[Top] [All Lists]

Re: [ontolog-forum] Levels

To: Christopher Menzel <cmenzel@xxxxxxxx>
Cc: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Pat Hayes <phayes@xxxxxxx>
Date: Thu, 15 Feb 2007 14:10:27 -0600
Message-id: <p06230912c1fa663c4d3e@[]>
>Pat, I think it will save you some work (and perhaps some 
>frustration :-) if you read this whole msg before starting to respond:    (01)

OK, I did. It didn't greatly help :-)    (02)

>On Feb 14, 2007, at 1:11 PM, Pat Hayes wrote:
>>>>>  then it seems to me that this confuses a
>>>>>  mechanism for saying what (some of) the axioms of an
>>>>>  ontology are with the ontology.
>>>>   Why do you call this "confuses"? The assertion of
>>>>   what some of the axioms are can be part of the
>>>>   theory.  You seem to be insisting on a 'levels'
>>>>   distinction here which I fail to see the reason
>>>>   for (and plenty of reasons to reject.)...
>>>  Here's why I said the two are being confused.  Intuitively,
>>>  an ontology is about a certain domain, say, faculty and
>>>  admin at TAMU.  The relevant objects, properties and relations
>>>  in that domain are indicated by the names in one's ontology
>>>  language.  If one of the names is now a URL pointing to your
>>>  ontology at IHMC, then, on this intuitive understanding of
>>>  what an ontology is, your ontology at IHMC becomes
>>>  part of the *subject matter* of my faculty and admin ontology.
>>  If by 'subject matter' you mean that it is in the universe of 
>>  discourse, yes. But all sorts of things are in the universe of 
>>  discourse of any reasonably complicated assertion: your TAMU 
>>  example may well refer to numbers, dates, time-intervals, 
>>  publications, etc.
>Of course -- those things might well all end up being among the 
>things we find in the faculty/admin structure at Texas A&M when we 
>write our ontology.    (03)

OK.    (04)

>  But surely YOUR ONTOLOGY (containing, we've 
>supposed, useful information about higher-level administrative 
>structures) is not.    (05)

? I am completely unable to see what motivates 
your obviously very strong intuition here. No, 
surely, if you refer to it, it IS.    (06)

>  That would be no different than asserting that 
>the entire faculty/admin ontology is itself part of its own subject 
>matter.    (07)

?? Does it import (or otherwise refer to) itself?    (08)

(I suppose it could refer to itself in a kind of 
recursive way, perhaps in order to say something 
like "nothing that cannot be inferred from me 
should be considered to be implied by me", a kind 
of explicit disavowal of any remote or indirect 
entailments; some legal documents have clauses 
like this in them. In which case I would say, 
yes, it is part of its own subject-matter.)    (09)

>  It is the CONTENT of your ontology, not the ontology itself, 
>that becomes part of the overall ontology's subject matter.    (010)

They both do.    (011)

>  Your 
>ontology just happens to be where we find that content, but, unlike 
>the other entities you mention, it is *not* among the objects, 
>properties, and relations that are constitutive of the faculty/admin 
>structure at Texas A&M.    (012)

I didn't say it was constitutive (not sure I know 
exactly what that means). I said it was referred 
to in the ontology, and I insist that it is.    (013)

>>>   That seems wrong;
>>  I have no idea what you mean by "wrong". It seems to me to simply 
>>  be a fact, and follows from the model theory in the specs. And I 
>>  have absolutely no intuitive problem with it (I suspect because I 
>>  don't take the notion of "subject matter" as being as salient as 
>>  you apparently do.)
>I think perhaps that is the nub of our disagreement -- though I'm not 
>sure we've distinguished carefully between that intuitive idea and 
>the semantic quibble about what "formal ontology" should mean; see 
>last paragraph below.    (014)

It may be that I simply don't have this intuitive 
idea (of 'subject matter') that is guiding your 
intuitions so strongly.    (015)

>>>  So, while I am most certainly not "insisting" on it,
>>>  I am indeed suggesting that there is an important conceptual
>>>  distinction to be made here between "object-level" statements
>>>  concerning the entities, types,  and relations of interest in
>>>  the ontology and "meta"-statements about the ontology itself,
>>>  notably statements about where certain object-level statement
>  >> can be found on the web.
>>  And I want to strongly deny this conceptual distinction. I think it 
>>  reflects a pre-Web way of thinking about knowledge. And PLEASE 
>>  don't say that imports is meta-level, even in scare quotes.
>As wise as you are in the ways of the web, I cannot follow you here.  
>While, logically, you are quite right -- "imports" statements are 
>object language statements with a clean semantics -- there is still a 
>distinction to be made between ontologies, qua things in their own 
>right, and their content, just as there is between names and their 
>referents.    (016)

