Re: a little aside: lojban dictionary?

[Logical Language Group <lojbab@GREBYN.COM>]

Yep, it's been discussed many times, with several different varieties of
answers to your question of 'how'.  John Parks-Clifford even wrote a sample
definition of some word or another.  I find your constraints a bit beyond
anything I've ever seen in lexicography, though.  I think you'd be hard
pressed to define any word in ENGLISH without recourse to synonyms, metaphors,
and negative terms, if you further insist (as seems likely) that all content
words used in the definition themselves also be defined under the same
constraint.

Lojban definitions, unlike standard English dictionary definitions, must define
not only the semantics of the gismu relation, but also each of the places that
is being related to show how it fits into the relation.  (This is akin to
requiring all English words to 1) be defined as verbs, and 2) define the
effect of ALL English prepositions when used with each verb, realizing that
most English prepositions have multiple meanings depending on the verb they
modify.)

There is no reason to believe that a Lojban-only dictionary COULD be done,
but an obvious question at this point is why someone would write one.  The
purpose of a book is to be read or referenced, not generally to prove that
it can be written.  There is as of yet no one that could effectively use a
Lojban-only dictionary.

I am curious as to why the high standards for such a dictionary.

lojbab