On 5 Feb 2015 21:14, "Jorge Llambías" <jjllambias@gmail.com> wrote:
>
>
>
> On Thu, Feb 5, 2015 at 4:35 AM, And Rosta <and.rosta@gmail.com> wrote:
>>>
>>>
>> So if I, at the age of 500 after a lifetime spent diligently on this task, present you with a full explication of the rules of English, what is the difference between English and that explication?
>
>
> First I'd have to know what English is, in order to compare,
So although you're not sure what it is, you have an idea of what it is that is good enough for you to know what it isn't?
> but it seems unlikely that English is the same as a full explication of its rules.
How about if English is the same as a full explication of a family of sets of rules, one set per idiolect? Or you feel that even an idiolect is not the same as a full explication of its rules? Is the game of chess different from a full explication of its rules? If Yes, is that because there are many different possible explications, or because chess, like tigers, is very different from a set of rules?
> If a zoologist presented me with a full explication of a tiger I would be able to tell it apart from the tiger immediately. I wouldn't be as scared of the explication as of the tiger.
This is easy to explain, for tigers are material and bite, whereas, like language, explications are abstract, immaterial and don't bite.
>
> But more importantly, if someone else presented me with their own full explication of the rules of English, using different terminology and different analytic tools, I don't think it would be necessarily the case that one explication had to be better than the other, they could just be two different explications.
Would it necessarily be the case that neither is better than the other?
Are preferential criteria such as simplicity and knownness (by the speaker's mind) valid?
>> So technically the "syntax" is separate from the phonology. But of course in fact the "syntax" isn't a syntax, for all it does is generate a labelled tree with selmaho as its leaves; it doesn't encode logical form.
>
>
> Right. But the question is whether it's an aid, an impediment, or neutral in our quest to encode logical form. My impression is that it is an aid.
I don't disagree with that; I've noted already that although the "syntax" isn't a syntax, much of it could be recycled into an actual syntax.
>> Furthermore, since every selmaho leaf (with the possible exception of terminators) corresponds to a phonological words, the "syntax" looks like it's driven by the sentence phonology.
>
>
> Either it's driven by, or drives it, or both.
Yes. The important point is that it's not first and foremost driven by the requirement that it should encode logical form; though, it is of course the case that the "syntax" has also been shaped by the idea that in some cloudily understood way it should contribute to the encoding of logical form, hence the pretty good job it does with the (very easy) task of encoding predicate--argument relations.
> The "syntax" is pretty good with predicate-argument relations, but poor with binding relations. One important type of binding relation is achieved by repetition of phonological form, but the "syntax" is completely blind to phonological form (in the sense that it can't tell "da" and "de" apart). But it can tell that a given KOhA is an argument of a given BRIVLA for example.
That's where you'd start with the work of converting "syntax" into syntax.
--And.
--