Re: [lojban-beginners] Re: Let's do something like this for Lojban!

29 views
Skip to first unread message

Jonathan Jones

unread,
Mar 3, 2012, 7:29:47 PM3/3/12
to loj...@googlegroups.com, rden...@gmail.com


On Sat, Mar 3, 2012 at 4:28 PM, Remo Dentato <rden...@gmail.com> wrote:
On Sun, Mar 4, 2012 at 12:10 AM, Jonathan Jones <eye...@gmail.com> wrote:
>> 2) I know parallelism is important. That is why I made 2 and 3 parallel.
>> It is true, though, that pictures 2 and 4 are parallel.
>
> Which is precisely why the text should be parallel in #2 & #4.

I agree one should preserve the orginal as much as possible but not
without stretching Lojban.

I English you say "My name is", in German you have a specific verb for
this: "ich heiße" while in French, Spanish and Italian you use the
equivalent of "I call myself" ("je m'appel", "me llamo" and "mi
chiamo"). What would you think if I would say "I call myself Remo"
instead of "My name is Remo" just to be faithful to the original
sentence? It wouldn't not be good English.

We are discussing the use of {mi'e.clalis.} against {mi se cmene zo.clalis.}

Colloquially,
{mi'e.clalis.} means "I'm Clalis"
{mi se cmene zo.clalis.} means "My name is Clalis."

Both of these are valid, used, Lojban. I agree that {mi'e.clalis.} is more common for self-introduction, but there is absolutely nothing wrong with using the longer form normally used for introducing others to introduce oneself.

Combined with the fact that it is important to repeat new words with varying combinations to more clearly show function and meaning, as well as the original's text following this pattern, give a lot of reasons to give preference to {mi se cmene zo.clalis.}

> I disagree. ko'a - vo'u are Lojban's he, she, and it. Unlike most languages,
> we get 8 of them, and they're all gender neutral. ko'a-vo'u aren't pointing
> words, they refer to a specific entity, which may be explicitly made using
> goi, or implicitly assigned from context.

I disagree here. While I'm a great support of "meaning by context", on
his own {ko'a} has no meaning and can't be used the same way the third
person is used in English and other languages.

You give no reasons for why you think that, so I have nothing to say on this other than that I don't share your belief.
 
>> > Yeah, I'm sticking with the translation that does have regard for the
>> > original, but thanks.
>>
>> Ok, but why?
>
> Because it has regard for the original, obviously.

Which is fine as long the translation is proper Lojban.

--
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 )

Remo Dentato

unread,
Mar 3, 2012, 8:02:19 PM3/3/12
to Jonathan Jones, loj...@googlegroups.com


On Sunday, March 4, 2012, Jonathan Jones <eye...@gmail.com> wrote:

>> > I disagree. ko'a - vo'u are Lojban's he, she, and it. Unlike most languages,
>> > we get 8 of them, and they're all gender neutral. ko'a-vo'u aren't pointing
>> > words, they refer to a specific entity, which may be explicitly made using
>> > goi, or implicitly assigned from context.
>>
>> I disagree here. While I'm a great support of "meaning by context", on
>> his own {ko'a} has no meaning and can't be used the same way the third
>> person is used in English and other languages.
>
> You give no reasons for why you think that, so I have nothing to say on this other than that I don't share your belief.

The reason is CLL 7.5 (or at least the way I read it). I thought it was not contentious that the KOhA series had not "obvious value"

Felipe Gonçalves Assis

unread,
Mar 4, 2012, 6:47:31 AM3/4/12
to loj...@googlegroups.com
[On the relation between 3rd person pronouns, demonstratives and the
ko'a-series]

There is an important example of a natural language that has no personal pronoun
for the 3rd person, i.e., Classical Latin, in which demonstratives
haec, ista, illa, etc
were used instead. In fact, it is from these demonstratives that 3rd
person personal
pronouns developed, in romance languages at least (taking their form
from ille/illa).

It is indeed as demonstratives of people that 3rd person personal
pronouns behave.
I can point to a girl and refer to her as "she", or refer to a guy
nearby mentioned in
the text as "he". These vague but noticeable rules of reference give
rise to the lojbo
ti-series and ri-series respectively.

The ko'a-series is a whole different thing. Its single rule of
reference is assignment.
The only parallel to that in the English language is the "henceforth
referred to as"
of legal documents, which does not include personal pronouns at all.

So, while the CLL does comment that the ko'a-series is the lojban version of 3rd
person pronouns, this is pretty insensitive to how natural languages work.

mu'o
mi'e .asiz.

v4hn

unread,
Mar 4, 2012, 8:04:33 AM3/4/12
to loj...@googlegroups.com
coi rodo

.i lo ti se mrilu ku pamoi se mrilu mi lo jbomriste
(this is the first mail I send to the list)

On Sun, Mar 04, 2012 at 08:47:31AM -0300, Felipe Gonçalves Assis wrote:
> The ko'a-series is a whole different thing. Its single rule of
> reference is assignment.
> The only parallel to that in the English language is the "henceforth
> referred to as"
> of legal documents, which does not include personal pronouns at all.
>
> So, while the CLL does comment that the ko'a-series is the lojban version of 3rd
> person pronouns, this is pretty insensitive to how natural languages work.

As far as I know, there is one natural example of the ko'a-series in Sign Language [0].
This is called "referent locus system". The speaker can assign referents to points
in space.

mu'o mi'e van

---
[0] - http://en.wikipedia.org/wiki/American_Sign_Language_grammar#Referent_locus_system

Felipe Gonçalves Assis

unread,
Mar 7, 2012, 11:38:09 PM3/7/12
to loj...@googlegroups.com, eye...@gmail.com
On 7 March 2012 16:57, Jonathan Jones <eye...@gmail.com> wrote:
> On Wed, Mar 7, 2012 at 8:10 AM, Felipe Gonçalves Assis
> <felipe...@gmail.com> wrote:
>>
>> >> zmadu3 is a dimension, so it should be {lo ka ce'u nanca makau kei}.
>> >> Since this is the last utterance, and it appears to be decipherable,
>> >> you might want to play with the reader and leave it like that. A more
>> >> friendly alternative is to use a tanru like {nanca zmadu}, as in 56.
>> >
>> >
>> > z3 is not a dimension. It is a property or quantity. lo se nanca is a
>> > quantity.
>> >
>>
>> z4 is a quantity and a {se nanca}. z3 is a quantity that is a function of
>> something. Otherwise, what could be an answer to the following questions?
>>  {la .sam. zmadu la .clalis. li xo}
>>  {la .sam. zmadu la .clalis. lo se nanca be ma}
>>
>
>
> Seriously, stop arguing things without making sure you're right first.
>
> zmadu: x1 exceeds/is more than x2 in property/quantity x3 (ka/ni) by
> amount/excess x4.
>
> z4 is amount by which x1 is more than x2, z3 is the property/quantity in
> which x1 is more.
>

I am not denying that you may call z3 a quantity, but there is a major
difference
between the natures of "my age" and "one's age". The first is a number, the
second is a mapping. nanca2 is a number, z3 is not.

la .lindar.

unread,
Mar 9, 2012, 5:20:22 AM3/9/12
to loj...@googlegroups.com, eye...@gmail.com
Can I have a link to this product so we can feature it on the Lojban social network pages?
Reply all
Reply to author
Forward
0 new messages