Chapter III.

Epistemology, Ontology, and Logic

9.

Dobrin Spassov.

Refutation of Linguistic Philosophy

 

I cannot see any logical objection to the application of the term “Linguistic Philosophy” to a broad tendency whose cohesion consists in considering language as the only place for finding philosophical problems or the grounds to solve them. Linguistic Philosophy thus conceived includes the following; (a) Logical Atomism which sees in the structure of language the key to the structure of the world; (b) Logical Positivism which reduces philosophy to the “logical syntax of language”; (c) Linguistic Analysis which explains philosophical misunderstandings by the confusion of different “language games”, and which also tries to clarify “the actual use” of language; (d) Linguistic Transformationism which treats their “theory of language” as an epistemology.

The basic common feature of these trends is sociologically and psychologically rooted in the contemporary impossibility of returning directly to classical philosophical subjectivism, and, on the other hand, in the unwillingness of these philosophers to turn to a “metaphysical” (objective) reality. There is no difficulty in discovering in language the desired hybrid of the subjective and the objective, the only possible field of philosophical reasoning and generalization. Whatever its historical and social roots are, linguistic philosophy in all its varieties is a great theoretical mistake, an intolerable dislocation of the fundamental and the derivative.

The philosophy of linguistics can be neither linguistics about linguistics, nor metaphilosophy. It cannot be a third science existing in an imaginary domain between these two sciences. It is a manifestation in the specific sphere of linguistics of general philosophical theory; its raison d’être springs from a principle which is nowhere so broadly applied as in Marxist Philosophy. I am speaking of the unity of the general and the particular, the first existing in the second, as Lenin saw, as its “part or aspect or essence”. That is why philosophical statements and categories can be converted into a philosophy of linguistics.

Here the general philosophical theory of relations is exceptionally important. The point is that almost all the definitions of language contain the words “means”, “vehicle”, “tool”, or “instrument”. True, their correlates, “aim”, “purpose”, etc., are psychological rather than philosophical categories. But from the most general point of view, the determination of something as a “means”, etc., entails understanding them as terms of some relation. This is the basis of the paramount importance of the questions; In what relations does language participate? How should relations and relata be investigated?

I think the right approach here can be found under the slogan; neither opposition nor identification of relations and relata.

It is a fact, for example, that in the prepositional form “a to the east of b” the range of suitable values of the term-variables (a and b) includes only names of geographical places. This peculiarity of the possible relata is undoubtedly determined by the character of the relation. On the other hand, it is not less a fact that, in the propositional form “Bulgaria R Yugoslavia”, words for geographical, historical, political, and cultural relations could be substituted properly for the relational variable R. In such a case the peculiarities of the possible relations obviously depend on the nature of the relata. Therefore, the unity of relations and relata is not an empty phrase but a complete reality.

Logical Atomism has been refuted on grounds similar to that of the refutation of epistemological representationism. The logical atomists have been asked: if you know the structure of facts independently of the structure of sentences, then why do you need the analysis of language as a means of solving classical philosophical problems? And if you hold that the existence of your own assertions alone are certain, then how do you know that these assertions picture the facts? Relational analysis of the speech situation could elaborate this criticism by bringing to light that language itself is impossible without the world, and that, among the relations in which it is interlaced, the human practical attitude occupies first priority. How is it possible to deduce the structure of facts from the structure of sentences, since the cognition of facts designated is a necessary condition for distinguishing between people that speak and people that merely make noise?

The refutation of Logical Positivism presupposes among other things the thesis that there are no purely syntactical words, that, consequently, pure syntactic constructions are not of a linguistic nature. This thesis could be defended as a corollary of the definition of language, according to which without the designatory function, graphic, and phonic materials could still remain, but there would be no linguistic facts. However, more concrete considerations are possible. If, for example, the word-combinations “both we and you”, on the one hand, and “either we or you”, on the other, are semantically different; if this were not due to their descriptive elements (“we” and “you”), which are identical in both cases, then we must assume (in spite of the authority of Russell and Carnap) that logical (syntactic) terms have a semantical content of their own.

What else, indeed, but old nominalistic spectacles or philosophical blindness to the dialectic of relations and relata could prevent the naming of objective coexistence by “both...and...” and also the objective incompatibility by “either...or...”? Of course, the meaning of these conjunctions is not self-dependent. But will a realistic mind look for independence of relations in respect to relata? And does the dependence of relations entail their unreality? This the reason why pure “logical syntax” could be a game with written forma, a calculus, or, suitably interpreted, a model at the service of some inquiry, but “logical syntax” in itself can be neither language nor philosophy.

