p Many philosophers argued that, inasmuch as a most important and probably the only task of epistemological analysis is substantiation of knowledge, it should obviously single out and dissect all premises of knowledge, including those connected with self-consciousness. Epistemological research must explicate what is implicit, thus implementing absolute reflexion.

p We may recall that one of the solutions to the problem offered in the past consisted in the assertion that the reflexive relation of "I" to itself constitutes the supreme foundation of any knowledge. The proposition formulating this reflexive relation was taken to be absolutely indubitable and irrefutable. The epistemological reflexion about knowledge was interpreted as reflexion of "I" about itself.

p We have endeavoured to show the cul-de-sacs and the insoluble difficulties to which the acceptance of this orientation in epistemology leads. In particular, we have tried to show that any knowledge, and in the first place the knowledge of the world of external objects, though it assumes the subject’s self-consciousness, cannot as a matter of principle be reduced to the subject’s reflexion about himself. And insofar as knowledge about external objects can never be absolutely unquestionable (such as not to allow 219 any further specifications and corrections), however reliable it may be in practice, natural doubts arise about the need for searching for absolute principles and absolutely indisputable foundations of knowledge.

p These doubts are redoubled as we take into account the experiences of modern science in substantiating certain special kinds of scientific knowledge. We have already noted, for instance, the impossibility of completely reducing arithmetic to set theory or of one physical theory to another, as well as the impossibility of reducing theoretical knowledge to a set of protocol utterances, propositions about "sense data" or laboratory operations. Different structures of knowledge are linked in ways other than reduction. This circumstance has to be taken into account in substantiating knowledge.

p The question remains, however: to what extent is absolute completeness of reflexion possible? To what degree can the premises of knowledge be singled out, elucidated, and dissected?

p In attempting to answer this question, let us recall Quine’s arguments about the problem of radical translation. Quine points out that the language in which we speak is given to us in a different manner than a strange language which we study. With regard to the latter, we consider the relation of its expressions to the real objects and actual situation, that is to say, we reflect about this language. As for our own language, it directly presents to us the picture of the world and not its own structure. We know our own language in the sense that we can use it for rendering some objective content. But that is not explicit knowledge. The language is inseparable for us from that objective knowledge which we obtain with its help, so that we do not even notice it, as it were: it is in the background of consciousness. (That does not rule out the possibility of reflexion about our own language, but we have to split our language into two in this case. One of them will be the object language, the one that is studied, that is, it will play a different role than before, functioning as an ensemble of theoretical hypotheses, idealisations, etc., rather than as implicit knowledge naturally given to consciousness. The second language, used as a tool for studying the first, retains the properties of implicit knowledge.) Assume that we study the structure of the theory of arithmetic trying to establish its ontology, thus performing an act of theoretical reflexion about this conceptual system. In this case, we use set theory as an instrument of reflexion. In the context of this study, set theory is not an object of reflexion and is accepted as something familiar and clear. The 220 reverse task is also possible: translation of the propositions of set theory into the language of the theory of arithmetic. Here set theory itself will be the object of reflexion and the theory of arithmetic will be accepted as something not subject to reflexion in the given context.

p In studying the history of various proofs of the stereometrical theorem concerning the correlation of the numbers of sides, apices.and faces of a polyhedron, Imre Lakatos showed that finding the weak points of arguments, that is, increasing their rigour, always assumes the existence of "foundational" knowledge. The latter serves as an instrument of analysis itself, that is,a mode of reflexion about proofs, taken as an intuitively clear and unreflected guarantee of rigour. "By each ’revolution of rigour’ proofanalysis penetrated deeper into the proofs down to the foundational layer of ’familiar background knowledge’ ... where crystal-clear intuition ... reigned supreme and criticism was banned."^^90^^ At the same time reflexion about the "foundational layer", i.e., knowledge assumed to be immediately clear (implicit knowledge, in our terminology) reveals the problematic character and even falsity of a whole series of its components.^^91^^ "The amount of assumed familiarity decreases as criticism turns background knowledge into knowledge."^^92^^

p Reflexion about the "foundational layer" assumes adopting some other type of knowledge as not subject to reflexion in the given context of the means of analysis.

