Where should I use sets and where should I use masses?

[gleki]

{loi prenu cu sruri lo dinju} - why not lo gunma be lo prenu cu sruri lo dinju?

{lo'i tirxu cu cmalu} - why not so'u tirxu?

Can you give better examples?

I'm not a mathematician and don't clearly understand the trick with sets and masses.

[Jacob Errington]

All sets have the same types of properties such as cardinality, but have non of the properties of the things of which they are a set.

That is to say that {lo'i gerku} and {lo'i jubme} have the same properties, such as cardinality again, but have none of the underlying properties of dogs or tables, respectively.

Sets are required in some places in lojban, such as simxu1 and cmima2 (these relationships don't actually involve the underlying properties of the elements of the set) 

e.g. lo'i ro jbopre cu simxu lo ka ce'u slabu ce'u -> All lojbanists know each other.

e.g. ko cmima lo'i jbopre -> Join the lojbanists!

Aside from that, there is an interesting, albeit unused in practice, thing we can do with sets which is cartesian product, pairing up two sets, with the JOI {pi'u}.

.i mi'e la tsani mu'o

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban-beginners/-/_ysdbRWQWVYJ.
To post to this group, send email to lojban-b...@googlegroups.com.
To unsubscribe from this group, send email to lojban-beginne...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban-beginners?hl=en.

[Jorge Llambías]

On Sun, Sep 2, 2012 at 11:47 AM, Jacob Errington <nict...@gmail.com> wrote:
>
> Sets are required in some places in lojban, such as simxu1 and cmima2 (these
> relationships don't actually involve the underlying properties of the
> elements of the set)
>
> e.g. lo'i ro jbopre cu simxu lo ka ce'u slabu ce'u -> All lojbanists know
> each other.
> e.g. ko cmima lo'i jbopre -> Join the lojbanists!

The disadvantage of defining "simxu" and "cmima" that way, is that you
can't easily say things like:

lo ro jbopre cu zvati ti gi'e simxu lo ka ce'u slabu ce'u
All lojbanists are here and know each other.

I don't see any advantage to defining "simxu" and the few other gismu
that are defined that way in terms of sets, so I use them as if they
were defined in a more user-friendly manner, in terms of ordinary
groups rather than mathematical sets.

mu'o mi'e xorxes

[Jacob Errington]

Of course, if we work under the assumption that {lo} produces the weird polymorphic type, then yes, you can get away with shenanigans like that, but if you don't, you need an experimental in the form of a sort of JAI that can do afterthought LAhE in the way JAI TAG does afterthought TAG.

.i mi'e la tsani mu'o

 

mu'o mi'e xorxes

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

[Jorge Llambías]

On Sun, Sep 2, 2012 at 12:09 PM, Jacob Errington <nict...@gmail.com> wrote:
> On 2 September 2012 11:03, Jorge Llambías <jjlla...@gmail.com> wrote:
>>
>> lo ro jbopre cu zvati ti gi'e simxu lo ka ce'u slabu ce'u
>> All lojbanists are here and know each other.
>>
>> I don't see any advantage to defining "simxu" and the few other gismu
>> that are defined that way in terms of sets, so I use them as if they
>> were defined in a more user-friendly manner, in terms of ordinary
>> groups rather than mathematical sets.
>
> Of course, if we work under the assumption that {lo} produces the weird
> polymorphic type, then yes, you can get away with shenanigans like that,

No, I don't think "lo" can produce sets. What I'm doing is using a
different definition of "simxu" such that it does not take a set as
its x1 but just an ordinary plural reference. Instead of meaning "x1
(set) has members who mutually/reciprocally x2" I take it to just mean
"x1 mutually/reciprocally x2", bypassing the unnecessary set step. The
x1 of "simxu" is no different than the x1 of "jmaji". "jmaji" could
have been defined as "x1 (set) has members who gather/collect at
location x2 from locations that are members of x3 (set)", but what
would be the point, other than making things more complicated?

