p Let us begin a critical analysis of the conceptions of alternative worlds with Quine’s theory. That is all the more convenient since the latter sets this problem against the broadest philosophical background.
p Let us note from the start that the existence of alternative theories in metamathematics, which was the premise of Quine’s theory, is in itself indubitable and requires serious philosophical study. Quine is also right when he says that the problem of substantiation of scientific theory cannot be solved through ontological and epistemological reduction (thus, it is impossible to reduce the theory of arithmetic to set theory; the problem of substantiation of mathematics cannot be formulated or solved within the framework of reductionism). But we are first of all interested in the analysis of the conception of ontological relativity in its general epistemological significance.
p It is easy to show, however, that its basic propositions can hardly be regarded as acceptable. Indeed, Quine notes that sensory information (that is precisely what his term "stimulus meanings" refers to) does not carry in itself any object dissection of the world. He is, of course, right here. But he errs on another score, namely, in asserting that the existence of objects is not inherent in the world outside man, and that the subject singles out external objects only at the language level, the grouping of "stimulus meanings" in the objects of a definite type being entirely determined by the structure of the given language system, so that it must be fundamentally different in different languages. According to Quine, the initial content information about reality is restricted to an ensemble of "stimulus meanings”, and human behaviour is determined by this information rather than the objective properties of the objects themselves. As a matter of fact, what we have here is a kind of revival of the subjectivist empirical theory of "sense data”, hard as Quine might try to dissociate himself from it. He is also close to the subjectivism of logical positivism (although he believes to have overcome the latter) at another point of his conception—in rejecting the existence in objective reality of objects that are referents of the given language expressions. For Quine, just as for Carnap, the subject 194 cannot in a certain sense break away beyond the confines of language to the objective world itself: the meanings through the relations of which the world is given to the subject are, in their view, only a system of intralinguistic relations. Therefore the world itself as an ensemble of objects is only the product of language.
p This line of reasoning, consistently and logically implemented, inevitably leads to conclusions which are in themselves enough to make one doubt such conceptions. (For instance, it is exactly that interpretation which Quine gives of subjective "stimulus meanings”, taken as the starting point of cognition, that determines the emergence in this conception, in a new variant, of the difficulty which subjective empiricism would never solve: the imaginary impossibility to cognize the state of consciousness of another person.) The main point, however, is that Quine’s conception cannot provide an adequate explanation of a whole series of important facts which the modern sciences of cognition cannot ignore.
p The basic fact is that the subject is capable of singling out the objects of the external world before he masters the language, though Quine asserts the opposite. The modes of singling out objects are directly correlated with definite forms of object>related practical activity worked out by society; assimilating these forms, man assimilates the specifically human cognitive relation to reality, that is, that relation which assumes the givenness of the world of objects to consciousness and differentiation of the latter from the inner world of the subject, his consciousness. The assimilation of language itself implies that the subject has mastered definite "reference mechanisms”, that is, the modes of referring knowledge to reality: these mechanisms are included in the basic perceptive structures. The main types of referents and systems of meaning are not constructed by the language system but are its premises.
p That is why expressions of different languages may have common referents and, moreover, common meanings in a narrow sense of the term, i.e., they may be synonymous. Quine’s conception compels one to reject the possibility of synonymy, which contradicts elementary language intuition. These fundamental facts are recognised in all more or less serious theories inquiring into cognitive activity in general and perception in particular.
p Inasmuch as perception structures are linked with definite forms of object-related practical activity, one may make judgments, from the knowledge of the given subject’s mastery of these forms, about the character of these structures and the degree in which they are formed, i.e., 195 judgements about the form in which the object dissection of the world appears to the subject. That is the method of inquiry used by Piaget, among others. (We ignore here the fact that Piaget deals with spontaneous development of structures and actions in the child and not with mastering socially evolved forms of object-related activity.) Piaget draws conclusions about the forms in which external objects appear to the subject from the child’s behaviour, e.g., from its searching for an object that passed beyond its field of vision. It is through external object-related activity that knowledge which the subject has is actually combined with the real objects. The forms of this activity are of course determined not only by the objects but also by the historical traits of social practice: those aspects are singled out in the object which became particularly significant for the given types of object-related activity. At the same time it is essential that, first, we are dealing here with the real aspects of the objective things themselves, and second, that the essential structure of human practical activity remains invariable, however diverse and historically changeable types of practice may be. Therefore, the principal types of objects with which humans deal in ordinary life are the same regardless of the languages they speak and the stage of cultural-historical development. (Only we must not confuse the types of objects given in knowledge with what the subject knows about these objects.)
