On Tue, Apr 28, 2009 at 4:05 PM, Don Watson <[email protected]> wrote:
>        You say:
>
> "So if [ and ] were exact replacements for x and y,
> Skip's example would be syntactically invalid.
>
> But y is a noun -- it is not a verb."
>
> Do you see the contradiction here?
>
>    I am not saying that y as a noun is replaced by another noun. I am
> saying that it is replaced by the noun phrase "]y" which produces the same
> result in this situation.

But not in all situations.

Unfortunately, we would probably need to have half a dozen different
situations to illustrate this point.  If we just have one situation, it's
probably possible to pick an interpretation which works in that situation
but not in others.

> To make the inclusion of y implicit, the y is
> removed, Thus the symbol "y" is replaced by the symbol "]" - but this
> doesn't mean that the noun y is replaced by a noun "]". This is identical to
> the way Skip is doing it in tacit J. He too is replacing a symbol "y" with a
> verb "]", but he can do that because a noun implicitly follows the verb "]".

Your description of what Skip did is an oversimplification -- it describes
the instance but does not describe the general case.

Here's a hasty example of a case where you can not replace y with ]

   '1 2 3/4 5/' 4 :'(y + ".);._2 x' 7
 8  9 10
11 12  0

Do you understand why you can not replace y with ] in this example?

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

Reply via email to