define

[Kirchner@CSNET-RELAY.ARPA, Roger <rkirchne%carleton.csnet@CSNET-RELAY.ARPA>]

Kent M Pitman said that there are other possible interpretations of
(DEFINE (((...) ...) ...) ...)
besides as an extended template for procedure definition.
What would be the objections to making this interpretation standard?

If
(define (comp f g) (lambda (x) (f (g x))))
is allowed, why shouldn't
(define ((comp f g) x) (f (g x)))
be allowed?

Evidently it is in MIT Scheme, and it is in PC Scheme, an
implementation of TI Scheme.  In what Schemes is it allowed,
which treat it as an error, and which give it a different
interpretation?

Roger Kirchner