From: Bill H. <ho...@gm...> - 2010-07-19 14:54:48
|
On Mon, Jul 19, 2010 at 9:45 AM, Michel_Dumontier <Mic...@ca...> wrote: >> I was not saying that physicists and chemists are basing things on >> pure conjecture. That accusation was leveled at you and Rob. I don't >> see any evidence that physicists and chemists are requesting these >> terms, or are saying that their science is being impeded by their >> absence from a given ontology. > > Perhaps they are not requesting these terms because the current policy does not recognize them. Perhaps? Again, no evidence here. And you admit the lack of requests. Progress. > >> Your premise is that scientific discourse mandates that we drop >> certain criteria from the OBO Foundry because we cannot reasonably >> carry out scientific discourse otherwise. I see no evidence that >> scientific discourse is being impeded or cannot be accommodated >> through other means. And I see no scientists claiming that their >> discourse is inhibited. > > I'm a computational scientist, and I think the discourse is not being accommodated. Uh, OK. I still haven't heard why the suggestion to use application ontologies, derived from reference ontologies is not sufficiently accomodating, though. > >> You have not provided any data that >> scientists are having trouble specifically due to the Foundry criteria >> you are attacking. > > Interesting. What precisely would be compelling *data*? "We received 100 requests on the Theoretical Physics Ontology request tracker who collectively co-authored 75% of the papers mentioning the Higgs Boson for a term 'Higgs Boson'. Detailed analysis of these requests revealed that these individuals wish to formulate hypotheses about the Higgs as logical axioms for use in an application rapidly gaining ground in the community, called 'The Theoretical Physics Analyzer', which makes inferences in the domain based on the mathematical formulae in which their theories are currently formulated". > > m. > > >> >> The leap from "criticism" to "scientific evidence" (in Rob's original >> post that launched this debate) was a breathtaking one. >> >> > >> >> >> >> Despite formally naming element 112 in February (copernicium with >> >> symbol Cn) on the basis of a rigorous process of assessing the >> >> evidence (see: http://www.iupac.org/publications/pac/82/3/0753/), >> the >> >> IUPAC has not felt any urgency to update the Periodic Table just >> yet, >> >> as it still shows element 111 as the highest confirmed (see: >> >> http://old.iupac.org/reports/periodic_table/, and if you can find a >> >> newer Periodic Table from IUPAC please point me to it). >> >> >> >> For certain, they don't need these terms to annotate data, because >> >> there are no instances of these things to annotate (or at least, >> they >> >> haven't observed any yet, in the case of the Higgs). So scientists >> >> don't need these terms to communicate data about their instances. >> >> >> >> So what other types of scientific communication do folks purport to >> >> serve with such (as yet) non-referring terms? The best case I've >> >> heard so far is NLP, but again, an application ontology would serve >> >> the purpose just fine. >> > >> > It's not just NLP - it's scientific discourse in general. >> >> Such as? Are all the ideas in all our scientific papers and >> presentations now to be machine processable? A noteworthy and >> certainly audacious goal, but not all such ideas and hypotheses and >> conjectures (no matter how well supported by evidence) belong in an >> ontology, nor can or should be represented in an ontology. >> >> > >> >> >> >> Finally, many theoretical physicists, who are taken quite seriously >> in >> >> the community, for example Lisa Randall, are formulating theories >> that >> >> have no Higgs Boson. Putting the Higgs Boson in your reference >> >> ontology could be detrimental to such work. >> > >> > As Rob has been arguing, this is only a problem if you entrench a >> particular world view on your ontology. There's no reason why Lisa's >> terminology can be part of a formalized scientific discourse, and >> appear in an ontology. >> >> And there's no reason the Higgs cannot go in an application ontology >> with a provisional status and support various (quite theoretical and >> very few actual) use cases just fine. I'd like to see one physicist >> actually using an ontology of leptons, fermions, and bosons at all, >> let alone one who needs the Higgs, to send over his latest hypothesis >> about the Higgs as a logical axiom to his colleague on another >> continent. >> >> > >> >> >> >> So, so far as the OBO Foundry has as its mission the creation of >> only >> >> reference ontology, the requirement that terms refer to entities >> known >> >> to exist is just fine. >> > >> > Only if you think that it should be restricted to such things, >> obviously ;-) >> >> That was not my idea. I saw it stated here at least twice. >> >> > >> > m. >> > >> >> >> >> Bill >> >> >> >> On Fri, Jul 16, 2010 at 11:35 AM, David Osumi-Sutherland >> >> <dj...