Since the publication of The Meaning of Meaning (1938) the connection of symbol and referent was recognized to be an “imputed relation” And truly, there is no direct, dyadic relation of designation.  Sign and designatum could for centuries stand face to face, they might or might not resemble one mother; they might be causally connected or separated. No matter what they are, only the interpreter could introduce semantic relations between them as far as his perceiving of a sign constantly gives rise to his conceiving a designatum. All this could well be commonplace. But such a commonplace means that the semantic connection presupposes the more fundamental and more general epistemological relation neglected by linguistic philosophers. Epistemology is inseparable from ontology, since the admission of cognoscibility is impossible without the admission of existence.

On the basis of such an analysis, a criticism of ordinary language philosophy can also commence. Perhaps it is the tacit refusal to bring language and reality into correlation that is responsible for the identification of meaning and use, for “language games” connected only by “family resemblances”, and for setting ordinary speech against the philosophical “confusion of its rules”.

The actual use of language is an historical phenomenon. As Maurice Cornforth remarks, even linguistic philosophers would be convinced of this if they were to become witch doctors in some primitive tribe. The combination of exalting the “actual use” of language and denying what is common in communication is a strange paradox. Without a basic, common character of signs and designata, no language is possible. What is a given word if nuances of its pronunciation and writing do not change it? Do not “individual” meanings slip out of the language-net woven into social life. Therefore, are not Hegel and Lenin right when they assume that language consists only of generalities?

No human treasure are buried in language. It does not live its own life. It is a generalized physical mediator of social beings. Language in itself contains neither thought, feeling nor will. Its charm or ugliness, grandeur or meanness, are nothing but the charm, ugliness, grandeur, or meanness of man.

Perhaps from a linguistic point of view philosophizing transformationalism surpasses logical positivism and Oxford philosophy. I am thinking mainly of the transformationalist attempt to combine abstract formalism and empirical concreteness, language theory and separate language descriptions. But when some transformationalists declare that language theory itself belongs to epistemology, a philosophical protest must not be delayed. Analysis of linguistic communication shows that the unity of language and thought does not justify melting them into a single event. People communicating are nowhere but in physical space; nothing but material facts could fulfill human linguistic connections. Of course, “language” is a specific relational characteristic of sounds or written forms; it is their capacity of directing these same thoughts and turning them to the same objects, otherwise language could not be a tool of mutual understanding, of coordinating actions, a result and condition of human (i.e. social) life. But the necessary connection of language with psychological states and epistemological processes does not justify treating them as linguistic facts, and should not lead us to confuse linguistics with psychology and epistemology.

There are some points in language theory itself, which, critically examined, reveal the actual priority of philosophical considerations. Can the role of the “syntactic component” as an input of semantical and phonological components really not be illusory? With­out taking into account thoughts and their “contents” is it possible to say that syntactical analysis deals with sentences? Without distinguishing characteristics and things characterized is it possible to discern elements of sentences? The confusion of real input and output in transformationalism can be seen in the fact that, contrary to the specific grammar, but in agreement with the epistemological approach to the matter, Katz repeats Chomsky's assertion that in “John is easy to Please” “John” is the object, while in “John is eager to please” “John” is the subject of the sentence. True, sometimes linguistic philosophy realizes the real priority of ontological and epistemological pro­blems, but in such cases it merely ceases to be linguistic philosophy. When, for example, R. M. Hare declares, “One cannot study language in a philosophical way without studying the world that we are talking about”, he actually crosses the borderline dividing linguistic philosophers and simply philosophers.

The refutation and self-refutation of linguistic philosophy will strengthen the view that the ground and scope of philosophical problems and solutions must not be confined within the narrow horizon of a separate special science. Philosophy remains an investigation of the most general features of all scientific fields.

Perhaps the most noticeable concrete achievement of this philosophical verbalism is the widespread linguistic treatment of logic. Correspondingly, one of the principal tasks of non-linguistic philosophy must be the reconquest of the fundamental logical problems.

I should like first of all to draw a distinction between the meanings of the expressions “interpretation of logic” and “interpretation of logic”. The first of them means nothing else but the clarification of the term “logic”. As to the “interpretation of logic”, it is possible only when logic itself is understood as a certain set of signs; such a conception is obviously engendered by the influence of linguistic philosophy. But is it correct?

