carna, my turn.

[Robin Lee Powell]

(Jesus fuck this is long; it's *really* not as bad as it looks.
It's certainly nowhere near as complicated as, say, the tense
system.)

This is my attempt to synthesize the thinking so far into an
alternate definition for {carna} that shouldn't invalidate any
current usage.

Note that between this and the problem we've had with {farna}, I'm
inclined to write up a full tutorial on directions, distances,
locations, vectors, and so on, add it to the CLL, and have these
words just have the basic definition and say "See chapter N". But
here it is anyway.

The basic mechanism here, in programming terms, is polymorphism in
the x2 and x4 places.

Since this *looks* crazy complicated, I'm going to start with
examples, to show that what comes out is quite reasonable.

{mi carna fi lo pritu}
I turn towards something to my right. The distance is
unspecified and left up to context; it could mean you spin in
place many times, or that you turn just enough to face the house
on your right, or anything in between.

{mi carna fi lo berti}
I turn towards the north [again, distance unspecified].

{mi carna (mi/zo'e) lo pritu (li pi re mu/lo julra'o be li so
no/lo radno be li pi mu pi'i pai)}
I turn a quarter turn to the right.

{do carna do lo pritu li pi re mu}
You turned a quarter turn to your right.

{mi carna fo lo pritu}
I turn in some direction unspecified such that I end up facing
something to my right (i.e. that was on my right before I started
turning). Since minimal distance is assumed, this is the shortest
way to say "I turn right" with almost no risk of being
misunderstood.

{mi carna fi lo pritu lo pritu}
I turn to the right. This is the same as the previous example,
just being extra clear.

{do carna mi lo pritu}
From my point of view, you rotated clockwise [which, assuming two
standing humans facing each other, implies a cartwheel or similar]

{do carna do lo pritu be mi}
You turned towards something on my right.

{do carna le zunle tupyjanco be fi do ce'o le pritu tupyjanco be
fi do le zunle be mi li pi mu}
You rotated over your hips towards something to my left, one half
a turn, i.e. you are now standing on your hands or something. A
situation that would make this perfectly reasonable is if you're
doing a flip into a handstand, and I'm your spotter, and I'm
standing to your right. Note that in this case, because the thing
being turned towards isn't relative to the x1, the direction of
the x2 vector isn't important. For the same reason, we don't know
if the gymnast did a front or back flip.

{do carna le zunle tupyjanco be fi do ce'o le pritu tupyjanco be
fi do lo zunle li pi mu}
Same thing, but now we know for certain that it was a back flip.
If you were willing to argue that the line through a person's hips
has a natural front, you could say {lo gapru} instead of {lo
zunle}, because if you imagine a person pointing an arm out of the
middle of their crotch, a back flip is equivalent to that arm
moving upwards.

{do carna mi lo zunle li pi mu}
If I'm your spotter and I'm on your right and you are sideways to
me facing to my right, this means exactly the same thing as the
previous sentence, and is what would be used in almost every case.

{le selkei cu carna fo so'i da}
The toy is spinning.

{le selkei cu carna mi lo pritu so'i da}
The toy is spinning clockwise from my point of view.

{le selkei cu carna (lo gapru/le drudi) lo pritu so'i da}
The toy is spinning clockwise if viewed from above.

{le selkei cu carna le loldi ce'o le drudi lo pritu so'i da}
The toy is spinning clockwise if viewed from above.

{le selkei cu carna le drudi ce'o le loldi lo pritu so'i da}
The toy is spinning clockwise if viewed from below.

And here's the actual definition plus notes/tutorial. This really
is *fundamentally* complex, as we've discovered in this discussion.
My previous version, which relied on the right hand rule, was very
easy to define, but led to *horrendous* sentences that were
extremely counter-intuitive. In this version, I've sacrificed some
initial learning curve for very easy use once you get used to it.

x1 turns about vector x2 towards direction x3, turning angular
distance / to face point x4

Notes: When reading these notes please give allowances for the
fact that facing in three dimensions is inherently complicated,
but note that the defaults make this actually very easy to use.
Each subsequent section gets more general and less likely to be
needed in practice.

Let's start with an x1 object that has both a top and a front,
i.e. you yourself. Your default x2 is a vector running along your
body from the bottom to point out of the top of their head.
Imagine that you start pointing your hand in some direction, and
then move your pointing hand towards the x3. The movement
describes a circle around you, perpendicular to the x2 vector;
we'll reference that later. For this case in the x2 you could use
{zo'e}, or {mi}, or just skip it entirely; all of those options
are shorthand for "the vector running through me pointing out the
top of my head". The x3 is anything you can point at. We'll
cover x4 seperately.

