## CONLANG@LISTSERV.BROWN.EDU

#### View:

 Message: [ First | Previous | Next | Last ] By Topic: [ First | Previous | Next | Last ] By Author: [ First | Previous | Next | Last ] Font: Proportional Font

Subject:

Re: Case logic introduction

From:

Date:

Tue, 2 Dec 2014 05:11:28 -0500

Content-Type:

text/plain

Parts/Attachments:

 text/plain (52 lines)
 ```On Tue, 2 Dec 2014 04:05:20 -0500, Patrik Austin <[log in to unmask]> wrote: >Thanks for comments! > >Alex: you raised many questions and I could answer them but I think it's best we first look at one core point of your criticism and hold it there for a minute. I'm sure many people (yet not all) here totally agree with you, so this goes out to the lot. As does anything on a mailing list, yes. >When I say in the video 1 + 2 = 3 is grammatical, what makes you decide it's not? Is it logically impossible? I didn't say 1 + 2 = 3 was ungrammatical. I said +(1,2) =(3) is ill-formed in any treatment of arithmetic in predicate calculus I'm familiar with. If you're using the notation +(1,2), you're using not exactly a predicate but a function (one of the "n-ary function symbols" of https://en.wikipedia.org/wiki/First-order_logic ); so the denotation of +(1,2) is an element of the universe. As for =(3), I've never seen anyone write that and don't know what it is meant to mean. One thing that people will write when using functions is   +(1,2) = 3 . But the status of "=" is very different to that of "+" here: the denotation of "+" is a function from pairs of elements to _elements_, that of "=" is a function from pairs of elements to truth values (if indeed the theory gives it a denotation in the ordinary sense at all, as opposed to building it in to the interpretation rules). Endowing "=" and "+" with the same apparent syntactic status in your notation is at the very least misleading, and is the sort of thing someone who didn't realise the difference between them would do. >The video also says that another dozen syntactic rules are needed, and there's a link to the old draft article which states: > >"S9. Each adverbial (cf. prepositional phrase) refers to the previous non-adverbial constituent". > >There's actually a parse tree of the following sentence (p. 20): > >'The talker to Ann about the seller goes.' I concede that this is solvable with extra rule(s). To do this you have to change the rules so that a preposition is not simply a marker of a case, where cases are defined in relation to predicates, but something more general; and the syntax gets more complicated too. If adverbial and non-adverbial constituents are different, you also have to account for how. I take this to be a greater injection of complexity than you let on. For one example of why this is: in your language, a case-marked constituent can now modify either a verb or another case-marked constituent. But you could get by with having them modify _only_ verbs, as formulations closer to predicate calculus do, which would get rid of one rule. (This is not meant to be a decisive rejection of your claims of simplicity in totality, note, but I do think your judgment is biased.) I was trying to take your video standing on its own. If it's meant to be an introduction to case logic, it should introduce case logic in a way that doesn't take your previous writings as prerequisite. Does case logic _exist_ as a freestanding entity, distinct from your proposed language? >I think now is time we do a little reality check. > >If you knew the first thing about Davidson, you'd have understood by now that doing logic and philosophy with disrespect to the principle of charity is of no value. This means in practice that while you're looking for a quick way to refute what I just pointed out above, you should realise that whatever yeah-buts you may find, it's too dangerously likely to be something I've figured out long ago. One could see it as charity that I was still trying to engage with you, when (e.g.) Ray thinks I'm only being too polite, and And seems to have given up. One could see it as charity that, some months ago, I understood your disappointment with challenges to your claims by reference to Xorban, which I agree _is_ poorly documented, and as such prepared a bare-bones predicate-calculus based grammar for you, attempting to label and count the rules the same way as you did for RNG. I hoped it would facilitate your comparison of the two, I hoped it would help bridge any gaps there might be between different sets of terminology vel sim that were getting in your way, but instead it seems to have gone entirely ignored, not even challenged but ignored. One could see it as charity to imagine that logicians and philosophers writing about predicate calculus and employing it in formal semantics of natural language had, themselves, many good reasons to go for that theory and not one much closer to their native languages in structure, which you'd think would be easier for them. Are your theories "yeah-buts" to the majority of extant work in formal semantics? >It's also too easy to call someone's theory names. "Strange." "Idiotic." "Stupid." As if: "Look at this guy: he's so dumb he can't even spell his own name right, and yet he's back waving his so called research like a red rag only to be beaten againâ€¦ Let's finish this nonsense for once and for all!" Perhaps that is easy to do. On the other hand, not all criticism is name-calling merely by virtue of being criticism. Of these words I did use the word "stupid" in my message: it wasn't in reference to your theory, it was in reference to a way one might try to translate English to a logical representation which was... stupid, in that it ignored polysemy and ignored the existence of multiple syntactic constructions for the same verb. You implicitly admitted as much yourself when translating into case logic, by doing it better there, e.g. sticking dictionary-sense tags onto the words. I called some of your analyses _wrong_; but in some cases you have rebutted that by reminding me of your adverbial constituent rules. (Not, for instance, the case of English "or": that's still wrong.) I trust you accept that "wrong" is a name that one should not hold back from applying to theories for which it is apt. >It may seem to some people as if you're having a strong case, but the reality is that the only single argument you've come up with so far is ad populum, repeated ad nauseam in different ways. ... or maybe there is no more light to be had out of this thread, only heat. That's all from me. Alex ```