> but
> if you don't, you need an experimental in the form of a sort of JAI that can
> do afterthought LAhE in the way JAI TAG does afterthought TAG.

Right, but what's the point of having to go through a set?

[tijlan]

On 2 September 2012 15:47, Jacob Errington <nict...@gmail.com> wrote:
> e.g. lo'i ro jbopre cu simxu lo ka ce'u slabu ce'u -> All lojbanists know each other.

simxu2 is defined as an event -- {nu} as opposed to {ka} --, which
renders the (equally official) set status of simxu1 even more
paradoxical. Can a set engage in an event?

If the x1 was a set, having {ka} for the x2 as in your example would
be less unreasonable. But I doubt that would be more practical than
"simxu1=non-set, simxu2=event".


mu'o

[ianek]

On 4 Wrz, 09:04, tijlan <jbotij...@gmail.com> wrote:

> On 2 September 2012 15:47, Jacob Errington <nicty...@gmail.com> wrote:
>
> > e.g. lo'i ro jbopre cu simxu lo ka ce'u slabu ce'u -> All lojbanists know each other.
>
> simxu2 is defined as an event -- {nu} as opposed to {ka} --, which
> renders the (equally official) set status of simxu1 even more
> paradoxical. Can a set engage in an event?

Not a set, but its members:
x1 (set) has members who mutually/reciprocally x2 (event [x1 should be
reflexive in 1+ sumti]).

"mutually" is not something similar to carrying a piano, together or
otherwise. It's a property of a set, not its members.
It's a property of being a clique http://en.wikipedia.org/wiki/Clique_(graph_theory).
That way it's more logical, and hey, Lojban was designed to be a
logical language, not an easy language (we already have Esperanto!).

Otherwise we can throw away sets completely, because now you can't say
{lo ro jbopre ku zvati ti gi'e se cmima mi}. If you can do with masses
everyting you can do with sets, we don't need them.

mu'o mi'e ianek

[tijlan]

On 2 September 2012 13:47, gleki <gleki.is...@gmail.com> wrote:
> {loi prenu cu sruri lo dinju} - why not lo gunma be lo prenu cu sruri lo dinju?

{loi prenu} means {lo gunma be lo prenu}. If you reckon the latter is
correct, you would think the former is correct too.

That's a comparison between different ways of expressing a mass,
whereas "Where should I use sets and where should I use masses?" is
asking for a comparison between set expressions & mass expressions in
the same places:

( loi / lo'i ) prenu cu sruri lo dinju -- which gadri?

> {lo'i tirxu cu cmalu} - why not so'u tirxu?

Do you mean
"Why {lo'i tirxu cu cmalu} and not {so'u tirxu}?"
or
"Why {lo'i tirxu cu cmalu} and not {so'u tirxu cu cmalu}?"
or
"Why {lo'i tirxu cu cmalu} and not {loi tirxu cu so'u mei}?"
?


mu'o

[Jacob Errington]

On 4 September 2012 04:18, tijlan <jbot...@gmail.com> wrote:

On 2 September 2012 13:47, gleki <gleki.is...@gmail.com> wrote:
> {loi prenu cu sruri lo dinju} - why not lo gunma be lo prenu cu sruri lo dinju?

{loi prenu} means {lo gunma be lo prenu}. If you reckon the latter is
correct, you would think the former is correct too.

{loi} and {lo'i} are just shortcuts for LAhE or longer bridi expressions.

 

That's a comparison between different ways of expressing a mass,
whereas "Where should I use sets and where should I use masses?" is
asking for a comparison between set expressions & mass expressions in
the same places:

( loi / lo'i ) prenu cu sruri lo dinju -- which gadri?

Sets have no representation in reality and exist in a somewhat ethereal realm alongside bridi (lo du'u broda) and text (lu ma drani danfu). Therefore they can't surround anything.

 

> {lo'i tirxu cu cmalu} - why not so'u tirxu?

