Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cornell!uw-beaver!rice!titan!dorai
From: dorai@titan.rice.edu (Dorai Sitaram)
Newsgroups: comp.lang.lisp
Subject: Re: Overloading of NIL (as empty list and logical falsity)
Message-ID: <2906@kalliope.rice.edu>
Date: 22 Mar 89 21:43:10 GMT
References: <8247@csli.STANFORD.EDU> <9840@megaron.arizona.edu>
Sender: usenet@rice.edu
Reply-To: dorai@titan.rice.edu (Dorai Sitaram)
Distribution: na
Organization: Rice University, Houston
Lines: 36

In article <9840@megaron.arizona.edu> mike@arizona.edu (Mike Coffin) writes:
$From article <8247@csli.STANFORD.EDU> (Mark Johnson):
$[Speaking of the overloading of nil and the empty list in Lisp]
$$ Does anybody else have any comments on the matter?  Do more advanced
$$ functional languages (ML?) incorporated the same overloading of NIL?
$
$Scheme has distinct values for false and the empty list: #F and '(),
$respectively.  Any Scheme value can be used as a boolean value for the
$purpose of a conditional test. Any value except #F and the empty list
$count as "true";  #F and the empty list count as "false".  This seems
$to be the best of both worlds to me...

Which Scheme is this?

In all the Schemes I've come across, #f and '() are just two ways of
writing the same object. They are indistinguishable from each other,
i.e., (eq? #f '()) ==> true. A unification algorithm written in Scheme
that used '() as the empty substitution list and #f to indicate
failure would bomb because of the false/empty-list schizophrenia.

It is perhaps helpful not to think of the-empty-list as a given in
Lisp/Scheme. Consider that we're given the numbers, booleans t and
nil, characters and other basic constants of your choice. Added to
this, we now are given the ability to form dotted pairs (cons), with
which we can form data-types of our choice. _One_ of the user
data-types that we can create from the above givens is something
called a list, which is distinguished by its having a nil finally when
one traverses down its dotted-pair representation in a rightmost
fashion. This way, nil is just a user-representation for the user's
empty list concept, and _we_ are doing the overloading of nil, _not_
Lisp. Feel better? ;-)

As an aside, Lisp would be more accurately though less euphonically
described as Dottedpaip.

--dorai