Philosophy and Ideology. Z. A. Jordan 1963
According to the materialist theory of truth, a statement is absolutely true if it is a wholly true statement and reflects in some respect fully and faithfully either the totality of reality or any of its fragments. Such truths are ‘immutable’, ‘eternal’, and incorrigible[872]. The relativity of truth is an absurd doctrine for it leads to contradictions; if it were valid, the same statement could be both true and false, and this is absurd[873]. There are singular and general or universal wholly true statements, but the singular are probably more numerous and certainly easier to discover. Singular wholly true statements are often poor in content; for this reason Engels called them Plattheiten. Such statements as ‘Napoleon died on May 5, 1821’ or ‘ Paris was the capital of France in the nineteenth century’ are absolutely or wholly true. They reflect a tiny bit of the objective reality, considered from a limited point of view[874].
Not every true statement is wholly true, there are also partially true statements, and the relation of the latter to the former constitutes the central problem in the understanding of an asymptotic progress towards absolute knowledge. Partially true statements are relative truths in some sense of this expression, for being neither wholly true nor wholly false they are corrigible and time-dependent. In the course of human practice they are constantly replaced by other and better partial truths. Since they are accepted at a given moment in virtue of the fragment of truth which they contain, they never become utter falsehood, though they might become otiose and pass into limbo[875].
It cannot be denied that the term ‘true’ is sometimes used in the manner suggested by Marxist-Leninists, that is, in the sense of ‘being partially true’. This use is familiar from everyday language. Some philosophers also felt that it was not illegitimate to use the term ‘true’ in that sense. They argued that a partially false sentence is not true, but its not being true is not incompatible with its being partially true. Similarly, a partially true statement is not false, but its being partially false is not inconsistent with this fact. In recent times Bradley has been the most eminent protagonist of this view, which he supplemented with the belief that propositions possess different degrees of truth. This idea was, however, closely related in his case with what has been described as a rather extreme form of the coherence theory of truth[876]. On the other hand, partially true statements of the materialist theory are true clearly in the sense of the correspondence theory of truth. It is not easy to see, how a proposition can be true in this sense and yet be partially true or differ in its degree of truth from some other true proposition. For according to the correspondence theory what makes a statement true is the fact which it asserts, and no fact can together make and unmake a statement true or perform this function to a varying degree.
The Marxist-Leninist theory of partial truths and degrees of truth is in one respect superior to that of the British idealist. For while maintaining that a partially true statement is true in virtue of being a partial reflection of objective reality, the materialist theory of truth recognises the existence of wholly true and wholly false statements and thus escapes some obvious objections against the views of Bradley and Joachim. For Bradley maintained that all beliefs without exception are both partially true and also partially false, and no belief is wholly true or wholly false. Joachim included the coherence principle of truth, itself being a human product, among partial truths. G. E. Moore had no difficulty in refuting the view that all statements are partially true and partially false[877].
The materialist theory of truth does not explicitly maintain that a partially true statement is both partially true and partially false, though it is difficult to see how it could deny it. But the psychological reason for ignoring this fact is perhaps understandable. If a statement is not only partially true but also partially false, one cannot escape the impression that it must be wholly false (it will be shown that this impression is right). Its confusing impact would obstruct the road which otherwise the concept of degrees of truth might seem to have thrown open. For if some scale or scales of comparison between various partially true statements could be established, a method of assessing the degree of truth and of getting closer and closer to the Truth would have been found. Something similar was suggested by John of St. Thomas, a medieval scholar, a commentator of Thomas Aquinas, and author of Ars Logica. John of St. Thomas held a doctrine not unlike Schaff’s. He also maintained that there are degrees of truth and tried to devise rules by means of which propositions containing truth and falsehood in different proportions could somehow be compared. Thus, for instance, he suggested that the proposition ‘every European is a Moslem’ is false to a higher degree according to quantity than ‘every African is a Moslem’, and ‘trees are stones’ is false to a lesser degree according to quality than ‘the human spirit is a stone’ [878]. As will be seen later, Schaff did think of some scale which would allow the grading of partial truths according to their approximation to absolute truth as their limit.
The semantical definition of truth makes it clear that the idea of finding a method of assessing the degrees of truth is unsound and that the concept of degrees of truth is incompatible with the correspondence theory of truth. According to the semantical definition of truth, ‘true’ is a one-place predicate, and, therefore, it cannot be said meaningfully that a proposition is true in some respect and untrue in some other. As long as the semantical definition of truth is adhered to, a statement is either true or false. This fact is obscured in the materialist theory of truth by the concept of reflection, for ‘reflection’ is a relational predicate as much as ‘probable’ or ‘confirmed’. It is very likely that the doctrine of degrees of truth owes its existence to the obscurity and ambiguity of this concept.