There is such a distinction, I agree, though Im 
not sure I really know what 'content' actually 
*is*, if you see what I mean. I presume you mean 
something like the appropriate semantic 
constraint on possible interpretations, however 
that is specified (or maybe independently of how 
it is.) Except of course that if were to take 
this seriously, then it would obviously be the 
case that the ontology *is* part of the content.    (017)

The only way I can see that this whole discussion 
makes sense is that you are appealing to a 
pre-theoretic notion of 'subject matter' or maybe 
'content' which is prior to the model theoretic 
analysis. I guess I simply don't share this 
intuition and so have no way to respond to it.    (018)

>  Ontological engineering has taken hold in part because 
>people have exceedingly strong intuitions about *what there is*, what 
>they talk about, in their various domains.  If I ask a seasoned 
>administrator about the things they talk about vis-a-vis faculty and 
>admin structure at Texas A&M, your ontology itself will not be among 
>those things; once again, the content of your ontology will be.    (019)

Hmm, that depends a lot on the particular 
administrator, I venture to guess. But in any 
case, judgements like this are very tricky. 
Domain experts themselves often have rather poor 
models of their own expertise.    (020)

>  I 
>think that is an intuition worth preserving, and I don't find it in 
>the least bit "pre-Web".  What would be pre-Web is to fail to realize 
>that one can exploit open networks to incorporate remote content like 
>your ontology into one's own ontology, with all the attendant 
>practical and semantic benefits.    (021)

I think it runs deeper than this. If B includes 
'import A' then there is a link from B to A. If 
we read this as 'endorses' (as I suggested a 
while ago, though its not really my idea) the 
global structure of these links to identify 
popularly endorsed ontologies, or for more 
sophisticated purposes. I believe Swoogle is 
already doing this. The global structure of links 
then becomes a source of information in its own 
right, and is more than just a kind of footpaths 
worn by Web traffic. Web mavens get very 
energetic about the Web not just being a huge 
collection of linked documents, but something 
more.    (022)

>   But I frankly think that it will 
>likely lead to confusion on the part of ordinary users to say that, 
>anytime a remote ontology is imported the ontology itself, rather 
>than its content,    (023)

I don't say RATHER THAN, I say IN ADDITION TO.    (024)

>  suddenly (and entirely accidentally)    (025)

accidentally?!!? Damn it, what could be less 
accidental? You put into your ontology an 
explicit reference to the other ontology, by 
USING its NAME. If this was an accident, you 
really should be more careful. :-)    (026)

>elevated to a fully-fledged element of my domain.
>>>  But really, we're simply involved in a quibble here.  The 
>>>  important (?) thing is to have a consistent understanding of what 
>>>  an formal ontology is.
>>  But I think it is a very important quibble. The fact that imports 
>>  statements are genuine assertions, not some kind of 'markup 
>>  metalevel' that can be handily ignored to protect the innocent 
>>  intuitions, is not merely a quibble: it is central to the SWeb 
>>  vision. So, I insist, imports statements *are* part of a formal 
>>  ontology.
>I am of course as enthusiastic about that vision as anybody.  And you 
>are transmogrifying my argument into a strawman if you characterize 
>me as saying that imports statements are not genuine assertions and 
>can be "handily ignored".  I am only saying that there is a 
>distinction that merits preservation between an ontology and its 
>content    (027)

Of course there is such a distinction, I have 
never denied it, though see above for worries.    (028)

>, just as there is a distinction between names and things 
>named -- I'd be *perfectly happy* with a definition of "formal 
>ontology" that includes the imports statements; and in fact, upon 
>reflection, I find myself rather persuaded that we should so define 
>them, given the OWL precedent.  I would simply argue that there 
>should also be recognized something like "the logical theory of O" 
>for a given ontology O, which comprises the DOMAIN axioms found 
>directly in O together (recursively) with all those domain and upper-
>level axioms that are pulled into O by imports statements.    (029)

Well, first you tell me what counts as a "domain 
axiom". Suppose for example we have an 
IKRIS-style structural bridging ontology which 
mostly comprises axioms which state equivalences 
between different notational styles for temporal 
information. Are these domain axioms? What if you 
import an IKL-style theory of lists, in order to 
make sense of some axioms translated from OWL 
which use RDF-style collection vocabulary to 
encode n-ary relations in binary OWL: is this all 
domain axioms?    (030)

Pat    (031)

>   If I'm 
>entirely alone here, fine, but I really do think this is important.
>Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/ 
>Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/ 
>Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
>Shared Files: http://ontolog.cim3.net/file/
>Community Wiki: http://ontolog.cim3.net/wiki/
>To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>    (032)

IHMC            (850)434 8903 or (650)494 3973   home
40 South Alcaniz St.    (850)202 4416   office
Pensacola                       (850)202 4440   fax
FL 32502                        (850)291 0667    cell
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes    (033)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (034)

<Prev in Thread] Current Thread [Next in Thread>