From pycyn@aol.com Mon Aug 12 13:54:45 2002 Return-Path: X-Sender: Pycyn@aol.com X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-8_0_7_4); 12 Aug 2002 20:54:45 -0000 Received: (qmail 41685 invoked from network); 12 Aug 2002 20:54:45 -0000 Received: from unknown (66.218.66.217) by m1.grp.scd.yahoo.com with QMQP; 12 Aug 2002 20:54:45 -0000 Received: from unknown (HELO imo-r06.mx.aol.com) (152.163.225.102) by mta2.grp.scd.yahoo.com with SMTP; 12 Aug 2002 20:54:44 -0000 Received: from Pycyn@aol.com by imo-r06.mx.aol.com (mail_out_v33.5.) id r.51.2247f927 (4402) for ; Mon, 12 Aug 2002 16:54:32 -0400 (EDT) Message-ID: <51.2247f927.2a897a88@aol.com> Date: Mon, 12 Aug 2002 16:54:32 EDT Subject: Re: [lojban] space tenses To: lojban@yahoogroups.com MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="part1_51.2247f927.2a897a88_boundary" X-Mailer: AOL 7.0 for Windows US sub 10509 From: pycyn@aol.com X-Yahoo-Group-Post: member; u=2455001 X-Yahoo-Profile: kaliputra --part1_51.2247f927.2a897a88_boundary Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit In a message dated 8/9/2002 6:14:06 PM Central Daylight Time, jjllambias@hotmail.com writes: << > The general way I understand BAIs is this: > > ko'a BAI broda = BAI ku ko'a broda = BAI zo'e ko'a broda >> > No problems so far << {mi pu klama} means {pu zo'e mi klama}, "in the past of the understood reference point (usually now) I go", i.e. "I went". >> This is about right right but the wording can cause trouble later on. If you read {pu zo'e} as "in the past of something" it is then easy to read {ba'o zo'e} as "in the perfective of something" and that will get matters backward. The significant thing is that {zo'e} is an axis (probably NOT the understood reference point, as sumti tcita seem to be used only with remote axes names -- a result of Lojban having no name for the understood axis) and all tense phenomena in Lojban are aligned with axes -- vectors depart from axes, larger structural features center on axes, and so on. So, {ba'o } makes the sumti name an axis on which the perfective centers --i.e., is in occurring at that axis -- and perfeective is of the event in the main clause to which the {ba'o} phrase attaches. (And it seems strictly to say "The prior I go", not "Priorly, I go" (((pu zo'e) mi) (klama VAU))-- but it is unclear how well the parser groupings reflect modification patterns.) << Unfortunately this analysis doesn't work for the way some people use {pu'o} and {ba'o} (see for example Nick's last post asking for proofreading. He wrote: {.i mi [...] cu ca'e cpedu lenu ro se cinri cu rore'u banlytroci pranygau lanli le cfacilre cmacku .e le velcli pu'o li 20 pi'e 9.} The way I read it he was asking for everyone to hold back on doing anything until september 20.) >> up, that is what he said -- he wanted {ba'o} instead of {pu'o}, or, at least, I assume he did: 20/9 whould be in the midst of the perfective aspect of of the editing. I presume he reads this as the editing should be in the inchoative aspect of 20/9, though the latter is not here presented as an event which has such aspects and the relation given is the reverse of all the other cases -- as you say. << But anyway, leaving aside those two exceptions pu'o and ba'o, the rest of the tags do have the same meaning as selbri tcita as they do as sumti tcita, don't they? >> Not exactly, but close. The selbri tcita always attach to the contextually given axis, the sumti tcita always attach to the axis named/described by the following sumti. If Lojban had words strictly for "here" and "now", selbri tcita would be just like sumti tcita but always with those words after. ({ca} and {bu'u} do not mean "now" and "here" but "simultaneous/coincident with the understood axis" -- 0-vectors. Of course, it is practically the same thing, but disallows some usages). << Space tenses in Lojban need a lot of discussion in my opinion. They are very messy. >> Not too bad, if you keep the axes in mind -- though the {mo'i} forms and even some of the stranger regular FAhA lke {ne'i} tend to leave me puzzled when used as selbri tcita (I don't think of axes as having insides). and later << Yes. In the case of the aspectuals it is not so much a placement relative to an axis as a selection of the phase we're concentrating on. But in any case it is a phase of the main event, not a phase of an event within the tagged sumti. >> It is still placement relative to an axis -- the phase the named axis is the center of (focus of? I don't