ИСТИНА

Theory undedetermination by data (TUD) entails existence of empirically equivalent, but postulating incompatible ontology theories. There is a definite link between concrete forms of such strong TUD and attitudes toward realism defense in the philosophy of science. Historical analysis reveals that some forms of TUD can be eliminated by future experiments and does not pose an unsolved problem for realist. But the other forms of TUD appear more than simple puzzle for realist. We can find such TUD in modern quantum-relativistic paradigm, that counters to realist`s claim that modern physics is closer to the reality then classical physics. Theory underdetermination by data (TUD) produces active polemics in modern philosophy of science. Different forms of such undedetermination are discussed and criticized in order to gain an adequate understanding of the process of theory-based construction of ontology. Some forms of TUD, such as Humean and holistic, neither historically nor logically were opposed to realism. But if TUD entails existence of empirically equivalent, but postulating incompatible ontology theories, it poses a big problem to realism. Such form of TUD I should call strong TUD. There is a definite link between concrete forms of such strong TUD and attitudes toward realism defense in the philosophy of science. The strategy which antirealists use to make strong TUD an argument against realism includes two stages: 1) vindicate the principle possibility of an equivalent theories construction; 2) find concrete examples of such theories in the history of science. The first problem was examined by Quine, who understand strong TUD as “the thesis about the world, not the thesis about theories”. The world itself produces empirical multiformity and we always can deduce the same set of “observation conditional” from alternative theoretical constructions. Some philosophers of science make big efforts to develop special algorithms in order to produce empirically equivalent theories (Kukla). The other (van Fraassen) construct theories which are equivalent in some trivial sense, like Newtonian mechanics with the assumption that the center of the universe is at absolute rest and versions of Newtonian mechanics which postulate that it is moving with constant velocity in particular direction. The vast majority of these approaches is opened for critique (Stanford) and can construct only “artificial” theories which scientists never interested in. Increasingly realist should be concerned about existence of empirically equivalent theories in the history of science. Realists and critiques of TUD work out some strategies aimed to demonstrate that strong TUD can be eliminated through the history of science. The first strategy is presented by Worrell, who suggests distinguishing empirical adequacy and empirical success. Two empirically equivalent theories will not pose a problem for realism, if one is successful in predicting phenomena, while the other just “saves the phenomena” by the use of ad hoc hypotheses. In this case we should realistically interpret only the ontology of the first theory. Trivial example of such unproblematic for realist equivalence is Ptolemean and heliocentric astronomy. The second strategy is more ambitious, it was discussed by Laudan and Leplin. It`s key idea is that scientific development makes the equivalence the local property of theories. This idea can be stated in the argument, which I called “optimistic meta-induction” (OMI). The argument has local applicability and suggests that there exist such form of equivalence that for every equivalent theories future experiments will eliminate equivalence and show which of these theories (as realist wants it) permits to represent the structure of the world. OMI is not the strong argument: it can be supported only by the history of classical physics. For example, Weber`s and Maxwell`s electrodynamics (long-range action electrodynamics and close-range action electrodynamics) were empirically equivalent until the discovery of electromagnetic fields in 1888 Hertz`s experiments. Maxwell exactly for a long time suggested these two types of ontologically different electrodynamics as empirically equivalent. Possible, until the investigation of the Poisson`s spot we can assert the equivalence of corpuscular and wave optics. Particular question is if we can find such form of underdetermination in modern quantum-relativistic paradigm. In the context of realism-antirealism debate OMI also can`t block rather strong argument against realism, based on the history of science - famous “pessimistic meta-induction”. OMI just fixes the special form of TUD which can be eliminated and does not contradict the realistic picture of scientific progress. Then the problems for realism arise if we can find in the history of science the other form of strong TUD which is nonlocal and can`t be eliminated in principle. Such form appears in the context of physical geometry choice in general relativity. It was investigated by Poincare, Reichenbach and Grunbaum. The same set of observational data can be described by equivalent combinations of different (Euclidean or Non- Euclidean) geometries. But this equivalence is extra linguistic, because different geometries require different physical laws. Reichenbach illustrates this state by the notion of the abstract “universal force” (gravitation is it`s physical analogue): only the combination “laws+geometry” is empirically verifiable statement. So this is the example of empirically equivalent theories, but not simply the different mathematical formulations of the same theory. So, the choice between them historically appeared as the matter of non-empirical (pragmatic) criteria of theory evaluation. Historical analysis reveals that realism faces different forms of strong TUD. Some of them can be eliminated by future experiments and does not pose an unsolved problem for realist. But the other forms of TUD appear more than simple puzzle for realist. We can find such TUD in modern quantum-relativistic paradigm, that counters to realist`s claim that modern physics is closer to the reality then classical physics.