One cannot merely take a generally accepted understanding of “logic”, because there is no such generally accepted understanding. Nevertheless, the long chain of logical works, from Aristotle’s Prior Analytics to contemporary meta-mathematics, suggests as a common assumption the truism that logic deals with logicality. One could also accept the statement that logicality is the necessary dependence of a given truth or falsehood on others. But what are these “logical values”? How are they connected? If they are properties, what is their substratum? Answers to such questions furnish the dividing line between philosophically different conceptions of logic.

Logicians usually assert that truth and falsehood are properties of propositions. Their incompatibility in one and the same proposition is treated as the most fundamental logical relation, which if neglected leads one to fall into contradiction. Perhaps even logical necessity is, so to speak, a theoretical superstructure of non-contradiction. It is necessary to recognize (under the threat of contradiction) that the truth of a given assertion follows from the truth of others. Of course, non-contradiction, inadmissibility of joining “a” and “non-a”, is not an ultimate datum. But it can hardly be explained if we speak of two possible properties of one and the same proposition. In my opinion, it is far more natural to admit that “a” and “non-a” are different assertions which refer to incompatible objects. Is it strange, indeed, to think that “non-a” means, at first, not that “a” is not true, but that “non-a” is true? Thus we avoid the doubling of “logical values”; we explain non-contradiction not by a vicious circle—the incompatibility of truth and non-truth—but by an objective relation. (We could set forth the foundation of negation and define logic in terms of truth.)

Propositions can be conceived in three different ways, at least. Its name could mean: first, a certain objective situation; second, a thought, a judgment' about this objective situation; and, third, the sentence or the sign that expresses the thought and designates the objective situation.

It is clear that proposition in the first sense of the word must be rejected as a candidate for the substratum of truth. The contents we are aware of could be real or unreal, but they can never be true or untrue. And so we have to decide: where is truth? In the propositional judgment, in the thought, or in the propositional sentence or sign?

Some will say, it is only the thought that can be true. But we attribute this value to the sign on the basis of a certain correspondence between it and the thought for the sake of easier manipulation; otherwise how will the logical calculus be possible?

On their part, others declare: taking into account the inseparable connection of thinking and language, it will not be a great mistake if we say that linguistic phenomena or, in general, signs are the upholders of truth. Does the sign possess its truth? Signs by themselves are always physical facts; sounds or written forms, little knots, bank­notes. But as I have already tried to show, the physical nature of such things is far from being a sufficient reason to call them signs; the sign is their relational property; it presupposes something else.

From Locke’s days up to now attempts have been made to clarify this “something else” by means of understanding designation as a dyadic relation: a given phenomenon is a sign if there is another thing - its designatum. That is all.

Let us consider in abstraction different pairs composed of signs and their designata. Under this condition it is impossible to determine the designation. While signs vary within the limits of the perceptible physical facts, designata are almost beyond every limit: they could be things, properties, relations. There is somewhere a causal connection between sign and designata, somewhere a resemblance catches the eye. According to Peirce’s terminology, it has to do with “indexical signs” in the first case, and with “icons” in the second. But with “symbols” neither causality nor resemblance is necessary. Then what is designation in general?

Designation is found to be indeterminable in abstraction from people or, in general, from some living beings. Signs turn out to be not only designating something but always designating something to somebody. More exactly speaking, the function of the sign consists in the following: while perceiving it, the living being, on the basis of some association, refers to a determinate object. That is why the function of a sign as a substitute is sometimes emphasized. But for our task it is more important to stress that designation is accomplished by means of the pragmatic relation of sign and interpreter and the epistemological relation of interpreter to the designated object. So where is that truth which by itself has an obviously relational nature?

Some “philosophical” syntacticians are convinced that it is possible to define truth in terms of relations of signs, without going beyond the sign-member of designation. But it is well known that the necessary truth, for example, of the assertion, “Jack is a bachelor because he is unmarried”, is reducible to the empirical fact that the words “bachelor” and “unmarried man” designate one and the same thing. Only in this respect (referring to the designatum) could they be substituted each for the other.

As far as a given syntactical link is in fact connected with truth, it is completely reducible to the pragmatic relation of people to different signs as determining the epistemological relation to one and the same object.

Thus we again approach the contention that truth is a semantical property of a sign. And I shall stress once more that there is no semantical relation as a direct link of the two things. The sign designates a determinate object only as far as perceiving the sign calls forth the conceiving of that object. This is an invariable situation. But if so, where is the root of the difference between truth and falsehood? What does it consist in?

