delayed response to Iain

[Logical Language Group <lojbab@GREBYN.COM>]

Iain asks some good questions in a posting back in Sept. that I don't think
were answered:
>Let's start with {zmadu}.  It seems to me that there's a canonical form
>for such as {zmadu}
>
>        leni broda cu zmadu leni brode
>
>        leni mi djica lo vanju cu zmadu leni go'i lo djacu
>
>I used this form in my "massification" comment, because it was the only
>construction I could think of at the time that actually seemed to work.
>(I still don't know if {go'i} works in these situations, but I think it
>should.)

It specifically does not: go'i would refer to the main bridi of the previous
sentence.  There is no obvious way to anaphorize djica as you have.  I might
avoid the issue with:

lo vanju cu zmadu lo djacu leni mi djica da po'u ri .a ra

I would use bare "djica" in place of "go'i" if you wanted to stick to your
phrasing.

There may be some anaphoric capability expressible through no'a (next
outer bridi, which in this case is the main bridi), with a "1" or "x1"
subscript attached.  You can also use forethought anaphorizing with
"djica cei broda" to assign it to "broda", something you obviously would NOT
do here, since it doesn't save anything, but which you might do if you
were talking about saving many syllables.  Finally, you could ellipsize
the selbri as 'obvious' using "co'e":

        leni mi djica lo vanju cu zmadu leni co'e lo djacu
        I want wine more than [I want] water.


>  This makes {le te zmadu} irrelevant, and if you want to
>specify {le ve zmadu}, you can always wrap another {ni} round it
>
>        leni leni mi djica lo vanju cu zmadu leni go'i lo djacu
>        cu barda
>
>(I suspect Colin's going to tell me {barda} is wrong, but I don't know
>any better - unless the answer's {mutce}) or use {la'u}
>
>        leni mi djica lo vanju cu zmadu leni go'i lo djacu
>        kei ku la'u lo mutce

barda seems right to me. mutce implies some degree towards an extreme.
Another set of possibilities is lo so'Vmei, which gives a greater
specificity of relative degree without having to explicitly say how you
are measuring that degree quantitatively.

>Other constructions are there for variety, but should all be
>transformable to the canonical form.
>
>        lo vanji zmadu lo djacu leni mi djica [ri]
>
>(I'm not sure {ri}'s quite right here, but what else?)

I think my formulation above is correct, but ellipsis is fine, too.

>        mi djica lo vanju .esemaubo lo djacu
>
>and others such as Colin's already mentioned.
>
>The problem with {zmadu} comes with more abstract, metaphorical
>comparisons, or when the two things you're comparing, though similar,
>are not quite the same.
>
>        leni ko'a certu tu'a loi zgike cu bancu leni mi kakne lenu ctuca
>        His expertise in music is beyond my power to teach.
>
>I can't quite imagine {zmadu} working there in place of {bancu}.  And it
>doesn't seem like the _amount_ by which it's "beyond" is vital -
>certainly nothing concrete.

I think this agrees with what I said at the time.  Even if there is a
value, it may be misleading as to the truth of the beyondness.

>Another way of looking at this is that perhaps {bancu} is to {zmadu}
>what {banli} is to {barda}.

Seems like an interesting comparison, though it opens up the question of
what exactly "banli" means, which in my mind is less clear than "bancu".

lojbab