p If we break the real ties between systems of knowledge and forms of practical object-oriented activity, it is, of course, impossible to assess the extent to which the object dissection of the world given in knowledge corresponds to what exists independently of cognition and consciousness. But that is exactly what Quine does, insisting that judgments cannot be made from behaviour about the system of referents to which the given language expressions relate, and that different systems of object dissection of the world implemented in different languages can be associated with one and the same type of behaviour, the latter being, in Quine’s view, oriented at an ensemble of "stimulus meanings" and not at a system of objects.
p It is obvious, however, that the most meaningful and essential connections of reality are implemented precisely in the object picture of the world. The knowledge of an object assumes cognitive mastery of a whole system of substantive connections in their complex mutual dependences, the connections being not only actual but also potential. If man’s activity were directed at an ensemble of "stimulus meanings" rather than at objects and their 196 mutual relations, his behaviour would hardly differ much from the behaviour of animals. It should be assumed that if a rabbit were perceived as a set of actually given aspects or as a phenomenon of “rabbitness” (to use Quine’s example), this would affect in one way or another the behaviour of the carriers of this kind of perception. True, Quine himself underlines that the anthropologist trying to translate a text from one language into another must deal with relatively short stretches of "stimulus meanings" and correspondingly with relatively small fragments of natives’ behaviour. But this restriction is hardly justified. It may be assumed that it is this restriction that makes it hard to define the object referents of some language expressions and that, consequently, its elimination does away with a whole series of imaginary difficulties to which Quine refers in substantiating the thesis that radical translation is indeterminate.
p Further, Quine’s conception of language itself also gives cause to certain objections. In Quine’s view, language is a definite ensemble of purely conventional (associative, in his terminology) links between separate sound complexes, some of which stand in conventional (associative) relations with "stimulus meanings”. Language is definite more or less stable connections of verbal behaviour, or predisposition to verbal behaviour of a definite kind. Alterations in verbal behaviour change language too, in Quine’s opinion. He believes that the relations of different sound complexes to one another and to external stimuli in the framework of the given language system are not determined by the object dissection o.f the world. Meanings are only mutual relations of language expressions. For Quine, it is therefore quite natural to infer that the existence of a common object environment, of a common neurophysiological apparatus in carriers of different languages, and even the substantive community of different kinds of practice associated with different languages are no indications at all of the existence of essentially common structures in different languages.
p However, the conception of language which becomes more and more firmly rooted in modern linguistics is based on fundamentally different notions. The point is that a language system is not determined simply by a set of sentences similar in certain respects and produced within a given time interval, that is, by verbal behaviour. This set will always be finite. Yet any language system contains the possibility of generation of an infinite set of acts of verbal behaviour, including those that do not reproduce any of the past acts. Generation of new 197 sentences does not .necessarily entail changes in the essential characteristics of the given language system (whereas, if language is a set of conventional connections between sentences the emergence of new sentences must change the language itself). An essential feature of language is the phenomenon of synonymy—that very phenomenon which Quine treats as a pseudoproblem. Modern linguistics works out theoretical models of generative grammar proceeding from the fundamental, premise that language serves to express a system of definite meanings and that this system is basically common to all the different languages (having an extralinguistic nature). Apparently, such grammatical categories are universal as names of objects (nouns and nominal phrases), names of situations-ysentences, and the so-called transformers, that is, linguistic objects changing linguistic objects of one class into other linguistic objects of the same or different class.^^74^^ And that means, for instance, that singling out objects in reality and distinguishing them from processes and actions, that is, from situations, is common to all languages, being determined by extralinguistic circumstances which inevitably affect all languages.^^75^^ The existence of invariant grammatical structures in all languages is also an argument against the universalist claims of the Sapir-Whorf hypothesis. The grammatical structure of a given language assumes a definite object dissection of the world specifying a definite very general sense of each sentence generated by the given system of grammatical rules. Therefore, if lexical semantic connections in a sentence are disrupted but the sentence itself is constructed grammatically, it is meaningful in a wider sense, though properly speaking nonsensical; at any rate, it is understandable. Noam Chomsky cites in this connection the sentence "Colorless green ideas sleep furiously".^^76^^