p Thus even in such a science as mathematics, where the problem of substantiating knowledge figures prominently, and reflexion about the existing systems of knowledge plays an enormous role, every procedure of reflective analysis implies a framework of implicit "foundational" knowledge that is not reflected upon in the given context. Implicit knowledge plays a much more important role in factual sciences, that is, in the disciplines dealing with explanation of empirical facts. In these sciences, research activity is, as a rule, aimed at the world of real external objects rather than at the theory itself. The elaboration and development of a theoretical system and its application to empirical data (the two are usually inseparable) are perceived by the researcher as establishment of the objective connections of reality itself.

p The theoretical conceptual system is not in this case considered separately from the knowledge about real objects formulated in its terms. In such disciplines, theories are usually left unformalised and often unaxiomatised. The rules for processing empirical data, the norms and standards of discourse, and the modes of selecting 221 significant problems are not formulated explicitly but are specified along with the basic paradigmal content premises of the theory, i.e., as implicit knowledge. Michael Polanyi and Thomas Kuhn, modern specialists in the history and theory of science, pointed out the importance of implicit knowledge ("tacit knowledge" in their terminology) for the development of natural science.^^93^^ That does not mean that theoretical reflexion plays no role in the development of natural scientific knowledge (although the theoreticians mentioned here are inclined to belittle this role in every way, distorting the actual state of affairs).

p The property of reflexion indicated here (the dialectical connection between reflected and unreflected knowledge) is fully manifested with regard to those kinds of knowledge which exist in unobjectified form, i.e., belong to the individual subject (perception, recall, etc.), and also with regard to individual consciousness itself. As we have stressed, each act of individual cognition assumes self- consciousness, that is, implicit knowledge of the subject about himself. One may try to transform this implicit knowledge into explicit one, that is, to translate self-consciousness into reflexion. In this case, the subject analyses his own mental experiences, observing the flow of his psychical life, endeavouring to find out the nature of his "I", etc. It appears that in this act of reflexion, "I" simply merges with itself. In actual fact that is not so. Every act of reflexion is an act of conscious realisation or understanding. The latter always assumes definite means of understanding, a kind of framework of semantic connections. Outside this framework, reflexion is impossible. At the same time, the semantic framework presupposed by the act of reflexion is not subject to reflexion in the act itself; "dropping out" of it, it is taken as an instrument of such an act, that is, as implicit knowledge. The dissection of the flow of psychical life and meaningful definiteness of the images coming to the surface of consciousness, the spatio- temporal reference of memories—all of this is given to consciousness in the act of individual reflexion. However, the modes themselves of semantic formation of this givenness are not reflected upon. Therefore, the question does not arise in subjective reflexion about the basic possibility of other semantic characteristics of psychical life, that is, of the possibility of the content and structure of psychical life other than that which is given in self-observation. "I" itself also drops out of the act of reflexion, at least partially, for if it makes itself the object of its reflexion, it must also perform this as the subject. And that means that "I" as the subject of reflexion is not reflected upon as long as 222 we remain within the framework of individual consciousness.

p This circumstance served as the basis for Kant’s and Sartre’s view that the true nature of the subject cannot be the object of knowledge or reflexion, being given to non- reflecting consciousness only. Ludwig Wittgenstein follows a similar pattern of argument: "5.631. The thinking, presenting subject; there is no such thing. If I wrote a book The World as I Found It, I should also have therein to report on my body and say which members obey my will and which do not, etc. This then would be a method of isolating the subject or rather of showing that in an important sense there is no subject: that is to say, of it alone in this book mention could not be made. 5.632. The subject does ndt belong to the world but it is a limit of the world. 5.633. Where in the world is a metaphysical subject to be noted? You say that this case is altogether like that of the eye and the field of sight. But you do not really see the eye. And from nothing in the field of sight can it be concluded that it is seen from an eye... 5.641. ...The philosophical I is not the man, not the human body or the human soul of which psychology treats, but the metaphysical subject, the limit—not a part of the world."^^94^^ The English philosopher Gilbert Ryle believes that the "systematic elusiveness of "I" in the course of reflexive analysis indicates the fictitious nature of the object of reflexion itself, i.e., of "I "as a specific structure irreducible to the physical, bodily characteristics of man.^^95^^

