Confirmation Theory And Achinstein's Definition Of Evidence

The confirmation relation question relating to the concept of evidence is an issue with great importance in philosophy of science.Among various confirmation theories that attempt to capture the nature of confirmation relation,there are three of them as mainstream.They are confirmation by instance,Hypothetico-Deductive confirmation,and probability confirmation theory based on Bayesian probability theory.The last one can be divided into high probability confirmation theory and positive relevance confirmation theory.These confirmation theories are confronted with some paradoxes or difficulties,including black raven paradox,irrelevance theory difficulty,underdetermination theorem and grue paradox.Peter Achinstein claims that all three confirmation theories will subject to two evidence assumptions,namely the weakness assumption and the a priori assumption.Therefore,they would not satisfy our requirement of the concept of evidence.An appropriate concept of evidence should provide scientists with a good reason to believe a hypothesis.Aiming that,he puts forward the concept of potential evidence,on which he then puts forward subjective evidence,epistemic-situation evidence and veridical evidence based.He considers veridical evidence as what scientists pursue.By importing a probability threshold and the concept of explanatory connection,indeed will potential evidence reject the counterexamples bothering three confirmation theories.However,the lack of some important standards such as the means to determine the correctness of an explanation and the means to assign probability will decrease the plausibility and utility of Achinstein's theory.Moreover,revising the debate between Kronz and Achinstein,I think admitting the role of intermediary hypothesis in determining a three-directions explanatory connection will inevitably incur some predicaments which disobey our intuition.And if a good reason were unable to work as the sufficient condition of evidence,a separation between a good reason and evidence is also inevitable.We may appeal to a constrained two-directions explanatory connection,or some reason rather than a good reason to reply the difficulty.In the end of the article,I point out that veridical evidence is so ideal that divorces from our requirement of an evidence theory.When pursing the utility of an evidence theory,we must consider the influence of epistemic-situation in determining the truth of an evidential claim.Consequently,epistemic-situation evidence,or potential evidence relative to epistemic-situation,maybe more important.