A.V. Chusov - Ph.D. (Philos.), Associate Professor, Science Philosophy and Methodology Chair, Philosophy Faculty, M.V. Lomonosov' Moscow State University. E-mail: anchusov@yandex.ru

We define method as a system of regulatives of transformations of data and knowledge in the perspective of a typical result (it-s typical components-regulatives: a typical beginning, categories, principles, logic, and a typical result). Let's note that: a) transformations are made by means and tools; b) in science, there are three levels of transformation - objective data, knowledge and representation; c) models in scientific research are the means of objectivization of representations; d) results are often used as beginnings. In neurophilosophical researches (NPR) typical beginnings and results are specific. The beginning of a scientific research may be different from many standpoints: in logic and mathematics - definitions and axioms, in natural sciences - experimental data, in humanities and social sciences - behavior and knowledge, et al. The beginning in modern complex mathematicized disciplines often consists of previously developed models and tools. But even in mathematics the choice of initial definitions has «subject-matter» motivations, rather than formal motivations. In our opinion, subject-matter motivation is defined in part - by solved problems, in part - by problems to be solved. In a case of NPR there are presented many directions, and the motivation of applying tools and models (which were developed in mathematics and engineering) includes three factors: a) the complexity of the object of investigation, b) the existence of such methods and corresponding models that allow to solve abstract and formal problems, c) simple metaphors that enable «approximation» of a complex object with simple «building blocks». And beginnings of different research directions cannot simply be combined, in particular due to different concepts and models of typical objects. Thus, studies where «neuro-« means a single neuron are not fully compatible with those where «neuro-» means a network. Results of scientific research also constitute an essential methodological problem. In every region of research (data, knowledge and models), the result should be smth whole/integral, relatively new and relatively autonomous. But explicit criteria for integrity, innovation and autonomy in the case of data often do not exist as well as in the case of knowledge and models are controversial. In our opinion, in NPR the common basis of such criteria could be the concept of «changing reproduction of interaction». Another aspect concerns the existence of the result and it's getting - the difference between the system of methods, leading to the result in scientific research and mode of existence of a result as a relatively independent object. The result is included in the system of scientific knowledge only in accordance to his interpretation. In addition, every complex research has dual problem of integration of different level results, including intermediate results - compatibility of ontologies of levels and transition between them.