p But even if we deal with a completely foreign language, the grammar of which is not known to us and may prove to be quite different from the grammar of our native tongue, the existence of language universals does not permit us to interpret that language as a simple set of sounds conventionally connected with stimuli from the external environment and allowing an almost unlimited spectrum of incompatible interpretations. Quine makes a universal and an absolute out of a definite procedure that is justified in metamathematical studies. In the latter, the need indeed arises to view a given theoretical system as a purely formal structure (a set of symbols on paper) functioning as an object language meaningfully discussed in metalanguage. Under ordinary circumstances, however, the 198 attitude to natural language and even those languages which express the content of theories in the factual sciences, i.e., those that differ from mathematics, is different. If I come across a language spoken by beings physically similar to me and interacting with the external world in a basically similar way (and that is expressed in their behaviour), I must assume at once that the semantic fields of our different languages have essential features in common.
p According to Quine’s conception, only those acts of verbal behaviour are ultimately given to me on the semantic plane whose agent I myself am. A dissected picture of the object world is given to me by the language which characterises my predisposition to definite verbal behaviour. As for the language of any other subjects, including those that speak apparently the same language as myself, this language functions, according to Quine, merely as an object language, i.e., a set of sounds allowing different interpretations. However, an individual that merely receives from the external world a set of impacts in the form of "stimulus meanings”, holding other individuals and their actions also to be mere external stimuli, cannot be the starting point of epistemological or even psychological and psycholinguistic study. Such a study must have for a starting point the subject included from the outset into real connections of communication with other subjects representing society and the accumulated social-historical experiences. From the very first days of its life, the child is involved in a meaningful interaction with an adult. At first, the ties of semantic communication are established directly through practical activity with external real objects (and not at all with a set of "stimulus meanings"! ) and later through the use of language, too. The latter thus actually expresses not only a definite relation of sound complexes to external objects but also the relation of the subjects using it to one another. The types of objects to which the expressions of the given language and the principal types of relations between these objects, i.e., the principal systems of meanings, relate are therefore common to all the carriers of the given language. The question, discussed by Quine, of what the ontology of the language is which my compatriot speaks cannot for this reason arise in reality and is a typical pseudoproblem.
p The difference between the object dissection of the world given to me through the language I speak and the ontology of this language does not have the character ascribed to it by Quine. Of course, this differentiation has some sense in formal analysis, when we have to discuss 199 an object language in terms of a metalanguage, but it can hardly be so essential in all the other cases. As far as natural languages and the languages of the theories of factual sciences are concerned, their ontology is substantively determined by the object dissection of the world given in them and cannot be viewed as something determined only by the relation of the given language or theory to another arbitrarily chosen language or theory. Therefore in non-formal contexts the ontology of a theoretical system may be regarded as inbuilt.
p That does not, of course, mean that all semantic shadings included in the perception of the given object by myself and another subject are absolutely identical. Indeed, if two subjects look at the same object, the latter will present to them its different aspects if only because their angles of vision may not coincide, as they occupy different positions in space. That means that the backgrounds against which the given object will be perceived will be different, to say nothing of the differences in perception determined by the specific traits of the personality and the life story of each of the subjects. The situation is essentially the same with their utterances. One and the same word necessarily calls forth different associations in them conditioned by their unique life experiences. For this reason, if we accept that understanding another person assumes complete and absolute comprehension of the entire system of the subjective M-mantle shadings of behaviour and speech essential for that person, we may conclude that such an understanding is in general impossible and that, consequently, semantic interaction or meaningful dialogue between two subjects are in general nonexistent. That is the conclusion to which Quine comes. But, as we have tried to show, the difference in semantic shadings characterising the experiences of different subjects is itself possible only against the background of essentially common semantic structures underlying real practical, objectdirected and language communication. Real understanding and dialogue do not at all rule out certain semantic differences in the details and shadings that are inessential for the needs of communication. Moreover, communication presupposes these differences, for no subject can cease being himself or transplant himself into the body and consciousness of another.
