> On Thu, Sep 6, 2012 at 4:39 PM, ianek <
jane...@gmail.com> wrote:
> > On 5 Wrz, 00:24, Jorge Llambías <
jjllamb...@gmail.com> wrote:
> >> On Tue, Sep 4, 2012 at 4:52 AM, ianek <
jane...@gmail.com> wrote:
>
> >> > It's a property of being a cliquehttp://
en.wikipedia.org/wiki/Clique_(graph_theory).
>
> >> If you restrict "simxu" to the cliques of graph theory, you can't use
> >> it for something like:
>
> >> lo zajba pu simxu lo ka ce'u ce'u sanli kei gi'e se morna lo remna pramide
> >> "The gymnasts stood on one another's shoulders and formed a human pyramid."
>
> > So what {simxu} really means? That the graph is connected?
>
> Maybe even that is not completely necessary. Probably something like
> most/almost all nodes must be connected to other nodes and the
> connections have to be dense enough.
>
> >Arguments
> > based on something that sounds good in a natural language are
> > suspicious for me. The definitions of Lojban words are (in most cases)
> > in a natural language, but it doesn't mean that they should be
> > ambiguous.
>
> Vague is not the same as ambiguous. You can always define a more
> precise word ("rolrelsi'u"? "simymu'o"? something else?) for the more
> specific meaning. If you make "simxu" require full pairwise
> distributivity, you make it practically unusable for most ordinary
> contexts.
This is exactly the problem with Lojban. It aims at being logical, but