Last night I was talking with an old friend of mine, an atheist Jew, who is now in the best relationship of her life with a devout Roman Catholic. We talked about the fact that she was more surprised than he was about the fact that their connection transcends their difference in metaphysics. He sees himself as a devout Roman Catholic; she sees him as a good human being.
This conversation reminded me of an older thought that’s been swirling around in my head for a few weeks: the disunity of reason.
It is widely held by philosophers (that peculiar sub-species!) that reason is unified: that the ideally rational person is one for whom there are no fissures, breaks, ruptures, or discontinuities anywhere in the inferential relations between semantic contents that comprise his or her cognitive grasp of the world (including himself or herself as part of that world).
This is particularly true when it comes to the distinction between “theoretical reason” and “practical reason”. By “theoretical reason” I mean one’s ability to conceptualize the world-as-experienced as more-or-less systematic, and by “practical reason” I mean one’s ability to act in the world according to judgments that are justified by agent-relative and also agent-indifferent reasons (“prudence” and “morality”, respectively).
The whole philosophical tradition from Plato onward assumes that reason is unified, and especially, that theoretical and practical reason are unified — different exercises of the same basic faculty. Some philosophers think of them as closer together than others — for example, Aristotle distinguishes between episteme (knowledge of general principles in science, mathematics, and metaphysics) and phronesis (knowledge of particular situations in virtuous action). But even Aristotle does not doubt that episteme and phronesis are exercises of a single capacity, reason (nous).
However, as we learn more about how our cognitive system is actually structured, we should consider the possibility that reason is not unified at all. If Horst’s Cognitive Pluralism is right, then we should expect that our minds are more like patchworks of domain-specific modules that can reason quite well within those domains but not so well across them.
To Horst’s model I’d add the further conjecture: that we have pretty good reason to associate our capacity for “theoretical reason” (abstract thinking and long-term planning) with the dorsolateral prefrontal cortex and also pretty good reason to associate our capacity for “practical reason” (self-control and virtuous conduct) with the ventromedial prefrontal cortex (and especially in its dense interconnections with the limbic system).
But if that conjecture is on the right track, then we would expect to find consistency between theoretical reason and practical reason only to the extent that there are reciprocal interconnections between these regions of prefrontal cortex. And of course there are reciprocal interconnections — but (and this is the important point!) to the extent that these regions are also functionally distinct, then to that same extent reason is disunified.
And as a consequence, metaphysics and ethics may have somewhat less to do with each other than previous philosophers have supposed.
possible 28-letter phrases, so it should take about 
different phrases before we found the target. That is without arranging that the phrase that replaces the parent is the one closest to the target. Once that highly nonrandom condition is imposed, the number of generations to success drops dramatically, from