{porsi}

[Ilmen]

coi ro do

I would like to hear your opinion about the following.

Can the predicate {porsi} be applied to a sequence, in the sense of an ordered list allowing multiple occurrences of the same element, such as the letter sequence [b,r,o,d,o] (wherein the element "o" appears twice at different positions)?

The second and third argument slots of {porsi} seem to entails that it is only appropriate to describe ordered sets generated from applying a sorting rule to an unordered set, which probably (though I'm not sure, it depends on the exact definition of {porsi}) excludes sequences with duplicate elements.

If you think that the letter sequence "brodo" can satisfy {porsi}'s x1, I'd be grateful if you could show me the Lojban definition of the sorting rule that can output the sequence "brodo" when applied to the unordered set of letters {b,r,o,d}. :)

Another way to claim in Lojban that [b,r,o,d,o] (for example) is a sequence could be {X te lidne da de}; however that wouldn't work with unary sequences (which are probably identical to their equivalent unordered sets anyway).

Any thought, comment or suggestion will be greatly appreciated.

mi'e la .ilmen. mu'o

[Jorge Llambías]

On Tue, Jan 27, 2015 at 6:47 PM, Ilmen <ilmen....@gmail.com> wrote:

If you think that the letter sequence "brodo" can satisfy {porsi}'s x1, I'd be grateful if you could show me the Lojban definition of the sorting rule that can output the sequence "brodo" when applied to the unordered set of letters {b,r,o,d}. :)

In my view, ro da poi porsi cu gunma lo te porsi be da, so porsi3 are all the members of porsi1, not just the member types. So for example the sequence of digits in the decimal expansion of pi has an infinite number of elements, not just ten.

Describing the porsi2 of [b,r,o,d,o] is a pain because it's an arbitrary sequence, not one that follows some simple rule. Since we're going to need to refer to two instances of the letter "o" we can use the pronouns ".o bu xi pa" and ".o bu xi re" for each instance, and "by", "ry" and "dy" for the other three letter instances. Then porsi2 could be something like "lo du'u ge by cu pa moi gi ge ry cu re moi gi ge .o bu xi pa cu ci moi gi ge dy cu vo moi gi ,o bu xi re cu mu moi".

mu'o mi'e xorxes

[Jorge Llambías]

On Tue, Jan 27, 2015 at 8:06 PM, Jorge Llambías <jjlla...@gmail.com> wrote:

In my view, ro da poi porsi cu gunma lo te porsi be da, 

* lo te porsi be fi da

[la durka]

I think there has to be a distinction between "ordered set" and "sequence", no? Lojban seemingly conflates the two, but they're not the same -- a set, ordered or not, can't have duplicate elements, but a sequence can.

- mu'o mi'e la durkavore

[Jorge Llambías]

On Wed, Jan 28, 2015 at 2:26 AM, la durka <dur...@gmail.com> wrote:

I think there has to be a distinction between "ordered set" and "sequence", no? Lojban seemingly conflates the two, but they're not the same -- a set, ordered or not, can't have duplicate elements, but a sequence can.

For me, porsi relates a sequence porsi1 to its members porsi3, so for me there are no sets involved.

Since an ordered set is a special type of sequence, we could have "rolcmipoi" or "cmipoi", "x1 porsi x2 lo ro cmima be x3" for ordered set, a sequence whose members are the members of a set.

[Alex Burka]

Ah, so then length(lo porsi) may be greater than length(lo te porsi). I suggested this solution on IRC but it wasn't that well received.

For an example, in what sense do we translate to Lojban the fact "O is the third letter of the word 'brodo'."?

- mu'o mi'e la durkavore

--
You received this message because you are subscribed to a topic in the Google Groups "lojban" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/lojban/pL3yVycaRSQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to lojban+un...@googlegroups.com.
To post to this group, send email to loj...@googlegroups.com.
Visit this group at http://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

[Jorge Llambías]

On Wed, Jan 28, 2015 at 1:21 PM, Alex Burka <dur...@gmail.com> wrote:

Ah, so then length(lo porsi) may be greater than length(lo te porsi). I suggested this solution on IRC but it wasn't that well received.

As I understand it, the length of lo porsi is equal to the number of lo te porsi.

For an example, in what sense do we translate to Lojban the fact "O is the third letter of the word 'brodo'."?

 pa me'o .o bu cu ci moi lo mu lerfu poi se gunma zo brodo

[Alex Burka]

How can this be, if here the porsi1 is b-r-o-d-o, but porsi3 is {b, r, o, d}.

[Jorge Llambías]

On Wed, Jan 28, 2015 at 1:48 PM, Alex Burka <dur...@gmail.com> wrote:

How can this be, if here the porsi1 is b-r-o-d-o, but porsi3 is {b, r, o, d}.

But I don't agree that porsi3 is {b. r, o, d}. As I keep saying, I don't think porsi3 is a set, I think porsi3 are the members of porsi1, and I think that the sequence b-r-o-d-o has five members (two of which happen to be two instances of the same type). So I consider "brodo" a five-letter word, not a four-letter word.

