Re: [Re] Place structures with co

[Logical Language Group <lojbab@ACCESS.DIGEX.NET>]

la veion. cusku di'e

> > i.e. the purpose of co is to make the sumti places of the modifier more
> > easily accessible.

la xorxes. cusku di'e

> These places are also easily accesible using {be}:
>
> brode co broda be da
>
> But the trailing places of brode are lost once the {co} appears.

Ah, now the true root of the confusion appears.  No, they aren't lost.
"be...be'o" constructs are tight-binding, not loose-binding; they bind
to the tanru-unit-2 (i.e. brivla or equivalent) immediately to the left.
So you can get the x2 place of "brode" with "brode be de broda [be] da".
I put the second "be" in brackets because (in selbri context) there is
no difference between "mi klama le zarci" and "mi klama be le zarci".

> My point is that it's not correct to say that {co} gives the possibility
> to use modificand-modifier word order. It seems more logical to me that
> the places of a tanru should always be those of the tertanru, without
> this exception.

The places of a tanru are those of the rightmost brivla, without exception.
This is not necessarily the logical tertanru.

--
John Cowan              sharing account <lojbab@access.digex.net> for now
                e'osai ko sarji la lojban.