Hmmm... I see your point, but I'm not sure using multiple ZAhO in a single
tense gives a meaningful tense, at least in CLL Lojban. I would interpret
that as, if anything, "At some glorked point in time, something both is
having broda-ed and is at the point of completion of broda-ing" - which
makes no sense.
> > - for states (and activities?), {le nu broda cu mulno} means something
> > along the lines of broda being as true as possible during the
> > state/activity.
>
> Yes, I suppose that it has to work for activities too:
>
> le nu mi bajra cu mulxadba ze'a le pamoi mentu gi'eku'i
> mulno ze'a le drata temci
> My running was half-hearted for the first minutes, but
> to the full for the rest of the time.
>
'K. Though again, your non-CLL use of tense seems strange to me, however
useful.
> > - for point-events...? Both interpretations make sense.
>
> Do you have an example?
No. But only because I don't really understand point events, except as
another way of looking at states/activities/processes - which is why I
said both the state and process version make sense.
>
> > - for objects, mulno's other, three-place structure applies.
>
> I suppose we could say that for events the full place structure
> applies too, with x2 being {le ka ce'u fasnu}:
>
> le nu mi klama le zarci cu mulno le ka fasnu}
> My going to the market is complete in its occurring.
>
> le nu mi gleki cu mulno le ka fasnu
> My being happy is complete in its occurring.
>
Sounds reasonable. Alternatively you could take the event definition to be
primitive, and define the object one in terms of it:
{ke'a mulno le ka [ce'u] broda} :<==> {le nu ke'a broda cu mulno}
But then we have the problem of whether {le nu ke'a broda} is a state or
a process - {mi ca mulno le ka jmive} could be either "I have completed
the process of living [and am dead]" or "I am completely alive".
And if not that, then how in general to interpret the x2 of mulno?
mu'o mi'e maten.
Let's see.
{mulno} is not problematic when the x1 involves a change of state.
{le nu klama cu mulno} says that the going is complete, i.e. that
the goer has reached the destination. Similarly {le nu binxo
cu mulno} and {le nu zbasu cu mulno} are clear.
The problem appears when x1 is a state. Does {le nu mi gleki
cu mulno} mean that my happiness is complete, or that I'm no
longer happy? States don't have a temporal completion point,
so a state is complete when it is in some sense saturated,
if I could not be happier then my happiness is complete.
That is what I take {mulno} to mean when x1 is a state.
Now, as to {co'u} and {mo'u}.
{sisti} does not quite correspond to {co'u}, because it has
an x1. The correspondence would be co'a-cfari, co'u-tolcfa
mi co'a gleki -- le nu mi gleki cu cfari
mi co'u gleki -- le nu mi gleki cu tolcfa
{mulno} does not describe a change, so it corresponds to {ba'o}
rather than to {mo'u}, for events that involve changes:
mi ba'o klama le zarci -- le nu mi klama le zarci cu mulno
{mo'u} corresponds (for change events) to {mulbi'o}:
mi mo'u klama le zarci -- le nu mi klama le zarci cu mulbi'o
States don't have a natural ending point, so {mi mo'u gleki}
is not that meaningful. I wouldn't mind interpreting it as
"I achieve complete happiness", though.
mu'o mi'e xorxes
__________________________________
Do you Yahoo!?
The New Yahoo! Shopping - with improved product search
http://shopping.yahoo.com
mi'e la xod na.e la xorxes u'i
> > Perhaps {lo nu broda cu mulno} <==> {ba'o mo'u broda}, since the
> > LHS is tenseless so the RHS should be tenseless too.
>
> Hmmm... I see your point, but I'm not sure using multiple ZAhO in a single
> tense gives a meaningful tense, at least in CLL Lojban.
It is grammatical, why would it not be meaningful?
CLL doesn't mention multiple ZAhO explicitly, but it does
have a multiple number ROI example, which follows the same
principle:
>>Note the difference between:
>>
>>10.13) mi pare'u paroi klama le zarci
>> I [first time] [one time] go-to the store.