It is impossible to answer such a question in abstraction from the epistemological relation to the object. In general we must reject the claims of the sign to be the upholder of truth. Nothing is left to us but to search for the truth in the subject, in the interpreter of designation, to search for it in his cognitive relation towards the external world. One could ask in this connections is it not possible for truth to be, in an indirect way, a property of the sign too, so far as the latter is a sign only in a determinate connection with human thinking? But there is no reason for assuming that the characteristics of a given thing which is necessarily connected with another are characteristics of this second thing too. A linguistic sign, for example, is necessarily connected also with human brains, organs of speech, etc.

The problem of knowledge as a problem of the relationship between the subject and the object is a complex one. We can approach it sociologically by trying to explain the direction of knowledge by means of human activity, which, in turn, is determined by our position in the system of social relations. Further, it is possible to investigate knowledge as a psychological fact, as a function of personality. It is possible to reveal its physiological aspect as well, and sometimes, for instance in the case of the visual perception, it is very useful to study the external, physical conditions of knowledge.

But the problem of truth is always connected with the question of the object known. Is it or is it not a reality independent of the process of knowing it? Acknowledgment of truth is possible only after we can answer this question in the affirmative. Truth is possible only if there is such a “content of our ideas which is independent of man and mankind” (Lenin). This statement is in accordance with the most widespread human conviction that serving truth means revealing real situations, regardless of one’s own or somebody else’s subjective preferences. Thus the problem of truth turns out to be inseparably connected with the classical philosophical questions of the nature of the date of experience, of appearance and reality, of the importance of practice for knowledge, and so on.

Since assuming knowledge means assuming the existence of its object, while assuming objective reality, presupposes acknowledgment of its cognoscibility, the recognition of truth is simultaneously an ontological and an epistemological question. That is why, in Marxist philosophy, ontology and epistemology are treated as two inseparable aspects of fundamental philosophical theory. So, there is a sound way of advancing logic from the bounds of semiotics and linguistics to materialistic philosophy. But how can we build the edifice of this “ontological-epistemological” logic without disregarding the achievement of logicians who have advanced it even without suspecting its existence? I have already emphasized the significance of non-contradiction for the explanation of logicality. Now I can say more firmly that the understanding of the incompatibility of propositions in respect to their truth presupposes, first, accepting that they are thoughts, and second revealing the objective contents that make the given propositions incompatible. The determining point here is the existence, if not of incompatible, then at least of disparate, objects.

Since “objects” constitute the end-point in the explanation of logicality, it becomes possible to place them into the theoretical foundation of logic. Maybe the term “object” is too ambiguous to serve as a solid ground of such an exact science. This term is, infinitely multi-significant; on the basis of its universal meaning, it can be specified as sign, as thought, as a concrete external thing, or as a class of objects : generally speaking, as everything that exists independently of the process of knowing.

Let us designate the object with “a”. In such a case the object incompatible with “a” can be designated with “non-a” (a). The properties of the relation of incompatibility which underlie non-contradiction and other analogous relations could be designated with the well-known formulae:

                                                           a . â, a V â

The first of them may be interpreted equally as incompatibility of truth and falsehood in thought, incompatibility of class and complementary class, etc. All these possibilities are due not to the alleged emptiness of the formulae in question but to the extremely general content designated by means of them.

Philosophical logic will succeed in approaching the problem of conjunction, disjunction, implication, equivalence, and negation, which are the basis for the linguistic conquest of the logical investigations. There is no profound reason to think that the logical connectives are purely verbal facts or facts peculiar only to thinking. Is there not, for instance, an obvious analogy between the disjunctive proposition and the parallel set of relays or the sum of two classes of things? I am convinced that the preference of four out of sixteen mathematically possible combinations of two propositional variables under two values of every variable could be explained only by the “logic of the objects”. The semantics of ordinary language is powerless to solve this problem because there is no complete coincidence of meaning between the signs of conjunction, disjunction, implication, and equivalence, on the one hand, and the corresponding connectives “and”, “or”, “if...then...”, “if and only if...”, on the other. (The conventionalistic solution of this difficulty also does not elucidate the special applicability of these connectives in comparison with the others.)