It is, indeed, not easy to find out what Marxist-Leninists mean when they speak of a statement being partially true. The examples given by Engels, for instance Boyle’s law, in which, as Engels puts it, a grain of truth is combined with a particle of error (for Boyle’s law does not hold good in certain cases), seem to suggest that what Marxist-Leninists have in mind is the concept of the degree of confirmation. But a statement confirmed to such and such a degree cannot be said to be either partially true or partially false. Like any other statement it is either true or false, though we are unable to decide finally and with certainty which is the case.
Engels’ argument was intended to make the position clear that the concept of confirmed statement is legitimate, though different from that of true statement. He emphasised the important place which statements approximately correct or correct within definite limits occupy in science, and exposed what he called ‘empty phrasemongering’ of some philosophers who from the fact that a statement cannot be shown to be true drew the inference ‘therefore it is not a truth at all, therefore it is an error’ [879]. The fallacy involved in this inference is obvious. No new concept is necessary to provide a safeguard against this kind of fallacy.
If ‘Partially true’ is synonymous with ‘confirmed’, we can, in principle, dispense with the former expression and this seems to be advisable in view of its being misleading. The concepts ‘true’ and ‘confirmed’ must be clearly distinguished and the fact that they have no equiform part in common helps to keep this distinction clear. This is not the case with the concepts ‘true’ and ‘partially true’. This increases the risk of confusion instead of eliminating it.
If ‘partially true’ does not mean ‘confirmed’ and is to apply to beliefs and sentences which are true in the sense of the correspondence theory of truth, though they cannot be said to be absolutely true, for like confirmed statements they are corrigible, the question arises whether the concept of partial truth, plausible as long as it remains unanalysed, does not involve contradiction. The fact that there are corrigible and incorrigible statements does not justify the distinction between partially and wholly true statements. For a confirmed statement is corrigible without being wholly or partially true in the sense of the correspondence theory of truth. Schaff makes it quite clear that a partially true statement is corrigible, but being partially true cannot simply consist in being corrigible. If the distinction between corrigibility and incorrigibility only were involved, the distinction between partially and wholly true statements would be redundant. The former must have some other property, apart from being corrigible, if they are to constitute a separate class, distinct from that of absolutely true statements on the one hand, and of confirmed statements on the other.
This property can only consist in the fact that them is a partial agreement with reality which does not belong either to absolutely true or to confirmed statements. But a partially true statement in this sense is at the same time partially false; otherwise it would be simply true. If this is the case, a partially true statement would be wholly false. For the statement “ ‘p’ is partially true and ‘p’ is partially false” is true implies that “’p’ is partially false” is true too. But a sentence is true when there is a corresponding state of affairs, and is false when there is no corresponding state of affairs designated by this sentence. Consequently, since there is no state of affairs corresponding to ‘p’, the conjunction “’p’ is partially true and ‘p’ is partially false” must be false, or wholly false, in the Marxist-Leninist terminology. If the partial truth of a statement means its being in partial agreement with reality, as understood in the correspondence theory of truth, all partial truths are falsehoods[880]. This must have escaped the attention of those who wish to distinguish wholly and partially true statements, unless their understanding of the meaning to be attached to the expression ‘partially true’ differs from that assumed.
The concept of partially true statements which seems to be either redundant or self-contradictory or obscure, is not an isolated fragment of the materialist theory of truth. To use Richard Mises’ expression, it is a connectable concept, whose ramifications reach far and wide to comprise much of what the materialist theory of truth has to say on the subject of truth in empirical science, on perfect knowledge, and on the validity of the whole Marxist-Leninist philosophy. It commands one of the central positions in the system of the Marxist-Leninist epistemology and theory of knowledge.
For instance, Schaff used it to dismiss summarily Twardowski’s view that ‘true’ cannot be predicated of inductive generalisations, empirical laws, hypotheses and theories. According to Twardowski, whose opinions on this matter were shared by many scientists and philosophers, including Engels, and continue to be upheld up to the present day, the terms ‘true’ and ‘false’ do not apply to laws of empirical science. They are more or less probable, they describe in a manner more or less approximate the regularity of phenomena. In the case of universal empirical statements the crucial question is whether they hold or not, and if so, within what limits.