This holds for any object with a top and a front; by default, if
the x2 is {zo'e} or the object itself, the x2 vector is from the
bottom pointing topwards, and the turning is from the front of the
object as though the object was a person describing eir own
movement.

If the object has a top but not the front, it's basically the
same, you just can't specify anything but {lo pritu} or {lo zunle}
for the x3. If it has a front you could talk about the front
turning "towards the north" or whatever, but with no natural front
you can only talk about the direction of rotation in general
terms.

Note that in all such cases, which is things like "x1 carna x1 lo
pritu", these are equivalent to "x1 rotates clockwise if viewed
from above", and similarily for "lo zunle" and
"counter-clockwise".

If the x2 is a single object or place other than the x1, then draw
the vector from the x1 to the x2. This leads to very natural ways
to talk about something you're looking at moving clockwise or
whatever; {le junla grana cu carna mi lo pritu} means "if the clock was
a person, with its head pointing towards me and its clock hand
sticking out in front of it, it would say that it is moving its
clock hand to the right", which happens to be what "clockwise"
is, as with viewing a person from above in the {do carna do lo
pritu} case.

Similarily, {le selkei cu carna lo gapru lo zunle} is "the toy
[perhaps a top] is spinning counter-clockwise when viewed from
above", because that's what it would look like if a human was in
place of that top, and standing up (i.e. head pointing to {lo
gapru}) and turning to what they called {lo zunle}.

For maximum specificity, you can specify the vector exactly, by
using "ce'o"; the vector points from the first element of the
"ce'o" to the last. This is unlikely to be useful very often.

The x4 is one of a number, in which case it is the (possibly
fractional) number of complete turns made; an exact specification
of radians or degrees or whatever (i.e. {lo julra'o li so no}), in
which case that's how far the turn went; or an object or place, in
which case the turn was the minimum distance required to end with
the x1 facing that object or place. That last is not OK if the x1
has no natural front.

Unless clearly specified otherwise, the x3 and the x4 from the point of
view of the circle of movement perpendicular to the x2 vector, as
everything was positioned before the movement started. This is
almost always the same as saying that the x3 and x4 are from the
point of view of the x1 before the movement started.

--
http://singinst.org/ : Our last, best hope for a fantastic future.
.i ko na cpedu lo nu stidi vau loi jbopre .i danfu lu na go'i li'u .e
lu go'i li'u .i ji'a go'i lu na'e go'i li'u .e lu go'i na'i li'u .e
lu no'e go'i li'u .e lu to'e go'i li'u .e lu lo mamta be do cu sofybakni li'u

[Jonathan Jones]

I think that's the best idea for fixing carna I've seen yet. I'm certainly for making this change. I do think the actual wording of the definition could use a little polish, but that's my only contention.
 

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.

--
mu'o mi'e .aionys.

.i.e'ucai ko cmima lo pilno be denpa bu .i doi.luk. mi patfu do zo'o
(Come to the Dot Side! Luke, I am your father. :D )

[Robin Lee Powell]

On Sun, Aug 26, 2012 at 03:02:00PM -0600, Jonathan Jones wrote:
> On Sun, Aug 26, 2012 at 2:35 PM, Robin Lee Powell <
> rlpo...@digitalkingdom.org> wrote:
>
> >
> > And here's the actual definition plus notes/tutorial. This really
> > is *fundamentally* complex, as we've discovered in this discussion.
> > My previous version, which relied on the right hand rule, was very
> > easy to define, but led to *horrendous* sentences that were
> > extremely counter-intuitive. In this version, I've sacrificed some
> > initial learning curve for very easy use once you get used to it.
> >
> > x1 turns about vector x2 towards direction x3, turning angular
> > distance / to face point x4
> >
>
> I think that's the best idea for fixing carna I've seen yet. I'm certainly
> for making this change. I do think the actual wording of the definition
> could use a little polish, but that's my only contention.

Feel free to suggest particular polishings.

-Robin

[Pierre Abbat]

On Sunday, August 26, 2012 13:35:14 Robin Lee Powell wrote:

> And here's the actual definition plus notes/tutorial. This really
> is *fundamentally* complex, as we've discovered in this discussion.
> My previous version, which relied on the right hand rule, was very
> easy to define, but led to *horrendous* sentences that were
> extremely counter-intuitive. In this version, I've sacrificed some
> initial learning curve for very easy use once you get used to it.
>
> x1 turns about vector x2 towards direction x3, turning angular
> distance / to face point x4

Suppose that x1 is an object with a natural axis of rotation, but no natural
top or front, such as an axle with a wheel on each end (which I'll call "lo
relxiljendu"). What does "lo relxiljendu cu carna ri lo pritu" mean?