p Does all of this mean that unreflected implicit knowledge cannot in general be regarded as an object of reflexion being doomed to remain forever on the periphery of consciousness, unamenable to analysis in principle? Not at all. The instrument of reflexion, that is, its semantic framework; can itself become the object of reflexive analysis, but for this purpose it must be interpreted in another semantic framework which will not be reflected upon in the new context. Implicit knowledge should not be understood as something irrational or as an arbitrary assumption unrelated to reality. In actual fact, this kind of knowledge always reflects, with a definite degree of precision, objective dependences, and in many cases practical and cognitive activity does not need special analysis of at least some of the cognitive premises on which they are constructed. There are situations, however, when this kind of analysis proves a necessity. As we have noted, such a situation exists, e.g., in the study of the foundations of mathematics.

p Let us consider the following very important point. 223 Where implicit knowledge becomes explicit, thus becoming the object of reflexion, it undergoes certain changes. Theoretical reflexion about a system of objectified knowledge means its dissection, formulation of a number of assumptions and idealisations and at the same time (which is particularly essential) specification of the knowledge itself, rejection of certain implicitly accepted premises (the procedure of reflexion is prompted exactly by the need for revising some premises of knowledge). What previously appeared clear, intuitively understandable and simple, proves to be complicated enough as a result of reflexion, and often problematic, sometimes even simply erroneous. The result of reflexion is not therefore some simple and self-obvious truths or a set of absolutely indisputable assertions forming an "absolute foundation" of the system of knowledge to which different kinds of knowledge can be reduced in one way or another. The result of reflexion is a theoretical system which is a relatively genuine reflexion of some real dependences in a definite context and which at the same time implies a whole series of assumptions, a certain implicit knowledge as a premise.

p Reflexion thus takes one beyond the framework of the existing system of knowledge, generating new knowledge, both explicit and implicit. What originally seemed (e.g., in mathematics) a purely substantiating procedure, is in reality a mode of development of the content of knowledge itself and one of the important ways of theoretical development. This procedure results in increasingly more precise reflection of the objective dependences of reality and exact reproduction of the structure and content of the scientific theories themselves. A study whose immediate goal was merely increasing the rigour of an argument generated in fact greater theoretical content in the given scientific field. Summing up his investigation of the history of proofs of the stereometrical theorem, Imre Lakatos writes: " ’Certainty’ is never achieved (the reference here is to metaphysical absolute certainty.—V.L.), ’ foundations’ are never found—but the ’cunning of reason’ turns each increase in rigour into an increase in content, in the scope of mathematics." ^^96^^

p As for the factual sciences, the links between the procedure of substantiating knowledge and the development of theoretical content are here even more explicit. We have noted already that in these sciences the problem of substantiation does not usually figure as an independent one. To the extent in which the existing system of theoretical notions allows the solution of scientific problems arising in this system, permitting at the same time definite 224 practical applications, this system is regarded as sufficiently well founded. The emergence of a substantively novel theoretical system and adoption of new paradigmal research premises reveal that the conviction of the adequacy of the old paradigm’s foundations was not quite justified. The new paradigm is not adopted through analysing the structure of theoretical knowledge within the framework of theoretical reflexion about science but in studying the real objects themselves, that is, it is accepted as a tool for a more adequate theoretical reproduction of the real dependences. At the same time, the adoption of a new paradigm implies a procedure for correlating it with an old paradigm. The latter figures in this case as an object of reflexion. Its postulates, concepts, and semantic connections are reconstructed and compared with the real objects and actual connections with the aim of retaining all that has objective real content in the old paradigm, and of eliminating everything that has no such content, that is, proves to be fictitious. Here, the new paradigm functions as an instrument for presenting the real objects and dependences. Thus, theoretical reflexion acts as an important element of transition from one paradigm to another (though Polanyi and Kuhn reject this), albeit it does not exhaust the content of the transition. This reflexion essentially means reconstruction’ of and inquiry into the old paradigm in the light and by means of the new one. Thus the theory of relativity allowed a clarification of the latent premises of classical mechanics which were not (and could not be) clear to its creators themselves. Galileo, in his turn, had to subject the system of premises and assumptions of Aristotelian physics to theoretical reflexion in laying the foundations of classical mechanics. But he could only solve this task successfully insofar as he went beyond the framework of the conceptual system of Aristotelian physics. Theoretical reflexion is the result of going beyond the limits of a given conceptual system and at the same time the means of such a step. As we see, in any case it proves to be closely linked with the development of the content of theoretical knowledge.

