The arguments of philosophy of science are directly based on the analysis of scientific argumentation. Thus it does not come unexpected, that the deep methodological changes which characterise the evolution of particle physics up to string theory have philosophical consequences in this field. Three fundamental messages can be identified:
The first message is not directly coupled to string theory. The empiricist claim that scientific theories can only be motivated by their measurable effects seems the less plausible the more the evolution of particle physics has proceeded. The minimalism of the visible consequences of modern theories (in current particle physics they are reduced to a few strange lines on a bunch of photos made in an accelerator experiment) contrasts with an enormous richness of the theoretical concepts. It seems absurd to degrade the complex world of elementary particle theories to auxiliary constructions for predicting a few patterns of lines which are per se entirely irrelevant and turn up only under the extreme conditions of modern particle experiments.
The second message is equally problematic for empiricism. The core commitment of empiricism to the under-determination of physical theories (see the section on the realism debate) is undermined by the altered characteristic of theory dynamics in string theory. The fact that string theory has been a flourishing and dynamical field for at least two decades by now without any experimental corroboration suggests to posit the increasing importance of a new principle of 'theoretical uniqueness'. This principle seems to be able reliably to give a direction to the theoretical evolution even without any connection to experiment. In the context of string theory arguments of theoretical consistency are in principle sufficient to make univocal structural statements. (An example would be the theoretical deduction of the number of space-time dimensions mentioned in the section on string theory.) The principle of theoretical uniqueness can be also identified as a structural property of string theory and in our opinion constitutes an important shift of the scientific paradigm. It contradicts the postulate of theoretical under-determination and thus is at variance with empiricism.
The third message however contradicts the conventional approach of scientific realism. In string theory the principle of duality acquires a crucial position. (See the section on string theory.) Dual Versions of string theory have identical physical implications but often comprise different elementary objects and have different spatial structure. In this light a theory's elementary objects appear as technical objects which depend on the choice of the theoretical scheme. They seem to evade any realist interpretation just like envisioned by empiricism.
The three described messages in their totality contradict both empiricism and classical scientific realism. A way out could be a non-ontological realism, which, as it turns out, can be based nicely on the phenomenon of theoretical uniqueness. The resulting position, named by us 'Consistent Structure Realism', attributes reality to the unique consistent theoretical structure without imputing ontological reality to its objects. Since the abstract debate in philosophy of science also searches for intermediate positions between straightforward realism and empiricism, the relevance of Consistent Structure Realism goes beyond giving satisfaction to physical necessities. Being initially suggested by recent physical developments, it might eventually contribute to the solution of longstanding problems in philosophy of science. An interesting interplay between arguments inspired by physics and purely philosophical arguments emerges, which can be interpreted as a sign for a rapprochement of the two fields. (The arguments sketched in this section are developed in the article "Scientific Realism in the Age of String Theory")