Re: knowledge and belief

[Logical Language Group <lojbab@ACCESS.DIGEX.NET>]

>Lojbab:
>>I was talking with Nora about the knowledge thread.  the xistence of the
>>epistemolgy place means that "djuno" probably means "is convinced that"
>>as much as "knows".  "djuno would overlap "birti" more than "krici"
>>except for the emotive nature of birti.  "birti" might more clearly
>>be glossed as "trusts that".
>
>Why would we want to change the meaning of {djuno} at this stage?

It is not a change in the meaning of djuno, but a clarification of the
English by adding an extra gloss.  The current argument, I am reading as
saying that to some English speakers, you can only use djuno when the
speaker finds the x2 to be true, which in Lojbanic terms probably means
finding the x4 of djuno to be truth-providing epistemology.  By contrast,
as Steven just pointed out, the "footnotes" beyond column 160 in effect
say that epistemology places are generally filled with references to brivla
with du'u places of which krici is one.  This is in keeping with my statement
of a few days ago that "krici" is suitable abstract form is an acceptable
x4 epistemology for djuno (so I didn't just make that one up on the fly -
it was already intended and stated  in the list).

The reason why I say that "be convinced of fact" fits djuno, is that to
be convinced of something, one must consider it true, AND have that
consideration be justified (thus bringing in the epistemology).  djuno thus
covers one sense of convinced just as it covers one sense of know.

>> Nora cited the example of PGP keys,
>--More--
>>where one may "know" a key for someone via some means, but the key
>>could still be wrong, with "birti" applyiong to the trustedness of the key.
>
>I don't know what PGP keys are, but in any case "to know a key" would
>not be translated with {djuno}, which means to know a fact. Probably
>you would have to use {selsau}.

I was using good-old English sumti-raising.  In this case, the complete
proposition is to know that the fact that the key value = <quoted text>
is valid.  My understanding is that key values obtained in person from
the owner (these are keys for encryption of net messages BTW) are the
most trusted.  Keys obtained from someone who got them directly from the
owner, where you personally know and trust the character of the provider
are of only slightly lesser trustedness, and so on, to the point that a
key obtained from a Net registry of keys that does not require proof
is probably of the lowest level of trustedness.  Using a less than
fully trusted key overs varying risks that the key might be compormised and
will therefore render the encryption transparent to analysis by suitably
high powered computers.

lojbab