Do you mean
"Why {lo'i tirxu cu cmalu} and not {so'u tirxu}?"
or
"Why {lo'i tirxu cu cmalu} and not {so'u tirxu cu cmalu}?"
or
"Why {lo'i tirxu cu cmalu} and not {loi tirxu cu so'u mei}?"
?

Sets can be big or small. They have a cardinality, i.e. a size, which is comparable. This leads to the drastic difference between {lo'i tirxu cu cmalu} and {lo tirxu cu cmalu}: {lo'i tirxu cu cmalu} says simply that the *set* of tigers is small and says *nothing* about the actual tigers, whereas {lo tirxu cu cmalu} (regardless of any quantifiers) says that the tigers are each small. You can have a small set of enormous tigers in the same way that you can have a large set of tiny ones.

mu'o

[Jorge Llambías]

On Tue, Sep 4, 2012 at 4:52 AM, ianek <jan...@gmail.com> wrote:
> On 4 Wrz, 09:04, tijlan <jbotij...@gmail.com> wrote:
>>
>> Can a set engage in an event?
>
> Not a set, but its members:
> x1 (set) has members who mutually/reciprocally x2 (event [x1 should be
> reflexive in 1+ sumti]).
>
> "mutually" is not something similar to carrying a piano, together or
> otherwise. It's a property of a set, not its members.

It's the members of the set, not the set, who mutually x2.

> It's a property of being a clique http://en.wikipedia.org/wiki/Clique_(graph_theory).

If you restrict "simxu" to the cliques of graph theory, you can't use
it for something like:

lo zajba pu simxu lo ka ce'u ce'u sanli kei gi'e se morna lo remna pramide
"The gymnasts stood on one another's shoulders and formed a human pyramid."

It's better to create a more specialized word for the more specialized
meaning and leave the basic word for the more general meaning.

> That way it's more logical, and hey, Lojban was designed to be a
> logical language, not an easy language (we already have Esperanto!).

How is it more logical? If a simxu is a set, you can always define:

ko'a simcmigunma ko'e = ko'a gunma lo cmima be lo simxu be ko'e

"simcmigunma" is no more nor less logical than "simxu". It's just more
useful, so it should get the simpler form.

> Otherwise we can throw away sets completely, because now you can't say
> {lo ro jbopre ku zvati ti gi'e se cmima mi}. If you can do with masses
> everyting you can do with sets, we don't need them.

Exactly! For every broda that for whatever reason was defined as
having a set in x1 you can define a new selbri "brodycmigunma" with
meaning "ko'a gunma lo cmima be lo broda (be ko'e bei ko'i ...), with
an x1 argument that is more tractable than a set, so sets are not
really needed. Sets are useful in a language without plural reference,
but if you have plural reference it doesn't make much sense to have
sets.

[gleki]

On Tuesday, September 4, 2012 6:48:22 PM UTC+4, tsani wrote:

On 4 September 2012 04:18, tijlan <jbot...@gmail.com> wrote:

On 2 September 2012 13:47, gleki <gleki.is...@gmail.com> wrote:
> {loi prenu cu sruri lo dinju} - why not lo gunma be lo prenu cu sruri lo dinju?

{loi prenu} means {lo gunma be lo prenu}. If you reckon the latter is
correct, you would think the former is correct too.

{loi} and {lo'i} are just shortcuts for LAhE or longer bridi expressions.

 

That's a comparison between different ways of expressing a mass,
whereas "Where should I use sets and where should I use masses?" is
asking for a comparison between set expressions & mass expressions in
the same places:

( loi / lo'i ) prenu cu sruri lo dinju -- which gadri?

Sets have no representation in reality and exist in a somewhat ethereal realm alongside bridi (lo du'u broda) and text (lu ma drani danfu). Therefore they can't surround anything.

 

> {lo'i tirxu cu cmalu} - why not so'u tirxu?

Do you mean
"Why {lo'i tirxu cu cmalu} and not {so'u tirxu}?"

