Reasoning by analogy

[scope845h...@icebubble.org]

As imprecise human beings saddled with the inheritance of natural
language, much of what we say involves reasoning by analogy. For
example:

"You can learn to walk on your hands just like you learn to walk on
your feet."

How would you express such a notion in Lojban? The English suggests
that the method of learning (perhaps {ta'i} or {xelcli}) is the same in
both instances, but there is also an implication: IF you can learn to
walk on your feet, THEN you can also learn to walk on your hands. In
Lojban, I might render this:

do cilre fi lonu do cadzu fi lei do jamfu kei ku fu da
.inaja do cilre fi lonu do cadzu fi lei do xance kei ku fu da

or even

ta'i da do cilre fi lonu do cadzu fi lei do jamfu
.inaja ta'i da do cilre fi lonu do cadzu fi lei do xance

But that only works if there happens to be a FA place or BAI cmavo that
expresses the common property of the analogy (in this case, the method
of learning). And it requires that I express that property explicitly.
In some instances, there might not be a BAI cmavo, or a {FIhO SE BRODA}
tag, suitable to express it. Or, I might just want to elide the
relationship, and leave the relationship implicit. I could use {do'e}:

do'e da do cilre fi lonu do cadzu fi lei do jamfu
.inaja do'e da do cilre fi lonu do cadzu fi lei do xance

But this still strikes me as kind of clumsy. Is there a neater way to
express reasoning by analogy, like this, in Lojban?

[mezohe]

2020-12-18 04:04 scope845h...@icebubble.org:

>
> "You can learn to walk on your hands just like you learn to walk on
> your feet."
>

A first, quick and dirty attempt:

{lo'e xance lo'e jamfu cu simsa lo ka ce'u goi ko'a zo'u ce'u tadji lo
ka ce'u ka'elre lo ka ce'u cadzu fi ko'a}

{simsa} could be substituted by something more specific, and the {lo'e}
creatures could be properties instead, of course.

On the other hand, there's the question of what a {tadji} really is;
specifically, whether a {tadji be lo ka co'e lo'e xance} itself refers
to hands in some way. I'd rather not get into that now.


Another vague, naturalistic way, similar to yours with {do'e}:

{pa'a gi lo'e xance gi lo'e jamfu zo'u ka'e ka'elre lo ka cadzu fi ri}

[Jacob Thomas Errington]

Often times when translating from English to Lojban, it turns out that a
conjunction such as "just like" winds up being the main claim of the
Lojban. In this case, I'd translate as:

.i lo xance lo jamfu cu simsa lo ka cilre da'i lo du'u ta'i makau cadzu
sepi'o ce'u

the hands the feet - are-similar-in the property learning hypothetically
the fact by-method what?-indirect walk using it

Here the idea is exactly the observation you made too: to abstract what
is common between the hands and feet, namely that there is a method by
which you can walk using them, and that you can learn this method. The
abstraction {lo ka cilre da'i lo du'u ta'i makau cadzu sepi'o ce'u}
represents exactly that common property.

(If you're familiar with functional programming or lambda calculus, it's
kind of like writing \X -> learn hypothetically the fact by-method
what?-indirect walk using X. The "what?-indirect" essentially acts to
say what the output of the function is, and in this case to identify the
thing that is similar between the hands and the feet.)

I'm not sure if trying to express it using .inaja is right. In
principle, if someone knew how to walk on their hands and not on their
feet, then the same analogy would also apply to them trying to learn to
walk on their feet! So really, I think the main claim here is that hands
and feet are similar, but in a very specific way.

.i mi'e la tsani mu'o

[scope845h...@icebubble.org]

Jacob Thomas Errington <ja...@mail.jerrington.me> writes:

> Often times when translating from English to Lojban, it turns out that
> a conjunction such as "just like" winds up being the main claim of the
> Lojban. In this case, I'd translate as:

I've suspected that for a while, but had never seen it mentioned
anywhere, outright. I'm glad to see it confirmed. :) The only way I've
been able to translate sentences such as "I'm stronger than you." has
been to (like you suggest) express the comparison as the main selbri,
i.e., {mi do zmadu le ni tsali}.

"You can learn to walk on your hands just like you learn to walk on
your feet."

lo xance ce'o da lo jamfu ce'o da cu simsa lo ka ka'e cilre fu da
fi lo nu do cadzu fi ce'u

Would that be right? I don't know how to pass multiple arguments into
an abstraction (or lambda expression) in Lojban... especially when they
occupy two different levels of (nested) abstraction.

> I'm not sure if trying to express it using .inaja is right. In
> principle, if someone knew how to walk on their hands and not on their
> feet, then the same analogy would also apply to them trying to learn
> to walk on their feet! So really, I think the main claim here is that
> hands and feet are similar, but in a very specific way.

Right, but most people who walk do so on their feet, and few people walk
on their hands. So, there is an implication that it's the upright
walkers who will be learning to walk upside down (on their hands).

do ka'e cilre cei broda fu da fe lo jai ta'i nu cadzu fi lo jamfu
.ije da simsa de
.ije seni'i bo broda fu de fe lo jai ta'i nu cadzu fi lo xance

That sort of expresses it, but it still seems kind of clumsy. Use of
{ni'i} is perhaps more accurate than {.inaja}.

