On Fri, Apr 22, 2011 at 6:21 PM, tijlan
<jbotijlan@gmail.com> wrote:
On 21 April 2011 12:58, MorphemeAddict <
lytlesw@gmail.com> wrote:
> 1) When is any language fully defined? Not even Lojban is fully defined.
> Being fully defined is not a well defined concept. This was discussed here
> several months back.
> 2) Many (all?) languages lack vocabulary for concepts in other cultures.
> Lacking words doesn't mean that a language (Klingon, in particular) can't be
> used to talk about those concepts.
SELBRI}, effectively yielding a Lojban-Lojban definition. If we
couldn't so define {ka'e}, either Lojban lacks explanatory vocabulary
or there's no sufficiently logical ground for the very idea of {ka'e}
to be expressed as a bridi component.
> On Thursday 21 April 2011 01:14:38 Jonathan Jones wrote:
>> Ideally, /every/ Lojban word should be defined in Lojban. And whenever a
>> word /is/, that most certainly /is/ the canonical definition.
>
> It's impossible to define every word in a language in the same language
> without circularity. There must be some definitions which refer to things
> outside the language. All words for species of organisms: "cinfo", for
> instance, is defined as any organism of breed x2 which belongs to the species
> Panthera leo, which is in turn defined by a type specimen.
Yes, circularity is unavoidable.
Circularity is avoidable. That's the whole point of Anna Wierzbicka's Natural Semantic Metalanguage, which assumes a small number (~62) of 'words' that can't be defined in terms of simpler words.
Here's the OED definition of "lion":
"a large powerful animal of the cat family, that hunts in groups and
lives in parts of Africa and southern Asia. Lions have yellowish-brown
fur and the male has a nane (= long thick hair round its neck)."
Every single word in it is circularly defined in other parts of the
same dictionary. But the text as a whole still describes, for speakers
of the language, what "lion" is.
Almost all dictionaries are unapologetic about circularity.
stevo