As a convinced Knightian/ Keynesian on risk and uncertainty, and someone who thinks stories are a hugely important part of how we understand things, your post gives me a double sense of confirmation. Which is perhaps either unbearably dogmatic or itself vindication of what you say.
At first sight, 'better than any of the others' looks more like plurality than majority (to put things in voting terms). Balance of probs seems more like majority (esp since I guess we want it to align with actual 'objective' probabilities when those are in evidence).
'Better than all the others put together' doesn't work for a number of reasons (non-additivity, proliferation of alternatives...); 'better than its complement' is obscure; 'better than its negation' doesn't engage the competing narratives framework.
You could say, by analogy with the final round in an AV ballot, plurality can be normalised as a majoritarian criterion since 'best' entails 'better than next best', but this seems like a trick.
I suppose the problem boils down to the 'aligning with objective probabilities' bit. Maybe that's wrong framing: perhaps objective prob can be assimilated to the category of 'story', or (very boldly) it can be dismissed on the never-very-satisfying basis that hard cases make bad law.
Great idea for SF novels not being written, I have a few too, but I don't see how narratives as we know them would even exists where an animal logic of causation does not exist. Logic is a hindsight. And as the Roman's said, truth is the daughter of time, so why bother finding/investigating guilt/credit/blame/reasons/ratio/closure.
You could assign (1) to lack of plausibility (US standard for dismissing a complaint on the pleadings, although technically it is not the same as “makes no sense”) and (2) to the standard to survive summary judgment and get to a jury in a civil case (a reasonable jury could accept this version of events).
I like this. I’ve always found attempts to quantify legal standards of proof maddeningly meaningless (barring “preponderance of the evidence”, which does work as a better than 50% chance).
As a convinced Knightian/ Keynesian on risk and uncertainty, and someone who thinks stories are a hugely important part of how we understand things, your post gives me a double sense of confirmation. Which is perhaps either unbearably dogmatic or itself vindication of what you say.
Yes, narrative plausibility is why the Linda problem in behavioral economics is so misleading and actually a set up.
I hadn't thought of that but, yes, what a great point
At first sight, 'better than any of the others' looks more like plurality than majority (to put things in voting terms). Balance of probs seems more like majority (esp since I guess we want it to align with actual 'objective' probabilities when those are in evidence).
'Better than all the others put together' doesn't work for a number of reasons (non-additivity, proliferation of alternatives...); 'better than its complement' is obscure; 'better than its negation' doesn't engage the competing narratives framework.
You could say, by analogy with the final round in an AV ballot, plurality can be normalised as a majoritarian criterion since 'best' entails 'better than next best', but this seems like a trick.
I suppose the problem boils down to the 'aligning with objective probabilities' bit. Maybe that's wrong framing: perhaps objective prob can be assimilated to the category of 'story', or (very boldly) it can be dismissed on the never-very-satisfying basis that hard cases make bad law.
I like the idea for the novel
Great idea for SF novels not being written, I have a few too, but I don't see how narratives as we know them would even exists where an animal logic of causation does not exist. Logic is a hindsight. And as the Roman's said, truth is the daughter of time, so why bother finding/investigating guilt/credit/blame/reasons/ratio/closure.
You could assign (1) to lack of plausibility (US standard for dismissing a complaint on the pleadings, although technically it is not the same as “makes no sense”) and (2) to the standard to survive summary judgment and get to a jury in a civil case (a reasonable jury could accept this version of events).
I like this. I’ve always found attempts to quantify legal standards of proof maddeningly meaningless (barring “preponderance of the evidence”, which does work as a better than 50% chance).