Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uunet!mstan!amull
From: amull@Morgan.COM (Andrew P. Mullhaupt)
Newsgroups: comp.lang.misc
Subject: Re: The Fundamental Concept of Programming language X
Summary: Operators in New APL's
Keywords: programming languages, abstractions
Message-ID: <666@s5.Morgan.COM>
Date: 11 Jan 90 05:50:53 GMT
References: <1470@mdbs.UUCP> <1782@aipna.ed.ac.uk> <2886@water.waterloo.edu> <1532@castle.ed.ac.uk>
Organization: Morgan Stanley & Co. NY, NY
Lines: 30

In article <1532@castle.ed.ac.uk>, nick@lfcs.ed.ac.uk (Nick Rothwell) writes:
> 
> Why are "operators" different to functions?
 The situation is much like that in APL2, to wit:
They can have functions as arguments. The result of an operator is
(can be) a function suitable for use as an argument to another
operator. The exceptions in APL are "jot" which is not any kind of
thing, (it only exists as part of the "jot-dot" operator), and slash,
which is the symbol for both "compress" (a function) and "reduce"
(an operator). (A mistake graven in stone by now...)


Functions, therefore, are not 'first class data' since the functions
which operate on them are not functions. (This is a lot like Lisp).

> 
> What do you do about full higher-order facilities (e.g. higher-order
> functions returning higher-order functions)? "adadadadadverbs"?
> 

I think here, the APL2 answer is that you don't. Operators are not
(so far as I can tell) legitimate areguments to other operators. The
reason for this is not unlike what you have guessed - you cannot
express the definition of an "operator on operators" in the present
syntax. English is not so limited; things which modify adverbs
are very frequently adjectives.

Later,
Andrew Mullhaupt
> 		Nick.