On Thu, Jul 30, 2009 at 4:00 PM, Jose Mario Quintana<[email protected]> wrote: > I did not refer to the parsing but to the execution of the fork. > The side effect is in the sense that, in a typical context, the behavior > of [: u v is equivalent to [ u@:] v (or [ u...@] v if you prefer).
Ok, but there need be no side effect here, unless you consider all execution to be "side effect". [: is visible at parse time, and ([: u v) can be used to create a different kind of derived verb than (w u v) where w, u and v are verbs and w is not [: > For example, before the introduction of [: no verb w "affected the > valence of u" during the execution of a dyadic fork w u v, as far as I know. >From my point of view, all verbs have both valences, though of course they can have empty domains in either or both valences. And, the parser selects the appropriate valence depending on context. But, I agree that, before the introduction of [:, the middle verb in a fork was always used dyadically. And, I agree that ([: U V) is equivalent o (U@:V) when U and V are verbs. -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
