Relating versus Replacing

“Centralization”. “Mapping”. These are overloaded concepts. They mean different things in different contexts. One hawk-eyed reader has rightly cried Ambiguous! on my usage of these terms in recent notes.

In Democracy in America, de Tocqueville “distinguishes between centralized government, which concentrates common interests like foreign relations or general laws, and centralized administration, which concentrates local interests into one place.” [ref] Centralization in governance, in coordination (including for tractability of certain analyses), and in registration/indexing/search, is not the same as centralization in administration and situation of resources.

The domain-domain “mappings” I had in mind in my last note are what are termed “functors” in category theory. In un-namespaced natural language, so-called synthetic names such as “functors” are more narrow than so-called natural names like “mappings”, but authors risk losing (a subset of) readers in any insistence on One Name for a concept.

A strength of Linked Data is in celebrating a plurality of narrow terms as namespaced URIs. A natural term needn’t be replaced with a synthetic canonical one – we have things, not strings, and we can relate things via the RDF model of a triple - , where each component is a URI – a thing, not a string.

I’m going to back away for now from attempting to explicate my thoughts on open- vs closed-world logic, in part because it seems particularly difficult to do so with brevity without opening serious gulfs of ambiguity, given the serious formal attention such concerns have garnered for decades. For a deeper dive relevant to this audience, I highly recommend a recent article by Chris Mungall.

Subscribe to get short notes like this on Machine-Centric Science delivered to your email.