p In this connection, one should consider the untenability of one interpretation of the special theory of relativity. This interpretation, which gained currency thanks to Bridgman, reduces the entire significance of the special theory of relativity to reflective analysis of primary concepts of physics (such as the concept of simultaneity). From this standpoint, Einstein created not so much a physical theory as a metatheory which deals with the problem of the meaning of physical concepts. The 225 laboratory operations of measurement referred to in the special theory of relativity are viewed as an absolutely reliable basis of science, the foundation on which physics must be built. In actual fact, the operations of measurement used in the special theory of relativity, in their turn, assume a number of theoretical premises that are not operational in nature. The task of this theory is by no means solution to the problem of meaning of scientific concepts but the discovery of new content dependences in actual reality. Reflective analysis, that is, discussion of the problems of the nature and meaning of the concepts of physics, indeed played an important role in the elaboration of the principles of this theory. But this analysis is intimately linked with comparing the old and new paradigms—classical mechanics and Einstein’s conception. Besides, theoretical reflexion was not and could not be the only tool of substantiating the new theory.

p Reflexion about knowledge thus proves to be closely linked with the development of its content and with going beyond the limits of the existing conceptual system. (That does not mean, however, that the reverse proposition is also true, that is, that any development of the content of knowledge appears as reflexion. For instance, the development of a theoretical system within the given paradigmal premises obviously cannot be taken as an example of reflexion.)

p If that is how things stand, the question arises, does the problem of substantiation of knowledge have any meaning at all? Classical philosophy and science presented the solution to the task of substantiating knowledge as finding a set of assertions which would be absolutely indisputable and unshakeable, assertions to which all other kinds and types of knowledge could be reduced in one way or another. Since such a task cannot be solved (and we have tried to show that that is so), should we not recognise that the problem of substantiating knowledge does not exist at aU? Many Western specialists in the foundations of mathematics, logic, methodology, and philosophy of science, in the theory and history of natural science come to this conclusion.^^9^^ ^^7^^

p One can hardly agree with this view. What is the meaning of the task itself of substantiating knowledge? Apparently, it is the establishment of the sphere of application of the given system of knowledge and separating that which is true knowledge from that which only lays an empty claim to this title. On the general epistemological plane, it is a question of finding general criteria for the solution of this task, which may be applied to different 226 cases, to various concrete systems of knowledge. If we assume that this task has lost all meaning, the conclusion will have to be accepted that there are no criteria in general which allow to draw a boundary line between knowledge and absence of it.

p In reality, the evolution of cognition is a dialectical process of delimitation of knowledge from absence of knowledge and at the same time a process of increasingly more precise demarcation of the objective sphere of application of the existing systems of knowledge. Substantiation of knowledge implies, first of all, correlating it with real objects through practical object-oriented activity. At the same time, not all kinds of knowledge can be directly included in practical activity. Besides, practice itself is always limited by the given concrete historical level of its development. Therefore, even the practical application of the given system of knowledge is not tantamount to full substantiation of the latter. Practice assumes the development of the systems of knowledge themselves. It is in the course of this joint development of mutually connected practical activity involving objects and cognitive activity that knowledge is substantiated. Substantiation must not thus be understood as an ensemble of procedures enabling one to provide an unshakeable basis for knowledge once and for all but rather as historical development of cognition, as emergence of new theoretical systems, discarding some old conceptions, establishment of new links between theories, revision of old theories, etc. Substantiating a given theoretical system means going beyond its framework, including it in a deeper context, and considering it against a broader background.

Thus, those procedures which were considered in the history of philosophy and science as methods for resolving the problem of substantiation are indeed relevant to the solution of this problem but in a sense different from the previously assumed. These procedures do not at all provide "absolute" substantiation, being merely elements in the historical process of substantiation coinciding with the development of knowledge itself. Substantiation as it actually takes place therefore includes elements of scientific research which classical pre-Marxian and nonMarxist philosophical and methodological literature did not consider in the context of the given problem (e.g., the origin of new theories). If substantiation of knowledge coincides with its development, and theoretical reflexion is only one of the elements of the latter, that means that actual substantiation is not reducible to reflexion, being much broader in scope.