Also, there's no need to copy me, as I am already on the mailing list.

[Jacob Thomas Errington]

On 2020-12-19 18:14, scope845h...@icebubble.org wrote:

> Jacob Thomas Errington <ja...@mail.jerrington.me> writes:
>
>> Often times when translating from English to Lojban, it turns out that
>> a conjunction such as "just like" winds up being the main claim of the
>> Lojban. In this case, I'd translate as:
> I've suspected that for a while, but had never seen it mentioned
> anywhere, outright. I'm glad to see it confirmed. :) The only way I've
> been able to translate sentences such as "I'm stronger than you." has
> been to (like you suggest) express the comparison as the main selbri,
> i.e., {mi do zmadu le ni tsali}.
>
> "You can learn to walk on your hands just like you learn to walk on
> your feet."
>
> lo xance ce'o da lo jamfu ce'o da cu simsa lo ka ka'e cilre fu da
> fi lo nu do cadzu fi ce'u
>
> Would that be right? I don't know how to pass multiple arguments into
> an abstraction (or lambda expression) in Lojban... especially when they
> occupy two different levels of (nested) abstraction.

So you want to emphasize that you _learn_ each one the same way? In
other words, walking on your hands and walking on your feet are similar
in the way that they are learned to do. We can express this with an
indirect question.

lo xance lo jamfu cu simsa lo ka ka'e cilre fu makau fi lo nu cadzu fi ce'u

Although I think the indirect question formulation is the most natural,
I think you could also formulate it in a similar way as your example
with {da}:

.i da zo'u lo xance lo jamfu cu simsa lo ka ka'e cilre fu da fi lo nu
cadzu fi ce'u

We don't need to pass in the variable {da}. We need to explicitly bind
it outside (otherwise it would be implicitly bound inside the
abstraction), but then we can simply refer to it inside. Then the
interpretation is that it's the same way of learning that's applying to
both instances of the abstraction (the one for feet and the one for hands).

For completeness, I want to say that actually passing in multiple things
to an abstraction isn't really supported in Lojban. I wrote an article
about a way to do it using a pattern-matching syntax:
https://mw.lojban.org/papri/Abstraction_Pattern_Syntax It requires the
experimental cmavo {ce'ai}.

>> I'm not sure if trying to express it using .inaja is right. In
>> principle, if someone knew how to walk on their hands and not on their
>> feet, then the same analogy would also apply to them trying to learn
>> to walk on their feet! So really, I think the main claim here is that
>> hands and feet are similar, but in a very specific way.
> Right, but most people who walk do so on their feet, and few people walk
> on their hands. So, there is an implication that it's the upright
> walkers who will be learning to walk upside down (on their hands).
>
> do ka'e cilre cei broda fu da fe lo jai ta'i nu cadzu fi lo jamfu
> .ije da simsa de
> .ije seni'i bo broda fu de fe lo jai ta'i nu cadzu fi lo xance
>
> That sort of expresses it, but it still seems kind of clumsy. Use of
> {ni'i} is perhaps more accurate than {.inaja}.

I'm still not sure whether this means the right thing.

exists X, exists Y:
(you can learn to walk on your feet by X) AND (X is similar to Y) ==>
(you can learn to walk on your hands by Y)

I'm not really sure I'd be able to interpret this to mean "You can learn
to walk on your hands the same way you can learn to walk on your hands.

> Also, there's no need to copy me, as I am already on the mailing list.

Sorry about that, I mistakenly clicked reply instead of reply list ^^"

[scope845h...@icebubble.org]

Jacob Thomas Errington <ja...@mail.jerrington.me> writes:

> For completeness, I want to say that actually passing in multiple
> things to an abstraction isn't really supported in Lojban. I wrote an
> article about a way to do it using a pattern-matching syntax:
> https://mw.lojban.org/papri/Abstraction_Pattern_Syntax It requires the
> experimental cmavo {ce'ai}.

What about something like this:

la lojban. bangu mi'o <-->

la lojban. ce'o mi'o ckaji loka ce'u bangu ce'u

ta'onai

On the original point (about reasoning by analogy), however, it could
very well be that Lojban was designed to make reasoning by analogy
difficult to express... if not explicitly on purpose, then as a
consequence of it's logical nature. It is, after all, the "logical"
language. If your goal is to make logical relationships explicit, and
avoid fuzzy reasoning, then it would be sensible to make reasoning by
analogy cumbersome to express in Lojban, or at least require its logic
to be expressed explicitly.

I was thinking about the possibility of using a {fi'o} modal tag, so I
checked the grammar for places where {fi'o} can be used. As it turns
out, there aren't very many, and there are even fewer where {fi'o} could
be used for this purpose. I found only two possibilities:

Using a {fi'o} tag in front of {tu'e}:

do ka'e cilre fi lonu do cadzu fi lo jamfu
.iseni'ibo fi'o simsa la'edi'u tu'e
do ka'e cilre fu lo xego'i fi lonu do cadzu fi lo xance

Using a {fi'o} tag in front of the selbri:

do ka'e cilre fi lonu do cadzu fi lo jamfu
.iseni'ibo do fi'o simsa la'edi'u je ka'e
cilre fu lo xego'i fi lonu do cadzu fi lo xance

These would seem to be fairly general solutions, too: connecting the two
claims using either {.iseni'ibo fi'o simsa la'edi'u tu'e} or {.iseni'ibo
<leading sumti> fi'o simsa la'edi'u fe'u} should work for whatever bridi
are being held in analogy, right?