Pierre
--
Don't buy a French car in Holland. It may be a citroen.

[Robin Lee Powell]

I would say that it's meaningless, and I don't see how you could
make that make sense in any language.

Without knowing the point of view from which I'm looking at the
axle, how can I say that it's rotating clockwise, in English? As
far as I know I can't. So the x2 needs to be {mi} or something
else.

-Robin

[selpa'i]

Am 26.08.2012 22:35, schrieb Robin Lee Powell:
> x1 turns about vector x2 towards direction x3, turning angular
> distance / to face point x4

How can an angular distance and a point faced be the same sumti place?

mu'o mi'e la selpa'i

--
pilno zo le xu .i lo dei bangu cu se cmene zo lojbo .e nai zo lejbo

^:i \jl /flr sen |ziu \su xn go kror
^:i \sym tfn /zu viw \xn jy ^jaiw

[v4hn]

On Tue, Aug 28, 2012 at 09:05:46PM +0200, selpa'i wrote:
> Am 26.08.2012 22:35, schrieb Robin Lee Powell:
> > x1 turns about vector x2 towards direction x3, turning angular
> > distance / to face point x4
>
> How can an angular distance and a point faced be the same sumti place?

On Sun, Aug 26, 2012 at 01:35:14PM -0700, Robin Lee Powell wrote:
> The basic mechanism here, in programming terms, is polymorphism in
> the x2 and x4 places.

That's just what polymorphism means...

The idea is that you can still disambiguate any sentence with {carna},
if the set of sumtis referring to possible angular distances
and the set of sumtis referring to possible points in space you can face
are always disjunct.

Is this concept already used somewhere in lojban?


mu'o

[Robin Lee Powell]

On Tue, Aug 28, 2012 at 09:05:46PM +0200, selpa'i wrote:

> Am 26.08.2012 22:35, schrieb Robin Lee Powell:
> > x1 turns about vector x2 towards direction x3, turning angular
> > distance / to face point x4
>
> How can an angular distance and a point faced be the same sumti
> place?

Polymorphism; they don't look remotely the same, so the word does
different things in the different cases. Look at the examples. No
confusion can result, as far as I can tell. And, in fact, it can
also take number of revolutions.

-Robin

[Robin Lee Powell]

I believe there are other sumti that work something like that, but
I'm not remembering any examples off the top of my head.

Briefly poking at the gismu list, ragve, danfu, zanru. Others, I'm
sure.

-Robin

[la gleki]

I wonder how selpa'i is gonna translate

"Rotate the screen 90 deg. clockwise"
"Rotate the screen 180 deg. clockwise" 

Using Robin's new carna it's pretty simple.

[v4hn]

On Tue, Aug 28, 2012 at 10:10:14PM -0700, Robin Lee Powell wrote:
> On Tue, Aug 28, 2012 at 10:28:41PM +0200, v4hn wrote:
> > The idea is that you can still disambiguate any sentence with
> > {carna}, if the set of sumtis referring to possible angular
> > distances and the set of sumtis referring to possible points in
> > space you can face are always disjunct.
> >
> > Is this concept already used somewhere in lojban?
>
> I believe there are other sumti that work something like that, but
> I'm not remembering any examples off the top of my head.
>
> Briefly poking at the gismu list, ragve, danfu, zanru. Others, I'm
> sure.

None of these is really polymorphic as far as I can see.

The only scene I can think of, that might cause an overlap, is the following:

You sit in front of a number line and want to say
"I turn to face the 3".

Could this theoretically be translated as {mi carna fo li ci}?
(I know there are better ways of saying this. I'm just looking for an
edge case of the Robins definition)
Because if it can, then we got a sentence with two rather different meanings:
- I turn to face the 3
- I turn around three times


mu'o

[Robin Lee Powell]

On Wed, Aug 29, 2012 at 11:25:36AM +0200, v4hn wrote:
> On Tue, Aug 28, 2012 at 10:10:14PM -0700, Robin Lee Powell wrote:
> > On Tue, Aug 28, 2012 at 10:28:41PM +0200, v4hn wrote:
> > > The idea is that you can still disambiguate any sentence with
> > > {carna}, if the set of sumtis referring to possible angular
> > > distances and the set of sumtis referring to possible points in
> > > space you can face are always disjunct.
> > >
> > > Is this concept already used somewhere in lojban?
> >
> > I believe there are other sumti that work something like that, but
> > I'm not remembering any examples off the top of my head.
> >
> > Briefly poking at the gismu list, ragve, danfu, zanru. Others, I'm
> > sure.
>
> None of these is really polymorphic as far as I can see.
>
> The only scene I can think of, that might cause an overlap, is the following:
>
> You sit in front of a number line and want to say
> "I turn to face the 3".
>
> Could this theoretically be translated as {mi carna fo li ci}?