p Therefore, if we interpret understanding another person as complete coincidence of semantic fields, as absolute merging with the subjective states of another through a kind of direct empathy, we may deduce that any understanding is in actual fact non-understanding. It is another 200 Emacs-File-stamp: "/home/ysverdlov/leninist.biz/en/1984/SOC279/20080104/279.tx" matter that this interpretation of understanding is untenable. To distinguish between the general structure of semantic fields essential for communication and the subjective system of semantic connections in which this structure is implemented in the consciousness of each individual subject, A. N. Leontyev draws a distinction between the categories of "meaning" and "personal sense".^^77^^ It does not follow from the above that the access to another subject’s system of personal senses is absolutely closed to me, and that I cannot to some «xtent assimilate his characteristic mode of interpreting the system of general meanings. Every real dialogue performs this task, enabling one to see the world through the eyes of another person and to allow the possibility of a standpoint different from mine. As for the perception of the object from the same angle of vision as it presents itself to another, it is enough to take up the same spatial position which was earlier occupied by another subject. It is a different matter that absolute merging of the systems of personal senses of two subjects is impossible, since they remain different.
p The situation is much more complicated, of course, in trying to understand the behaviour and speech of the carriers of a foreign language absolutely unknown to us. We have touched on the existence of grammatical universals common to different languages. But a great deal in the grammatical structures of different languages is indeed different, and this fact, though raised to an absolute by Sapir, Whorf, and Quine, cannot be ignored. For instance, there are three cases in Arabic, fifteen in Estonian, and no declension in some languages at all. Systems of lexical meanings vary particularly strongly from language to language. If we recognise the community of different types of objects assumed by different languages, we shall have to consider the fact that the ways of grouping these objects in classificatory systems (and it is the latter that are expressed in the lexical units of the given language) distinguish one language system from another and are determined by the specific traits of that kind of practice which is characteristic of the carriers of the given language. "As studies in most diverse languages have shown, the visible spectrum is ’distributed’ in different ways by different languages. Let us consider the designation of colours: ’green’, ’dark-blue’, ’light-blue’, ’grey’, ’brown’. In Welsh, three words correspond to this part of the colour spectrum: qwyrdd, glas, and elwyd. The last word denotes that part of the spectrum which is termed in English ’brown’ and ’grey’ or, to be more precise, ’dark grey’. The word glas covers the part named in English ’light-grey’, ’blue’, 201 and ’green’. The word qwyrdd also refers to that part of the spectrum which we call ’green’... And in the language of one of the Negro peoples living in Liberia, all colours of the rainbow are designated by two words only: one refers to the colours which painters call ’warm’ (red, orange, yellow, etc.), and the other, ’cold’ (blue, violet, etc.)."^^78^^ Inasmuch as language is directly connected with thought, not only serving as a means of expressing cognitive structures evolved in object-related practical activity before language but also creating for the first time the possibility of the emergence of new cognitive structures, it may be assumed that the difference between the grammatical and lexical means indicates a difference in definite cognitive schemes (although these differences do not involve the principal schemes of reasoning expressed in language universals). One may even go still further, assuming that the difference in language structures determines to some extent the difference in perception. It is, for instance, possible that the Negro tribe living in Liberia referred to above perceives colours in a different manner from the carriers of modern European languages.
