I’ve been talking about languages known to me. Yes, there was induction when I generalized. Yes, this might not be logically justified. You knew all of this.
You, too, use all these fuzzy words all the time even though you have never been told a rigorous definition. And we don’t need them. We understand each other without analyzing sentences and remembering definitions. cf L. Wittgenstein (I already thought about mentioning his Sprachspiele in my last post) Concerning your questions: Those crafted by humans can: They’re known to be artificial. (so Esperanto and Volapük are artificial) Those that seem to historically have developed by routine everyday communication by everyone in a society are at least said to be natural. So that’s what natural means. It’s just the word we use for those languages. (maybe you can find better characteristics but I guess you know what I am talking about) But for any language unknown yet: I don’t think so. But I don’t think it matters at all. Am 17.01.21 um 16:49 schrieb Justin Paston-Cooper: > All languages are fixed over a given Planck time. What is it for a language > to be artificial or not? Can it be objectively proved either way? > > On Sun, 17 Jan 2021 at 18:43, Hauke Rehr <hauke.r...@uni-jena.de> wrote: > >> Natural languages are flexible. Recipients of messages are >> forgiving, trying to understand what you meant. >> The rules are dynamic and at times even local or personal. >> >> This is much different from many artificial languages, >> in particular from programming languages. >> They have one set of fixed rules* (even if they are rules >> for declaring rules); the interpreter/compiler can only >> be told to handle a list of common mistakes but cannot >> intelligently try to understand anything never seen before. >> >> Therefore I think learning should be at least somewhat different, too. >> (And I used to learn even foreign languages by first studying >> their grammar, then learning a thesaurus and then applying them, >> building hopefully correct sentences. When a Spanish teacher began >> talking to us in Spanish from the start, I was overchallenged.) >> >> * yes, they are evolving – but for any version, they’re fixed >> >> Am 17.01.21 um 16:27 schrieb Henry Rich: >>> It gives them a wrong mental model of rank, which they must unlearn >>> later. This can have serious consequences, particularly if they get >>> the idea that u"n is 'like u with the rank set to n' (if that were true, >>> u"1"_1 would be the same as u"_ 1, which it isn't). >>> >>> Ken thought you should learn J like you learn a natural language, by >>> seeing and saying, and creating your own rules internally. I think he >>> was wrong when it comes to verb rank. The idea is so new, and so >>> subtle, that users left to themselves get it wrong. I had one very >>> bright student who, discovering that (,1) + 1 2 3 gave an error, found >>> that +/ would not give an error, and ever after applied / to every >>> verb. He created his own rule, you see. >>> >>> Henry Rich >>> >>> On 1/17/2021 12:24 AM, Raul Miller wrote: >>>> Does it really cost them that much? >>>> >>>> Given that beginner problems generally do not involve multi-megabytes >>>> of data, I mean... >>>> >>>> Thanks, >>>> >>> >>> >> >> -- >> ---------------------- >> mail written using NEO >> neo-layout.org >> >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm >> > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > -- ---------------------- mail written using NEO neo-layout.org ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
