[FRIAM] the role of metaphor in scientific thought

[┣glen┫]

Ha!  I struggled to come up with "single" as an alternative name and you had 4 waiting in the wings.

I'm going to skip ahead a bit and state that my entire line of rhetoric about circularity goes back to the complexity jargon discussion we were having and whether or not, as Nick put it, a system has a say in its own boundary.  It's all about _closure_.  This particular tangent targets closure from the functional programming perspective (or maybe from the procedural one, depending on how you look at it).  When you execute a loop in a "systems" language like C, you have a good chance that whatever you do in there could have side effects.  But when you do something like that in a purely functional language, you're very unlikely (never) going to leave side effects laying around.

If the unmarried person in the just-so story were somehow "closed", then there would be no side effects left lying around as a result of walking _any_ path from the name "unmarried" to/from any other name like "widow".  But people aren't ever "closed" in any vernacular sense (never mind Rosen's or Kauffman's parsing of agency for a while).  That's why I asserted that the existence of _any_ other name (bachelor, single, widow, whatever) opens up an entirely new world of side effects (including what Peirce should call practical) to the unmarried patient.  The fact that the condition even has _names_ opens it up to nomothetic generality.  An entirely unique condition, showing up nowhere else in space or time will not have a name and is not generalizable, by definition.

FWIW, in his introduction, Nick does distinguish 3 types of implication important to analogical reasoning: "basic", "surplus-intentional", and "surplus-unintentional".  And the latter 2 types are, I think, directly related to computational side effects, where type 3 would be a bug, type 2 might be considered sloppy, and type 1 is the ideal.  This is a fantastic way to talk about this sort of thing.  But it would be easier to discuss if we either avoided discussion of circularity _or_ gave it the full analytic context it needs (starting from a relatively complete definition of closure).

You may be asking: If Nick's talking about analogs and implications, how does that relate to a computational procedure?  Well, simulation has several meanings, the 2 main ones being: mimicry vs. implementation.  I'd say 90% of simulation is about implementation.  E.g. an ODE solver numerically implements (simulates) an ideal/platonic mathematical declaration.  So, when you write a program, the computer that executes it (only during the execution) is an analog to whatever other (physical or platonic) construct might also be described by such a mathematical declaration.  Either of these two analogs can leave (surplus) side effects lying about as they reify their analogous (basic) behaviors.

I hope that's not tl;dr. 8^)


On 06/23/2017 06:52 AM, Frank Wimberly wrote:
> Has anybody mentioned that there are lot of unmarried men that you usually
> wouldn't call bachelors?  There are widowers, priests, and nineteen
> year-olds, for example.  I learned the word because my father's brother was
> a thirty-five year old Major in the Air Force with no wife. He eventually
> got married and had children. Late bloomer?

-- 
␦glen?