> From: bill lam
>
> mer, 13 Jan 2010, Sherlock Ric skribis:
> > > From: Tracy Harms
> > >
> > > I think of inflections as occurring to the right of characters
> other than whitespace. I consider null, newline, and tab to be
> whitespace,
> > > along with the space character. I don't see the count of characters
> as significant. Longer primaries such as ( {:: ) are not subordinate.
> >
> > This is a nice and simple way of thinking of it. In other words the
> > primitives ( p.. ) and ( {:: ) have a double inflection.
>
> I think the role of . and : in p. {:: are word formation but not
> inflection. If it is a inflection as in human language, the new
> word should be a derivative of the root. eg,
>
> think -> think-able
> friend -> friend-ly
>
> so that the speaker who know the root can understand the meaning of
> the derivative without consulting a dictionary, but this is not the
> case of J primitive.
Although there are some instances of this in the J vocabulary (eg. if you know
( * ), ( *: ) and ( % ) then you can probably predict what ( %: ) will do), I
agree that this is often not the case for J primitives.
I took Tracy's use of "inflection" to be as described by Dan earlier in the
thread. If so I suppose we generally agree on what is going on, but need a
better term that isn't encumbered with an incompatible existing meaning.
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
