Some cmavo comments, if y'all please.

[cbmvax!uunet!ee.mu.oz.au!nsn]

My comments on the cmavo list, Bob's countercomments, and my snide asides,
open for evaluation by the community:

>Message-Id: <m0k8NJG-0001Z4C@snark.thyrsus.com>
>Date: Thu, 8 Aug 91 23:14 EDT
>From: lojbab@snark.thyrsus.com (Bob LeChevalier)
>To: nsn@munagin.ee.mu.oz.au
>Subject: response on cmavo proposals
>
>>>fa'e      BAI
>>For the record, I still can't find a sensible use for this one.  This is
>>(to me anyway) nothing more than a tanru-maker in the wrong place.
>Using it rather more metaphorically, I might say

>la djim. kartr fa'e jimpe le lojbau platu
>Jim Carter has the Lojban design arse-backwards.

>This can serve as the basis for understanding rather less metaphorical uses

pace Jim %^/, this does even less to convince me. What does that say that
{fatne jimpe} doesn't? What do we have BAI for - would-be tanru? BAI is intended
to bring in supplementary arguments to predications, and only secondarily
to act modally. {fa'e} is very much the odd-man-out in BAI, the glaring 
exception, the one that doesn't do anything BAIish (music pi'o la betoven,
pu'a la betoven, du'i la betoven, cu'u la betoven.... fa'e la betoven?! And
this is not spurious: I defy anybody out there to use fa'e as a sumti tcita
in a nonstupid way.) Let us, I exhort you, give {fa'e} the old heave-ho.


>>>fa'u      JOI
>>Essay please.  How does this differ from {.e} or {jo'u}?
>Well .e is a logical connective and this isn't.

>mi fa'u la nik cu cinba la noras. fa'u la djein.
>does not imply you kissing Nora or me kissing Jane.  jo'u and .e would have
>each of the two of us kissing each of the two of them.

(I didn't touch her, I swear %^)

As John pointed out, termsets can be considered to make this redundant. How
many people know what a termset is? Well, you can work it out from the BNF
with some effort...

>>>ji'i      PA       approximately        digit/number: approximately (default 1) (number)
>>>ji'ima'u  *        rounded up           rounded up (appended digits)
>>>ji'ini'u  *        rounded down         rounded down (appended digits)

>>as I've said to John, I think this system barbarous.  For starters,
>>{ji'i} defaulting to zero ("epsilon") is much more convenient; for
>>seconds (gee, I sound like a JCB ultimatum %^) ji'ima'u is not
>>semantically the sum of its parts.  With ma'u defaulting to "small
>>positive number", ji'ima'u meaning "and a bit" => "rounded up" makes
>>much more sense.

>since ma'u is more like the 'plus sign', this in digits looks like 3.14159~-
>or 3.1416~+
>But really, what I did was try to come up with a way of saying that a number
>was rounded - surely an important point.  I don;t undersatnd the second part of
>your comment.

Anyone else?

>>>ju'a      UI       I state              evidential: non-specific basis
>>>ju'e      UI       I conclude           evidential: I conclude
>>Too close, I think.
>Chance happening.  If it causes a problem, we can consider a change.
>Nora thinks that if we make a change, now is a good time for it.

>>>ne'a      FAhA     next to              location tense relation/direction; next to ... ; direction modal
>>>ne'e      FAhA     edged by             location tense relation/direction; edged by ... ; direction modal
>>>ne'i      FAhA     within               location tense relation/direction; within/inside of ... ; direction modal
>>>ne'o      FAhA     adjacent to          location tense relation/direction; adjacent to/touching ... ; direction modal
>>>ne'u      FAhA     away from point      location tense relation/direction; directly away from specified point ... ; direction modal

>>Again, too close. This is not a distinctive cv'x family.
>Weeelll, yes and no. ne'a/ne'e/ne'o are intentionally closely related -
>they are fairly similar in meaning but sometimes the difference is
>important.  ne'e and ne'o were added in the last 6 months after my comparative
>language study and usage demanded them.
>Originally all there were were ne'a and ne'i - the latter clearly tied to
>nenri, and the former having a hint of vi/va, as well as the obvious mnemonic
>for engish speakers of NE'a  NExt to.  Then we added the other two, and
>finally it was pointed out that fa'anai did not mean away from point as in
>the published cmavo list, but merely not-towards.  With four of the five in this
>bunch in FAhA, we made it five.  So they are a family by design, though the
>semantic ties are less than grmmatical ones.

As Jim'll tell you, give the defining gismu explicitly, whether they match
the form of the FAhA or not.

>>>pi'u      JOI      cross product        non-logical connective: cross product; Cartesian product of sets
>>One of the least useful operators.  Some operators, Lojbab, have to lie
>>outside cmavo space.  How about this one?

>mi ce do pi'u la noras. ce la djein. dansu

>quite useful in linguistic non-mathematical contexts

Hm.

>>>vu'i      UI       virtue               attitudinal modifier: virtue - sin
>>>vu'inai   *        sin                  attitudinal modifier: virtue - sin
>>Change it to {vu'e}, in line with {vrude}
>Probably a good suggestion, since that is what the word came out to, but
>we have a strong bent against changing what has been published when
>people may actually have learned a word (I doubt it in this case, but we
>have a principle we try to follow).  This one will be considered, but don;t
>be surprised if inertia wins, since by happenstance the cmavo we chose came
>out closer to ideal than most.

>Given that Nora thinks these worth considering, please post your comments
>on ju'a/ju'e, the ne'a series, and this one for vu'i to lojban-list, along
>with my responses, for comment by the community.

I included a couple more; hope that was OK.

Anyone saying anything?