[Ilmen]

On 28/01/2015 00:06, Jorge Llambías wrote:
>
> Describing the porsi2 of [b,r,o,d,o] is a pain because it's an
> arbitrary sequence, not one that follows some simple rule. Since we're
> going to need to refer to two instances of the letter "o" we can use
> the pronouns ".o bu xi pa" and ".o bu xi re" for each instance, and
> "by", "ry" and "dy" for the other three letter instances. Then porsi2
> could be something like "lo du'u ge by cu pa moi gi ge ry cu re moi gi
> ge .o bu xi pa cu ci moi gi ge dy cu vo moi gi ,o bu xi re cu mu moi".

"N moi" is a claim stating a relation between three arguments and a
number. What are the x2 and x3 of moi there so that the bridi's claim
holds true?

I think {moi} is totally inappropriate there, as we need a sorting rule
for defining another sorting rule; porsi-2 should at least be an unary
property of porsi-1, and I guess that {lidne} or {li'erla'i} is the way
to go for building up the sequence.

{porsi lo ka fi ce'u li'erla'i fa by fe ry gi'e li'erla'i fa ry fe .obu
gi'e.....}

pei

[selpa'i]

la .xorxes. cu cusku di'e

I agree with all of that, but to me there is at least a slight oddity
about the duplicate letter "o".

Do we agree that {ro da zo'u da jo'u da mintu da}? In the abstract, the
letter "o" exists only once, just like there is only one number "1".
When speaking abstractly about the letters in a word, there would then
however be two of the same "thing", and that would mean there are only
distinct four referents. Talking about instances of the abstract letter
is the only easy way out of this, but then there still remains the
slight problem that both of those instances are identical to one
another. They differ in nothing, except - you might argue - in the
property {lo ka du ma kau}. However, does this solve the problem? The
word "brodo" is spelled correctly no matter which of those two "o"
letters you place first, and I think that's because there aren't two
distinct "o" letters, since there is nothing really to tell them apart.

If you say that {me'o .obu} has two referents, which you must if you
claim {ci moi lo mu lerfu}, then you claim that the expression "o"
exists more than once. This is fine by me, but it requires a special
domain to work, one in which there are only instances of letters (even
though each instance can itself be a kind). I would probably feel more
at ease (for a general solution) if it involved {mupli}, but then the
aforementioned problem still remains: The two instances differ in
nothing, and either of them can be the third letter in "brodo". And
these so-called instances of the letters are still very abstract, since
words aren't always written down; they can simply be in our minds.

This intricacy is what kept me from proposing a solution similar to
yours yesterday (I had thought about the expression {by jo'u ry jo'u re
boi .obu jo'u dy} and found it very strange (with or without {me'o})!).

mi'e la selpa'i mu'o

[Jorge Llambías]

On Wed, Jan 28, 2015 at 2:12 PM, Ilmen <ilmen....@gmail.com> wrote:

On 28/01/2015 00:06, Jorge Llambías wrote:

Describing the porsi2 of [b,r,o,d,o] is a pain because it's an arbitrary sequence, not one that follows some simple rule. Since we're going to need to refer to two instances of the letter "o" we can use the pronouns ".o bu xi pa" and ".o bu xi re" for each instance, and "by", "ry" and "dy" for the other three letter instances. Then porsi2 could be something like "lo du'u ge by cu pa moi gi ge ry cu re moi gi ge .o bu xi pa cu ci moi gi ge dy cu vo moi gi ,o bu xi re cu mu moi".

"N moi" is a claim stating a relation between three arguments and a number. What are the x2 and x3 of moi there so that the bridi's claim holds true?

x2 is the same as porsi3. x3 can be zi'o'ed out, since all we need is an appropriate mapping to the naturals, which we are giving.

I think {moi} is totally inappropriate there, as we need a sorting rule for defining another sorting rule; porsi-2 should at least be an unary property of porsi-1, and I guess that {lidne} or {li'erla'i} is the way to go for building up the sequence.

{porsi lo ka fi ce'u li'erla'i fa by fe ry gi'e li'erla'i fa ry fe .obu gi'e.....}

pei

Taking porsi2 to be a property of porsi1 ? I guess.

My first thought, before giving an explicit mapping to (an initial segment of) the naturals, was to give an ordering relation, but that's even longer than yours, something like "lo ka ce'u xi pa goi ko'a ce'u xi re goi ko'e zo'u tu'e ganai ge ko'a du by gi ko'e du ry gi ko'a lidne ko'e ;i je ga nai ge ko'a du ...". But then you will say that you want to know what lidne3 is, and so I have to zi'o it out again.

[Ilmen]

Also, I used to think that {porsi}-x2 was a binary comparison function,
like {lo ka ce'u ce'u bramau} etc. (The word "comparison" is present in
the English definition of {porsi}).

