Message-Id: <m0pjFwh-0000PZC@xiron.pc.helsinki.fi>
Date:         Tue, 22 Mar 1994 18:18:20 EST
Reply-To:     Jorge Llambias <jorge@PHYAST.PITT.EDU>
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         Jorge Llambias <jorge@PHYAST.PITT.EDU>
Subject:      Mad MEX Proposal
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 2617
Lines: 57

This is a proposal to make a minor change in the interpretation
of the cmavo {ji'i}, of selmaho PA.

MAD MEX PROPOSAL

CURRENT LANGUAGE:

The current meaning is that all digits following {ji'i} are approximate,
and if no digits follow, then the number preceding it is taken to be rounded.
{ji'ima'u} means rounded up, and {ji'ini'u} rounded down.

PROPOSED CHANGE:

Interpret {ji'i} at the beginning of a number to mean that the complete
number is approximate, rather than specifying which digits are approx.,
and when ji'i is between two numbers, that the two numbers specify an
approximate number between them. i.e. ji'i would not separate digits but
numbers. When {ji'i} is at the end, the current meaning could be maintained.

RATIONALE:

In the case of leading {ji'i}, this usually will mean that just the last
digit is approximate, (or the last string of zeroes in the case of an integer)
so {ji'i63549} would mean approximately 63549, probably plus or minus five
or at most ten. Now it could be anywhere from 50000 to 80000, since the 6 is
already approximate. To get what we want we have to say {635ji'i49}, which
separates the digits that are exact from those that are approximate.

I don't like numbers with things between the digits, but the reason for my
proposal is not that, but that I think the proposed new function is much more
useful.

In English (and in Spanish and in Esperanto, and I suppose in most languages)
one can give an approximate number by giving two numbers, that give an idea of
the range of the uncertainty. For example, I could say that "there were thirty
or forty houses in that town". This is certainly not a logical connection, I'm
not saying that there were either thirty or forty or both, but no other number.
I'm saying that the number was probably between 30 and 40, or maybe it was 29
or 41 or close to that.

There is no easy way to do this now. We can say {ji'i35}, but this misses the
whole point, because 35 is a more complicated number than 30 or 40. If no other
method was available, a native Lojbani in the same circumstances would probably
say {ji'i30}, or {ji'i40}, depending on which seemed closer, but the point of
giving two numbers is to give an idea of the error bars.

With the new interpretation this would be {30ji'i40}, "some 30 or 40".

CONFLICT WITH EXISTING TEXT:

{ji'i} has never been used in text (at least in that at the ftp archive)
with its current interpretation, so there would be no problem with that.

(It has been used twice in translation with the proposed meaning, though,
but by an unreputable translator who better remain unnamed.)

Jorge