Feyerabend I 284
Lakatos/Feyerabend: his case studies contain:
(A) a discussion of certain sociological regularities
(B) the proposal of arbitrary standards which have no methodological force
(C) the submission that the regularities are not merely in the facts, but are the traits of reason.
>
Regularities.
I 285
Lakatos/Feyerabend: all theories of knowledge start with the question: what is knowledge? The conventional answer contains a definition of knowledge. The answer of Lakatos is a huge improvement. His standards are much closer to science. The methods of revision refer to history. One can now discuss all the rules in a realistic context.
>
Knowledge, >
History/Lakatos, >
Progress.
I 286
FeyerabendVsLakatos: he has not shown that his standards are those of sciences, and that they lead to important results. He neglects the "external" influences or distorts the history of the sciences by the imputation that the progress of the observed deviations from his standards is not necessary. (sic).
---
Hacking I 193
Science/Research programs/Lakatos:
1. Specific plan of problem solving. (general).
2. Lakatos: abstract and historical: sequence of evolving theories over centuries.
Definition "Heuristics"/Lakatos/Hacking: heuristics defines the importance of problems.
I 198
Lakatos: "hard core" ("negative heuristics": this must remain untouched): the Gravitational law and the three laws of dynamics are irrefutable.
I 199
"Protective Belt"/Lakatos: one only encounters a selection of problems with which one deals. Further objections are then ignored.
LakatosVsPopper: therefore the verification has still a place! The researchers are choosing a few problems, refutations can be completely uninteresting!
I 199
Degenerative/Lakatos: bad research programs: E.g. instead of malnutrition one mistakenly assumed a virus disease of the population. Instead of Beriberi epidemic. Malnutrition by new process of steam peeling of rice.
Degenerative/Lakatos: any modification of the theory was not made before, but only after the observations.
I 202
HackingVsLakatos: it does not help to choose new programs without proof of past performance. E.g. Is the attempt to find cancer viruses progressive or degenerate? We'll know that later.
