Dimitri Mollo replied to uh's status
@uh @UlrikeHahn@fediscience.org @dcm@social.sunet.se
Indeed this is a much longer chapter! I stopped at p. 73 for now.
I share your concern about the ontology here: Juarrero talks constantly about constraints 'doing' things, and even being a form of causality (though she doesn't say how). But her examples suggest that what she calls constraints are just ways of describing patterns that appear when certain entities interact with each other in specific organised ways. This impression is reinforced by the apparently circular treatment of context-dependent constraints on p.70: they are characterised by appeal to constrained interactions...
But then, rather than being something ontologically additional that does things, constraints are just ways of talking about features of such patterns, which are in their turn constituted by the familiar kinds of causal interactions between entities. So, nothing ontologically new, just, at most, new-ish alternative explanatory tools.
(This connects, I think, to the Deacon vs …
@uh @UlrikeHahn@fediscience.org @dcm@social.sunet.se
Indeed this is a much longer chapter! I stopped at p. 73 for now.
I share your concern about the ontology here: Juarrero talks constantly about constraints 'doing' things, and even being a form of causality (though she doesn't say how). But her examples suggest that what she calls constraints are just ways of describing patterns that appear when certain entities interact with each other in specific organised ways. This impression is reinforced by the apparently circular treatment of context-dependent constraints on p.70: they are characterised by appeal to constrained interactions...
But then, rather than being something ontologically additional that does things, constraints are just ways of talking about features of such patterns, which are in their turn constituted by the familiar kinds of causal interactions between entities. So, nothing ontologically new, just, at most, new-ish alternative explanatory tools.
(This connects, I think, to the Deacon vs Heil discussion we had last week. Hat tip: @NicoleCRust@neuromatch.social, @awaisaftab@mastodon.social) (Also connected, perhaps, to the traditional criticisms of Dynamical Systems Theory, when applied to CogSci, in terms of it offering just redescriptions, rather than explanations)
Though I promised not to complain any more about the general obscurity and lack of rigour, I can't help myself. Juarrero talks about enabling constraints 'locking in' (??) information and energy flow to 'real-world traits/characteristics' (what others are there? what do such constraints do that has to do with real-world traits rather than... what exactly?). Then on p.71 she seems to give 3 non-equivalent characterisations of enabling constraints in the space of two paragraphs...