- From: pat hayes <phayes@ai.uwf.edu>
- Date: Fri, 13 Dec 2002 14:30:07 -0600
- To: "Wagner, G.R." <G.R.Wagner@tm.tue.nl>
- Cc: Sandro Hawke <sandro@w3.org>, www-rdf-rules@w3.org, timbl@w3.org, pfps@research.bell-labs.com
> > >and KR is not a branch of classical/mathematical logic but rather >> >the other way around. >> >> My goodness, ... I absolutely do not accept your premis here. > >I know :-) > >> >In real world domains, and in natural language, >> >we use the term "rule" for denoting many different things with very >> >different semantics. >> >> I take it then that 'rule' in Rule ML can have no single semantics, >> and that (very much in the XML spirit) , RuleML is not intended to >> convey a uniform semantic framework, but instead to be all things to >> all men? > >Having different semantics for different types of things does not >mean to have "all things to all men"! Im afraid it does, in practice, on the semantic web. Unless there is some way to make the intended choice of semantics explicit, there is no way for a software agent to know what semantics is intended, ie how to treat the formal sentences it is reading. This is a very important and basic point: normal human use can accept a degree of ambiguity which is not acceptable in the SW. >We need different semantics >for different types of derivation rules, and still different >semantics for reaction rules and for deontic assignment rules, >in the same sense as there are different semantics for different >types of implication (such as material implication, intuitionistic >implication and relevant implication). You speak as someone whose business is 'rules', which is fine. But my primary concern is the semantic web, and the primary aim is to find technologies which may be useful, not to use this as a domain of application for ones favorite technology. I am not at all convinced, yet, that 'rules' in your sense are particularly useful for the SW, to be honest; they may do more harm than good unless their semantics are carefully delineated. > > Yes, the discussion on this thread has been about rules in two >> senses: inference rule as in logic, and 'rule' as in logic >> programming, where it refers to a Horn clause. > >Rules in normal (and even less in extended) logic programs are >not Horn clauses! More's the pity. But yes, I was oversimplifying slightly. > > I note that all the >> examples you cite are systems which are considered to be a kind of >> programming language, and those, of course, typically do not have a >> normal assertional semantics. > >Logic "programming" languages have never been very successful as >programming languages, but rather as specification and KR languages. They have had almost no impact whatever as serious KR languages. So if they aren't successful as programming languages, then they would seem to be a niche technology of no particular importance. And in the larger scheme of things, I think that is in fact a reasonable summary. Fascinating as it is to its enthusiasts (among whom I number myself, BTW), I don't think logic programming has been a serious contender as a programming platform since the demise of the 5th Generation initiative. >Extended/intuitionistc/fuzzy/etc. logic programs do have an >"assertional" (model-theoretic) semantics. But unlike that of >RDF and OWL, also because they have additional connectives, >these semantics are non-classical. I would draw a distinction between various kinds of logic, on the one hand, and various kinds of logic programming, on the other. The latter, unlike the former, typically have a minimality condition in their semantics which relies upon the recursion theorems, and makes them inherently nonmonotonic. Such languages are basically different from assertional logics, in spite of their superficial appearance to the contrary; and this difference is more or less orthogonal to the classical/nonclassical distinction between various styles of logic. Pat -- --------------------------------------------------------------------- IHMC (850)434 8903 home 40 South Alcaniz St. (850)202 4416 office Pensacola (850)202 4440 fax FL 32501 (850)291 0667 cell phayes@ai.uwf.edu http://www.coginst.uwf.edu/~phayes s.pam@ai.uwf.edu for spam
Received on Friday, 13 December 2002 15:30:22 UTC