[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lojban-beginners] Does the argument limit lead to half-ass words?

On 5/16/2015 10:48 AM, TR NS wrote:

On Saturday, May 16, 2015 at 2:22:37 AM UTC-4, la gleki wrote:
    2015-05-16 5:10 GMT+03:00 TR NS <tran...@gmail.com>:

        In my recent studies of Lojban (and Loglan) I've started to
        question the efficacy of the argument system. One the things
        that struck me was the word for "run".

             barja x1 runs on x2 using limbs x3 with gait x4


    {bajra} ({lo barja} is a tavern, bar)


Ha! I was wondering why the rafsi was `baj` ;-)
bajra could have been assigned either baj or bar, but there were usually multiple candidates for each rafsi, and the tradeoffs were complex (probably too complex to be worth explaining at this point). 


        That seems a strange definition. I can't really think of single
        time I ever needed to express that the running was done with
        anything other then legs.
bajra is intended to cover more than one kind of human locomotion, as the gait place indicates. 


    I watched many videos on Youtube where people were able to run using
    their hands. Besides, horses don't necessarily have hands and feet.

        Perhaps it would be useful when talking about Oscar
        Pistorius Olympic races, but that's a rather rare case!
Rare for whom? Not that rarity of expressing a given place was of prime importance, since that is often associated with the particular language and lexical item (and bajra is not the same as the English "run") 


    So we'll use all places of this verb very seldom! Not all places are
    always needed.
Correct. All places are implicit to the concept, but context often makes explicit usage unnecessary. Humans have only 2 legs and seldom "run" using anything but both of their legs. We also rarely talk about human gaits, unless they are abnormal. 


        When I think of running, it tends to be *to* some place or at
        least *via* some path.
Then you are thinking of runningly-going, not running itself. To/from are not necessarily part of running, or it would be meaningless to talk of "running in place". 


The word "barja" really doesn't seem like
        the idea of running. It seems more akin to "treading",
"tread" is more strongly associated with walking, rather than running, and is ambiguous between a particular kind/gait of walking and the walking itself. 


    Yes, indeed. Conciseness of English definitions can sometimes lead
    to their incorrect interpretations.
    Here is one of my examples (not sure where I got it):

    xu do su'o roi senva lo ka bajra lo jdika grana lo xance be do —
    Have you ever dreamt of running on your hands on a narrow rod?

Yea, I am not saying the limbs isn't a valid argument. It is. It's just
that it seems a less useful than a "to", "from" or "via".
We have "klama" and "muvdu" and "benji" for referring to interactions between origins and destinations. "bajra" is not such an interaction, but is an interaction between an animate "actor", the limbs being used and the surface that the limbs are being used on (and gait, which is more important with more than 2 limbs). bajra is thus conceptually closer to cadzu and cpare than to klama. 


        But I suspect that is not what the definer really had in mind.
I was the definer. Usefulness was not always the main consideration. Conceptual similarities and differences from other words/concepts, coverage of semantic space, usefulness in compounding are just a few of the other considerations. 


I
        think rather, those arguments were left out (as if we could
        sensibly talk about running without them) because the definition
        needed to stay under five arguments and the definer already knew
        that a lujvo could be formed with "klama". And so we find the
        word "bajykla".
No. There is no magic about the number 5. For a short period, I think it was the word pikta that had 6 places, and it made perfect sense at the time.
There was a period where I considered adding standards places to a lot of words, but when I tried I realized that my criteria were too subjective, and I backed most of them out. Places were also added/removed in contrast to other words of similar meaning. 


             bajykla k1 runs to destination k2 from origin k3 via route
        k4 using limbs b3 with gait b4

        This word strikes me as what running is really all about. But
        notice we lost the surface (x2) argument. Moreover, I could
        easily imagine an additional speed argument.

        That lead me to wonder if the ordinal argument system is really
        sufficient. "Running" is a concept and everything that can be
        reasonably associated with the concept should be accounted for
        in the possible arguments.


