On Thu, Apr 14, 2011 at 4:16 AM, Boyko Bantchev <[email protected]> wrote:
> On 14 April 2011 04:23, Don Guinn <[email protected]> wrote:
>> I was simply trying to point out that you chose to apply the meaning of
>> "between" not from the J terminology of "between" but from some other
>> context where it is not clear of order. J documentation makes it very clear
>> for J context.
>
> Does it?  How?  I don't believe `between' is some notion specific
> to J.  I don't see a definition of it.  Between is between, let's
> not put a halo of specialness around it.

The only thing "special" about insert is that J evaluates the sentence
that is obtained.

Anyways, I take back my earlier comment about you asking J to be
formal in the way other language references are formal -- when I went
out to look at how c#, perl and javascript standards define join, I
was not able to find any that gave special treatment to the case where
the array length is 1.  But I was using web searches, and perhaps I
overlooked something?

> On 14 April 2011 05:13, Raul Miller <[email protected]> wrote:
>> That was exactly my point:  People need to learn arithmetic before
>> they can understand mathematical definitions of arithmetic.
>
> No, you maintained quite a different stand (to which I objected):
> that in DoJ -- which for J is supposed to be the analogue of the
> formal definition of arithmetic -- it is normal to have circular
> definitions; that I should consider the definitions of \ and
> determinant and who knows what else as part of the definition of /.

My claim would have been that use of the definitions is inductive, and
not circular.  Specifically, I was saying that you need some exposure
to all of a group of them before you could *master* any of that group.

> The structure and content of a defining document have nothing to do
> with `learning arithmetic'.  There is /, and there are \, \., and
> many others.  If the examples accompanying the definitions of the
> latter happen do use /, that's ok, but these examples are not
> *the* definition of /, not even `parts of it', as you insist.

They are, nevertheless, a part of the reference work I refer to as the
dictionary of J.

And, I see no reason to not consider them to provide relevant information.

And, I see no reason why I should not use that information in building
my mental model of what / means.

> Or, to repeat what I've already said,
>
>> In all more or less formally presented systems (which, I assume, a
>> programming language definition ought to be) these things work in the
>> opposite direction: first establish a definition, then use it in further
>> definitions and applications.

And Godel has long since shown that only trivial axiomatic systems can
be complete.

Anyways, we are iterating on your statements here and perhaps I shall
eventually acquire enough knowledge of what you are talking about to
have some clue as to why you think these are important issues.

>>> All I have been proposing is replacing the not sufficiently clear
>>> definition of / with a clear one -- just what is expected from a
>>> definition.
>>
>> But, of course, the dictionary's illustrations clear up that issue for
>> people with enough background.
>
> I am forced to repeat myself again:
>
>>> Examples are not a substitute for definitions -- they are just an
>>> illustrative aid.

Why draw a distinction between "illustrative" and "clear"?

Specifically, if you feel that the definition is "unclear" on one
point, why do you feel that "illustration" is not a sufficient
mechanism for making it "clear" on that point?

-- 
Raul
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to