>> For the first time, I go to the store once.
>>
>>and
>>10.14) mi paroi pare'u klama le zarci
>> I [one time] [first time] go-to the store.
>> There is one occasion on which I go to
>> the store for the first time.
So {ba'o mo'u broda} would be the aftermath of the completion
of brodaing, and {mo'u ba'o broda} would be the completion of the
aftermath, if that ever makes sense.
> I would interpret
> that as, if anything, "At some glorked point in time, something both is
> having broda-ed and is at the point of completion of broda-ing" - which
> makes no sense.
That might be more like {ba'o je mo'u}.
> > > - for states (and activities?), {le nu broda cu mulno} means something
> > > along the lines of broda being as true as possible during the
> > > state/activity.
> >
> > Yes, I suppose that it has to work for activities too:
> >
> > le nu mi bajra cu mulxadba ze'a le pamoi mentu gi'eku'i
> > mulno ze'a le drata temci
> > My running was half-hearted for the first minutes, but
> > to the full for the rest of the time.
> >
>
> 'K. Though again, your non-CLL use of tense seems strange to me, however
> useful.
What's non-CLL about that?
CLL:
>>12.12) loi snime cu carvi
>> ze'u le ca dunra
>> some-of-the-mass-of snow rains
>> [long time interval] the [present] winter.
>> Snow falls during this winter.
> > > - for point-events...? Both interpretations make sense.
> >
> > Do you have an example?
>
> No. But only because I don't really understand point events, except as
> another way of looking at states/activities/processes - which is why I
> said both the state and process version make sense.
Right.
> > > - for objects, mulno's other, three-place structure applies.
> >
> > I suppose we could say that for events the full place structure
> > applies too, with x2 being {le ka ce'u fasnu}:
> >
> > le nu mi klama le zarci cu mulno le ka fasnu}
> > My going to the market is complete in its occurring.
> >
> > le nu mi gleki cu mulno le ka fasnu
> > My being happy is complete in its occurring.
>
> Sounds reasonable. Alternatively you could take the event definition to be
> primitive, and define the object one in terms of it:
>
> {ke'a mulno le ka [ce'u] broda} :<==> {le nu ke'a broda cu mulno}
>
> But then we have the problem of whether {le nu ke'a broda} is a state or
> a process - {mi ca mulno le ka jmive} could be either "I have completed
> the process of living [and am dead]" or "I am completely alive".
>
> And if not that, then how in general to interpret the x2 of mulno?
Indeed {le nu mi jmive cu ca mulno}, "my life is now full/finished" is
ambiguous, too, depending on whether we are thinking of life as a
process or as a state. (The gi'uste definition of jmive suggests the
state as the basic interpretation, though.)
> So can that be summarised as -
> - whenever {le nu broda} is of an event type which has a mo'u (i.e a
> process), {le nu broda cu mulno} <==> {pu mo'u broda} ==> {ba'o
> broda}
Perhaps {lo nu broda cu mulno} <==> {ba'o mo'u broda}, since the
LHS is tenseless so the RHS should be tenseless too.
> - for states (and activities?), {le nu broda cu mulno} means something
> along the lines of broda being as true as possible during the
> state/activity.
Yes, I suppose that it has to work for activities too:
le nu mi bajra cu mulxadba ze'a le pamoi mentu gi'eku'i
mulno ze'a le drata temci
My running was half-hearted for the first minutes, but
to the full for the rest of the time.
> - for point-events...? Both interpretations make sense.
Do you have an example?
> - for objects, mulno's other, three-place structure applies.