No; that needs to be {me'o ci}.

-Robin

[selpa'i]

Am 29.08.2012 07:35, schrieb la gleki:
> I wonder how selpa'i is gonna translate
> "Rotate the screen 90 deg. clockwise"
> "Rotate the screen 180 deg. clockwise"

How about:
[ko gasnu lo nu] lo vidni cufi'u vo si'e carna lo pritu
[Make it so that] the screen turns a quarter in clockwise direction.

The x2 of si'e would be used to refer to a single complete rotation,
which I would understand to be the default value for si'e2 when used
with carna. The default could be defined as "PA si'e lo carlai be li
pa", so that PA tells us how many rotations (or what fraction of a
rotation) occur.

[selpa'i]

Excuse the typo:

Am 30.08.2012 16:56, schrieb selpa'i:
> [ko gasnu lo nu] lo vidni cufi'u vo si'e carna lo pritu

It should say:
[ko gasnu lo nu] lo vidni cu fi'u vo si'e carna fi lo pritu

[la gleki]

On Thursday, August 30, 2012 6:57:02 PM UTC+4, selpa'i wrote:

Am 29.08.2012 07:35, schrieb la gleki:
> I wonder how selpa'i is gonna translate
> "Rotate the screen 90 deg. clockwise"
> "Rotate the screen 180 deg. clockwise"

How about:
[ko gasnu lo nu] lo vidni cufi'u vo si'e carna lo pritu
[Make it so that] the screen turns a quarter in clockwise direction.

The x2 of si'e would be used to refer to a single complete rotation,
which I would understand to be the default value for si'e2 when used
with carna. The default could be defined as "PA si'e lo carlai be li
pa", so that PA tells us how many rotations (or what fraction of a
rotation) occur.

si'e says nothing about full rotation. It's {carna} that must do the job.

And if we assume that {carna} imples one full revolution by default then... well...

[Sebastian Fröjd]

well, it seems useful, but still I'm rather confused.

How do you express the following examples with the new definition?

1. He did a 900° (a skateboard trick, rotation to the right in the transverse plane, person joi skateboard (xilta'o) same rotation)

2. The car rotates around me

3. The car is spinning (around its own axis) and rotates around the moose (on the left side of the moose)

4. She did a back flip (axis her own head, rotation feet to the front) and landed lying face down

5. Tony kickflips his board 90° left in the sagittal plane, and at the same time rotates his body 180° to the right in the coronal plane

Is it even possible to describe movements like this in a convenient way with the new definition?

What's the difference between gunro and carna, except that gunro needs a surface in the 2nd place?

mu'omi'e jongausib

2012/9/2 la gleki <gleki.is...@gmail.com>

--
You received this message because you are subscribed to the Google Groups "lojban" group.

To view this discussion on the web visit https://groups.google.com/d/msg/lojban/-/su_9XKeKKNEJ.

[Sebastian Fröjd]

also see http://en.wikipedia.org/wiki/Anatomical_terms_of_location for the different planes

2012/9/2 Sebastian Fröjd <so.co...@gmail.com>

[Jorge Llambías]

On Sun, Sep 2, 2012 at 1:37 PM, Sebastian Fröjd <so.co...@gmail.com> wrote:
>
> 1. He did a 900° (a skateboard trick, rotation to the right in the
> transverse plane, person joi skateboard (xilta'o) same rotation)

ko'a soznonynoncarna (to carna lo sedjmajendu lo pritu be vo'a ze'a lo
julra'o be li so no no toi)

> 2. The car rotates around me

lo karce cu mluni mi

> 3. The car is spinning (around its own axis) and rotates around the moose
> (on the left side of the moose)

lo karce cu carna (vo'a) gi'e mluni lo mirli (bu'u lo zunle be lo mirli)

> 4. She did a back flip (axis her own head, rotation feet to the front) and
> landed lying face down

ko'a ti'ercarna (to carna lo kerkerjendu lo trixe toi) gi'e lolfa'u
co'u lo nu lo crane cu cnita

> 5. Tony kickflips his board 90° left in the sagittal plane, and at the same
> time rotates his body 180° to the right in the coronal plane

What's left in the sagittal plane? That plane has no left or right.
Assumng you mean the transverse plane:

la tonis cu tipcargau lo ty xilta'o lo sedjmajendu lo zunle gi'e ca bo
carna lo zulprityjendu lo pritu ze'a lo julra'o be li pa bi no

mu'o mi'e xorxes