@ge...> wrote: >> >> > Hi Raven, >> >> > >> >> > On 16 Jul 2010, at 16:06, Ravensara Travillian wrote: >> >> > >> >> >> insisting on instances sounds like--to me--to insist on proving >> >> >> every single entity existed >> >> > >> >> > I think its just insisting that we have good scientific reasons to >> >> > believe that instances of the class exist/can exist/have existed. >> If >> >> > someone at the LHC claims to have found an instance of Higg's >> Boson, >> >> > what they'll really be saying is that the all their measurements >> of >> >> > bubble trails (or whatever methods they now use for these >> >> experiments) >> >> > are consistent with the presence of a particle with the properties >> >> > predicted for Higg's Boson. I don't see this as proof - someone >> else >> >> > might come along with a different and highly plausible >> interpretation >> >> > of the same data. But it provides a good reason to believe that >> >> > instances of Higg's boson exist. I don't see assertions about >> >> > evolutionary history as so massively different from this. We >> can't >> >> > go back in time and identify instances of common ancestors. All >> >> we'll >> >> > ever have is the clues of extant species and fossils. Still, >> there >> >> > are many assertions we can make about evolutionary history, based >> on >> >> > this evidence, that are boringly uncontentious. >> >> > >> >> > There seems to be general agreement here that the OBO Foundry >> rules >> >> > need to be clarified - particularly with respect to time. But I >> >> don't >> >> > see any prospect of coming up with guidelines for what types of >> >> > evidence counts. Nor should there be. That's the job of science. >> >> > >> >> > Finally - can we take this discussion of LCAs off the main list. >> I >> >> > think the details are a bit of a branch from the main thread here. >> >> > I'll try to reply to some of your points directly. >> >> > >> >> > Cheers, >> >> > >> >> > David >> >> > >> >> > >> >> > David Osumi-Sutherland, PhD >> >> > Ontologist / Curator >> >> > Virtual Fly Brain / FlyBase >> >> > Department of Genetics >> >> > University of Cambridge >> >> > Downing Street >> >> > Cambridge, CB2 3EH >> >> > UK >> >> > +44 (0)1223 333 963 >> >> > >> >> > >> >> > >> >> > >> >> > ------------------------------------------------------------------ >> --- >> >> --------- >> >> > This SF.net email is sponsored by Sprint >> >> > What will you do first with EVO, the first 4G phone? >> >> > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> >> > _______________________________________________ >> >> > Obo-discuss mailing list >> >> > Obo...@li... >> >> > https://lists.sourceforge.net/lists/listinfo/obo-discuss >> >> > >> >> >> >> -------------------------------------------------------------------- >> --- >> >> ------- >> >> This SF.net email is sponsored by Sprint >> >> What will you do first with EVO, the first 4G phone? >> >> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> >> _______________________________________________ >> >> Obo-discuss mailing list >> >> Obo...@li... >> >> https://lists.sourceforge.net/lists/listinfo/obo-discuss >> > >> > --------------------------------------------------------------------- >> --------- >> > This SF.net email is sponsored by Sprint >> > What will you do first with EVO, the first 4G phone? >> > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> > _______________________________________________ >> > Obo-discuss mailing list >> > Obo...@li... >> > https://lists.sourceforge.net/lists/listinfo/obo-discuss >> > >> >> ----------------------------------------------------------------------- >> ------- >> This SF.net email is sponsored by Sprint >> What will you do first with EVO, the first 4G phone? >> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> _______________________________________________ >> Obo-discuss mailing list >> Obo...@li... >> https://lists.sourceforge.net/lists/listinfo/obo-discuss > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by Sprint > What will you do first with EVO, the first 4G phone? > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first > _______________________________________________ > Obo-discuss mailing list > Obo...@li... > https://lists.sourceforge.net/lists/listinfo/obo-discuss > |