- From: Pat Hayes <phayes@ihmc.us>
- Date: Wed, 2 Mar 2016 22:17:44 -0600
- To: Reto Gmür <reto@wymiwyg.com>
- Cc: David Booth <david@dbooth.org>, semantic-web@w3.org
On Mar 2, 2016, at 2:14 AM, Reto Gmür <reto@wymiwyg.com> wrote: > > > On Wed, Mar 2, 2016, at 05:45, Pat Hayes wrote: >> >> On Feb 29, 2016, at 2:50 AM, Reto Gmür <reto@wymiwyg.com> wrote: >> >>> On Mon, Feb 29, 2016, at 03:04, David Booth wrote: >>>> On 02/26/2016 06:04 AM, Reto Gmür wrote: >>>>> Sure, still I think that schema:rangeIncludes is not meaningless (as it >>>>> restricts the rdfs:range statements that are possible) and that >>>> >>>> Under the standard open world assumption (OWA) I do not think it is >>>> correct to say schema:rangeIncludes *restricts* anything. Bear in mind >>>> that given the statement: >>>> >>>> :p schema:rangeIncludes :Cat . >>>> >>>> one could always add an arbitrary additional class to the property's >>>> "expected type(s)" by adding another statement like: >>>> >>>> :p schema:rangeIncludes :Dog . >>>> >>>> Therefore, the original statement cannot be *restricting* anything >>>> (under the OWA). >>> >>> I did not say that it restricts the possible values of the properties, >>> but I'm saying that it restricts the possible rdfs:range statements that >>> are possible without creating a contradiction. >>> >>>> >>>> Personally, I think a reasonable way to interpret its meaning is that it >>>> says 'there exists an individual :d such that :d rdf:type :Dog'. >>>> >>>>> it has >>>>> some pragmatic usefulness such as when building editors that suggest >>>>> values for a specific property. >>>> >>>> Agreed. And it's also useful if you're doing closed world reasoning. >>> >>> Well, even if you're closing the world I'm not sure you can do reasoning >>> about the instance data based on this property. >>> >>> I claim that for something to be expected it must be possible, based on >>> this one can create a contradiction with statements of necessity >>> expressed with rdfs:range. >> >> Nothing in the RDFS namespace can express anything about necessity. RDFS >> is not a modal logic. > > Well, According to the Necessitation Rule, any theorem of logic is > necessary (⊢ p →⊢ ◻ p). You have to be very careful when citing this rule. It is a valid rule of *some* *modal* logics, not a generally valid principle of reasoning. It is not an inference rule in all modal systems. And even for those systems in which it is used, it applies to logical derivability *in that same system*, not derivability in (for example) RDFS. > So if - as you do - agree that p rdfs:range t and x p y together entail > y rdf:type t, They do so in RDFS and OWL, yes. > you cannot at the same time state that it is not > *necessary* for y to be of rdf:type t when p rdfs:range t and x p y, I can, and do, claim this; since (to repeat) neither RDFS nor OWL are modal logics, so the question of whether some RDFS or OWL statement is or is not *necessary* is meaningless. Or at any rate, it is meaningless until you or someone else provides a semantics for the appropriate modal extension of RDFS or OWL in which these assertions can be meaningfully stated. Pat > > Reto > >> >> Pat Hayes >> >>> >>> However, I don't think that only what is expected is possible. So even >>> if we know that only :Cat and :Dog are expected the unexpected :Mouse is >>> still possible. >>> >>> Reto >>> >>> >> >> ------------------------------------------------------------ >> IHMC (850)434 8903 home >> 40 South Alcaniz St. (850)202 4416 office >> Pensacola (850)202 4440 fax >> FL 32502 (850)291 0667 mobile (preferred) >> phayes@ihmc.us http://www.ihmc.us/users/phayes >> >> >> >> >> >> > > ------------------------------------------------------------ IHMC (850)434 8903 home 40 South Alcaniz St. (850)202 4416 office Pensacola (850)202 4440 fax FL 32502 (850)291 0667 mobile (preferred) phayes@ihmc.us http://www.ihmc.us/users/phayes
Received on Thursday, 3 March 2016 04:18:16 UTC