Provided that porsi2 is an unary property of porsi1, the below is a
possible way to define the sequence [b,r,o,d,a] using {porsi}:

• { porsi fa B ce'o R ce'o O ce'o D ce'o O fi B ce R ce O ce D fe
loka ce'u poi'i ge ke'a pormei li mu gi fi ke'a li'erla'i fa B fe R gi'e
li'erla'i fa R fe O gi'e li'erla'i fa O fe D gi'e li'erla'i fa D fe O }

And building an ordered set from the set {A, B, C} by sorting them by
size would require putting something like the below into porsi2:

• { lo ka ro da ro de zo'u ganai da de ce'u lidne gi da de branalme'a }

However then the place structure of {porsi} seems a little awkward, as
porsi3 is not necessary anymore and claiming "porsi1 cu ckaji fe porsi2"
is enough.

[Jorge Llambías]

On Wed, Jan 28, 2015 at 2:15 PM, selpa'i <sel...@gmx.de> wrote:

Do we agree that {ro da zo'u da jo'u da mintu da}?

"ro da zo'u da jo'u da du da", yes.

 

In the abstract, the letter "o" exists only once, just like there is only one number "1".

At the most abstract level, yes, but we talk about letter instances all the time.

 

When speaking abstractly about the letters in a word, there would then however be two of the same "thing", and that would mean there are only distinct four referents.

But that's not the most common way of talking about the letters of a word. If we were to talk like that we would most likely talk about "different letters", not just "letters".

 

Talking about instances of the abstract letter is the only easy way out of this, but then there still remains the slight problem that both of those instances are identical to one another. They differ in nothing, except - you might argue - in the property {lo ka du ma kau}.

Well, their most important difference is their position the word. One important difference is that one carries stress and the other one doesn't.

  

However, does this solve the problem? The word "brodo" is spelled correctly no matter which of those two "o" letters you place first, and I think that's because there aren't two distinct "o" letters, since there is nothing really to tell them apart.

But what's the problem? If you switch around the wheels of a car, you still have the same car. The differences among the wheels are mostly irrelevant. (I know they can be relevant, but let's say they are all brand new, or pick a better example.) You can still talk about "the left front wheel" even though its molecules are not the same molecules they used to be. In the case of the "o" you don't even have to worry about molecules, "the first o" is just "the first o". You can say that it carries the stress, whether you have switched it or not, whatever that may mean. 

If you say that {me'o .obu} has two referents, which you must if you claim {ci moi lo mu lerfu}, then you claim that the expression "o" exists more than once. This is fine by me, but it requires a special domain to work, one in which there are only instances of letters (even though each instance can itself be a kind). I would probably feel more at ease (for a general solution) if it involved {mupli}, but then the aforementioned problem still remains: The two instances differ in nothing, and either of them can be the third letter in "brodo". And these so-called instances of the letters are still very abstract, since words aren't always written down; they can simply be in our minds.

But they do differ in some things. 

This intricacy is what kept me from proposing a solution similar to yours yesterday (I had thought about the expression {by jo'u ry jo'u re boi .obu jo'u dy} and found it very strange (with or without {me'o})!).

I think it's fine, although you may want to do "pa boi by jo'u pa boi ry jo'u pa boi dy jo'u re boi ,o bu" to make it more uniform.

[Alex Burka]

cinri nuncasnu .i mi te sidbo lo drata mupli

.i mu'a da'i ca lo cabdei mi pu klama lo mi briju ca lo cerni gi'e ba bo xrukla lo mi zdani gi'e ba bo xrukla lo mi briju

.i mi ponse lo pa briju .i mi na ponse lo briju xi pa .e lo briju xi re

.i ja'o lo bi'unai briju cu xo moi lo se klama be mi

- mu'o mi'e la durkavore

--

[Jorge Llambías]

On Fri, Jan 30, 2015 at 3:15 PM, Alex Burka <dur...@gmail.com> wrote:

cinri nuncasnu .i mi te sidbo lo drata mupli

.i mu'a da'i ca lo cabdei mi pu klama lo mi briju ca lo cerni gi'e ba bo xrukla lo mi zdani gi'e ba bo xrukla lo mi briju

.i mi ponse lo pa briju .i mi na ponse lo briju xi pa .e lo briju xi re

.i ja'o lo bi'unai briju cu xo moi lo se klama be mi

zo'o .e'u ko seltcitygau lo do lisri lu lo nabmi pe lo briju poi zilkancyze'a li'u

.i ru'a do tavla fi lo pa briju poi do ponse ke'a gi'e klama ke'a re roi lo cabdei to ku'i mi na mulno jimpe lo du'u lo re briju ku poi do ponse su'e pa ke'a cu srana ma kau toi

.i pe'i lo briju ge pa moi gi ci moi vau lo ci se klama be do bei lo cabdei .i lo ci se klama zo'u lo pa moi jo'u lo ci moi cu mitsi'u lo ka se stuzi ma kau kei gi'e ku'i ficysi'u lo ka se klama ca ma kau .i ca lo nu do tavla fi lo se ponse do pensi lo briju poi pa mei lo se ponse .i ca lo nu tavla fi lo se klama porsi do pensi lo briju poi re mei lo se klama