p To evaluate this line of reasoning, let us consider some facts. A difference in the grammatical structure of two given languages does not by itself predetermine the possibilities of rendering certain senses with their aid. It is well known, for instance, that the category of determinationindetermination is very essential for Romance and Germanic languages and is expressed in them grammatically, through definite and indefinite articles. There is no system of articles in Russian, but the category of determinationindetermination can be expressed in this language too—by lexical rather than grammatical means (through the pronouns etot, tot in one case and kakoy-to, nekotoriy in the other). Generally speaking, identical senses are expressed in different ways in different languages: in some the sense is expressed grammatically, in others, lexically (the phenomenon of lexico-grammatical synonymy). The vocabulary, being continually renovated, is the most flexible and dynamic part of the language system, a kind of complement to grammar. Inasmuch as there are no essential obstacles to the development of the vocabulary, one may keep introducing new senses and types of senses while remaining within the framework of the given grammatical structure, which is the most conservative part of the language system. It would therefore be rash, to say the least, to infer the characteristic world picture of a language and the cognitive schemes specific for its carriers directly from the grammatical structure, neglecting the study of lexical 202 systems. The study of actually existing natural languages rather provides arguments in favour of recognising the community of their semantic fields in their basic essential features.
p As for the undoubtedly existing differences in vocabulary, and consequently in certain classificatory and semantic systems, their influence on the perception of the world needs careful investigation. In any case one must bear in mind that the main perceptive structures take shape already before language is mastered. Therefore the absence in the given language of words for certain objects and their aspects does not necessarily mean that the latter are not perceived at all. Indubitably, language also affects the semantic characteristics of perception structures, although the nature and extent of this influence have been quite insufficiently studied.^^7^^ ^^9^^
p Let us note, finally, that the social-historical changes and the actually growing affinity between the cultures of different regions necessarily lead to the addition of words to the vocabulary of the given language which allow the expression of new systems of meanings, which results in the affinity of the semantic fields of different languages thus growing.
p However, as long as differences between cultures and the underlying types of practical activity continue to exist, certain differences between the semantic fields of language systems continue to exist, too. All of this creates actual difficulties both for translation and for understanding. True, these difficulties are not at all insurmountable, being eliminated in the course of social progress and cultural interaction. At certain stages in history, however, they still exist and have to be taken into account.
p One-to-one translation from one language into another is in general impossible. Separate elementary meanings of one language often have no equivalents in another. But combinations or systems of meanings of different languages may on the whole correspond to each other. If the languages are very dissimilar (owing to the difference in cultures), translation of some meanings at a given stage of social, cultural and language development is sometimes simply impossible, but that does not mean that this possibility cannot arise in the future. Thus the actual difficulties of translation are quite different from those outlined by Quine in his theory of the impossibility of radical translation and, which is the main thing, they do not warrant Quine’s philosophical conclusions.
p Let us now imagine that we have to deal with a reasonable being whose physical make-up, the modes of 203 obtaining and processing information from the surrounding environment, and the type of interaction with the world are essentially different from the human (extraterrestrials are favourite characters in science fiction, as we know). Assumedly, it will be extremely difficult to understand the language of this being. It is this case, rather than what we usually observe in ethnolinguistic studies, that is closest to Quine’s view of the situation of an anthropologist studying the language of an unknown tribe. Yet even this case does not fully answer Quine’s interpretation. Assume that the extraterrestrial’s system of perception of the world was formed under conditions essentially different from terrestrial ones, that his environment did not include solid bodies, that is to say, it was something like liquid or gas. (Of course, this assumption is highly hypothetical if not improbable. We temporarily accept it entirely as a kind of "mental experiment".) In this case, the extraterrestrial will either have no means of perceiving the world of objects with which we deal or will perceive these objects in a specific manner different from ours. If we observe, however, that our guest out of space fairly successfully orients himself in our terrestrial world, we must conclude that he perceives, in one way or another, our system of objects. And if we consider as well that object dissection of the world characterises definite systems of dependences of reality itself, far from merely expressing the properties of our language (and we tried to show the necessity of exactly this interpretation of the facts), we inevitably come to the conclusion that a reasonable being different from ourselves perceives, under terrestrial conditions, essentially the same types of objects as we do. This conclusion may serve as the basis for the search for the modes of understanding the language of extraterrestrials. It also allows the assumption that we shall be able to translate a certain part of this language, though this apparently does not obtain with reference to the extraterrestrials’ language as a whole, for the modes of existence of the Earth’s inhabitants and of the guests from space differ too greatly. Success is more likely if we deal with messages containing scientific information: it is through science that we acquire knowledge about real objects and their dependences regardless of their being included in some form or other in direct practical activity at the given historical stage. It is not accidental that it is hoped to establish communication with extraterrestrial civilisations (if they exist! ) through transmitting scientific texts.
