Re: [engelang] Xorban compounding revisited - engelang messages

On Sat, Sep 29, 2012 at 8:58 AM, John E Clifford <kali9putra@hidden.email> wrote:

From: Mike S. <maikxlx@gmail.com>
But green things share a universal property that's more or less absolute, namely having a surface perceptibly reflecting light of approximately 520–570 nm wavelength, within a certain amount of tone/hue/brightness variation. You don't have an infinitude of discrete, non-overlapping localized extensions of green things, you have one vast massively overlapping aggregated extension. In contrast "small" things are small strictly by whatever extension is being applied. (I'd rather avoid talking about "greenness" and "smallness"; I don't know what you mean unless you mean the intension. I do know what green things are.)

But that reflecting is neither necessary nor sufficient for something to be green, as any Psych 1 class will point out. The point is just that, as being small is relative to the object involved, so is being green enough to be called that: the "-ness" is about how green something is. From where I sit I can see a house which reflects a significant quantity of green light but is called the blue house because it reflects a quite unusually large amount of blue light (for a house). Similarly, when I look at my lawn, I can clearly distinguish the green grass (mainly oat grass at this time of year) from the other grass, which is still green by your definition (and probably others).

The simplest answer to this is that "crn-" does not encode, and by itself was not intended to encode, these finer distinctions. Naturally that is not the same thing as saying these distinctions are not important, salient and real, it's just that "crn-" doesn't encode them, but rather it lumps things into one bunch satisfying what we call "green". If you want to make another predicate that allows an x2 to predicate to x1 the shades of green typically exhibited within a certain class of things x2, that would be fine, and I doubt anyone would misunderstand "crnake" as meaning anything but "A is of a shade of green typically displayed by E". But I don't think that the basic one-place "crn-" needs that machinery in the same way that "lnt" and "cml" depend on their standard place-argument. In the former case x2 is a refinement, in the latter an essence.

I may be missing your point, because I fail to see any mystery here: "je mmto'eka'a mmto'eke'e" is true and means we both have mothers, even though we don't have the same mother. Likewise "je sa trca lntako'e sa mlta lntako'e" is also true and means that some stars and some cats both are unmassive, though obviously not by the same standard.

The point is that. when you infer from the (putatively) true claim that it's a small star to that it is small, you lose the context and come up false, while the context is not relevant for the mother case.

Maybe not relevant for the example I gave, but consider "le mlte mmteka" and "le grke mmteka" from the point of view of "a".

As far as "black" -- again these are basic color terms and 10-12 of them are intended to cover most of the color cone, perhaps even overlapping slightly. Very dark brown cats pass for black.

Precisely, but they are not black except by cat standards.

We'll have to disagree on this. "Black" just means having a very dark shade with no respect to hue. If you want to refer to a more absolute color, then you can say "100% black" or "glossy black" or "cat-black" for that matter. But we lose nothing essential by just saying "black".

"Pretty little girls' school" presents no problems in Xorban. Logical conjunction stands for itself. I am going to use "ju" to crudely approximate the tanru relationship, except that "girl type-of school" will invoke the proper predicate place. Normally I wouldn't use "ju" like this. There is one exception, #39, which forces me to coordinate unlike logical structures. Another interesting one is #3, which I think is ambiguous in Lojban, and is the only one that gets translated with the group operator.

Tell me if you see any mistakes:

Using ju saves you from the logical problem, so, of course, gets you off that hook. But it still does not seem to me to give an accurate reading even at #1. We get things that are pretty and little and girls at the same time, but there is no reason on that basis alone to think that they are pretty girls or even little girls, let alone that they are girls whose being little is pretty, as the intended meaning is here (it seems, for example, to give, at best, that their prettiness is in a little sort of way).

Tanru are far more glorkative than "ju" expressions were ever dreamt of being, so I would say that Xorban come closer to the (most likely intended) bull's-eye. You're right that the standard of prettiness and smallness is left open by the translations, but pragmatically it's clear what they all mean.