Well, yes I meant this one. ^ ^ ^ 

[ianek]

On 5 Wrz, 00:24, Jorge Llambías <jjllamb...@gmail.com> wrote:

> On Tue, Sep 4, 2012 at 4:52 AM, ianek <jane...@gmail.com> wrote:
> > On 4 Wrz, 09:04, tijlan <jbotij...@gmail.com> wrote:
>
> >>  Can a set engage in an event?
>
> > Not a set, but its members:
> > x1 (set) has members who mutually/reciprocally x2 (event [x1 should be
> > reflexive in 1+ sumti]).
>
> > "mutually" is not something similar to carrying a piano, together or
> > otherwise. It's a property of a set, not its members.
>
> It's the members of the set, not the set, who mutually x2.
>

> > It's a property of being a cliquehttp://en.wikipedia.org/wiki/Clique_(graph_theory).

>
> If you restrict "simxu" to the cliques of graph theory, you can't use
> it for something like:
>
> lo zajba pu simxu lo ka ce'u ce'u sanli kei gi'e se morna lo remna pramide
> "The gymnasts stood on one another's shoulders and formed a human pyramid."

So what {simxu} really means? That the graph is connected? Arguments
based on something that sounds good in a natural language are
suspicious for me. The definitions of Lojban words are (in most cases)
in a natural language, but it doesn't mean that they should be
ambiguous.

mu'o mi'e ianek

[selpa'i]

Am 06.09.2012 21:39, schrieb ianek:
>> If you restrict "simxu" to the cliques of graph theory, you can't use
>> it for something like:
>>
>> lo zajba pu simxu lo ka ce'u ce'u sanli kei gi'e se morna lo remna pramide
>> "The gymnasts stood on one another's shoulders and formed a human pyramid."
> So what {simxu} really means? That the graph is connected? Arguments
> based on something that sounds good in a natural language are
> suspicious for me. The definitions of Lojban words are (in most cases)
> in a natural language, but it doesn't mean that they should be
> ambiguous.

It's not ambiguous. The individuals in simxu1 do simxu2 to each other.
The only difference is that instead of saying "the members of set x1
reciprocally (do) x2" you have the individuals in simxu1 do it directly.
There is really no use for sets in such cases, it's just a waste of time
and makes things unnecessarily inconvenient.

mu'o mi'e la selpa'i

--
.i pau mi me ma .i pa mai ko mi jungau la'e di'u
.i ba bo mi va'o lo nu nelci lo nu me ma kau cu barkla
.i va'o lo nu na nelci cu denpa ti lo nu mi drata

[ianek]

On 6 Wrz, 21:50, selpa'i <m...@plasmatix.com> wrote:
> Am 06.09.2012 21:39, schrieb ianek:
>
> >> If you restrict "simxu" to the cliques of graph theory, you can't use
> >> it for something like:
>
> >> lo zajba pu simxu lo ka ce'u ce'u sanli kei gi'e se morna lo remna pramide
> >> "The gymnasts stood on one another's shoulders and formed a human pyramid."
> > So what {simxu} really means? That the graph is connected? Arguments
> > based on something that sounds good in a natural language are
> > suspicious for me. The definitions of Lojban words are (in most cases)
> > in a natural language, but it doesn't mean that they should be
> > ambiguous.
>
> It's not ambiguous. The individuals in simxu1 do simxu2 to each other.
> The only difference is that instead of saying "the members of set x1
> reciprocally (do) x2" you have the individuals in simxu1 do it directly.
> There is really no use for sets in such cases, it's just a waste of time
> and makes things unnecessarily inconvenient.

