[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >

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

Re: [engelang] Xorban: Semantics of "l-" (and "s-" and "r-")

On Mon, Sep 10, 2012 at 10:05 AM, John E Clifford <kali9putra@hidden.email> wrote: 

In well used Lojban, every argument (sumti) is either intensional or extensional, never both.  The problem (aside from bad usage) is then twofold.  Intensional sumti contain thing that look like extensional sumti (and, indeed, are -- in the alternate worlds to which the intensions lead us).  Further, while (probably) most predicates take only extensional arguments and a few take only intensional ones, quite a few can take either.  The result is mistakes like {mi djicu lo xorma} rather than {mi djicu tu'a lo xorma} when there is no particular horse I want, any will do.  'fV' looks like a plausible way to go here, although I don't quite understand exactly how it functions, theoretically.

For the very reason you cite, I don't know why you regard intensions as such bugaboos.  Intensions are foundational to predicate semantics and extensions are subsidiary.  In Montague's system, given an intension you can always get to the extension if you know the world.  In fact, intensions are formally simply functions from worlds to extensions, plain and simple.  The opposite is not true; given an arbitrary extension, you cannot get back to the intension, because several intensions might have contain the exact same extension for a given world.  That is why one of the rules in Montague's framework if you read the fineprint is that you are not allowed to apply the up- and down-operators in any order.  In particular, you can't apply "up" after you've already applied "down".

This is also why Old Lojban's "tu'a" particle seems so strange to me.  You supposedly start with an extension, which is just a set of individuals with no "meaning" holding them together, and somehow get back to an intension.  Obviously this intension has to be glorked, and obviously it's going to be glorked based on the complement of "tu'a", and no human is going to resist reading that complement intensionally, so the idea that you actually started with an extension is a charade.  It would have made more sense to start with "lo mlatu" or maybe "lo'e mlatu" involving an intension and interpret "su'o" and "ro" as invoking Montague's down operator to yield a quantification over the extension in the appropriate world, and allow "lo mlatu" or maybe "lo'e mlatu" to stand for themselves when intensional arguments are needed.

None of that is to suggest that "lo" or "l-" necessarily bind intensional objects in the languages under consideration.  It's just that if you need to choose, it makes more sense to start with intensions and get extensions as you need them than vice versa.