I suppose we could say that for events the full place structure
applies too, with x2 being {le ka ce'u fasnu}:
le nu mi klama le zarci cu mulno le ka fasnu}
My going to the market is complete in its occurring.
le nu mi gleki cu mulno le ka fasnu
My being happy is complete in its occurring.
mu'o mi'e xorxes
I would take mulno to mean "finished" & hence no state can be
a mulno. IOW either {le nu mi gleki cu mulno} is false, or
le nu mi gleki is not a state but a finishable event.
--And.
> How would you say "x1 consoles/comforts x2", in Lojban?
>
> I have used {surgau} for this, but I'm not convinced
> this is quite it. Any better ideas?
toldrigau?
mu'o mi'e .filip.
--
filip.niutyn. <pne...@gmx.de>
si faurmu'o
I have used {surgau} for this, but I'm not convinced
this is quite it. Any better ideas?
mu'o mi'e xorxes
__________________________________
Do you Yahoo!?
The New Yahoo! Shopping - with improved product search
http://shopping.yahoo.com
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
Printer at MyInks.com. Free s/h on orders $50 or more to the US & Canada.
http://www.c1tracking.com/l.asp?cid=5511
http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/GSaulB/TM
---------------------------------------------------------------------~->
To unsubscribe, send mail to lojban-un...@onelist.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
> la maten cusku di'e
>
> > How about {dristigau}? Or {drimulgau}, which might have different
> > connotations?
>
> I like {dristigau}. Also maybe {drijdikygau} or {drivi'u}.
>
> But {drimulgau} I would interpret as making someone completely sad.
> {mulno} is somewhat ambiguous.
Hmmm, that's true. Personally I think overloading gismu was a mistake,
but I guess it's one we'll have to live with.
So what brivla unambiguously corresponds to mo'u, in the way sisti does to
co'u?
I guess a lujvo along the lines of faurmo'u would do.
mu'o
> On Tue, 7 Oct 2003, Philip Newton wrote:
> >
> > toldrigau?
>
> Maybe, but "console" isn't really {to'e badri gasnu}, nor is it quite
> {to'e ke badri gasnu}.
>
> How about {dristigau}? Or {drimulgau}, which might have different
> connotations?
I like {dristigau}. Also maybe {drijdikygau} or {drivi'u}.
But {drimulgau} I would interpret as making someone completely sad.
{mulno} is somewhat ambiguous.
ki'e mi'e xorxes
> I would take mulno to mean "finished" & hence no state can be
> a mulno. IOW either {le nu mi gleki cu mulno} is false, or
> le nu mi gleki is not a state but a finishable event.
That would make the meaning of mulno more precise, but the
gi'uste seems to give it a much broader meaning, more like
"not lacking any of its constituents", so that for processes
it means all stages have been actualized, but for states
it would mean that the state is completely actualized,
not a marginal or dubious "incomplete" case.
If mulno does have the general meaning I read from the gi'uste,
it would be useful to have lujvo for the more restricted and
precise meanings, of course. Perhaps {rucmu'o}, process-complete,
for the mo'u sense.
mu'o mi'e xorxes
> On 7 Oct 2003 at 8:31, Llamb�as wrote:
>
> > How would you say "x1 consoles/comforts x2", in Lojban?
> >
> > I have used {surgau} for this, but I'm not convinced
> > this is quite it. Any better ideas?
>
> toldrigau?
Maybe, but "console" isn't really {to'e badri gasnu}, nor is it quite
{to'e ke badri gasnu}.
How about {dristigau}? Or {drimulgau}, which might have different
connotations?
mu'o mi'e maten.
OK, that's clear, and it looks like I'd managed to misread the
definitions of both mulno *and* sisti.
So can that be summarised as -
- whenever {le nu broda} is of an event type which has a mo'u (i.e a
process), {le nu broda cu mulno} <==> {pu mo'u broda} ==> {ba'o
broda}
- for states (and activities?), {le nu broda cu mulno} means something
along the lines of broda being as true as possible during the
state/activity.
- for point-events...? Both interpretations make sense.
- for objects, mulno's other, three-place structure applies.
mu'o