WikiDiscuss

Posted by rab.spir on Thu 08 of Jul, 2004 16:15 GMT posts: 152

The proposal about assigning a direction to goi has been brought up before, and considered unnecessary.

goi unifies its two arguments so that they refer to the same thing. If one is defined and the other isn't, the undefined one gets the meaning of the defined one. If both are undefined, they are linked to refer to the same undefined thing, which may be assigned later. If both are defined, using goi is an error.

Posted by xorxes on Thu 08 of Jul, 2004 20:45 GMT posts: 1912

> goi unifies its two arguments so that they refer to the same thing. If one is
> defined and the other isn't, the undefined one gets the meaning of the
> defined one. If both are undefined, they are linked to refer to the same
> undefined thing, which may be assigned later. If both are defined, using goi
> is an error.

I assume {ko'a goi le prenu} is equivalent to {le prenu ku goi ko'a}.
Are {le prenu goi ko'a ku} and {le goi ko'a prenu ku} also equivalent
to the first two?

I assume that in {ci le mu prenu goi ko'a ku} and in
{ci le goi ko'a mu prenu ku}, {ko'a} gets assigned {le mu prenu ku}.

What happens with {ko'a goi ci le mu prenu} and {ci le mu prenu ku
goi ko'a}? Presumably ko'a gets assigned the three people which
fulfill whatever will be claimed about them?

What if the assignment occurs under the scope of another quantifier?

ro le ze gerku cu batci ci le mu prenu ku goi ko'a
"Each of the seven dogs bites three of the five people, from now on ko'a"

If we use ko'a in the next sentence, does this force an interpretation
on the first sentence that every dog bit the same three people? Or does
ko'a refer to all of the five people that were bitten, even if there's
more than three in all?

Is {ko'a goi no le mu prenu} at all meaningful?

Does {su'eci le mu prenu ku goi ko'a} force existential import
into {su'eci} ("at most three")?

mu'o mi'e xorxes

__
Do you Yahoo!?
New and Improved Yahoo! Mail - Send 10MB messages!
http://promotions.yahoo.com/new_mail

wikidiscuss@lojban.org scripsit:

> goi unifies its two arguments so that they refer to the same thing. If
> one is defined and the other isn't, the undefined one gets the meaning
> of the defined one. If both are undefined, they are linked to refer
> to the same undefined thing, which may be assigned later. If both are
> defined, using goi is an error.

The difficulty with this view (which is CLL's) is that it doesn't say how
to redefine anything. If ko'a and ko'e were both defined some time back,
then ko'a goi ko'e could mean "redefine ko'e as ko'a" or vice versa.
We need to provide for at least a weak preference order in such a case.

--
Eric Raymond is the Margaret Mead John Cowan
of the Open Source movement. jcowan@reutershealth.com
--Bruce Perens, http://www.ccil.org/~cowan
some years ago http://www.reutershealth.com

Posted by pycyn on Thu 08 of Jul, 2004 20:45 GMT posts: 2388

If things work as you say, then the directional rule is unnecessary, but also harmless and used only when conflicts arise. And at least the last point is open to question: after a realy screwed up nip and tuck I start my comments on it wit {lemi mikce goi le malpinji} to cover the fact that I am going to call him a schmuck for the rest of the discourse. Maybe I don't understand what is meant by "defined" in this case; in the ordinary sense of "define" both expressions are. I guess you mean "assigned"?
wikidiscuss@lojban.org wrote:Re: Relative Clauses and Phrases
The proposal about assigning a direction to goi has been brought up before, and considered unnecessary.

goi unifies its two arguments so that they refer to the same thing. If one is defined and the other isn't, the undefined one gets the meaning of the defined one. If both are undefined, they are linked to refer to the same undefined thing, which may be assigned later. If both are defined, using goi is an error.

Posted by xorxes on Thu 08 of Jul, 2004 20:45 GMT posts: 1912

John Cowan:
> The difficulty with this view (which is CLL's) is that it doesn't say how
> to redefine anything. If ko'a and ko'e were both defined some time back,
> then ko'a goi ko'e could mean "redefine ko'e as ko'a" or vice versa.
> We need to provide for at least a weak preference order in such a case.

{da'o} could probably be adapted for this:

ko'a da'o goi ko'e: cancel ko'a and redefine as ko'e
ko'a da'onai goi ko'e: don't cancel ko'a and redefine ko'e as ko'a

It wouldn't hurt to have a default, though.

mu'o mi'e xorxes

__
Do you Yahoo!?
Take Yahoo! Mail with you! Get it on your mobile phone.
http://mobile.yahoo.com/maildemo

Posted by pycyn on Thu 08 of Jul, 2004 20:45 GMT posts: 2388

I am not delighted to learn that someone would actually use such a gloppy expression as {le goi ko'a prenu}. On the other hand, things like that might provide some help for issues oof which of teh several coterminal sumti is meant; maybe {le goi k'oa ci lo prenu} v. {le ci lo goi ko'a prenu}. And maybe not; we would still need the afterthought cases. But this problem turns up all over the place apparently (what ever happened to "uniquely parsible"? I guess it means "has exactly one parse, which may or may not have ans significance for the meaning of the sentence.) As does the problem of what pronouns (assigned or ad hoc) refer to when they refer to things referred to by functions of some quantifier (the intended reading apparently of the dog biting sentence is that {re nanmu} is a function from dogs — at least *these* dogs --to pairs of men (all particular quantifiers can be treated that way). Does {ko'a} attached then pick up the function or does it pick one of the pairs or
all three of them or the three to six men or ...?
I don't know of any proposals on any of these issues and don't have any of my own at the moment. Mark that down on the GOI to do list.

