"Free will" is worthy of extended treatment. Whether a sentient being is a human or god or fictional robot, I last asserted that free will is an inadequate excuse for failing to demonstrate its qualities through definitive actions. (The truth of having a quality is verified pragmatically.) I continue in this belief, but I've noticed a counterargument which needs to be addressed.
The potential problem with any real test or experiment of anything is an unavoidable entanglement with external facts. Interference should be minimized. Output data should be tied as closely as possible to the tested thing and as loosely as possible to all untested things. Setup should aim for isolation. Overall, tests should be carefully designed to confirm a single supposition without introducing extraneous factors.
In the case of a decision, the entangling external facts take the form of a "frame of reference". The decision-maker evaluates each choice by referring to the frame around the decision. Information, prediction, and inclination are all included. The frame of reference is how the decision reduces to a solvable puzzle. Apart from it, the decision-maker is lost and bewildered.
Its importance is felt in the common regret, "If I'd only known then what I know now." A decision evaluated through a disparate frame produces a disparate selection. Hence the frame is a crucial ingredient in judging what someone decides and reconstructing the rationale. It's indispensable in the assignment of blame or praise. An observer could commit grave mistakes without knowing more about the decision-maker's frame.
During a test decision, the frame of reference of the sentient is one of the test's complicating external factors. Simply put, the sentient may have greater or lesser knowledge than the tester! Then the disparity in knowledge leads to a disparity in frame and, crucially, a disparity in checking the result. Perhaps the tested sentient had the quality X all along but a divergent frame overrode/concealed quality X or demanded a convoluted expression of it. In order to dodge a test failure and continue to maintain that the tested sentient has tested quality X, an objector may propose that the sentient decides through a poorly-understood frame of reference. Checkmate!
As I hinted earlier, I think the "frame of reference counterargument" is shrewd but not devastating to my main point about testing sentients who have free will. Frames of reference certainly can be distinct, so I agree that there can't be an absolute guarantee that tester and tested concur in basic analysis of the test decision. My disagreement centers on how insignificant a problem it is.
To interpret an apparent "bad" decision as "secretly good" requires an alarming mismatch between frames of reference. If preserving the link between the sentient and quality X involves radically disconnecting the sentient's frame from the tester's, then I believe that even this solution throws quality X into question. For example, I'm stymied by the very meaning of "mercy" when a sentient's frame considers the killing of an unbeliever to be "merciful".
But not every gap needs to be that wide. As a matter of procedure, a tinier and less consequential test presumably avoids the frame-based inaccuracy that plagues weighty and thorny tests. A sentient with genuine quality X should show it in tests, no matter the size. Arguably, the most trivial decisions are the most revealing and the least able to afford excuses. Someone fixated on punctuality will arrive early to unimportant events. Yet there's a ready explanation too for failed tiny tests, which have barely any scope for frame ambiguity: the tiny "doesn't matter enough" to the sentient, who only bothers to act on the big stuff.
Finally, I'm dumbfounded by the apogee of the counterargument by frame of reference. The most extreme discrepancy is between the tester and a sentient whose frame is all-inclusive of space and time. A sentient with consciousness of everything is inherently not testable by decisions. The sentient's decisions are not understandable by the tester. Nobody can estimate a decision made by reference to everything in every time. Once again, instead of saying, "The god with quality X chose not to pass the test due to a frame of reference incomprehensible to us", the more honest summary is "The god with a frame of reference incomprehensible to us chose not to pass the test due to the utter worthlessness of characterizing decisions in that frame as having 'quality X'." Faith is trusting a god when a god's decision-making is irresistibly opaque. It may be true that the god isn't "beyond" good and evil, but how would anyone know which choices are which, when measured from that perspective?
The potential problem with any real test or experiment of anything is an unavoidable entanglement with external facts. Interference should be minimized. Output data should be tied as closely as possible to the tested thing and as loosely as possible to all untested things. Setup should aim for isolation. Overall, tests should be carefully designed to confirm a single supposition without introducing extraneous factors.
In the case of a decision, the entangling external facts take the form of a "frame of reference". The decision-maker evaluates each choice by referring to the frame around the decision. Information, prediction, and inclination are all included. The frame of reference is how the decision reduces to a solvable puzzle. Apart from it, the decision-maker is lost and bewildered.
Its importance is felt in the common regret, "If I'd only known then what I know now." A decision evaluated through a disparate frame produces a disparate selection. Hence the frame is a crucial ingredient in judging what someone decides and reconstructing the rationale. It's indispensable in the assignment of blame or praise. An observer could commit grave mistakes without knowing more about the decision-maker's frame.
During a test decision, the frame of reference of the sentient is one of the test's complicating external factors. Simply put, the sentient may have greater or lesser knowledge than the tester! Then the disparity in knowledge leads to a disparity in frame and, crucially, a disparity in checking the result. Perhaps the tested sentient had the quality X all along but a divergent frame overrode/concealed quality X or demanded a convoluted expression of it. In order to dodge a test failure and continue to maintain that the tested sentient has tested quality X, an objector may propose that the sentient decides through a poorly-understood frame of reference. Checkmate!
As I hinted earlier, I think the "frame of reference counterargument" is shrewd but not devastating to my main point about testing sentients who have free will. Frames of reference certainly can be distinct, so I agree that there can't be an absolute guarantee that tester and tested concur in basic analysis of the test decision. My disagreement centers on how insignificant a problem it is.
To interpret an apparent "bad" decision as "secretly good" requires an alarming mismatch between frames of reference. If preserving the link between the sentient and quality X involves radically disconnecting the sentient's frame from the tester's, then I believe that even this solution throws quality X into question. For example, I'm stymied by the very meaning of "mercy" when a sentient's frame considers the killing of an unbeliever to be "merciful".
But not every gap needs to be that wide. As a matter of procedure, a tinier and less consequential test presumably avoids the frame-based inaccuracy that plagues weighty and thorny tests. A sentient with genuine quality X should show it in tests, no matter the size. Arguably, the most trivial decisions are the most revealing and the least able to afford excuses. Someone fixated on punctuality will arrive early to unimportant events. Yet there's a ready explanation too for failed tiny tests, which have barely any scope for frame ambiguity: the tiny "doesn't matter enough" to the sentient, who only bothers to act on the big stuff.
Finally, I'm dumbfounded by the apogee of the counterargument by frame of reference. The most extreme discrepancy is between the tester and a sentient whose frame is all-inclusive of space and time. A sentient with consciousness of everything is inherently not testable by decisions. The sentient's decisions are not understandable by the tester. Nobody can estimate a decision made by reference to everything in every time. Once again, instead of saying, "The god with quality X chose not to pass the test due to a frame of reference incomprehensible to us", the more honest summary is "The god with a frame of reference incomprehensible to us chose not to pass the test due to the utter worthlessness of characterizing decisions in that frame as having 'quality X'." Faith is trusting a god when a god's decision-making is irresistibly opaque. It may be true that the god isn't "beyond" good and evil, but how would anyone know which choices are which, when measured from that perspective?
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.