want someone to say "Nah, it's a good two thrids of the way along in it"). But yes, it is the aspect of the main event, not the one introduced by the sumti tcita. << it has been the official interpretation, and there are even examples in the Book: {mi klama le zarci pu'o le nu mi citka} is interpreted as {mi klama le zarci ca le nu mi pu'o citka} instead of as {mi pu'o klama le zarci ca le nu mi citka}. As you say, the blurb tends to force that interpretation, but it is a weird usage if you analyze it carefully. >> Well, it is a perfectly good interpretation except that it is at odds with virtually all the other tenses uses in the book. All of the ones that anyone has a clear sense of for sure. The whole system could have been set up that way, but setting it up as it is and then shifting in these two cases (and I am sure there are others here and there) seems very odd in deed (though easy to understand in retrospect how it happened). << The most intense discussions on this, as I remember, I had with Lojbab, and they were before you joined the list, and also before the Book was published. Obviously I did not manage to convince him. >> I was probably involved third hand or so, and, at a guess, as confused then as I was a couple of days ago. << I use it to introduce the time when the overshooting is produced, the point when the "natural end" of the event should have taken place. For example: {ko'a zvati le purdi za'o le nu co'a carvi}, "he keeps being in the garden as it begins to rain", i.e. a kind of {za'o zvati}. >> That is a convenient way of doing things, I admit, and I think I came down for it at one time or another. BUT it is using Nick's {pu'o} rule, rather than the usual one again. The natural end is not in the superfective area of an event, so introducing it as an axis, makes a false claim. It is the later point, where it is clear he has overshot, that correctly indicates that. He doesn't keep on running at the mile post, his natural goal; he keeps on running at the mile and a furlong or whatever, that is in the overshot, even if he stops there. (But we do need a good way to get the natural end in and we don't need the overshot axis near as often -- thence the slippery slope to using {pu'o} and {ba'o} backwards -- ably aided by the similarlities to {pu} and {ba}, of course). and & << {mi klama le zarci pu'o le nu mi citka} I go to the shop in the runup to my eating. {mi klama le zarci ca le nu mi pu'o citka} I go to the shop at the time of the runup to my eating. {mi pu'o klama le zarci ca le nu mi citka}. It is the runup to my going, at the time of my eating, to the store >> 1ought to be "when I'm eating, I am about to go to the store," like 3 but not 2. The axis is my eating, the aspect is of the main clause event. << Oh, I see: the issue is whether pu'o means "runup to" or "inchoative" or "until", since all 3 are different but equally sanctioned by the ma'oste, just like ba'o (aftermath v. since v. perfective). >> I'm not sure that this gets the problem, though it gets some of it. The aspect, which is focused on the subordinate event, is of the main clause event. So, part of it is the reading "until, in anticipation of" that makes it look like an aspect of the sumti'd event rather than the main. It seems the "run up to" also makes it look like the aspect is of the sumti'd event rather than the main. But in fact, it works just like {mi klama ba le nu me citka}: the going is after the eating, not the other way around. So here the anticipating is for the going, not the eating. xorxes again << I start from the premise that: broda = ku broda = broda (these might be strict equalities or approximations, it doesn't matter, but they surely hold at least roughly for all tags except for the official interpretation of {pu'o broda} and {ba'o broda}. These should be more precise forms of {pu'o broda} and {ba'o broda}, but they get more or less the opposite meaning. >> The first two are strict, the third analogical, but that {pu'o} violates in in a number of honored cases is correct and suggests that those cases are wrong. Not a lot here about space tenses, but it is likely that cases like the {pu'o} have slipped into those, too, fgetting lefts and rights mixed up along with whose left and right (stage directions model). --part1_51.2247f927.2a897a88_boundary Content-Type: text/html; charset="US-ASCII" Content-Transfer-Encoding: 7bit In a message dated 8/9/2002 6:14:06 PM Central Daylight Time, jjllambias@hotmail.com writes:

<<
The general way I understand BAIs is this:

   ko'a BAI broda = BAI ku ko'a broda = BAI zo'e ko'a broda

>>
No problems so fa
r


<<
{mi pu klama} means {pu zo'e mi klama}, "in the past of the
understood reference point (usually now) I go", i.e. "I went".
>>

This is about right right but the wording can cause trouble later on.  If you read {pu zo'e} as "in the past of  something" it is then easy to read {ba'o zo'e} as "in the perfective of something" and that will get matters backward.  The significant thing is that {zo'e
} is an axis (probably NOT the understood reference point, as sumti tcita seem to be used only with remote axes names -- a result of Lojban having no name for the understood axis) and all tense phenomena in Lojban are aligned with axes -- vectors depart from axes, larger structural features center on axes, and so on. So, {ba'o <sumti>} makes the sumti name an axis on which the perfective centers --i.e., is in occurring at that axis -- and perfeective is of the event in the main clause to which the {ba'o} phrase attaches.
(And it seems strictly to say "The prior I go", not "Priorly, I go"  (((pu zo'e) mi) (klama VAU))-- but it is unclear how well the parser groupings reflect modification patterns.)

<<
Unfortunately this analysis doesn't work for the way some people
use {pu'o} and {ba'o} (see for example Nick's last post asking
for proofreading. He wrote: {.i mi [...] cu ca'e cpedu lenu ro
se cinri cu rore'u banlytroci pranygau lanli le cfacilre cmacku
.e le velcli pu'o li 20 pi'e 9.} The way I read it he was asking
for everyone to hold back on doing anything until september 20.)
>>

up, that is what he said -- he wanted {ba'o} instead of {pu'o}, or, at least, I assume he did: 20/9 whould be in the midst of the perfective aspect of of the editing. I presume he reads this as the editing should be in the inchoative aspect of 20/9, though the latter is not here presented as an event which has such aspects and the relation given is the reverse of all the other cases -- as you say.

<<
But anyway, leaving aside those two exceptions pu'o and ba'o,
the rest of the tags do have the same meaning as selbri tcita
as they do as sumti tcita, don't they?
>>
Not exactly, but close.  The selbri tcita always attach to the contextually given axis,  the sumti tcita always attach to the axis named/described by the following sumti.  If Lojban had words strictly for "here" and "now", selbri tcita would be just like sumti tcita but always with those words after.  ({ca} and {bu'u} do not mean "now" and "here" but "simultaneous/coincident with the understood axis" -- 0-vectors.  Of course, it is practically the same thing, but disallows some usages).  

<<
Space tenses in Lojban need a lot of discussion in my opinion.
They are very messy.
>>
Not too bad, if you keep the axes in mind -- though the {mo'i} forms and even some of the stranger regular FAhA lke {ne'i} tend to leave me puzzled when used as selbri tcita (I don't think of axes as having insides).

and later

<<
Yes. In the case of the aspectuals it is not so much a placement
relative to an axis as a selection of the phase we're concentrating
on. But in any case it is a phase of the main event, not a phase
of an event within the tagged sumti.
>>
It is still placement relative to an axis -- the phase the named axis is the center of (focus of?  I don't want someone to say "Nah, it's a good two thrids of the way along in it").  But yes, it is the aspect of the main event, not the one introduced by the sumti tcita.