p We recall, however, that Thomas Kuhn believes that in science itself the assertions about laws assume essentially 204 different senses in different paradigms, so that adherents of such paradigms in science do not understand one another. So, if the inhabitants of the Earth engaged in one and the same undertaking—scientific study of the given phenomenal domain, accepting a whole series of assertions as true, and using the same apparatus, still do not fully understand one another, according to Kuhn, how is science to ensure understanding between ourselves and the hypothetical reasonable inhabitants of remote worlds differing from ours?
p Can we accept Kuhn’s thesis about the incommensurability of different paradigms?
p To answer this question, we shall have to do a certain amount of analytical work.
p Let us start by stating that the existence, at different stages of the development of science, of various ways of semantic organisation of systems of knowledge implemented in different paradigms appears quite likely. The irreducibility of one paradigm to another is expressed not only in that identical formulas are given different meanings in them, as emphasised by Kuhn himself. Even if we ignore the problem of semantic interpretation of assertions expressed in symbolic form and forming scientific theories, it is apparently impossible to perform the operation of formal deduction of all propositions of one theory from the propositions of that theory which came to replace it and which, it would appear, must fully supplant the previous one. (We have in mind here sufficiently global theories, that is, close to what is termed paradigms by Kuhn.) Reduction of one theory to another mostly proves impossible not only on the content plane but also on the formal one.
p Mario Bunge shows that even thermodynamics is not fully reducible to classical mechanics, although the relation between these two theories is often cited in philosophical literature as a striking example of reduction. "In fact, no rigorous derivation of the second principle of thermodynamics is known: only the thermodynamics of the ideal gas—a very special case—has so far been reduced to molecular dynamics. As to rigid bodies, particle mechanics cannot account for their existence, since the ’particles’ concerned are quantum-mechanical systems and they are glued by fields, which are extraneous to particle mechanics. Nor does quantum mechanics yield classical mechanics in some limit: it retrieves only some formulas of particle mechanics, none of continuum mechanics, which is the bulk of classical mechanics. Finally, some relativistic theories have no nonrelativistic limits while others have more than one."^^80^^
205p Does that mean that communities of scientists adhering to different paradigms live in different worlds and cannot adequately communicate?
p The fact itself of the existence of paradigms hardly proves that the mode of vision of the world is entirely restructured in their successive replacement. Of course, the framework of what is observed in scientific experiment is determined by the content of the theory adopted. But the principal structures of perception, just as interpretation of the world in terms of natural everyday language, take shape at the pre-scientific level and hardly change to any essential degree throughout successive scientific theories. One may rather assume that many semantic systems characteristic of pre-scientific knowledge constitute, in a transformed shape, part of science essentially determining its content aspects. The replacement of fundamental scientific theories or paradigms thus takes place against the background of definite constant strata of knowledge implemented, at any rate, in the structures of perception and in the propositions of the so-called common sense expressed in ordinary language.
p Let us note further that in the actual practice of scientific research theory is not, as a rule, applied directly to experience but through the mediation of another (" interpretant") theory, as has been indicated above. The replacement of one substantive theory by another does not, as a rule, coincide with the replacement of interpretant theories. Besides, as we have just noted, new theories never fully oust out old ones. The actual multilevel structure of scientific knowledge, the existence in it of a number of systems (not a single one!) at each given stage, changing in different ways and at different rates, and finally the "immersion" of scientific theories in everyday pre- scientific knowledge, allow actual comparison and assessment of different paradigms in terms of external criteria, so that the assertion of their incommensurability has no basis at all.
p The existence of a common background for different paradigms makes it possible to apply common measuring rods or standards to them. That does not mean that they are mutually fully translatable, since that would imply the existence of common referent systems and common meanings. But paradigms are characterised precisely by different contents, by giving different interpretations to identical formulas and sometimes even by different referents. Even if we assume that there is no complete semantic break between paradigms but merely a certain difference (we shall touch on this point somewhat later), complete 206 mutual translatability of different paradigms is impossible. Under criticism, Kuhn gave a less rigid formulation of the thesis about that draws a parallel between paradigms and "alternative worlds", asserting in the "Postscript-1969" that, although different paradigms are mutually translatable, they are still incommensurable.^^81^^ In actual fact, the reverse is true, as we have tried to show: paradigms are commensurable but not mutually translatable.