[Jacob Thomas Errington]

On 2021-01-01 19:56, scope845h...@icebubble.org wrote:

> What about something like this:
>
> la lojban. bangu mi'o <-->
>
> la lojban. ce'o mi'o ckaji loka ce'u bangu ce'u

The problem there is using a binary relation {lo ka ce'u ce'u bangu}
where ckaji2 should be a unary relation. If we allow this, then we can't
unambiguously interpret {ko'a ce'o ko'e ckaji lo ka broda}. Is it
unpacking the tuple or not?

> I was thinking about the possibility of using a {fi'o} modal tag, so I
> checked the grammar for places where {fi'o} can be used. As it turns
> out, there aren't very many, and there are even fewer where {fi'o} could
> be used for this purpose. I found only two possibilities:

You can put fi'o tags in more places than that. Try these in a parser.

.i do jamfu cadzu cilre kakne .iseni'ibo fi'o simsa la'e di'u do xance
cadzu cilre kakne

.i do jamfu co'e .i fi'o simsa bo do xance co'e

.i ko'a jai fi'o broda fe'u brode

More generally, {fi'o broda} is pretty much equivalent to a BAI cmavo,
so you can use it to form a connective with {bo}, you can use it with
{jai}, you can use it before a selbri, you can connect it with other
tags, and you can use it as a term if you follow it with a sumti or the
terminator {ku}.

> Using a {fi'o} tag in front of {tu'e}:
>
> do ka'e cilre fi lonu do cadzu fi lo jamfu
> .iseni'ibo fi'o simsa la'edi'u tu'e
> do ka'e cilre fu lo xego'i fi lonu do cadzu fi lo xance
>
> Using a {fi'o} tag in front of the selbri:
>
> do ka'e cilre fi lonu do cadzu fi lo jamfu
> .iseni'ibo do fi'o simsa la'edi'u je ka'e
> cilre fu lo xego'i fi lonu do cadzu fi lo xance
>
> These would seem to be fairly general solutions, too: connecting the two
> claims using either {.iseni'ibo fi'o simsa la'edi'u tu'e} or {.iseni'ibo
> <leading sumti> fi'o simsa la'edi'u fe'u} should work for whatever bridi
> are being held in analogy, right?

This seems okay to me, but I don't think it's as precise as it would be
to directly go for using {simsa}.

The usual strategy to interpret a {fi'o} clause is to rearrange to make
its selbri the top-level selbri. For example, I would interpret {mi fi'o
simsa do se bangu lo lojbo} as

  mi do simsa lo ka lo lojbo cu bangu
  + a claim that {mi se bangu lo lojbo}

The same idea applies to BAI, so {.i broda .i seni'i bo brode} is
interpreted as

  lo du'u broda cu nibli lo du'u brode
  + the fact that both {broda} and {brode} are claimed

When there are multiple BAI, tenses, fi'o-clauses, quantifiers, etc.
(generally called bridi operators) at the same level in a sentence,
they're interpreted from left to right.

Anyway, using the usual interpretation strategy, we don't get the right
meaning from your examples. They mean "it is possible that you learn to
walk on your feet ==> similar to that, it is possible that you learn by
the method of learning to walk on your hands to walk on your feet."

What strikes me about that is that it's saying that the similarity
between the possibility of learning to walk on your hands and on your
feet is implied by the possibility of learning to walk on your feet. But
I don't think there's a logical implication there at all, is there?

An analogy is not an implication. It's an observed similarly that is
used to make inferences. So the order of the bridi operators seems a bit
backwards in the fi'o examples.

Taking a step back and working from this idea of what an analogy is, we
want to say "walking on your hands is similar to walking on your feet,
which implies that _learning_ to walk on your hands is similar to
_learning_ to walk on your feet." Here's my take on that in Lojban:

.i lo xance lo jamfu cu simsa lo ka kakne co cadzu fi ce'u kei .e ja'e
bo lo ka makau xe cilre co cadzu fi ce'u
Hands and feet are similar in that one can walk on them, and therefore
similar in what way one can learn to walk on them.

This approach also generalizes: {ko'a ko'e simsa lo ka broda [kei] .e
ja'e bo lo ka brode}. For example,

  lo najnimre lo plise cu simsa lo ka farvi bu'u lo tricu kei .e ja'e
bo lo ka makau tadji co kurji
  Oranges and apples are similar in that they grow in trees, and
therefore similar in what way one cares for them.

[scope845h...@icebubble.org]

Jacob Thomas Errington <ja...@mail.jerrington.me> writes:

> On 2021-01-01 19:56, scope845h...@icebubble.org wrote:
>
>> What about something like this:
>>
>> la lojban. bangu mi'o <-->
>>
>> la lojban. ce'o mi'o ckaji loka ce'u bangu ce'u
>
> The problem there is using a binary relation {lo ka ce'u ce'u bangu}
> where ckaji2 should be a unary relation. If we allow this, then we
> can't unambiguously interpret {ko'a ce'o ko'e ckaji lo ka broda}. Is
> it unpacking the tuple or not?

