green numbers give full details | back to texts | unexpand these ideas
17596 | Coherence problems have positive and negative restraints; solutions maximise constraint satisfaction |
Full Idea: A coherence problem is a set of elements connected by positive and negative restraints, and a solution consists of partitioning the elements into two sets (accepted and rejected) in a way that maximises satisfaction of the constraints. | |||
From: Paul Thagard (Coherence: The Price is Right [2012], p.42) | |||
A reaction: I'm enthusiastic about this, as it begins to clarify the central activity of epistemology, which is the quest for best explanations. |
17597 | Coherence is explanatory, deductive, conceptual, analogical, perceptual, and deliberative |
Full Idea: I propose that there are six main kinds of coherence: explanatory, deductive, conceptual, analogical, perceptual, and deliberative. ...Epistemic coherence is a combination of the first five kinds, and ethics adds the sixth. | |||
From: Paul Thagard (Coherence: The Price is Right [2012], p.43) | |||
A reaction: Wonderful. Someone is getting to grips with the concept of coherence, instead of just whingeing about how vague it is. |
17598 | Explanatory coherence needs symmetry,explanation,analogy,data priority, contradiction,competition,acceptance |
Full Idea: Informally, a theory of explanatory coherence has the principles of symmetry, explanation, analogy, data priority, contradiction, competition and acceptance. | |||
From: Paul Thagard (Coherence: The Price is Right [2012], p.44) | |||
A reaction: [Thagard give a concise summary of his theory here] Again Thagard makes a wonderful contribution in an area where most thinkers are pessimistic about making any progress. His principles are very plausible. |
17602 | Verisimilitude comes from including more phenomena, and revealing what underlies |
Full Idea: A scientific theory is progressively approximating the truth if it increases its explanatory coherence by broadening to more phenomena and deepening by investigating layers of mechanisms. | |||
From: Paul Thagard (Coherence: The Price is Right [2012], p.46) |
17601 | Neither a priori rationalism nor sense data empiricism account for scientific knowledge |
Full Idea: Both rationalists (who start with a priori truths and make deductions) and empiricists (starting with indubitable sense data and what follows) would guarantee truth, but neither even begins to account for scientific knowledge. | |||
From: Paul Thagard (Coherence: The Price is Right [2012], p.46) | |||
A reaction: Thagard's answer, and mine, is inference to the best explanation, but goes beyond both the a priori truths and the perceptions. |
17600 | Bayesian inference is forced to rely on approximations |
Full Idea: It is well known that the general problem with Bayesian inference is that it is computationally intractable, so the algorithms used for computing posterior probabilities have to be approximations. | |||
From: Paul Thagard (Coherence: The Price is Right [2012], p.45) | |||
A reaction: Thagard makes this sound devastating, but then concedes that all theories have to rely on approximations, so I haven't quite grasped this idea. He gives references. |
17599 | The best theory has the highest subjective (Bayesian) probability? |
Full Idea: On the Bayesian view, the best theory is the one with the highest subjective probability, given the evidence as calculated by Bayes's theorem. | |||
From: Paul Thagard (Coherence: The Price is Right [2012], p.45) |