Положителната роля на субективния фактор в науката: защита срещу Попъровата критика на субективистката вероятност
The Positive Role of the Subjective Factor in Science: A Defence against Popper's Criticism of Subjective Probability
Author(s): Anton DonchevSubject(s): Philosophy, Epistemology, Logic, Philosophy of Science
Published by: Институт по философия и социология при БАН
Keywords: Popper; probability; probability theory; Bayesian probability; Karl Popper; subjective factor in science
Summary/Abstract: Karl Popper developed his own axiomatization and interpretation of mathematical probability, commonly known as the propensity theory of probability. In arguing for his theory, he strongly criticized the so-called „subjectivist“ interpretation, which views probabilities as representing personal degrees of belief in the truthfulness of a proposition. Popper maintained that the subjectivist interpretation turns probabilistic statements into empirically unverifiable tautologies. I propose three arguments against this critique. First and foremost, the revision of personal beliefs, when available empirical evidence is given, is in fact the core of the subjective interpretation of probability. Secondly, it is relatively easy to show that the complete elimination of the subjective factor in scientific research is an utopian ideal. No philosophical interpretation of mathematical probability is able to shield scientific research from the inset of prior information – at least in the form of auxiliary hypotheses or theories. However, the subjective interpretation is the only one that explicitly includes the subjective factor in its calculations. Thirdly, the subjective factor may have a positive role and be extremely useful in science. Finally, I propose that we avoid using the relatively outdated terms „subjective probabilities“ and „subjective interpretation“ in probability theory. These elicit bad connotations and evoke scepticism in scientists from the natural sciences. In order to avoid ambiguity, it would be well to employ the commonly used modern terms „prior probabilities“ and „Bayesian interpretation“.
Journal: Философски алтернативи
- Issue Year: XXV/2016
- Issue No: 2
- Page Range: 16-22
- Page Count: 7
- Language: English, Bulgarian
- Content File-PDF