as a sumti, you take it as referring to the last sumti whose selbri starts with that letter. So in

la suzyn. cusku lu coi li'u lo nanmu .i ny. cisma
Susan says "Hello" to a man. The man smiles

ny. stands for nanmu. There is no need to explicitly assign ny. with goi; but you can, and indeed if you assign it to a sumti which doesn't start with that letter, then that assignment will be the one that counts ("A certain Lojbanist, let's call him N, dislikes KOhA cmavo..."). Some Lojbanists dislike this usage because it, too, seems a little calculated (and initials and acronyms have decidedly non-literary associations in most natural languages!) Only time will tell which of the two usages will become more commonplace.