But the gymnasts don't stand on each other shoulders. More, there's no
pair standing on each other shoulders, there's no reciprocity. So pe'i
it's really bad example for {simxo}. Either that, or {simxo} is very

ambiguous.

mu'o mi'e ianek

[Jorge Llambías]

On Thu, Sep 6, 2012 at 4:39 PM, ianek <jan...@gmail.com> wrote:
> On 5 Wrz, 00:24, Jorge Llambías <jjllamb...@gmail.com> wrote:
>> On Tue, Sep 4, 2012 at 4:52 AM, ianek <jane...@gmail.com> wrote:
>>
>> > It's a property of being a cliquehttp://en.wikipedia.org/wiki/Clique_(graph_theory).
>>
>> If you restrict "simxu" to the cliques of graph theory, you can't use
>> it for something like:
>>
>> lo zajba pu simxu lo ka ce'u ce'u sanli kei gi'e se morna lo remna pramide
>> "The gymnasts stood on one another's shoulders and formed a human pyramid."
>
> So what {simxu} really means? That the graph is connected?

Maybe even that is not completely necessary. Probably something like
most/almost all nodes must be connected to other nodes and the
connections have to be dense enough.

>Arguments
> based on something that sounds good in a natural language are
> suspicious for me. The definitions of Lojban words are (in most cases)
> in a natural language, but it doesn't mean that they should be
> ambiguous.

Vague is not the same as ambiguous. You can always define a more
precise word ("rolrelsi'u"? "simymu'o"? something else?) for the more
specific meaning. If you make "simxu" require full pairwise
distributivity, you make it practically unusable for most ordinary
contexts.

[ianek]

*{simxu}, of course

[ianek]

On 7 Wrz, 00:18, Jorge Llambías <jjllamb...@gmail.com> wrote:

This is exactly the problem with Lojban. It aims at being logical, but
the more logical, the more unusable it is, and vice versa. Some time
ago I thought it was possible to translate from Lojban to formal
logic...

[Jorge Llambías]

On Thu, Sep 6, 2012 at 7:24 PM, ianek <jan...@gmail.com> wrote:
>
> This is exactly the problem with Lojban. It aims at being logical, but
> the more logical, the more unusable it is, and vice versa. Some time
> ago I thought it was possible to translate from Lojban to formal
> logic...

If simxu has the precise technical meaning of "clique", there's
nothing stopping us from having another selbri brodysi'u with the more
vague meaning, and if simxu has the more vague meaning, there's
nothing stopping us from defining another selbri with the more
specific technical meaning. The fact that both predicates can coexist
in the language doesn't affect how logical the language is or how easy
or difficult it is to translate into formal logic. Whatever the
meaning of simxu, it will go into formal logic as something like S( ,
) anyway, so its meaning does not make the language any more or less
logical. If what you are looking for is a precise definition of
"simxu", then it will probably be in terms of "so'a" or "so'e" rather
than in terms of "ro", and not as straightforward as the mathematical
definition of "clique", but this doesn't make it less logical, just
more difficult to define. But if you consider "so'a" and "so'e"
illogical, then yes, Lojban is irredeemably illogical.

[Jacob Errington]

It's as logical as you want it to be. If you believe simxu to be fully pairwise, then so be it. I mean, what xorxes and selpa'i have said is simply their interpretation of what it means; doesn't mean it's that way and no way else. I believe that simxu is fully pairwise, and therefore I would never have given the previous gymnasts example, but I'm not going to misunderstand if someone uses it differently. This is an important aspect of Lojban: not everyone speaks the same way (in fact, it's unlikely that any two people speak the same way) and no one must try to impose their will on another. 

Lojbanists like to give away their interpretation as if it were the law without acknowledging that there be any other interpretation. However, it is definitely simpler to do so than to explain the various different interpretations and their differences, although doing so would be more "right" in my opinion.

latro'a and I have a strict(er) view of Lojban in that regard and believe than simxu1 must be a set, and that the simxu action is fully pairwise. {mi ce do simxu lo ka cinba} therefore has the obvious meaning. Likewise, {lo'i nanmu ce lo'i ninmu cu simxu lo ka cinba} (I'm using ce as a cheap set addition because I can't be bothered to really look up how to do it) doesn't mean that all the men kissed all the women and vice-versa; it means that each of the men kissed each of the women and also *was kissed* by each of them. That would be my intended interpretation of that, for instance.