An example of a sentence exhibiting that ambiguity might be {ko'a ce'o
ko'e ckaji loka porsi}. Would that mean {ko'a ce'o ko'e porsi} or {ko'a
porsi ko'e}? It would be ambiguous.

A way to resolve such ambiguity might be to treat {lo ka broda} as a
solitary sumti (not a sequence) when the {ka} phrase doesn't contain
{ce'u}, but as a sequence when it contains multiple {ce'u}. Such a rule
could become unwieldy when the property contains many arguments, i.e.
{lo ka ce'u broda ce'u ce'u ce'u ...}. So, I might propose the
following rule:

(1) The property is interpreted as a sequence when two or more {ce'u}
are explicity expressed within it.

(2) Otherwise (when there are zero or one {ce'u}), it is interpreted
as a solitary sumti (as opposed to a sequence).

(3) Any unexpressed places following the second {ce'u}, if any, are
assumed to be {ce'u}.

This rule would introduce two questions: (A) How would one speak about a
sequence with just one element?, and (B) How would one differentiate
between properties with different numbers of arguments (arities)?

If I say {ko'a ce'o zi'o ckaji lo ka ce'u porsi}, I'm speaking about a
sequence containing one element, {ko'a}, but I'm not saying that {ko'a
porsi}. On the other hand, if I say {ko'a ckaji loka ce'u ce'u porsi},
then it'd be true that {ko'a porsi}; it'd be the sequence {lo porsi ce'o
lo se porsi be ri}. The problem of specifying a sequence with just one
element (A) is a challenged posed by Lojban's infix syntax, not by
abstractors like {ka}.

Issue (B), however, does not suggest any easy solutions. Would {lo ka
ce'u broda ce'u} be a property with two arguments, or a property with
more than two arguments, such as {lo ka ce'u broda ce'u ce'u} or {lo ka
ce'u broda ce'u ce'u ce'u}, with the trailing {ce'u} unexpressed? It
might be possible to quantify the number of arguments.

A ternary property, {lo ka ce'u broda ce'u ce'u}, could be expressed as
something like {lo ci ka broda} or {lo cimei ka broda}. However, the
tanru form would introduce the semantic ambiguity inherent in tanru.
And the form with an inner quantifier might run afoul of Lojban's normal
interpertation of quantifiers.

This could be solved, however, by treating quantified abstractions as
"magic", kind of like the way quantified variables of selma'o GOhA are
treated magically in the terms of a prenex (preceeding {zo'u}). I.e.,
{ro bu'a zo'u broda} doesn't mean the same thing as {roda zo'u broda},
because {ro bu'a zo'u} is magic. So, if we consider {lo su'o ka} magic,
could we use this mechanism to speak about multi-variable properties?

[scope845h...@icebubble.org]

Jacob Thomas Errington <ja...@mail.jerrington.me> writes:

> You can put fi'o tags in more places than that. Try these in a parser.
>
> .i do jamfu cadzu cilre kakne .iseni'ibo fi'o simsa la'e di'u do xance
> cadzu cilre kakne

Hm. I hadn't thought of using a {fi'o} tag on a sumti of selma'o KOhA.

> .i do jamfu co'e .i fi'o simsa bo do xance co'e

I don't think that's grammatical. jufra can only be connected by simple
tags, not full-fledged sumti tcita. Otherwise, I could simply say:

do ka'e cilre fi lonu do cadzu fi lo jamfu

.i fi'o simsa la'edi'u fe'u je semu'i bo

do ka'e cilre fu lo xego'i fi lonu do cadzu fi lo xance

That would be nice and clean (and could further be simplified by use of
{go'i}, et al.), but the grammar doesn't allow it. .uu

> The usual strategy to interpret a {fi'o} clause is to rearrange to
> make its selbri the top-level selbri. For example, I would interpret
> {mi fi'o simsa do se bangu lo lojbo} as
>
>   mi do simsa lo ka lo lojbo cu bangu
>   + a claim that {mi se bangu lo lojbo}

No, I don't think that's how modal tags are interpreted. At least
according to CLL, a modal tag adds an additional place to the underlying
selbri, expressing something that fills the x1 place of the {BAI} or
{FIhO} construct. Your interpretation of that sentence would properly
be expressed: {lo me mi be fi'o simsa do me'u cu se bangu lo lojbo}.
Adding a modal tag to a sumti is not equivalent to using that tag in a
non-logical connection between sentences.

> What strikes me about that is that it's saying that the similarity
> between the possibility of learning to walk on your hands and on your
> feet is implied by the possibility of learning to walk on your
> feet.

Ah, yes, you have a point, there. Implication of similarity wasn't the
meaning I'd intended. My mistake. This could be fixed using prophor:

fi'o simsa la'edi'e fe'u do ka'e cilre fi lonu do cadzu fi lo jamfu
.iseni'ibo do ka'e cilre fu lo xego'i fi lonu do cadzu fi lo xance

> An analogy is not an implication. It's an observed similarly that is
> used to make inferences. So the order of the bridi operators seems a
> bit backwards in the fi'o examples.