> goi unifies its two arguments so that they refer to the same thing. If one is
> defined and the other isn't, the undefined one gets the meaning of the
> defined one. If both are undefined, they are linked to refer to the same
> undefined thing, which may be assigned later. If both are defined, using goi
> is an error.
I assume {ko'a goi le prenu} is equivalent to {le prenu ku goi ko'a}.
Are {le prenu goi ko'a ku} and {le goi ko'a prenu ku} also equivalent
to the first two?

I assume that in {ci le mu prenu goi ko'a ku} and in
{ci le goi ko'a mu prenu ku}, {ko'a} gets assigned {le mu prenu ku}.

What happens with {ko'a goi ci le mu prenu} and {ci le mu prenu ku
goi ko'a}? Presumably ko'a gets assigned the three people which
fulfill whatever will be claimed about them?

What if the assignment occurs under the scope of another quantifier?

ro le ze gerku cu batci ci le mu prenu ku goi ko'a
"Each of the seven dogs bites three of the five people, from now on ko'a"

If we use ko'a in the next sentence, does this force an interpretation
on the first sentence that every dog bit the same three people? Or does
ko'a refer to all of the five people that were bitten, even if there's
more than three in all?

Is {ko'a goi no le mu prenu} at all meaningful?

Does {su'eci le mu prenu ku goi ko'a} force existential import
into {su'eci} ("at most three")?

mu'o mi'e xorxes

__
Do you Yahoo!?
New and Improved Yahoo! Mail - Send 10MB messages!
http://promotions.yahoo.com/new_mail

Posted by pycyn on Thu 08 of Jul, 2004 20:45 GMT posts: 2388

Right. And why use an extra word, even a little one like {da'o} if we can make do with less.
Jorge Llambías <jjllambias2000@yahoo.com.ar> wrote:John Cowan:
> The difficulty with this view (which is CLL's) is that it doesn't say how
> to redefine anything. If ko'a and ko'e were both defined some time back,
> then ko'a goi ko'e could mean "redefine ko'e as ko'a" or vice versa.
> We need to provide for at least a weak preference order in such a case.

{da'o} could probably be adapted for this:

ko'a da'o goi ko'e: cancel ko'a and redefine as ko'e
ko'a da'onai goi ko'e: don't cancel ko'a and redefine ko'e as ko'a

It wouldn't hurt to have a default, though.

mu'o mi'e xorxes

__
Do you Yahoo!?
Take Yahoo! Mail with you! Get it on your mobile phone.
http://mobile.yahoo.com/maildemo

Posted by xorxes on Thu 08 of Jul, 2004 20:45 GMT posts: 1912

pc:
> I am not delighted to learn that someone would actually use such a gloppy
> expression as {le goi ko'a prenu}. On the other hand, things like that might
> provide some help for issues oof which of teh several coterminal sumti is
> meant; maybe {le goi k'oa ci lo prenu} v. {le ci lo goi ko'a prenu}.

There are three afterthought versions:

A1- le ci lo prenu ku ku goi ko'a}
A2- le ci lo prenu ku goi ko'a ku}
A3- le ci lo prenu goi ko'a ku ku}

two middlethought versions:

M1- {le goi ko'a ci lo prenu}
M2- {le ci lo goi ko'a prenu}

and one forethought:

F1- {ko'a goi le ci lo prenu}

I wonder which one corresponds with which.
I would hazzard

F1=A1, ko'a = le ci lo prenu;
M1=A2, ko'a = ?
M2=A3, ko'a = lo prenu.

mu'o mi'e xorxes

__
Do you Yahoo!?
Yahoo! Mail is new and improved - Check it out!
http://promotions.yahoo.com/new_mail

Posted by pycyn on Thu 08 of Jul, 2004 20:45 GMT posts: 2388

Why do ugly things have to be so damned handy? I take it that M1 = A1 = le ci lo prenu. the generalization is that these inserts go to the sumti whose gadri they (more or less) immediately follow. How do we get {ci lo prenu}, a phrase we are more likely to want to repeat (referentially) than {lo prenu} probably. To do for anaphora again.
Jorge Llambías <jjllambias2000@yahoo.com.ar> wrote:
pc:
> I am not delighted to learn that someone would actually use such a gloppy
> expression as {le goi ko'a prenu}. On the other hand, things like that might
> provide some help for issues oof which of teh several coterminal sumti is
> meant; maybe {le goi k'oa ci lo prenu} v. {le ci lo goi ko'a prenu}.

There are three afterthought versions:

A1- le ci lo prenu ku ku goi ko'a}
A2- le ci lo prenu ku goi ko'a ku}
A3- le ci lo prenu goi ko'a ku ku}

two middlethought versions:

M1- {le goi ko'a ci lo prenu}
M2- {le ci lo goi ko'a prenu}

and one forethought:

F1- {ko'a goi le ci lo prenu}

I wonder which one corresponds with which.
I would hazzard

F1=A1, ko'a = le ci lo prenu;
M1=A2, ko'a = ?
M2=A3, ko'a = lo prenu.

mu'o mi'e xorxes

__
Do you Yahoo!?
Yahoo! Mail is new and improved - Check it out!
http://promotions.yahoo.com/new_mail