p Recently, specialists in Scientology have been greatly interested in the so-called thematic analysis of scientific theories, that is, the study of those content components of theoretical constructions which are passed on from one stage in the history of scientific thought to another, linking up different paradigms and ensuring continuity of development of scientific cognition. For example, the concept of force has certain characteristics invariant both with regard to the Aristotelian and Newtonian paradigms. The theme of conservation (of matter, motion, electricity, etc.) is passed on from one paradigm to the next. Some themes, accompanying scientific thought from its inception, are grouped in relations of antithetical couples: atomism vs continualism, holism vs reductionism, etc.^^82^^ The existence of such common themes would be clearly impossible if different paradigms indeed implemented " alternative worlds".
p The emergence of a new paradigm certainly changes the semantic interpretation of a number of scientific concepts. However, this change should hardly be understood as complete replacement of the old meaning. If we recognise the existence of common themes in the history of cognition, this kind of replacement is apparently impossible. Besides, the changes obviously do not involve all concepts. In general, it is not any appearance of a given concept in a new context that entails the replacement of one meaning by another or others, otherwise we would be unable to communicate and to understand one another, since language involves, among other things, generation of utterances which cannot have been made previously. In the theory of relativity, the interpretation of mass differs in several important points from that of classical mechanics. It does not follow, however, that two paradigms using one and the same word operate with different concepts, as Kuhn asserts. The systems of objects to which these paradigms refer are sometimes common for the two.
p Finally, we must not forget that a new paradigm may only be adopted if, apart from everything else, it explains why the paradigm that is replaced could function successfully, until a certain moment, in a domain 207 that is common to both.
p This explanation is only possible if there exists a meaningful interpretation of the old paradigm, which is ensured by the fact that some sense units and separate senses of the old paradigm are immersed or form part of the new content structure expressing the new paradigm. Kuhn’s error stems from his failure to distinguish between paradigm as an integral structure and the separate semantic systems that form part of it. In his view, destruction of a paradigm is tantamount to completely discarding all systems of old meanings. In reality, it is the comprehensive incorporation of the semantic systems of one paradigm in the integral structure constituted by the new paradigm that makes mutual understanding and real communication between their representatives possible at an inter- paradigmal level. Importantly, not all the systems of meanings which are ascribed to identical terms and formulas coincide: that is excluded since different paradigms cannot be fully translatable into each other’s languages. It is sufficient for inter-paradigmal understanding and communication that meanings forming part of different paradigms should coincide in certain essential components. The existence of a common constant background of knowledge allows the comparison of different paradigms and a choice between them.
p Therefore a scientist studying the history of physics can understand not only the Newtonian but also Aristotelian paradigm. To do that, it is not at all required to forget the theory of relativity and quantum mechanics and, through a kind of mystic empathy, to grasp precisely the same meaning of all the concepts in these paradigms as was attributed to them in times long gone. On the contrary, it is in the light of modern scientific theories that the historian can see that content in the old paradigms of which their carriers themselves were not aware (e.g., to establish the fact that Newton did not distinguish between inertial and gravitational masses). The psychologist studying the stages of the formation of perceptive structures undoubtedly cannot see the world in the way a small child sees it. The researcher does not only describe the child’s external behaviour but also surmises how the world looks to the child. The psychologist has the right to formulate these surmises (and they are of considerable significance to him) because the child’s perceptual structures, different as they are from the corresponding structures of an adult, are not discarded in psychical development but, after restructuring, become components of mature perception structures. The psychiatrist’ in an interview with 208 a mental patient attempts to reconstruct his subjective world from the latter’s behaviour and speech. The success of these attempts does not at all mean that the doctor must in some way assume a condition similar to that of the patient. That is impossible as long as the doctor remains a sane person, and as soon as he ceases to be one, he can no longer be a doctor. The point is that the difference between a sane person and one with abnormal mentality does not exclude the existence of common psychical structures and functions in the two. Here the doctor apparently understands the patient better than the patient the doctor, and then the doctor understands himself better than the patient understands himself.