That depends on what is essential to the concept.


    I'm sure, you are not going to add places for what was the weather
    while the person was running or what was the political situation in
    Berguland at that time.
Weather and politics are irrelevant to whether a person is running or not; a surface and limbs are not. 


All of those factors could make running
    somewhat different and result in different results (the weather
    could change the route etc.)
I suspect you are too strongly thinking of the English word. "running" has a meaning in politics (whether in Berguland or not), but that kind of has nothing to do with limbs, and usually not much to do with weather. 


Right. The weather isn't integral to the idea of running. That's really
what I am getting at. It seems like the idea of running has been broken
up too much so that integral aspects of the concept have been divided
across two words, not because they make sense in themselves, but just to
fit a grammar limitation.

Let me give an example in the opposite direction to clarify what I mean.
Why is "klama" defined as:

x1 comes/goes to destination x2 from origin x3 via route x4 using
means/vehicle x5.

Why not instead have four simpler words for:

x1 goes to destination x2

     x1 comes from origin x2

     x1 traverses route x2

     x1 uses/employs x2 for purpose x3
Because it is not the case that one can have a destination without also having an origin and a route. It is possible that you only care about the origin or the route, because others are implicit, but they all are part of the concept. 


        While it's kind of neat how "bajykla" can be composed form
        "barja" and "klama", being *neat* isn't high in my list of
        criteria for being well defined.


My priorities were not necessarily the same as yours.


        On top of this, reading about Modal Tags, that really hammered
        home to me that the argument system has some holes. I don't see
        how a well defined predicate could ever make sense with
        dynamically added arguments. If they made sense they should
        already be part of the predicate's definition. (Of course, some
        modals are basically short-cuts for making relative clauses and
        not so much case tags at all. These stand out b/c they are
        universally applicable to just about any predicate.)
Modal tags are primarily NOT case tags, but the boundary between the two was rather fraught (and tied up in the history of Loglan/Lojban) 


        In short, it seems like the limitation of keeping the number of
        arguments within a small range (generally five) is an arbitrary
        provision


Yes, it would be.


        that causes some concepts to be chopped-up into
        equally arbitrary partial concepts.


Arbitrary in a sense, but not equally arbitrary.


Of course, the converse
        issue would be how to handle predicates with potentially a dozen
        arguments when it is already difficult enough to recall the
        fourth or fifth?
That would certainly be a consideration. If it is difficult to recall a place, it does suggest that it may not be necessary. But you have to be sure you are working with the correct concept. If you are thinking about running involving a destination, you are not really thinking of bajra, but rather bajykla. 


    Just use several verbs. You can easily say {mi bajra lo jdika grana
    lo xance gi'e klama do ...}
Lojban does not have verbs (or nouns). It has predicates. That is a fundamental distinction. 


Sure. But I am not asking about the technical how to deal with it in the
current structure of the language. Rather, I am wondering about a more
philosophical question.  i.e. Is "bajra" a real concept? Or is it merely
a "partial-concept" that exists only because of limitations of the
grammar? And if we were to make it complete, something more like
`bajykla`, but with even a few more arguments, e.g.

     x1 runs to destination x2 from origin x3 via route x4 at speed x5
on surface x6 using limbs x7 with gait x8

How could such long predicates be manageable?
If they were important, they would probably be managed and therefore manageable.
In English, verbs have an arbitrary and pretty large number of possible prepositions that they can be linked with. In a sense, all of those prepositions are "places" (or they might be thought of as modals or case tags), although there is a lot of polysemy. When you learn a word, you implicitly have to learn all the meanings of all the prepositions that may be linked with it, yet no one ever memorizes such a list for any word. 

lojbab

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban-beginners+unsubscribe@googlegroups.com.
To post to this group, send email to lojban-beginners@googlegroups.com.
Visit this group at http://groups.google.com/group/lojban-beginners.
For more options, visit https://groups.google.com/d/optout.