Re: BEGINNER'S QUESTION: internal sumti

["R.M. Uittenbogaard" <reneuit@SCI.KUN.NL>]

Nick Summers wrote on Fri 13 dec, my answer is dated Thu 12 :-)
I am not sure what you mean exactly with some of your questions,
but I will try to answer them:

> Taking an example from the Diagrammed Grammar:
>
>     ta cu tavla be do bei le melbi ku be'o vecnu
>
> Am I correct in the following assumptions:
>
>   1  < tavla vecnu > is a tanru for 'salesperson'
>      (is there a list of tanru anywhere?);

> ... a confirmation somewhere further down the page
> that < tavla vecnu > really was a tanru for salesperson, ...

I don't know where the diagrammed grammar says that "tavla vecnu" is
a salesperson, but it must be wrong.
"vecnu" means "is a seller", "sells", or "is a salesperson"
and "tavla" means "is a speaker" or "speaks"
so "tavla vecnu" means "is a speaker-type-of seller" or: "is a seller
who is speaking" or "is a speaking salesperson".

For finding out the meaning of a tanru, it has always helped me a
lot to formulate the tanru in terms of "X-type-of Y". This is
necessary for some unusual tanru which appear in the lessons like
"cmavo djedi" (structure-word day!?) and "cutci sutra" (shoe-type-of
fast thing!?). Although these are correct tanru, their meaning is
quite vague and requires a lot of imagination.

As you may already know, tanru are always ambiguous - by definition!
This is because "the sky is the limit of your imagination" when
interpreting tanru. A "gerku zdani" (dog-type-of house) might be a dog
kennel, but it may also be the White House, as long as there is some
connection between the White House and the dog (e.g. the dog was seen by
the daughter of Bill Clinton, who lives in the white house).

Actually, when the last meaning is intended, you may expect the writer
or speaker to use a more expanded tanru. Normal tanru which appear in
texts mostly do not require such a great deal of imagination. They are
relatively simple to understand, such as "barda zdani" (big house).
Reading texts and trying to discover the meaning of the tanru used there
is the best practice you can get when trying to get feeling for tanru.

By the way, the first part (modifying part) of the tanru is called
seltanru and the second part (modified part) tertanru. These words appear
in the reference grammar without explanation of their meaning.
For a three-part tanru: "mutce barda zdani" (very big house), the modified
part (tertanru) is "zdani", and the modifier (seltanru) is "mutce barda".
This is because the tanru is evaluated left-to-right (left-grouping rule).

>   2  ta is the x1 place of vecnu (place structures being
>      determined by the second-placed word in the tanru);
>   3  within tavla (which implicitly takes x1),
>      < be do > is x2,
>      < bei le melbi ku be'o > is x3,
>      and x4 is unspecified and thus ellided;

These assumptions are correct.

> placement of the x1 for vecnu above implies that
> "That..." is indeed the salesperson, but also that the concept of
> salesperson has been modified in some way, as in "The salesperson..."
> versus "The big salesperson...".

Right. "ta barda vecnu" vs. "ta tavla vecnu"

> What I don't understand is how the
> (supposedly) modifying first half of the tanru suddenly separates off
> from the (supposedly) modified second half and becomes its own almost
> clausal part of the (part) sentence.

This is because there must be a way by which tanru-internal sumti can
be placed in a sentence when there are no "places" available. When you
want to refer to a salesperson who is speaking you want to say
"ta tavla vecnu". The place structure of the sentence is identical to
the place structure of the tertanru (vecnu). So in order to be able to
specify that the salesperson is not just plain speaking, but
speaking to *you*, you must "open up" places somewhere in the sentence,
because the places of "tavla" are not available at the "bridi level".
This is done by the "be...bei...be'o" mechanism. These words simply
link sumti to a brivla, they do not "end" the selbri. So when reading:
"ta cu tavla be do bei le melbi ku be'o vecnu", when arriving at the word
"vecnu", you find out that the selbri has not ended yet, but continues with
a second part which is to be modified.

> Are the two halves of the tanru
> not somehow implicitly joined, as for example in blari'o, blue-green?
> (Does the fact that blue and green have been explicitly joined into a
> lujvo have anything to do with it?) Is my understanding of metaphora
> incomplete?

In the case of lujvo, the places of "blanu" and "crino" have been combined
to a single place-structure of "blari'o". It is not possible to place sumti
in individual places of "blanu" or "crino" once they have been joined into
a lujvo.

> In the second translation above, isn't the "...who..." misplaced?
> Shouldn't it be something like
> "That is a talker.to.you.about.beautiful.things-salesperson who...
> (performs some as yet unspecified action)"?

I don't think the "who" is misplaced, it is just a "freeer" translation.
My personal opinion is that lojban has such an unusual way of expressing
usual sentences and bridi relationships, that it is not possible to give
one exact translation of most lojban sentences. Actually, can you tell
me the difference in meaning between:

"That is a salesperson which is talking to you about beautiful things"
and
"That is a talker.to.you.about.beautiful.things-salesperson who performs
an unspecified action"

> Can someone shed some light on this for me?

I hope I have.

> Nick

Rene