p Thus, the problem of continuity of cognitive experiences proved to be more complicated than Kant once believed. In the course of the development of cognition conceptual structures emerge which cannot be reduced to each other, and that means that that process really includes semantic gaps.
p Even ordinary natural languages express systems of meaning somehow differing from one another. Therefore, there is no unambiguously determined translatability in this case either. On the whole,, however, translation from one natural language into another is fully realisable, and it is all the easier the closer the cultures of the carriers of these languages. The explanation lies in the basic community of the conditions of life and practical activity of communities employing ordinary national languages. As for the relations of different theories emerging in the development of scientific knowledge, the situation here is quite different. A new scientific theory, and, still more so, a fundamentally novel paradigm emerge precisely because they carry substantively different content inexpressible in terms of old conceptual instruments. Naturally there can be no complete translatability in this case. As we have seen, it is even absent where special attempts are made to express one theory in the language of another to attain greater precision in expressing the content of the former (e.g., in putting the theory of arithmetic in the language of set theory).
p At the same time there are relations of continuity and cohesion of definite meanings between different theories and paradigms, a general background of knowledge, so that paradigms cannot be equated with absolutely different "alternative worlds". Being mutually intranslatable, different paradigms are nevertheless commensurable.
p The problem of continuity and alterations in the meanings of concepts in the course of the development of 209 science has so far been but little studied. However, the understanding of the meaningful side of scientific theoretical knowledge largely depends precisely on the solution of this problem. It should be recognised that, although there are certain gaps in the development of the conceptual systems of science, they can hardly exist in perception structures—at least in adults. True, if we should believe, as Sartre does, that a cognitive orientation does not underlie experience, and that the latter is not marked by the division into subject and object, we shall have to admit that there are complete gaps not only between the experiences of any two subjects but also in the experience of each of them. The subject continually manifests himself in unique situations, Sartre believes, and separate situations have nothing in common. The subject, too, is each time unlike himself. This interpretation can only be accepted if one agrees with Sartre’s general philosophical and epistemological conception. We showed its untenability in the second chapter of Part One of the present work. Abstractly speaking, certain perception structures (though by far not all!) could, of course, change if the type of man’s object environment should radically change, or the type of his practice, or the set of sensory modalities inherent in his perceptual system. In this case gaps could appear in the structures of direct experience. We may recall that Gregory admits the theoretical possibility of man creating a supercomplex artificial environment that will demand the formation of new ways of perceiving objects. If man does not solve this task, he will be unable to see. And if he solves it successfully, gaps will emerge between old and new experience.
p As for the results of changing the set of modalities, they can be assessed by the results of actual successful operations of removing cataracts owing to which men begin to see. The perception of the world in such individuals is originally formed on the basis of the tactile sense. At the same time, an amodal objective scheme of the world was built in the system of perception of the formerly blind person. The blind man is confidently and correctly oriented in the system of objects, but the appearance of a new (visual) modality disrupts the well-formed system of orientation: the formerly blind person cannot at first correlate the visual information with the tactile perceptive images he has, and the earlier developed modes of tactile orientation cannot function as successfully as before. Only gradually new perceptive structures are developed which link up visual and tactile information. Apparently there is a gap (though not complete here either) between 210 old and new experience. The amodal objective scheme of the world -remains constant, and the new experience structure is formed on its basis.
The development of cognition is thus characterised by extension and deepening of the content of knowledge, the emergence of new semantic systems and the singling out of new types of objects. In this process, the characteristics of objective reality itself existing independently of cognition and consciousness are reflected and reproduced in more and more precise and differentiated ways; that is to say, objectively true knowledge is produced. As we see, the complex dialectical interrelations between discreteness and continuity in the development of cognition are one of the modes of concrete expression of the dialectics of absolute, relative, and objective truth, a classical philosophical analysis of which was given by Lenin.^^83^^
Notes