On Wed, 18 Nov 2015 00:16:20 +0000, Alex Fink <[log in to unmask]> wrote: >I'd also compare >something like the component level in the Chinese script, above >strokes but below whole characters, though it's not clear if there are >any combinations which are completely arbitrary as opposed to just >metaphorical like Chinese has. I think most 'arbitrary' combinations in Chinese would be phonetic. There's nothing like that in U2D, but I suspect that it would be necessary to use that to build a really large lexicon, say, 10,000 words or more. Although not interesting in the 2D-native sense, the component approach does allow for ready comparison to a spoken language, either with predictable rules or with a semantic/phonetic compound system. >It's interesting to see that the suggested reading order is the causal >order, which you also suggest is the one of the main mechanisms for >achieving an effect like tense. Does it take especial contortions to >narrate events in reverse-chronological order, then? Perhaps not. The causal order may be clearly defined, and that marks the time flow. If there were some other way to unambiguously define a starting position, that start being at the end of the causal order indicates a reverse-chronological reading. More difficult might be a disjointed time flow. Having said that, I suspect that the causal order might only be useful on a small scale; say, in a novel, on the scale of a scene. Where there are significantly larger time, distance or cause gaps, it might be clearer to mark those explicitly, in which case each region of closely related actions is connected through causal relations, and separate from other groups of actions. Time flow between those larger groups could be more easily marked explicitly. I don't know, though, because I haven't tried to translate anything of that scale. Although nothing is grammatically wrong with a single novel-length causal trunk of writing, it might be rather difficult to read, just as a single novel-length paragraph of spoken language might be. In the writing examples I've done, I've tended towards using more space, rather than less, and even on such a small scale, there tend to appear different 'sections' of the text. >In your example with the clock hands, it is not so clear to me that >you've broken the rule of "separate verbs for separate actions": the >two hands do move in concert, and it would be very natural to consider >that one action. But for the general case, maybe you're in want of >something like a distributive derivation on the verb >[which is sorta what UNLWS might use a "stack" for. You're right; the clock example is not the best example of rule-breaking. In the general case, a distributive could work. Though, it seems like a fairly uncommon usage, since it could only refer to one or two configurations of subjects and objects, and presumably not bind to any separate nouns. That alone should imply spatial and temporal cohabitation, or at least conceptual, so I don't think rule-breaking in that manner is ever going to be confusing to a reader. Though, at that point, it might turn from a rule violation into a rule itself. >UNLWS referents are not unmarked for number, they are always >singular unless group morphology is present, though they can be >unmarked for the generic vs. specific contrast. So, from its >perspective, by adding unmarkedness for number you've introduced the >means to obviate the more general problem... but then by limiting this >unmarkedness to nouns you've not taken it up.] I'm not sure what general problem that would solve. Insofar as verbs are part of the same broad 'action', like nouns are part of the same 'group', I can't see how it would solve the problem above. >I see you've allowed reflection of nouns. Given that pronouns "I" and >"you" differ only in orientation I see they're not subject to this; >but then the form of "I" is borrowed for a noun (basically for >reflexives?). So that noun could be drawn to look like "you"... Grammatical elements in general are not; verbs, for instance, would have their subjects and objects flipped. The nouns are reflected primarily for aesthetic reasons, and because the space of components is quite large; there's little chance of needing to use both mirrors, and I think that would be quite confusing in any case. Besides which, given the relative orientations of some of the components, they would need to be either mirrored or seem 'upside-down' compared to their partners. The noun form of "I" really means "self-". Although simple reflexives are quite natural, just by splitting the subject and binding it to the object as well nearby, I supposed that in many cases, the action would be qualitatively different, and have a separate concept, if done to oneself. In that case, the verb would probably not be reflexive at all, but you would bind to the object only. I don't think that a similar semantic role really exists for a hypothetical "you" component. >A loglanger would point out that the "and" that makes a group of nouns >is not quite the boolean "and" -- is a group containing True and False >the same as just False by itself? Unfortunately. Though, if you are directly asserting that something is true and false at once in an adjectival sense, by binding them both to one side of an "and" gate, that in itself may be somewhat nonsensical. And a group containing one Truth and one Falsehood should really just behave as if those are normal nouns. U2D is sadly not really Turing-complete. >As for >"or", one might ask what kind of a thing "you or I" is, if "you and I" >is a group: is it purely shorthand for a disjunction of two sentences? >or is it a single person whose identity is unknown to ___? or? The latter; more generally, if you have a things bound to one side (the large side) of "or", then the other side refers to one of those things, unknown to the writer which. This is much more than a disjunction, because the unknown noun can be treated like any other noun. In practice, it's rarely useful to be able to do that, except that it's really the only convenient way to ask, "Which of X fulfils Y?"