Books on Amazon
|Theories||Feyerabend I 238
Lakatos/Feyerabend: also Lakatos' insightful attempt to establish a methodology that takes the historical reality of the sciences seriously, but which nevertheless subjects them to a control on the basis of regularities discovered in itself, is not excluded from this conclusion.
1. There are not the regularities to which Lakatos refers to, he idealizes the sciences just as his predecessors.
2. If the regularities were regularities of the sciences, and therefore useless to the "objective" judgment.
3. Lakatos' regularities are only a finery behind which an anarchic process is basically concealed.
Falsification/LakatosVsPopper/Feyerabend: some of the most famous falsifications were anything but that. And, moreover, completely irrational.
Lakatos/Feyerabend: Thesis: one should grant theories a "breathing space": in the evaluation counts the development of theories over a long period of time and not the current form. Moreover, methodological standards are not beyond criticism.
Hacking I 206
Theories/Knowledge/HackingVsLakatos: Instead of increase of knowledge, it should mean: increase of theories!
Feyerabend/VsLakatos: his "methodology" is of no use when one needs advice on current research.
Cover I 196
Theory revision/Lakatos/Schurz: (Lakatos 1974, 129ff) Methodology of scientific research programs: two assumptions:
1. "Immunization": it is always possible to save the core of a theory in the event of a conflict with the experience by making adjustments to the periphery.
2. "Protective Belt": every (physical) theory needs auxiliary hypotheses (excluding ceteris paribus hypotheses) to provide empirical predictions. These lie like a protective belt in the outer periphery around the center and core. Conflicts with experience can then be eliminated by replacing or dropping an auxiliary hypothesis.
Definition Anomaly/Lakatos: an observation date which contradicts the entire theory (core + periphery).
Definition ad hoc hypothesis: assumes more complex system conditions in which unknown disturbing factors are postulated.
Vs: Problem: this does not explain the different date. That is, it remains an anomaly even after the introduction of the ad hoc hypothesis!
Ad hoc/Lakatos: such adjustments are only legitimate if they are scientifically progressive. They must have new empirical content.
Falsification/LakatosVsPopper: a theory version is only falsified when there is a progressive new version (with new empirical content). That is, there is no "immediate rationality" (instant decision) which theory is better. This can only be seen in historical development.
Definition Research Program/Lakatos: hard theoretical core along with a negative and a positive heuristics.
Definition negative heuristics/Lakatos: Adaptations are not made in the core, but only at the periphery. However, in the course of a degenerative development the modus tollens hits can also be directed against the core.
Definition positive heuristics/Lakatos: a program that allows more and more complex theoretical models or system conditions for the core to deal with unruly data.
Theory version/Schurz: core plus periphery.
Definition Falsification/Schurz: a theory version is falsified, iff. some of the phenomena derived deductively from it were falsified by actual observational sentences. (s) Schurz always speaks of sentences instead of observations.
Verisimilitude/SchurzVs/Failure/Success/Theory: the concept of failure has the advantage that it is not the epistemological-conflicted consequences of the theory that are understood, but the phenomena. The concept of truth is based only on the consequences. ((s) phenomena cannot be true success: > "object").
Definition tacking paradox/Lakatos/Schurz: the possibility to increase the empirical content of a theory version by the mere conjunctive addition of some empirically unchecked assertion.
Solution/Lakatos: the connection of an auxiliary hypothesis creating a new empirical content with the previous theory must be more intimate than that of a mere conjunction.
Solution: the theory T must be homogeneous with respect to the empirical content:
Definition Homogeneity/Theory/Schurz: a factorization ((s) division) of T with respect to E (T) is not possible. Logical form: subdivision of T and E(T) into two disjoint subsets
T1UT2 = T and
E1UE2 = E (T) so that T1 implies all phenomena in E1 and T2 implies all phenomena in E2. If this is possible, the theory is heterogeneous.
Any theory obtained by irrelevant amplification can be factored in this sense. A connection of the theory T with this gain H is empirically not creative.
The Methodology of Scientific Research Programmes: Volume 1: Philosophical Papers (Philosophical Papers (Cambridge)) Cambridge 1980
Wider den Methodenzwang Frankfurt 1997
Erkenntnis für freie Menschen Frankfurt 1979
Einführung in die Philosophie der Naturwissenschaften Stuttgart 1996