Re: [engelang] reformulating the core grammar - engelang messages

On Thu, Oct 4, 2012 at 4:02 PM, John E Clifford <kali9putra@hidden.email> wrote:

From: Mike S. <maikxlx@gmail.com>

One idea that has not been floated would yet be reversing the conventional order of arguments; in other words, the rightmost variable would be the x1, and x2 would be the next leftward one, etc. It would make some sense in this language which tends to be strongly head-final to make the most salient argument (x1) final. e.g.

la mlta vska = The cat sees [something]
la mlta le grke vskeka = The cat sees the dog

Any takers?

That just seems to be adding to confusion. We are so used to left-to-right readings that we will take the first one as x1, even if the subject turns out to be x3 or so.

We'd get used to it eventually. It's no more arbitrary than the system in place. It would make things look (slightly) tidier, as usually the "subject" is involved in most simple formulas in the predication, and they're slightly more visible at the end of everything, where usually they'd go as x1. The x2s and x3s would shift a little closer to their non-subject binders/restrictions, as in the simple example I gave. So instead of

Ba Ra Be Re ju Bi Ri Paki ju Bo Ro Pakeko Pake
A A E E ( I I A I ) ( O O A E O ) ( A E )

we'd have

Ba Ra Be Re ju Bi Ri Pika ju Bo Ro Pokeka Peka
A A E E ( I I I A ) ( O O O E A ) ( E A )

as a common pattern. It's not a huge issue, though.

Does this help? q{formula}qa <=> ma {formula}

That is pretty much what I was getting at, except that I didn't think that what was in the quotes had to be a formula.

It doesn't. That's the only difference between q...q- and m-.

IMHO: A quantifier does not affect already bound variables under its scope. In other words, inner bindings take precedence over outer ones.

I agree, but the discussion of binding is either incomplete or in flux so that other approaches might be used (l overriding internal s, for example).

O, that way madness lies. Luckily, no one has suggested we go that way. I really hope that no one will.

Well, something strange at least. I should add the restri9ction on quantifiers that at least the first formula complement whoudl contain a free occurrence of the indicated variable (lacking that doesn't lead to madness, necessarily -- we could always throw in a je smV , say -- but certainly a lack of clarity).

Yes, I proposed "je psV" myself -- "V pertains to something". That or your (not entirely different) "je smV" is probably the intended meaning for humans anyway. But we can't go fiddling with binding precedence. I don't want that madness considered even in passing.

I see you scrapped your conjoined variables in your idea. I was going to mention, while you had them, that if we really needed to do so, we could assign some C for conjoining. Then the apostrophe could go about its own business of forming discrete variables.

Yes, superimposition is much tidier and shorter and seems to work as well, so far. But I have only been messing with l; whether r and s (as they are now) will be as amenable is not quite so clear (they ought to be, but sad experience has taught me not to trust my intuitions on these things very far).