Exactly! That's the whole idea of reasoning by analogy: A and B both
have property X, and A has property Y, therefore (by analogy), B has
property Y. As an example: Ducks and geese both have webbed feet.
Ducks have feathers. Therefore, geese have feathers. It's not logical.
It's reasoning by analogy. It's illogical. Yet it's quintessentially
human, and very common in natural language usage.

> .i lo xance lo jamfu cu simsa lo ka kakne co cadzu fi ce'u kei .e ja'e
> bo lo ka makau xe cilre co cadzu fi ce'u
> Hands and feet are similar in that one can walk on them, and therefore
> similar in what way one can learn to walk on them.

I think you're using {co} incorrectly. The place structure of {broda co
brode} is that of {broda}, not that of {brode}. Using {ni'i} instead of
{ja'e}, I might express your example as:

.i lo xance lo jamfu cu simsa lo ka kakne cadzu fi ce'u kei
.e ni'i bo lo ka cilre be fu makau cadzu fi ce'u

Or, using {co}:

.i lo xance lo jamfu cu simsa lo ka cadzu co kakne fi ce'u kei
.e ni'i bo lo ka cadzu co cilre be fu makau fi ce'u

>   lo najnimre lo plise cu simsa lo ka farvi bu'u lo tricu kei .e ja'e
> bo lo ka makau tadji co kurji

I think you mean {be}, not {bu'u}:

lo najnimre lo plise cu simsa lo ka farvi be lo tricu kei .e
ja'e ja seni'i bo lo ka makau kurji tadji

[Hugh O'Byrne]

Oh, man.  It's like walking past an expensive restaurant, and getting the smell of... oh, just sumptuous.  I love it, but all I can do is smell it.  I don't have the brainpower to be in the circle, but I absolutely love that this kind of discussion happens, and it's fun even to be on the periphery catching a whiff.

My respect to Lojban enthusiasts and scholars.  And my thanks, for the opportunity to vicariously experience such intellectual sport!

Hugh.

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lojban/86im89g450.fsf%40cmarib.ramside.

[Jacob Thomas Errington]

On 2021-01-06 16:13, scope845h...@icebubble.org wrote:
> Jacob Thomas Errington <ja...@mail.jerrington.me> writes:
>
>> You can put fi'o tags in more places than that. Try these in a parser.
>>
>> .i do jamfu cadzu cilre kakne .iseni'ibo fi'o simsa la'e di'u do xance
>> cadzu cilre kakne
> Hm. I hadn't thought of using a {fi'o} tag on a sumti of selma'o KOhA.
>
>> .i do jamfu co'e .i fi'o simsa bo do xance co'e
> I don't think that's grammatical. jufra can only be connected by simple
> tags, not full-fledged sumti tcita. Otherwise, I could simply say:
>
> do ka'e cilre fi lonu do cadzu fi lo jamfu
> .i fi'o simsa la'edi'u fe'u je semu'i bo
> do ka'e cilre fu lo xego'i fi lonu do cadzu fi lo xance
>
> That would be nice and clean (and could further be simplified by use of
> {go'i}, et al.), but the grammar doesn't allow it. .uu

What I wrote, {.i fi'o broda bo} is grammatical. Try it in a parser!

http://camxes.lojban.org/?text=.i%20fi%27o%20broda%20bo

You wrote something slightly different: {.i fi'o broda ko'a bo}. That's
indeed ungrammatical.

>> The usual strategy to interpret a {fi'o} clause is to rearrange to
>> make its selbri the top-level selbri. For example, I would interpret
>> {mi fi'o simsa do se bangu lo lojbo} as
>>
>>   mi do simsa lo ka lo lojbo cu bangu
>>   + a claim that {mi se bangu lo lojbo}
> No, I don't think that's how modal tags are interpreted. At least
> according to CLL, a modal tag adds an additional place to the underlying
> selbri, expressing something that fills the x1 place of the {BAI} or
> {FIhO} construct. Your interpretation of that sentence would properly
> be expressed: {lo me mi be fi'o simsa do me'u cu se bangu lo lojbo}.
> Adding a modal tag to a sumti is not equivalent to using that tag in a
> non-logical connection between sentences.

It turns out that in the decades since the CLL was published that the
language has continued to evolve and that interpretations have been
refined. The CLL interpretation for fi'o falls a bit flat because it
doesn't explain the connection between this new place and the old
selbri, whereas the 'new' interpretation is for the most part compatible
with the old one (gives essentially the same interpretations) while
being more precise.

No, I'm using {co} correctly. It's just a tricky beast.

{.i ko'a ... broda co brode fo'a ...} has the sumti {ko'a ...} filling
places of {broda} starting at x1, and has the sumti {fo'a ...} filling
the places of {brode} starting at x2. The upshot is that one can use
{co} to elide abstractors using tanru.

.i mi djica co cliva ti

I forget the exact CLL reference, but this is indeed in the CLL.