Schaff considered this view as unsatisfactory and, in the final analysis, absurd. For if the universal empirical statements are not true they must be false, and if they are false, practically our whole knowledge consists of falsehoods. But what we should call errors and false opinions, if Twardowski’s view is accepted, has enabled engineers to take advantage of the forces of Nature to construct all kinds of machinery and to change the conditions of life on the Earth. This success of natural and technical science refutes the claim of the fallibility of empirical knowledge and of its being nothing but a collection of falsehoods. Since the consequences are absurd, the premiss, i.e. regarding natural laws as neither true nor false, must be rejected[881].
The fallacy of this reasoning, against which Engels warned, is transparent. If we cannot decide with certainty whether a statement is true and refrain from the assertion that it is true, that does not mean that it is not true and, therefore, that it must be false. The premiss does not imply, either, that if it is not possible to determine with finality the truth-value of empirical propositions, the whole history of science would consist exclusively of falsehood. Popper’s methodological rule that it should be possible for an empirical scientific system to be refuted by experience does not mean that nothing but falsehood can be discovered in science. For to say that a system is confirmed with respect to a definite class of its verifiers is itself a time-independent statement and it does not become false in virtue of a falsifier of this system having been discovered. The pragmatical and semantical concepts of confirmed statements should not be confused. The allegedly absurd consequences of the view that ‘true’ and ‘false’ cannot be predicated of empirical propositions do not follow, if the distinction of these two concepts is made and applied[882].
The standpoint which Schaff himself adopted is that of Lenin. Schaff felt that many difficulties might be cleared up, if the concept of partial truth is accepted, universal empirical statements are recognised to be partially true and approximately faithful reflections of the regularities observed in the external world, gradually approaching objective truth. For truth admits of degrees and the concept of partial truth involves the concept of degrees of truth. Partial truths can be differentiated and ordered in accordance with a uniform scale, one partial truth being more or less true than some other partial truth (this does not imply that the degree of truth is a metrical concept no more so than that of the degree of confirmation, and Schaff does not make this claim). The corrigibility of empirical statements, the elimination and replacement of one by another, marks the progress from less to more true partial truths[883].
It has been already emphasised that if the concept of the degree of truth does not differ from that of the degree of confirmation, the advantage of using it is dubious. Marxist-Leninist thinkers see it otherwise. They say that the concept of the degree of truth serves an important purpose, namely, of bridging the gap between a wholly and a partially, between an absolutely and a relatively true statement (in the Marxist-Leninist sense) and reveals dialectic connections binding the opposites together[884]. If by this is meant that the degrees of truth (of confirmation) can be ordered on a continuous scale extending between the two truth-values of classic logic – truth and falsehood – this is an illusion. For when this suggestion was seriously taken up by Reichenbach and Zawirski with a view toward establishing probability logic as a generalised multi-valued logical system, it became clear that the differences between truth – and probability values did not disappear. An example may make clearer what this means. Let us assume that according to probability logic the degree of confirmation of a hypothesis H with respect to its evidence E can be calculated and is equal to a rational number m/n in the interval 0 and 1. This does not mean that the truth-value of H is equal to m/n. The hypothesis H still remains either true or false. Probability logic does not allow us to dispense with the dichotomy of truth and falsehood, but to measure the degree of confirmation and possibly to substitute definite probability values for its different degrees. Probability logic is not a generalised logic; it operates, as it were, within two-valued logic to provide a technique of dealing with statements which can only, if at all, be either confirmed or falsified and which remain either true or false irrespective of whether this is or is not known by anyone[885].
The function of the concept of partial truth is not only to provide the bridge between error and truth but also between ignorance and perfect knowledge. Marxist-Leninists argue very much like William James, with whom otherwise they radically disagree, that ‘truth is largely made out of previous truths’ [886]. This applies not only to the partial truths of to-day but also to the absolute truth as an ideal goal of the whole historical development of mankind. Perfect knowledge, which no further experience can ever alter or improve, is not something ready-made and waiting to be discovered all together and at once. It is rather like the limit of a converging series to which its partial sums approximate more and more. Perfect knowledge is something in the course of becoming; it is a process in virtue of which perfect knowledge itself is coming into being[887].
The metaphor of an infinite series, applied to the way in which knowledge is gained and extended, does more harm than good when it is taken literally. For we do not know the law in accordance with which the present partial truths are made out of previous partial truths and in the absence of this knowledge no series is determined. Similarly, the concept of limit and convergence has no meaning as applied to statements. The idea that each truth, as well as knowledge as a system of wholly true statements, is produced by the summation of partial truths, transformed from a metaphoric image into a descriptive phrase with factual implications and regarded as a faithful reflection of the actual process of acquiring knowledge, is bound, as will be seen, to make havoc of what reliable knowledge is.