.i mi'e la tsani mu'o

 

>

> mu'o mi'e xorxes

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

[Jorge Llambías]

On Thu, Sep 6, 2012 at 8:02 PM, Jacob Errington <nict...@gmail.com> wrote:
>
> latro'a and I have a strict(er) view of Lojban in that regard and believe
> than simxu1 must be a set, and that the simxu action is fully pairwise.

But in what sense does that definition make Lojban more strict? Are
you saying that a predicate with the more vague meaning is simply not
a possible predicate in your strict version of Lojban?

> {mi
> ce do simxu lo ka cinba} therefore has the obvious meaning. Likewise, {lo'i
> nanmu ce lo'i ninmu cu simxu lo ka cinba} (I'm using ce as a cheap set
> addition because I can't be bothered to really look up how to do it)

Set union is "jo'e" (but maybe that's not what you want either). "ce"
would create a set whose two members are each a set.

>doesn't
> mean that all the men kissed all the women and vice-versa; it means that
> each of the men kissed each of the women and also *was kissed* by each of
> them. That would be my intended interpretation of that, for instance.

And also each men kissed and was kissed by each of the other men,
right? Otherwise you don't want the union there but something more
complicated. If you don't include the same sex kissing pairs, that
wouldn't match ianek's interpretation.

[Jacob Errington]

On 6 September 2012 19:57, Jorge Llambías <jjlla...@gmail.com> wrote:

On Thu, Sep 6, 2012 at 8:02 PM, Jacob Errington <nict...@gmail.com> wrote:
>
> latro'a and I have a strict(er) view of Lojban in that regard and believe
> than simxu1 must be a set, and that the simxu action is fully pairwise.

But in what sense does that definition make Lojban more strict? Are
you saying that a predicate with the more vague meaning is simply not
a possible predicate in your strict version of Lojban?

Perhaps more "rigid" rather than "strict", but in general, it means less intuitive interpretation. To a beginner, {lo verba cu simlu lo ka kelci} is probably intuitively correct. Indeed, it is understandable, but it doesn't have that rigid correctness that I adhere to. Similarly, this overall rigidness involves dislike for {kakne lo nu broda} (should be {ka}) and {zmadu fi lo ka broda} (should be {ni}). Although I have little evidence that actually supports this, this interpretation probably makes things simpler to formally define in lojban.

In my honest opinion, {lo verba cu simxu lo ka ce'u kelci kansa ce'u} is just nonsense, because it simply isn't distributive. I strongly dislike that {lo} can produce non-individuals and therefore use loi and lo'i accordingly (please don't supply the gi'e example; a "better" solution to that problem in my opinion is either a jai-like LAhE-cast or, if we aren't allowed to make up any new cmavo, to just use {ije} (and if the problem occurs inside an abstraction, it isn't my fault that there isn't an true afterthought bridi connective in the form of {vauJA} or some such)). Each individual of the description distributes into the predicate, but it *is not* true that each of the children {simxu lo ka kelci kansa}. In fact, if it's okay to just use definitely separate individuals like that, ignoring distribution completely, then {.i mi .e do simxu lo ka cenba} makes perfect sense, which again in my opinion, it most certainly should not, as {mi simxu lo ka cinba .ije do simxu lo ka cinba} is complete nonsense.

 

> {mi
> ce do simxu lo ka cinba} therefore has the obvious meaning. Likewise, {lo'i
> nanmu ce lo'i ninmu cu simxu lo ka cinba} (I'm using ce as a cheap set
> addition because I can't be bothered to really look up how to do it)

Set union is "jo'e" (but maybe that's not what you want either). "ce"
would create a set whose two members are each a set.

Right. I do know what the real effect of {ce} is in that situation, which is why I specified the laziness. It also happens to be why I have an issue with {ce}-strings.