>>   lo najnimre lo plise cu simsa lo ka farvi bu'u lo tricu kei .e ja'e
>> bo lo ka makau tadji co kurji
> I think you mean {be}, not {bu'u}:
>
> lo najnimre lo plise cu simsa lo ka farvi be lo tricu kei .e
> ja'e ja seni'i bo lo ka makau kurji tadji

Sure

[Jacob Thomas Errington]

On 2021-01-06 14:13, scope845h...@icebubble.org wrote:

>
> An example of a sentence exhibiting that ambiguity might be {ko'a ce'o
> ko'e ckaji loka porsi}. Would that mean {ko'a ce'o ko'e porsi} or {ko'a
> porsi ko'e}? It would be ambiguous.
>
> A way to resolve such ambiguity might be to treat {lo ka broda} as a
> solitary sumti (not a sequence) when the {ka} phrase doesn't contain
> {ce'u}, but as a sequence when it contains multiple {ce'u}. Such a rule
> could become unwieldy when the property contains many arguments, i.e.
> {lo ka ce'u broda ce'u ce'u ce'u ...}. So, I might propose the
> following rule:
>
> (1) The property is interpreted as a sequence when two or more {ce'u}
> are explicity expressed within it.
>
> (2) Otherwise (when there are zero or one {ce'u}), it is interpreted
> as a solitary sumti (as opposed to a sequence).
>
> (3) Any unexpressed places following the second {ce'u}, if any, are
> assumed to be {ce'u}.

Rule 3 only really works in simple abstractions; how does one deal with
situations like {lo ka ce'u broda lo pendo be ce'u}?

>
> This rule would introduce two questions: (A) How would one speak about a
> sequence with just one element?, and (B) How would one differentiate
> between properties with different numbers of arguments (arities)?
>
> If I say {ko'a ce'o zi'o ckaji lo ka ce'u porsi}, I'm speaking about a
> sequence containing one element, {ko'a}, but I'm not saying that {ko'a
> porsi}. On the other hand, if I say {ko'a ckaji loka ce'u ce'u porsi},
> then it'd be true that {ko'a porsi}; it'd be the sequence {lo porsi ce'o
> lo se porsi be ri}. The problem of specifying a sequence with just one
> element (A) is a challenged posed by Lojban's infix syntax, not by
> abstractors like {ka}.

I think a pretty reasonable interpretation of {ko'a ce'o zi'o} is a
1-tuple. Reminds me of Python's syntax (x,) for a 1-tuple.

>
> Issue (B), however, does not suggest any easy solutions. Would {lo ka
> ce'u broda ce'u} be a property with two arguments, or a property with
> more than two arguments, such as {lo ka ce'u broda ce'u ce'u} or {lo ka
> ce'u broda ce'u ce'u ce'u}, with the trailing {ce'u} unexpressed? It
> might be possible to quantify the number of arguments.
>
> A ternary property, {lo ka ce'u broda ce'u ce'u}, could be expressed as
> something like {lo ci ka broda} or {lo cimei ka broda}. However, the
> tanru form would introduce the semantic ambiguity inherent in tanru.
> And the form with an inner quantifier might run afoul of Lojban's normal
> interpertation of quantifiers.

I'm not in favour of introducing an exception to the interpretation of
inner quantifiers to be able to specify the number of arguments,
especially when we're just cooking up new interpretations without
consulting any existing usage. Right now, there is no situation where
the number of arguments to a ka-abstraction is completely unclear. The
overt {ce'u} in a ka abstraction each represent separate sumti places
without any implicit tuple unpacking.

> This could be solved, however, by treating quantified abstractions as
> "magic", kind of like the way quantified variables of selma'o GOhA are
> treated magically in the terms of a prenex (preceeding {zo'u}). I.e.,
> {ro bu'a zo'u broda} doesn't mean the same thing as {roda zo'u broda},
> because {ro bu'a zo'u} is magic. So, if we consider {lo su'o ka} magic,
> could we use this mechanism to speak about multi-variable properties?

I don't agree that the existence of a wart in the language is a good
justification for the introduction of new warts. Not to mention that
following the shift in the last ten years to understand ka-abstractions
as reified selbri, there isn't really any reason to use {bu'a} except
maybe a slightly cleaner syntax. Ignoring syntax cleanliness, why say

  .i ro bu'a zo'u da bu'a

when you could say

  .i ro da poi selkai (zi'o) zo'u de da ckaji


I think that tuple unpacking / pattern matching is interesting, which is
why I already solved this exact problem by leveraging the experimental
cmavo {ce'ai}.

The way {ce'ai} works in general is that each sumti appearing in it is
understood as a binder. We concocted {ce'ai} back in ~2013 when
discussing how to reuse {ce'u} within a ka-abstraction. Here's a
contrived example:

  without ce'ai: {.i mi ckaji lo ka ce'u goi ko'a batci ko'a}
  with ce'ai: {.i mi ckaji lo ka ko'a ce'ai ko'a ko'a batci}

The idea is to have something specifically for introducing ce'u-like
bound variables, similar to the way {da} works in {zo'u}. My pattern
matching idea extends the original definition of {ce'ai} to allow more
complex sumti to appear in the {ce'ai}-clause, understanding them as
patterns. For example:

  {ko'a ce'o ko'e ckaji lo ka X ce'o Y ce'ai X broda Y} interprets to
{ko'a broda ko'e}

(I'm writing capital letters to stand for the equivalent BY cmavo.)

