The general antipathy toward using an appropriate level of simplification can be seen here however, in this quote from Barry Smith. As you know, I do not have a philosophical issue with “realism” in general — I just don’t care — but with the particular strand of it that is inherent in BFO, which Barry pushes.

“I am beginning to suspect that for you everything is a simplification (model) — for me, functions are part of reality; they are not simplifications; I am not interested in simplifications. ”

http://groups.google.com/group/bfo-discuss/msg/865e601864fbc2dc

I agree with you,though. Anyone can get tied up in overly complicated arguments about some minor point (I hold my hand up here). We need to work on strategies to get around this; I think many of these do and will come from software engineering, and in particular from the agile and decoupled methodologies, rather than the waterfall approach that realism suggests.

I agree completely, but what realist principle says you need to give something the most detailed classification you can come up with? It’s not unheard of for conceptualists to get tied up in arguments about dependent continuants and their various flavours. Whether you are a conceptualist or a realist, you’re not using your time well if you spend most of it worrying about abstruse classification issues that are neither of interest to scientists nor useful for basic sanity checking.

I consider myself a pragmatic realist (I was recently taken to task by a realist for being a pragmatist, if I was forced to choose sides I would be on the side of the pragmatists).

I would say the problem here is “realist overreach” into the murky world of properties. Unlike with physical entities (isotopes, rats, pasta) or processes (radiation, fluorescence) we have a harder time pointing to things in order to resolve debates about what should go in our ontology or which distinctions to make. This leads to non-useful debates and analysis paralysis.

It’s particularly hard to point to a disposition (In fact, I have a particular problem with dispositions being “real” – BFO asks me to believe there are an infinite number of real but unrealized and perhaps wildly improbable dispositions floating around me every second: my disposition to raise my right eyebrow in the manner of Roger Moore, rub my tummy, fart if you pull my left pinkie, quit my job and join the circus, sing an ABBA song in a comedy Slovenian accent, throw custard pies at presenters during the next ICBO).

Trying to assert dubious sub-categories of DependentContinuant in advance is not useful and a recipe for pointless discussion, because it’s harder to use objective reality as a guide. A pragmatic approach would be to always model the physical entities and processes as named classes (taking a hardheaded ultra-realist approach – unicorns? NOT ON YOUR LIFE MATE; hallucinations/cheesy Athena posters of unicorns? SURE, KNOCK YOURSELF OUT, I DON’T CARE. Sorry Robert & Rob), and then to introduce properties where they are needed for modeling purposes (taking a hard-headed pragmatic approach – e.g. avoid weirdo classes that don’t correspond to a term a normal scientist would use; introduce distinctions that give you the desired results to queries and inferences).

If distinctions such as disposition vs quality turn out to be important in the long run, then we infer them. For example, if this particular entity is always radiating then we can infer there’s a “quality” there. If this particular entity periodically radiates contingent on presence of some input, then we infer there’s a “disposition” lurking around. If I am singing an ABBA song right now, then I have the quality of singing an ABBA song (perhaps other qualities, such as being a terrible singer, and looking like a bit of an arse). If you always sing ABBA songs after drinking three pints of lager, then we can model this precisely as such (e.g. “P(Phill-sings-ABBA|>=3 pints)=1″), and, if we so choose, infer that this is a “disposition” (based on the “if” or “P(X|Y)”). No need to argue about the category in advance. Focus on the important things (e.g. building a useful probabilistic model that predicts ontologists singing ABBA, so we are forewarned and can get out the pub).

Nothing could be more realist than this (at least according to my simplistic boneheaded sub-Newtonian can-you-point-at-it flavour of realism).

I would perhaps go further and relabel the ugly “dependent continuant” as “abstract”, “property” or “reified relationship”, but that’s enough heresy for one day..

TL;DR – realism works in general but “dependent continuant”s are dubious with regards to their claims to carve reality at the joints.

Cheers Chris

I wanted to highlight the content negotiation work because I thought it might be useful to you.

Your question may be rhetorical but I must say that I don’t understand the distinction that is made between URIs / DOIs and URIs / some other identifier scheme (for example, the accession numbers you refer to in the blog post).

What does one do with a DOI if it is not presented in the form of a URI? The answer is manually convert it into a URI using specific knowledge of the dx.doi.org service. But the same is true of accession numbers and the uniprot service, or any other identifier scheme and it’s accompanying service(s).

