From Lojban
Jump to: navigation, search

It's a cmavo class. Previously incorrectly glosed as "part of speech". "Part of speech" actually means noun, verb, adjective; which in Lojban are analogous to gismu, tanru, selbri, etc.

selma'o has been being used for lexeme (non-terminal in the formal grammar), so it ends up including non-cmavo: BRIVLA, CMENE. As pointed out by John Cowan, however (,

It's certainly bogus to call CMENE a selma'o, though the internal code of the parser does (as well as BRIVLA). In Lojban, a better term such as vlalei would be preferred. Then there are 3 vlalei and 120+ selma'o.

Naming selma'o in Lojban is a perennial problem. To say that .ui belongs to selma'o UI, you can try

  • zo .ui cmavo la'e zo .ui
  • zo .ui cmavo le .ui zei klesi
  • zo .ui cmavo la'o gy UI gy
  • zo .ui cmavo la .uik. (cf. jek, gihek, etc.)
  • zo .ui cmavo zo .ui (cmavo: x1 is a cmavo whose syntax is the same as that of word x2)
  • How about saying that zo .ui cmavo le selcinmo? after all, the se cmavo (hence selma'o) is the function it performs - .ui is an attitudinal. I don't think all "selma'o" really do se cmavo, so we should not try to say in Lojban that xu has the same function as .ui. It does not. Simply, I feel that the x2 of cmavo should be a short version of the function it performs, though I suppose I'd be satisfied with the cmavo the selma'o is named after - zo .ui cmavo zo .ui - mi'e. .kreig.daniyl.
    • Nobody is saying that xu has the same function as .ui. Only that they have the same syntax. (syntax != semantics) --Jay
  • You can try coming up with new words to describe each of the hundreds of selma'o; but I suggest something that sticks with the current established names is slightly saner. What's the word description of GUhA, in less than ten words? And I think using cmavo xy. referring to Lojban cmavo as anything other than "is of selma'o X" is going to be utterly uncommunicative; what else do people expect you to say of .ui? -- nitcion
    • le gimsmu be zo cmavo cuska le cmavo .e le smuni be le cmavo .i le smuni be zo .ui du le smuni be lu mi gleki li'u .i lenu gleki se cinmo .ija'e zo .ui cmavo le selcinmo .i zo xu no'u le drata cmavo be la .uik. ku cmavo le datnycpedu .i drata cmavo ka'e cmavo drata smuni ni'o ro cmavo be la gu'ak. na cmavo le naldrata smuni .i mi'e. .kreig.daniyl.
  • ta'o zo .ui cmavo le selma'o be zo .ui -mi'e tsali


Ugh. These are all blyeech. I can't find the thread this was discussed on in the List. -- nitcion

(I vote for something figurative like pevyseljimca.)

selma'o have been subdivided into numbered classes (e.g. UI1, UI2, UI3...), for semantic reasons.

nitcion propose such a scheme for BAI.

For a list of every selma'o, consult chapter 20 of The Complete Lojban Language.