Pinker contrasts "greedy" reductionism, which he rejects, with "good" reductionism , which he advocates (note the moralizing grade-school teacher language).
Good reductionism as he sees it is the sort that admits of different explanatory levels. Citing Hilary Putnam, he says that you can't explain why a round peg doesn't fit in a square hole by talking about what happens at the quantum level. The whole may be the sum of its parts, but we can grasp the one explanation of why the square peg doesn't fit into the round whole only by thematizing the whole (in this case, the peg and the hole) as such.
Digression: actually, a more adequate explanation must also talk about how forces at a micro level are responsible for the resistance we feel when pressing the non fitting parts together. An adequate explanation must include BOTH how the parts as such tend to interact and the configuration of the whole made of these non-matching parts (the square/round shapes [but aren't 'square' and 'round' really short-hand for the way many, many parts are related?]). I think Pinker and I would agree about this much.
But I would disagree with him about ontological reductionism. Or rather, Pinker doesn't acknowledge the distinction between epistemic reductionism (which he rejects) and ontological (which he accepts). He is a full-blown ontological reductionist (shall we call this "ontologically greedy reductionism"?), and this position has inherent problems. For example, if every whole (regardless of how we may conceptualize it) is the sum of its parts, then doesn't this lead to a kind of infinite regress (of wholes that are merely the sum of parts of parts of parts of parts... ad infinitum) BUT he seems to think he can deflect this and other possible criticisms of ontological reductionism by adopting an anti-reductionist epistemology. That is, he thinks that wholes in nature are nothing more than the sum of their parts (ontological reductionism); but recognizes that in order to understand how nature works one must think in terms of wholes (epistemological anti-reductionism).
It seems to me that the only way to avoid the problems of onto-reductionism is to affirm that some wholes are not simply the sums of their parts. Not every whole is machine-like in nature, even though some are. The central problem in the study of nature is to distinguish one from the other. And Pinker is not only fails to offer a solution; he seems clueless about the problem.
I have to relisten to chapter 4, as he also talks about the social/individual language use as a whole/part phenom. I am not clear about this, but he seems to me to treat language qua existing in society as a proper ontological whole rather and definitely NOT as being the sum of its parts (the parts being language as it is proper to individual communicators) This whole/part contrast seems to have a somewhat Cartesian pallor; for he treats the societal level as externally observable, while treating language in the individual is private, internal.
I must re-listen to make sure that I got this right, but it seems to me that he doesn't think in terms of language as in some sense being "amongst" commuicators. No sense of unity in manifold, of the common good, of the public yet non-measurable. Stuck in positivistic and categories, with a strong tendency toward atomism, even in treatment of linguistic activity. But recognizes that there is a need for contrast. Thus ends up with a dualism of sorts.
If my interpretation of him is right, then the question of how the public/private are related to each other in his schema would surely be embarrassing to him: how does the internal individual linguistic activity relate to the public, external one? A materialistic cognate of Cartesian dualism (we need a materialistic equivalent of the ghost in the machine how about the gnome inside the machine?).
Good reductionism as he sees it is the sort that admits of different explanatory levels. Citing Hilary Putnam, he says that you can't explain why a round peg doesn't fit in a square hole by talking about what happens at the quantum level. The whole may be the sum of its parts, but we can grasp the one explanation of why the square peg doesn't fit into the round whole only by thematizing the whole (in this case, the peg and the hole) as such.
Digression: actually, a more adequate explanation must also talk about how forces at a micro level are responsible for the resistance we feel when pressing the non fitting parts together. An adequate explanation must include BOTH how the parts as such tend to interact and the configuration of the whole made of these non-matching parts (the square/round shapes [but aren't 'square' and 'round' really short-hand for the way many, many parts are related?]). I think Pinker and I would agree about this much.
But I would disagree with him about ontological reductionism. Or rather, Pinker doesn't acknowledge the distinction between epistemic reductionism (which he rejects) and ontological (which he accepts). He is a full-blown ontological reductionist (shall we call this "ontologically greedy reductionism"?), and this position has inherent problems. For example, if every whole (regardless of how we may conceptualize it) is the sum of its parts, then doesn't this lead to a kind of infinite regress (of wholes that are merely the sum of parts of parts of parts of parts... ad infinitum) BUT he seems to think he can deflect this and other possible criticisms of ontological reductionism by adopting an anti-reductionist epistemology. That is, he thinks that wholes in nature are nothing more than the sum of their parts (ontological reductionism); but recognizes that in order to understand how nature works one must think in terms of wholes (epistemological anti-reductionism).
It seems to me that the only way to avoid the problems of onto-reductionism is to affirm that some wholes are not simply the sums of their parts. Not every whole is machine-like in nature, even though some are. The central problem in the study of nature is to distinguish one from the other. And Pinker is not only fails to offer a solution; he seems clueless about the problem.
I have to relisten to chapter 4, as he also talks about the social/individual language use as a whole/part phenom. I am not clear about this, but he seems to me to treat language qua existing in society as a proper ontological whole rather and definitely NOT as being the sum of its parts (the parts being language as it is proper to individual communicators) This whole/part contrast seems to have a somewhat Cartesian pallor; for he treats the societal level as externally observable, while treating language in the individual is private, internal.
I must re-listen to make sure that I got this right, but it seems to me that he doesn't think in terms of language as in some sense being "amongst" commuicators. No sense of unity in manifold, of the common good, of the public yet non-measurable. Stuck in positivistic and categories, with a strong tendency toward atomism, even in treatment of linguistic activity. But recognizes that there is a need for contrast. Thus ends up with a dualism of sorts.
If my interpretation of him is right, then the question of how the public/private are related to each other in his schema would surely be embarrassing to him: how does the internal individual linguistic activity relate to the public, external one? A materialistic cognate of Cartesian dualism (we need a materialistic equivalent of the ghost in the machine how about the gnome inside the machine?).
Comments