I suppose if I’m making a point, it is that identifier schemes and the services that support them are not separable. Knowledge of both components is required to make them useful. In that sense, I don’t see how DOIs are different from any other identifier scheme, regardless how a user consumes them – as URIs or naked identifiers.

That said, it is surely better to present identifiers in a usable form, and CrossRef has now updated their DOI display guidelines to recommend the dx.doi.org form.

Phil

$ curl -LH “Accept: application/rdf+xml” “http://dx.doi.org/10.1126/science.1157784″

$ curl -LH “Accept: text/turtle” “http://dx.doi.org/10.1126/science.1157784″

These requests return representations of RDF graphs for the metadata of a DOI.

See the announcement here:

http://www.crossref.org/CrossTech/2011/04/content_negotiation_for_crossr.html

I think the cost issues are hugely obviously overblown. Unis could all serve all scientific content in journals without much, if any, of a dent in their IT budget.

It’s a scam. A horrible horrible scam. But half-open access just moves the scam around. At some point, google’ll will crash that market, I expect.

Ironically, you can reach much of the content on this blog. You should also be able to reach the same content on archive.org but unfortunately the original was behind a registration wall, so the content has not been captured. This nicely demonstrates the problems that subscription (even free subscription!) based models have.

I’m afraid I don’t have the wikipedia email — it was on Nupedia-general I think, from Larry Sanger. I only archive my personal email. Nowadays I probably could archive all my email, because disk drives have got cheap, but this wasn’t always the case.

The link to the scientist article is broken…BTW

you should print out and frame what wikipedia email :-)

or blog it

The reviewers didn’t think it was a good enough idea. Such is life.

Phil

Seems a little strange to me – are we saying that Amazon Mechanical Turk isn’t crowd-sourcing? Seems to be to me!

Si

Although abstracts, there are some engineering principles for developing ontologies. As a matter of fact there are several methodologies for developing ontologies. Most of them, if not all, follow in one way or another these principles. However, very few of them do mention the need for a singular “view” of the world, on the contrary, methodologies tend to embrace ontological pluralism.

First design principle: “The conceptualization should be specified at the knowledge level without depending on a particular symbol-level encoding.” Second design principle: “Since ontological commitment is based on the consistent use of the vocabulary, ontological commitment can be minimised by specifying the weakest theory and defining only those terms that are essential to the communication of knowledge consistent with the theory.” Third design principle: “An ontology should communicate effectively the intended meaning of defined terms. Definitions should be objective. Definitions can be stated on formal axioms, and a complete definition (defined by necessary and sufficient conditions) is preferred over a partial definition. All definitions should be documented with natural language.” Fourth principle: “An ontology should be coherent: that is, it should sanction inferences that are consistent with de definitions. […] If a sentence that can be inferred from the axioms contradicts a definition or example given informally, then the ontology is inconsistent.” Noy and McGuinness’s first guideline: “The ontology should not contain all the possible information about the domain: you do not need to specialise (or generalise) more than you need for your application.” Noy and McGuinness’s second guideline: “subconcepts of a concept usually i) have additional relations that the superconcetp does not have, or ii) restrictions different from these of superconcepts, or iii) participate indifferent relationships than supperconcepts. In other words, we introduce a new concept in the hierarchy usually only when there is something that we can say about this concept that we cannot say about the superconcept. As an exception, concepts in terminological hierarchies do not have to introduce new relations”. Noy and McGuinness’s third guideline: “If a distinction is important in the domain and we think of the objects with different values for the distinction as different kinds of objects, then we should create a new concept for the distinction”. Noy and McGuinness’s fourth guideline: “A concept to which an individual instance belongs should not change often”.

As for the several definitions of ontologies I would say: “An ontology is a non-necessarily complete, formal classification of types of information structured by relationships defined by the vocabulary of the domain of knowledge and by the canonical formulations of its theories”

Guarino and Smith heavily influence this definition. It complies with Guarino in that an ontology is, possibly, an incomplete agreement about a conceptualisation and not a specification of the conceptualisation. Ontologies should therefore be understood as agreements amongst people within a community sharing interest in a common domain. By “incomplete” it is understood that the classification of types of information should be left open for interoperability purposes. By “formal” it is meant that the ontology specification can be easily translated into a machine-readable code, as the ontology should support inference processes within those information systems using it. However, it should be noted, that the latter is not mandatory when defining ontologies on an abstract level.

More on this on my PhD thesis http://www.alexandergarcia.name/node/19