[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >
re: having it both ways I've put some thought into this and I think my basic premise is right because there is more than one way to be redundant. For example, when I read a simple sentence in toki pona, I fire up every tool & engine that I have to comprehend it. insa mi li jo e ala. mi moku e kili. I use use discourse (what would typically follow a sentence like "I'm hungry"?) I use common sense. People need food, fruit is a common thing people eat. I know that mi is always animate, insa, moku and kili are inanimate and jo, ala are abstract. I know that li and e don't actually refer to anything concrete. I use most subjective statistics about part of speech. kili is usually an noun, moku is a verb half the time, a noun the other half of the time. mi is always an anaphora or pronoun. I know that subject stuff is first, object stuff is last and verbs are in the middle. I know that verbs are prefixed by li, sometimes suffixed by e, objects prefixed e, and subjects usually suffixed by li. I also know that all elements are in the lexicon and that none are proper modifiers. I know that there really is only one parse tree for the example, but if there were multiple parse trees, I can use common sense to discard the crazy sounding ones because I know that my colocutor isn't crazy. Some subset of all those things are telling me the same thing over & over. I'm not even sure how one would count this sort of redundancy, but 4 of the above let me know what is subject verb and object. Now if I'd said mi jo ala e olin. mi wan. mi jo ala e olin. mi wan. That is redundancy of a very countable sort, the entire message is sent twice. One still can't tell if the speaker is complain about being alone or being married, because of all the clues, some overlapping and some not, none of them resolve what "wan" means. I'm awash in clues about what is the subject verb and object. I'm really short on clues about what is an adverbial modifier of a verb, and many other things. > This, however, I disagree with: > > >>>I think that the reading gotchas and the effort-less to read mistakes are signs of mistakes or limitations in the conlang's fundamental design? underspecification and overspecification respectively. > > I don't think you can have it both ways. If Toki Pona's syntactic > redundancies make it easier to recover from common mistakes, why is > that a bad thing? > > Certainly there are many semantically ambiguous of the language that > can make valid sentences hard to interpret. I'm not sure those are > design flaws, more a matter of design limitations. The syntactic > ambiguities that make it hard to parse valid sentences are more > arguably design flaws. > > When you say > > >>things and things in the environment > > how do you mean to distinguish them? Is the first short for "unseen > things (= not in the local environment)" or for "things recently or > soon to be mentioned (= in the discourse environment, not necessarily > the physical environment)"? > > -- > Jim Henry > http://www.pobox.com/~jimhenry/ >