It's also aligned with usual programming language theory because we use
the introduction form for tuples {ce'o} as a means of deconstructing the
tuple.

Personally, I see many pros, and very few cons. In fact, the only con I
can see is the use of an experimental cmavo. But using an experimental
cmavo for something that's admittedly a niche application seems right on
brand to me.

.i mi'e la tsani mu'o

P.S. are you a member of roljbogu'e discord server?

[scope845h...@icebubble.org]

Jacob Thomas Errington <ja...@mail.jerrington.me> writes:

>> (3) Any unexpressed places following the second {ce'u}, if any, are
>> assumed to be {ce'u}.
> Rule 3 only really works in simple abstractions; how does one deal
> with situations like {lo ka ce'u broda lo pendo be ce'u}?

.a'onairu'e la'edi'u jetnu

> The way {ce'ai} works in general is that each sumti appearing in it is
> understood as a binder. We concocted {ce'ai} back in ~2013 when
> discussing how to reuse {ce'u} within a ka-abstraction. Here's a
> contrived example:
>
>   without ce'ai: {.i mi ckaji lo ka ce'u goi ko'a batci ko'a}
>   with ce'ai: {.i mi ckaji lo ka ko'a ce'ai ko'a ko'a batci}

ki'a .i zo GOI xu selma'o zo ce'ai
.i zo ZOhU xu go'i .i mi stidi tu'a lu

mi ckaji lo ka ce'u zo'u ce'u ce'u batci

li'u

> P.S. are you a member of roljbogu'e discord server?

na go'i .i le skami ciste na ciste fi le mi zdani .uu

[scope845h...@icebubble.org]

Jacob Thomas Errington <ja...@mail.jerrington.me> writes:

> What I wrote, {.i fi'o broda bo} is grammatical. Try it in a parser!
>
> http://camxes.lojban.org/?text=.i%20fi%27o%20broda%20bo

I'm looking at the EBNF grammar in the CLL, but not seeing it. Are you
using the rule statement-2_13? Odd... the on-line version of the CLL
contains the EBNF, but not the YACC grammar. According to the YACC:

lexer_K_955 : lexer_K_711 I_root_956 BO_508
| lexer_K_711 I_root_956 simple_tag_971 BO_508
;

The "simple_tag_971" doesn't permit {fi'o} clauses. Are you using
grammar.300?

> It turns out that in the decades since the CLL was published that the
> language has continued to evolve and that interpretations have been
> refined. The CLL interpretation for fi'o falls a bit flat because it
> doesn't explain the connection between this new place and the old
> selbri, whereas the 'new' interpretation is for the most part
> compatible with the old one (gives essentially the same
> interpretations) while being more precise.

The CLL's description of the meaning of {fi'o} clauses isn't expressed
particularly clearly, but the meaning expressed is pretty clear. The
meaning of a {fi'o} clause is determined the same way as for BAI cmavo:
it specifies something which fills the x1 place of the BAI or FIhO
clause. Tenses are interpereted slightly differently (exchanging the x1
and x2 places) for historical reasons. What is this "new", modern,
interpretation of {fi'o}, as you understand it?

> No, I'm using {co} correctly. It's just a tricky beast.
>
> {.i ko'a ... broda co brode fo'a ...} has the sumti {ko'a ...} filling
> places of {broda} starting at x1, and has the sumti {fo'a ...} filling
> the places of {brode} starting at x2. The upshot is that one can use
> {co} to elide abstractors using tanru.
>
> .i mi djica co cliva ti
>
> I forget the exact CLL reference, but this is indeed in the CLL.

You're right. It's in CLL section 5.8. I'd totally forgotten about
that.

[Jacob Thomas Errington]

coi

On 2021-01-13 16:17, scope845h...@icebubble.org wrote:
> Jacob Thomas Errington <ja...@mail.jerrington.me> writes:
>> The way {ce'ai} works in general is that each sumti appearing in it is
>> understood as a binder. We concocted {ce'ai} back in ~2013 when
>> discussing how to reuse {ce'u} within a ka-abstraction. Here's a
>> contrived example:
>>
>>   without ce'ai: {.i mi ckaji lo ka ce'u goi ko'a batci ko'a}
>>   with ce'ai: {.i mi ckaji lo ka ko'a ce'ai ko'a ko'a batci}
> ki'a .i zo GOI xu selma'o zo ce'ai
> .i zo ZOhU xu go'i .i mi stidi tu'a lu
>
> mi ckaji lo ka ce'u zo'u ce'u ce'u batci
>
> li'u

.i zo ce'ai cmavo zo zo'u .e nai zo goi .i pilno zo goi lo ka ciksi lo
po'o du'u makau smuni

.i la'e lu lo ka ce'u zo'u ce'u ce'u li'u mutce lo ka simlu lo ka selbri
fi lo cimei .e nai lo pamei .i mukti lo nu mi stidi tu'a zo ce'ai

[Jacob Thomas Errington]

On 2021-01-13 16:51, scope845h...@icebubble.org wrote:

I'm looking at the EBNF grammar in the CLL, but not seeing it.  Are you
using the rule statement-2_13?  Odd... the on-line version of the CLL
contains the EBNF, but not the YACC grammar.  According to the YACC:

lexer_K_955             :  lexer_K_711  I_root_956  BO_508
                        |  lexer_K_711  I_root_956  simple_tag_971  BO_508
                        ;

The "simple_tag_971" doesn't permit {fi'o} clauses.  Are you using
grammar.300?