<<
it has
been the official interpretation, and there are even examples
in the Book: {mi klama le zarci pu'o le nu mi citka} is
interpreted as {mi klama le zarci ca le nu mi pu'o citka}
instead of as {mi pu'o klama le zarci ca le nu mi citka}.
As you say, the blurb tends to force that interpretation,
but it is a weird usage if you analyze it carefully.
>>
Well, it is a perfectly good interpretation except that it is at odds with virtually all the other tenses uses in the book.  All of the ones that anyone has a clear sense of for sure.   The whole system could have been set up that way, but setting it up as it is and then shifting in these two cases (and I am sure there are others here and there) seems very odd in deed (though easy to understand in retrospect how it happened).

<<
The most intense discussions on this, as I remember, I had with
Lojbab, and they were before you joined the list, and also before
the Book was published. Obviously I did not manage to convince
him.
>>
I was probably involved third hand or so, and, at a guess, as confused then as I was a couple of days ago.

<<
I use it to introduce the time when the overshooting is produced,
the point when the "natural end" of the event should have taken place.
For example: {ko'a zvati le purdi za'o le nu co'a carvi}, "he keeps
being in the garden as it begins to rain", i.e. a kind of
{za'o zvati}.
>>

That is a convenient way of doing things, I admit, and I think I came down for it at one time or another.  BUT it is using Nick's {pu'o} rule, rather than the usual one again.  The natural end is not in the superfective area of an event, so introducing it as an axis, makes a false claim.  It is the later point, where it is clear he has overshot, that correctly indicates that.  He doesn't keep on running at the mile post, his natural goal; he keeps on running at the mile and a furlong or whatever, that is in the overshot, even if he stops there. (But we do need a good way to get the natural end in and we don't need the overshot axis near as often -- thence the slippery slope to using {pu'o} and {ba'o} backwards -- ably aided by the similarlities to {pu} and {ba}, of course).

and &
<<
{mi klama le zarci pu'o le nu mi citka}

I go to the shop in the runup to my eating.

{mi klama le zarci ca le nu mi pu'o citka}

I go to the shop at the time of the runup to my eating.

{mi pu'o klama le zarci ca le nu mi citka}.

It is the runup to my going, at the time of my eating, to the store
>>

1ought to be "when I'm eating, I am about to go to the store," like 3 but not 2. The axis is my eating, the aspect is of the main clause event.

<<
Oh, I see: the issue is whether pu'o means "runup to" or "inchoative"
or "until", since all 3 are different but equally sanctioned by the
ma'oste, just like ba'o (aftermath v. since v. perfective).
>>

I'm not sure that this gets the problem, though it gets some of it.  The aspect, which is focused on the subordinate event, is of the main clause event. So, part of it is the reading "until, in anticipation of" that makes it look like an aspect of the sumti'd event rather than the main.  It seems the "run up to" also makes it look like the aspect is of the sumti'd event rather than the main.  But in fact, it works just like {mi klama ba le nu me citka}: the going is after the eating, not the other way around.  So here the anticipating is for the going, not the eating.

xorxes again
<<
I start from the premise that:

   <tag> broda = <tag>ku broda = <tag> <sumti> broda

(these might be strict equalities or approximations, it
doesn't matter, but they surely hold at least roughly for
all tags except for the official interpretation of
{pu'o <sumti> broda} and {ba'o <sumti> broda}. These should
be more precise forms of {pu'o broda} and {ba'o broda}, but
they get more or less the opposite meaning.
>>
The first two are strict, the third analogical, but that {pu'o} violates in in a number of honored cases is correct and suggests that those cases are wrong.

Not a lot here about space tenses, but it is likely that cases like the {pu'o} have slipped into those, too, fgetting lefts and rights mixed up along with whose left and right (stage directions model).
--part1_51.2247f927.2a897a88_boundary--