But it is quite another thing to approach the problem on the ground of the possible combinations of two objects and their negations (by “negation” I understand every member of every incompatibility in respect to the other member of this relation). For example such a scheme

a.b1    a.b  a1.b   a1.b1

makes it possible to explain why it is the conjunction, the disjunction, the implication, and the equivalence that deserve special attention. It is possible to exemplify the objective content of the conjunction with student-sportsmanship which is absent in students not being sportsmen, in sportsmen not being students, and when there are neither students nor sportsmen. Further, it is possible to illustrate the implication with the relations of the type of the tragic marriage of man and mortality, or equivalence with every one, one relation, in short, the materialistic “translation” of the logical values defining the typical molecular propositions explains which of them have specific objective content and, consequently specific logical significance; a.b ; a.bxa.b1xa1.b ; a.bxa1,bxa.b1 ; and a.b x a1.b1 (cf. the corresponding sectors of the geometrical model)

At any rate, it is far from obligatory to think that the problem under consideration is monopolized by the so-called logic of statement-connections. It could be treated as a problem of determinate objective relations which are the common essence of molecular statements and the corresponding points of Class Logic, Technical Logic, and so on.

In an analogous way we could approach the abundance of logical laws cultivated by Symbolic Logic. Since they are true for all the combinations of the values of their elements, they allude to all the combinations of the corresponding objects. We could establish, for example, that the designatum of the analytical formulae containing two variables is the logical sum;

a.b. x a.b1 x a1.b x a1.b1;

under three variables we have;

a.b.c x a.b.c1 x a.b1.c x a1.b.c a.b1.c1 x a1.b1.c a1.b1.c1

 and so on.

It is possible to explain the reducibility of all analytical formulae to the designation of the relation of incompatibility detailed to different degrees. This fact is the foundation of the well-known possibility of reducing all the laws of symbolic logic to the Law of Excluded Middle as the shortest representation of their “normal disjunctive form”. (When admiring the so-called free logical constructions, we must not forget that they differ in elegance, acquire the splendor of truth, and the power of practical application, only as much as they are built on the ground of the simple logic of objects.) Written forms are the most suitable means for constructing models in aiding investigation of abstract, objective relations. But from a general logical point of view, signs are the means and not the objects of investigation.

I tried to demonstrate that linguistic philosophy can neither abolish ontology and epistemology, nor deliver logic from traditional philosophical problems. This failure is deeply rooted in the very nature of the relations between philosophy and the special sciences.

One of the queerest features of contemporary philosophical life is the growing willingness of philosophers and scientists to turn their tables. This tendency by itself is neither a plus nor a minus. Aside from the philosophical snobs’ flirtation with mathematics, cybernetics, and so on, the movement from philosophy to the special sciences is completely justifiable; it is a natural way of accomplishing the cultural function of philosophical knowledge. That is why the many contemporary philosophies of the various sciences must be greeted, provided these philosophies are applications and not dissolutions of the one, unified philosophy. As for the opposite movement, that is somewhat different. Undoubtedly, it can express the legitimate desire of the scientist to clarify the theoretical foundations of the special sciences; it can result as well in the correction of some philosophical generalizations. But when a separate special investigation pretends to be the substitute for, or the point of departure of philosophical knowledge, our protest must not be delayed; such a positivistic pretension neglects the real relation of theoretical problems, and besides it removes practice from its basic position. The ground for the refutation of every kind of positivism can easily be found, for example, in Lenin’s criticism of so-called physical idealism. Lenin emphasized that the transformation of physical concepts need not necessarily result in new fundamental philosophical statements. This is so because philosophy has a wider scope and a more general content than physics (or any other special science). Philosophical generalizations must be founded on all of history and the totality of human knowledge. And as for the basic problem of objective reality or objective truth, seeking solutions in the sciences leads one to fall into an obvious vicious circle. That is why Marxist Philosophy includes practice in epistemology. (Of course, a definite scientific discovery can abolish a given statement of philosophy, but can never replace it by another philosophical generalization.)

In conclusion, I cannot conceive of man-in-general somewhere outside of actual people. But suppose I know only a single human being? Shall I be able to discriminate his human essence from his individual peculiarity? Of course I shan't, because every generalization presupposes the comparison of many particular cases. The universalization, furthermore, always bears a risk. That is why the movement from philosophy to a separate special science is logically sounder than the opposite movement. For example, the features of man in general are my features as well, but my own traits do not necessarily characterize man in general. So linguistic philosophy must not be defended, even as a mere philosophical generalization of linguistic facts. Linguistic philosophy is inadmissible to everybody who understands the real nature of the unity between universal and particular, between philosophy and science.

Linguistic philosophy cannot be accepted by anyone who realizes that language is neither the only nor the fundamental reality, and consequently, linguistics is neither the only nor the fundamental science.

 

SOFIA UNIVERSITY, BULGARIA