[FRIAM] Graph/Network discursion.

[gepr ⛧]

Yes, absolutely! The arguments about the ambiguity of terms like complex, model, layer, and the capitalization of words in programming languages fall squarely in the ontologies domain. And that means they fall under graph and network theory, though I think "labelled transition systems" might be better.

The trouble with reduction to a unified ontology is also critical, because I think the majority of the problem we're struggling with (writ large) is reductionism, or more generally, monism/non-duality.  I think Aaronson makes the point nicely here:

  Higher-level causation exists (but I wish it didn’t)
  http://www.scottaaronson.com/blog/?p=3294

In microcosm, Nick's _latch_ onto the onion as metaphor for unorderable complexes is a symptom of the underlying problem that we use language (or conceptual structures) according to our temporally- and proximally-bound _purpose_.  Anyone who claims to work only with some sort of universal, Platonic truth is delusional or disingenuous.  A unification of that language is not only impossible, but if it were possible, it would be a kind of order-death (opposite of heat death).  Perfect and universal normalization to a single norm would paralyze us all.

But, obviously given my crybaby tantrum about "level" vs. "layer", I believe _some_ resolution/alignment of language is necessary for any sort of progress/produce.  To me, a collaboratively produced document about complexity that comes from a small subset of this community that intuitively agrees already, with no friction in the process, would be a useless "yet another jargonal paper about complexity".

So far, the useful friction I see is:

  Russ: information is required
  Stephen: nearly any physical system squeezed in the right way
  Nick: gen-phen map
  Eric: cumulative hierarchy

I don't think pressurizing this plurality into a unified "system of thought" will produce anything interesting.  But I _do_ think allowing them to flower/flesh out from a bare, common skeleton would be interesting _IF_ the fleshing out didn't lose the skeleton amongst the flowers or lose the flowers by over-emphasizing the skeleton.


On June 9, 2017 1:49:45 PM PDT, Steven A Smith <sasmyth at swcp.com> wrote:
>... how to explicitely *superpose* multiple
>graphs/networks, and in particular ontologies, rather than try to 
>*compose* and then resolve the contradictions among them.   It is 
>ancient enough work that I don't remember exactly what I was thinking, 
>but it was revisited in the Faceted Ontology work in 2010ish...  but 
>that was MUCH more speculative since we didn't actually HAVE a specific
>
>ontology to work with.   "If we had some rope, we could make a log 
>raft.... if we had some logs!"
>
>I sense that both you (Glen) and Marcus have your own work (or 
>avocational) experience with ontologies and I'm sure there are others 
>here.  For me it is both about knowledge representation/manipulation
>AND 
>collaborative knowledge building which is what I *think* Nick is going 
>on about, and what is implied in our bandying about of "concept/mind 
>mapping".