 

>doesn't
> mean that all the men kissed all the women and vice-versa; it means that
> each of the men kissed each of the women and also *was kissed* by each of
> them. That would be my intended interpretation of that, for instance.

And also each men kissed and was kissed by each of the other men,
right? Otherwise you don't want the union there but something more
complicated. If you don't include the same sex kissing pairs, that
wouldn't match ianek's interpretation.

Uh, yes, of course. Too tired at this point I guess. Naturally it includes the same-sex kissing as well. That's exactly what I meant to say.

.i mi'e la tsani mu'o

 

[ianek]

On 7 Wrz, 02:42, Jacob Errington <nicty...@gmail.com> wrote:
> On 6 September 2012 19:57, Jorge Llambías <jjllamb...@gmail.com> wrote:
>
> > On Thu, Sep 6, 2012 at 8:02 PM, Jacob Errington <nicty...@gmail.com>

If we'd want to exclude it, we could use the never ever used Cartesian
product {pi'u}. I don't know what is the type of a pair, but assuming
it's just a set (as hinted in CLL 14.15 http://dag.github.com/cll/14/15/
which makes it not-exactly-Cartesian product) and, again after CLL
14.15, assuming that the Cartesian product isn't itself a set, but a
plural reference to pairs (not sure whether distributive or
otherwise), we can just say:

lo'i nanmu pi'u lo'i ninmu cu simxu lo ka cinba

and it means that each pair kisses mutually (at least under the
fficial definition of {simxu}).

mu'o mi'e ianek

[selpa'i]

Am 07.09.2012 02:42, schrieb Jacob Errington:

On 6 September 2012 19:57, Jorge Llambías <jjlla...@gmail.com> wrote:

On Thu, Sep 6, 2012 at 8:02 PM, Jacob Errington <nict...@gmail.com> wrote:
>
> latro'a and I have a strict(er) view of Lojban in that regard and believe
> than simxu1 must be a set, and that the simxu action is fully pairwise.

But in what sense does that definition make Lojban more strict? Are
you saying that a predicate with the more vague meaning is simply not
a possible predicate in your strict version of Lojban?

Perhaps more "rigid" rather than "strict", but in general, it means less intuitive interpretation. To a beginner, {lo verba cu simlu lo ka kelci} is probably intuitively correct. Indeed, it is understandable, but it doesn't have that rigid correctness that I adhere to. Similarly, this overall rigidness involves dislike for {kakne lo nu broda} (should be {ka}) and {zmadu fi lo ka broda} (should be {ni}). Although I have little evidence that actually supports this, this interpretation probably makes things simpler to formally define in lojban.

In my honest opinion, {lo verba cu simxu lo ka ce'u kelci kansa ce'u} is just nonsense, because it simply isn't distributive.

It's not distributive until you add an outer quantifier. lo verba is a collection of one or more individual children, how they act on the selbri is not specified.

I strongly dislike that {lo} can produce non-individuals and therefore use loi and lo'i accordingly (please don't supply the gi'e example; a "better" solution to that problem in my opinion is either a jai-like LAhE-cast or, if we aren't allowed to make up any new cmavo, to just use {ije} (and if the problem occurs inside an abstraction, it isn't my fault that there isn't an true afterthought bridi connective in the form of {vauJA} or some such)).

A bit off-topic, but I do wonder when and if we'll ever get such an afterthought connective for use within abstractions. Who *hasn't* wanted that at least once? (As well as non-logical forethoughts, I want those too.)

Each individual of the description distributes into the predicate, but it *is not* true that each of the children {simxu lo ka kelci kansa}.

Same as above, they don't distribute.

In fact, if it's okay to just use definitely separate individuals like that, ignoring distribution completely, then {.i mi .e do simxu lo ka cenba} makes perfect sense, which again in my opinion, it most certainly should not, as {mi simxu lo ka cinba .ije do simxu lo ka cinba} is complete nonsense.

