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

[lojban-beginners] Re: For those into diagramming...

Jorge Llambías wrote:


The one with {co} may need some reworking.

For example, if you have {ko'a ko'e broda co brode ko'i ko'o}
then ko'a and ko'e fill the x1 and x2 of broda, but ko'i and
ko'o fill the x2 and x3 of brode. All the sumti that appear
before the selbri {broda co brode} will be arguments of broda,
all the sumti that appear after the selbri will be arguments of
{brode}.


I'm too ignorant to understand that, but I see that this *will*
need reworking.

I've never used broda at all -- is something like {ko'a ko'e
broda co brode ko'i ko'o} an actual complete sentence?



Attitudinals and {xu} are all in the same selma'o UI, so they
should all be treated the same way. The meaning of the
word should not matter for structuring a sentence, only the
syntax class (selma'o) it belongs to.


I'm a little uncertain about this. Bear with me here:

It's a little hard (for me) to draw a firm line between "diagramming
for syntax" and "diagramming for semantics."

In any language, syntax and semantics are obviously related; perhaps
more so in Lojban that most others.

Perhaps I am wrong, but I thought that the division of selma'o was
mostly 1) a convenience for the parser and the writing of the formal
grammar; 2) an aid in manually checking syntax at a glance; and 3)
an aid in learning grammar.

Plainly, many selma'o are grouped together even though they are
not strictly related semantically.

But getting back to the diagramming: If a diagram doesn't (in some
way) illuminate the sentence, it is useless as an analytical tool.
So I would say that ideally it should help in comprehending not
just syntax but also semantics. (That is why I draw a line from {ri}
to the thing to which it refers -- that is not strictly a syntactic
issue.)

I read just today that this has been one classic criticism of Reed-
Kellogg diagrams -- that they don't communicate enough of the
semantics. For example, predicate nominatives and predicate adjectives
are diagrammed the same way. "Bob is happy" and "Bob is a woodchuck"
would be diagrammed identically but for the article. (Of course, in
Lojban we don't make real distinctions between "noun-like" brivla
and "adjective-like" brivla, just as we don't distinguish much between
"doing" and "being.")

Consider these two English sentences:

   Bob seems willing to help.
   Bob seems difficult to help.

Superficially (syntactically) they are similar. But in the first, we
have Bob (as subject) helping an unspecified person. In the second,
we have an unspecified person (as subject) helping Bob (as object).

Ideally, I would like this kind of semantic information to be captured
in any diagramming technique I came up with. (RK diagrams don't capture
that distinction, though.)


How about doing something like this: Instead of lines use
filled-in coloured boxes. 

[snip description]

I like that, and it is obviously a viable alternative to what I am
doing. Someone already did something of that nature in a PDF I saw --
was it you or someone else?

In fact, it might be fairly easy to take jbofi'e output and HTMLize it
into the colored-box form.

I'm very much in favor of any kind of visual aids in describing Lojban
or Lojban sentences. (I enjoyed Desquilbet's article he did for IBM.)

However, I'm not going to abandon my own approach for these reasons:

1. I just like mine better. It fits my brain.
2. I learned to diagram English sentences from the age of ten. (At that
time and place, it was stressed.) We didn't quit till I was 16. :)
3. Since Reed-Kellog diagrams are nearly 130 years old, and are the most
popular technique ever used in the US, millions of living people have
already seen it. (See disclaimer below.)
4. Nested colored boxes involve an awful lot of horizontal and vertical
lines. Eventually it starts to look "busy" to me.
5. Printing colored boxes takes up my color ink. :)
6. It's hard to draw nested colored boxes on the back of a napkin in a
restaurant.
7. And finally, as I said, I want to illuminate semantics to whatever
extent I can, not just syntax.

Disclaimer on R-K diagrams: If you're under 40, there's only a 50-50
chance you ever learned to diagram. My friend who is 26 says they did
it "a little" in his English courses. Diagramming has been greatly
de-emphasized in American schools in the past thirty years.

Some other things have also been de-emphasized in the last thirty
years, such as grammar, spelling, punctuation, pronunciation, vocabulary,
etymology, and composition. :-/


Hal