I just use camxes. Perhaps the CLL doesn't allow fi'o clauses as sentence connectives due to a limitation in the parser (it would require too much lookahead, maybe?) Camxes essentially allows unlimited lookahead.

Anyway, there's no compelling reason that it should be forbidden, if we believe that fi'o clauses should be equivalent to BAI tags.

It turns out that in the decades since the CLL was published that the
language has continued to evolve and that interpretations have been
refined. The CLL interpretation for fi'o falls a bit flat because it
doesn't explain the connection between this new place and the old
selbri, whereas the 'new' interpretation is for the most part
compatible with the old one (gives essentially the same
interpretations) while being more precise.

The CLL's description of the meaning of {fi'o} clauses isn't expressed
particularly clearly, but the meaning expressed is pretty clear.  The
meaning of a {fi'o} clause is determined the same way as for BAI cmavo:
it specifies something which fills the x1 place of the BAI or FIhO
clause.  Tenses are interpereted slightly differently (exchanging the x1
and x2 places) for historical reasons.  What is this "new", modern,
interpretation of {fi'o}, as you understand it?

Here's the explanation I gave in my message on 2 January:

The usual strategy to interpret a {fi'o} clause is to rearrange to make its selbri the top-level selbri. For example, I would interpret {mi fi'o simsa do se bangu lo lojbo} as

  mi do simsa lo ka lo lojbo cu bangu
  + a claim that {mi se bangu lo lojbo}

The same idea applies to BAI, so {.i broda .i seni'i bo brode} is interpreted as

  lo du'u broda cu nibli lo du'u brode
  + the fact that both {broda} and {brode} are claimed 

I do need to mention thought that I don't particularly _like_ this example using simsa, since you have to "split" the sentence to fill two separate places of simsa (namely x1 and x3). It's cleaner when the whole enclosing bridi moves directly into a nu or du'u abstraction, e.g.

  .i fi'o djuno mi lo bruna cu djica co sonci binxo

And I would interpret as

  .i mi djuno lo du'u lo bruna cu djica co sonci binxo + a claim that "lo bruna cu djica co sonci binxo".

This lines up with the way tenses are interpreted, except as you mentioned, tenses are backwards.

  .i mi pu lo nu sipna cu cadzu lo klaji
  .i lo nu mi cadzu lo klaji cu purci lo nu sipna + a claim that "mi cadzu lo klaji"

[scope845h...@icebubble.org]

Jacob Thomas Errington <ja...@mail.jerrington.me> writes:

> .i zo ce'ai cmavo zo zo'u .e nai zo goi .i pilno zo goi lo ka ciksi lo
> po'o du'u makau smuni

.i na simlu fi mi .ini'ibo mi se slabu zo zo'u
.i ku'i piso'iroi ciska lo'u zo'u: le'u
.i na lojbo lerfu .i ku'i sidju fi lonu ticdu ja jimpe zo zo'u

[scope845h...@icebubble.org]

Jacob Thomas Errington <ja...@mail.jerrington.me> writes:

> I just use camxes. Perhaps the CLL doesn't allow fi'o clauses as
> sentence connectives due to a limitation in the parser (it would
> require too much lookahead, maybe?) Camxes essentially allows
> unlimited lookahead.

AFAIK, the latest official grammar is grammar.300. All those PEG
grammars are unofficial, as I understand. The PEG grammars may
introduce new features (such as full "tag"s where "simple-tag"s used to
be), but that introduces the question of WHICH grammar is being used.
Whenever you ask whether something is grammatical, you must specify:
grammatical with respect to which (official or experimental) grammar?

That said, it would be VERY nice if "tag"s replaced all the
"simple-tag"s in the grammar. It would also be nice if cmavo of NAhE
could be joined by joiks/jeks. Lojban has several asymetries like this
which could be resolved by adopting a new official grammar.

> The usual strategy to interpret a {fi'o} clause is to rearrange to
> make its selbri the top-level selbri. For example, I would interpret
> {mi fi'o simsa do se bangu lo lojbo} as
>
>   mi do simsa lo ka lo lojbo cu bangu
>   + a claim that {mi se bangu lo lojbo}

I still think you're mis-stating this. I think what you mean to
describe is {mi fi'o simsa BO do se bangu lo lojbo}. However, modal
connections between sumti are not allowed in grammar.300. They might be
allowed in a camxes-like grammar. (I haven't checked.)

Still, even if grammatical, this wouldn't imply any different (or
updated) interpretation of modal tags than appears in CLL. I would
still interpret {jai fi'o gasnu fe'u broda} as {jai gau broda}. The
strange constructs appear when you use {fi'o} clauses as tenses, i.e.:

.i ti fi'o bredi fe'u cutci

As I understand it, this would mean:

.i ti cutci
.i bredi lonu go'i

This would be the interpretation of modal tags as explained in the CLL.
What I think you're saying is that {fi'o} tags are just allowed in more
places (in PEG grammars) than they used to be (in YACC), not that there
are necessarily any new interpretations of what a modal tag means.