That is definitely nonsense. ".e" produces two different bridi, the connected sumti are rather unrelated. This is quite a different case from lo verba which is a single sumti, a single collection of individuals.

mu'o mi'e la selpa'i

-- 
fi'o co'e ko'a ki'e soi la'e vei jo'i pe'o su'i by 
   lo nu lu tu'e ne zu'i zi'e noi toldi 
   nu'i li rau ke me dei to be zi'o ce'u du zo'e bu'a

[Jacob Errington]

On 7 September 2012 07:59, selpa'i <m...@plasmatix.com> wrote:

Am 07.09.2012 02:42, schrieb Jacob Errington:

On 6 September 2012 19:57, Jorge Llambías <jjlla...@gmail.com> wrote:

On Thu, Sep 6, 2012 at 8:02 PM, Jacob Errington <nict...@gmail.com> wrote:
>
> latro'a and I have a strict(er) view of Lojban in that regard and believe
> than simxu1 must be a set, and that the simxu action is fully pairwise.

But in what sense does that definition make Lojban more strict? Are
you saying that a predicate with the more vague meaning is simply not
a possible predicate in your strict version of Lojban?

Perhaps more "rigid" rather than "strict", but in general, it means less intuitive interpretation. To a beginner, {lo verba cu simlu lo ka kelci} is probably intuitively correct. Indeed, it is understandable, but it doesn't have that rigid correctness that I adhere to. Similarly, this overall rigidness involves dislike for {kakne lo nu broda} (should be {ka}) and {zmadu fi lo ka broda} (should be {ni}). Although I have little evidence that actually supports this, this interpretation probably makes things simpler to formally define in lojban.

In my honest opinion, {lo verba cu simxu lo ka ce'u kelci kansa ce'u} is just nonsense, because it simply isn't distributive.

It's not distributive until you add an outer quantifier. lo verba is a collection of one or more individual children, how they act on the selbri is not specified.

Yes, I'm already aware of what you (and certainly others like xorxes) believe about {lo}. What I'm trying to say is that I dislike that (I think I said that already didn't I..). I use {lo} only for distribution. I enjoy keeping things separate. Sure you can go for insane genericness, but that's just not what I want.
 

I strongly dislike that {lo} can produce non-individuals and therefore use loi and lo'i accordingly (please don't supply the gi'e example; a "better" solution to that problem in my opinion is either a jai-like LAhE-cast or, if we aren't allowed to make up any new cmavo, to just use {ije} (and if the problem occurs inside an abstraction, it isn't my fault that there isn't an true afterthought bridi connective in the form of {vauJA} or some such)).

A bit off-topic, but I do wonder when and if we'll ever get such an afterthought connective for use within abstractions. Who *hasn't* wanted that at least once? (As well as non-logical forethoughts, I want those too.)

Naturally, but don't we have joigi...gi... for non-logical forethought? Still, I like A -> JA, JA -> JOI ;)
 

Each individual of the description distributes into the predicate, but it *is not* true that each of the children {simxu lo ka kelci kansa}.

Same as above, they don't distribute.

Same as above, that's what you say and use, not what I do or believe.
 

In fact, if it's okay to just use definitely separate individuals like that, ignoring distribution completely, then {.i mi .e do simxu lo ka cenba} makes perfect sense, which again in my opinion, it most certainly should not, as {mi simxu lo ka cinba .ije do simxu lo ka cinba} is complete nonsense.

That is definitely nonsense. ".e" produces two different bridi, the connected sumti are rather unrelated. This is quite a different case from lo verba which is a single sumti, a single collection of individuals.

Of course, I'm exaggerating. It looks like you aren't taking into account my view of it, which is that if {lo} only produces individuals (which is what I believe and not what you believe, let me say that again) and that using individuals is fine in situations like {lo verba cu kelci kansa simxu} then it implies that it's fine to use {.e} as well as {.e} clearly implies individuals.