What Makes One Idea Better Than Another?
Log in or sign up to participate in this discussion.
With an account, you can revise, criticize, and comment on ideas, and submit new ideas.When we reject explanation B in favour of explanation A, there must be some property of the explanations that comprehensively distinguishes A as better than B.
One fatal flaw in B is enough to reject B. A does not also need to possess some opposite property that makes it the winner.
A surviving the criticism does not show that A is progress over B. We could reject one error while moving even further from reality.
Explanation A surviving criticism does not establish that explanation A is objective progress over explanation B. But the claim that one fatal flaw in explanation B is enough to reject explanation B does not depend on explanation A being objective progress. Rejecting explanation B and establishing explanation A as objective progress are separate claims.
Progress does not require a complete ranking of every live explanation. If B has a fatal flaw while no fatal flaw is currently known in A or C, the flaw in B is enough to reject B while A and C both remain live. There does not need to be one property that all three explanations possess to different degrees.
We can rationally prefer A over B if A has no pending criticisms whereas B has at least one.
The property you mention is the existence of pending criticisms.
An explanation can be harder to vary than a rival and still contain a fatal flaw of another kind. Unless every possible fatal flaw is itself a form of ETV, HTV cannot comprehensively determine which explanation should be rejected.
The more HTV an explanation is, the more closely it corresponds to reality.
The idea ‘I’m located in a hemisphere’ is ‘easy to vary’ as to which hemisphere but it’s true, meaning it corresponds to reality.
HTV and correspondence are orthogonal.
Take the same explanation in two possible realities. The explanation has the same content and the same HTV in both, but it can correspond to one reality and fail to correspond to the other. HTV alone therefore cannot determine correspondence.
There’s a risk in posting follow-up thoughts as nested comments. The risk is that criticism chains won’t bubble up far enough.
If you want to add more information to an idea, I suggest revising it.
They are separate ideas, and the latter ideas depend on their formers. If there is a flaw in former ideas, they apply to latter ideas, but if there is a flaw in latter ideas, they don’t necessarily apply to the former. So arranging them like this was intentional, to respect that hierarchy.
This approach breaks search functions, where the system only goes off the ‘Criticized’ flag. Now the system shows some ideas as rationally adoptable even though they aren’t.
But the UI won’t reflect that automatically. Now you have to remember which threads have this structure and which don’t. Will you remember that, say, six months from now?
Popper says to prefer ideas that are better corroborated, more truthlike. I wrote about this recently, pasting it here:
Corroboration was meant to help us form rational preferences between ideas. From The Logic of Scientific Discovery (pp. 281-282):
We can sometimes rationally justify the preference for a theory in the light of its corroboration, that is, of the present state of the critical discussion of the competing theories, which are critically discussed and compared from the point of view of assessing their nearness to the truth (verisimilitude). The current state of this discussion may, in principle, be reported in the form of their degrees of corroboration.
Specifically, corroboration helps you form a rational preference between two theories that have both been shown to be false while there are no known alternatives. From Conjectures and Refutations (p. 318):
… even after t2 has been refuted in its turn, we can still say that it is better than t1, for although both have been shown to be false, the fact that t2 has withstood tests which t1 did not pass may be a good indication that the falsity-content of t1 exceeds that of t2 while its truth-content does not. Thus we may still give preference to t2, even after its falsification, because we have reason to think that it agrees better with the facts than did t1.
That “t2 has withstood tests which t1 did not pass” means t2 is better corroborated than t1.
Pavel Tichý and David Miller independently proved in 1974 that Popper’s definition of verisimilitude leads to a contradiction. Explanation:
Let us suppose that A and B are both false, and that A’s truth content exceeds B’s. Let a be a true sentence entailed by A but not by B. Let f be any falsehood entailed by A. Since A entails both a and f the conjunction, a&f is a falsehood entailed by A, and so part of A’s falsity content. If a&f were also part of B’s falsity content B would entail both a and f. But then it would entail a contrary to the assumption. Hence a&f is in A’s falsity content and not in B’s. So A’s truth content cannot exceeds B’s without A’s falsity content also exceeding B’s.
Suppose now that B’s falsity content exceeds A’s. Let g be some falsehood entailed by B but not by A, and let f, as before, be some falsehood entailed by A. The sentence f→g is a truth, and since it is entailed by g, is in B’s truth content. If it were also in A’s then both f and f→g would be consequences of A and hence so would g, contrary to the assumption. Thus A’s truth content lacks a sentence, f→g, which is in B’s. So B’s falsity content cannot exceeds A’s without B’s truth content also exceeding A’s. The relationship depicted in Diagram 4 simply cannot obtain: in this sense, that picture is an “impossible” one, like those in Escher’s famous drawings.
Preference formation is a computational task. (#5255) To claim we’ve understood a computational task, we must have programmed it (Deutsch). Popper never programmed it, so he didn’t understand it.
(To his credit though, he got pretty technical in defining verisimilitude, logical vs. empirical content, etc.)
There’s no need to single out corroboration or verisimilitude as special kinds of criticism.
Preference formation using either can easily be reduced to the method of counting pending criticisms:
Idea A fails to survive two tests.
Idea B fails to survive only one of the two.
Idea C says: I have only A and B; I don’t have time to create another; although both are false, I have to act; so for now I prefer B since at least it passes a test A doesn’t.
If idea C has no pending criticisms, then we should adopt it. (And thus ‘transitively’ adopt B.)
Consider that preference formation is an engineering problem:
https://libertythroughreason.com/rationality-as-computation/
And if it is, consider also that maybe it’s worth staying away from concepts like HTV, which we have no idea how to program:
https://blog.dennishackethal.com/posts/hard-to-vary-or-hardly-usable
https://libertythroughreason.com/simple-refutation-of-david-deutschs-hard-to-vary/
The task is to write an algorithm that takes two ideas (eg in the form of strings of characters) and returns the preferable one.
How would one go about writing that?
I think it’d be fine to start with obviously wrong answers to that question and then incrementally correct errors. For example, an algorithm that favors the idea with fewer words. (A bit like Occam’s razor.) That’d be fine as long as one wrote working code.