Marxism and the Linguistic Philosophy

by
Maurice Cornforth

III
Marxism

CHAPTER TWO

The Laws of Thought

1. FORMAL LOGIC

IN discussing the linguistic philosophers' investigations of the actual use of language I came to the conclusion that what distinguishes language as a means of communication is its expression of propositions. Language has the unique functions of negating, abstracting and generalising. These, by the way, should not be conceived of as separable or co‑exclusive functions—like walking, running and jumping as achieved by moving the legs. For purposes of logical exposition we can, of course, speak about or represent symbolically negation as distinct from generalisation, or generalisation as distinct from abstraction. But all propositions combine negating, abstracting and generalising functions. Every significant sentence negates, abstracts and generalises.

Operations involved in making propositions can be considered purely formally. That is to say, we can investigate the proposition‑making operations performed in sentences in abstraction from their meaning. For a formal investigation is one which abstracts from meaning. For example, consider such sentences in English as "This rose is red", "This pig has wings", "This triangle is equilateral", and equivalent sentences in English expressed in other idioms (such as "The rose I am calling your attention to has the colour denoted by the word 'red’”), and equivalent sentences in other languages, and then abstract the logical operations exemplified by all those sentences by disregarding the differences of both meaning and verbal construction. To do that is to start a formal investigation of logical operations. This is what is done in formal logic, and to do it the specialised symbolic techniques of formal logic have been worked out (starting from the rather crude and inadequate techniques of Aristotle's Analytics, his genius having consisted in inventing a technique but not in perfecting it).

The special business of formal logic is to work out the symbolic techniques of representing logical operations (of key importance was the recent invention of quantification technique); and to construct a logical calculus, with techniques for testing it and demonstrating consistency and completeness. This is a specialised science, allied to mathematics. What it does is to demonstrate how to conduct with consistency logical operations of any degree of complexity.

Those responsible for the great recent advances in formal logic were perhaps motivated mainly by personal interest and curiosity: they wanted to make these demonstrations because they can be made, much as mountaineers wanted to climb Everest "because it is there" (though some, like Russell, were apparently inspired also by a notion that they would achieve a revelation of the ultimate structure of the universe). Certainly, the science that has been achieved is far removed from the paltry practical aim sometimes claimed for formal logic—of being an aid to the art of disputation or catching other people out in logical fallacies. But the fact remains that formal logic, like various branches of mathematics, was not developed until advances of empirical sciences and of technology provided a stimulus and a demand; and though it is not of very much use to human beings as a compendium of instructions on how to think consistently (it is far easier to think consistently than to master such a complicated science as formal logic), it is of the greatest use, and indeed indispensable, for the purpose of programming computers. Machines do not think for themselves, and so a calculus of logical operations has to be available before we can make them think for us.

Formal logic, then, has become a formal science in its own right, with a status like that of mathematics. It is no part of philosophy, but, something philosophy has to take into account. In the past, its development was hampered by being treated as a part of philosophy—as a result of which quite different sorts of questions were continually mixed up together and the quite different procedures required for dealing with them were not observed. While Boole, Frege and others were turning formal logic into a science, a whole crop of rival and incompatible philosophical "logics" emerged, ranging from J. S. Mill's to F. H. Bradley's. These can now be written off. But as with other sciences, including mathematics, questions remain about what formal logic does and how it does it, and about its connections with other sciences; and these can properly be regarded as coming within the sphere of philosophical investigation.

2. CATEGORIES

Professor Ryle distinguished what he called the logical investigation of categories from formal logic. While his account of both is open to criticism (and was duly criticised above), the distinction is of profound and far‑reaching philosophical importance.

To start investigating categories is to become concerned with meaning, and to depart from purely formal inquiry. For this reason there is not and cannot be anything like a calculus of categories, nor are symbolic techniques like those of formal logic and mathematics applicable. At the same time, while not a formal inquiry, an inquiry about categories is not an empirical inquiry either—for its findings are not subject to empirical test. As Ryle has said, a "category proposition" is a generalisation to contradict which leads to absurdity; and so it is certainly not empirically falsifiable, like an empirical generalisation.

Hume proclaimed that whatever was neither formally certified nor empirically verified should be committed to the flames. But the very fact that he said so showed that he could not have meant what he said, for such a principle is self‑destructive. The conclusions he himself arrived at were neither formally certified nor empirically verified, nor could such conclusions be. Hume was, in fact (in a rather unmethodical way, but he deserves perpetual respect as a pioneer), undertaking an investigation of categories—especially in his famous discussions about material objects, minds and causality. There are no good reasons for committing his philosophical writings to the flames, though there are good reasons for concluding that he made mistakes in them.

It is an old prejudice in philosophy that all statements must be either "empirical statements" or else be “known a priori", that is, independent of experience. This dichotomy goes back to Hume, and was further fixed in the minds of philosophers by Kant. But it is totally false and misleading, like so many other hard and fast antitheses. Thus mathematical principles are not established empirically like the generalisations of the empirical sciences, but they are certainly derived from experience and not known independently of experience. And this is true also of the principles concerning "categories" which are of primary interest to philosophy.

The systematic or methodical investigation of categories has to do with the word‑function of abstraction. Formal logic deals with abstraction only in the way of calculating the logical operations involved in abstracting a conclusion from premises. For instance, from information "All A's that are B are C”, plus information that "This A is C”, can be abstracted the conclusion "This A is C”. This is the sort of abstraction a computer can be programmed to perform. And of course, when very complicated information is fed into it, it can abstract conclusions, as required, which human brains would make very heavy work of abstracting.

But there are procedures of abstraction depending not on the logical form but on the meaning. For instance, from information "This rose is red" may be abstracted "This rose is not white"; and that abstraction does not depend on the logical form of propositions but on meaning. Suppose a rose grower had a computer to tell him how many roses of different colours he had sold, so that for each sale it registered the colour of the rose. It would not work unless in programming it he had been guided by the principle: "These are the colours of roses . . . and a rose of one colour is not of any other colour". But that is not a proposition of formal logic, and is not demonstrable or certifiable by procedures of formal logic. Suitably generalised (for it applies not only to roses), it may be called, after Ryle, a "category proposition".

Propositions stating colour may be termed a category of propositions, distinct from, for example, those stating shape or weight. And obviously, there are principles or rules for abstraction ("category propositions") dealing not only with abstracting one proposition of a given category from another of that category (such as deriving "This is not white" from "This is red"), but also for abstracting propositions of other categories (such as deriving "This has a shape" from "This has a colour"). To go against category rules, or "category  propositions", always results in absurdity or non‑informativeness—such as saying that "So and so's red roses are white", or "So and so's roses have no shape or size".

Such examples or the conclusions of an investigation of categories may well be regarded as trivial and obvious, so that such an investigation may be thought trivial and unnecessary. However, two points are already worth noting.

The first is that a computer has to deal with definite categories of information, and has to be instructed and programmed according to the rules for those categories. The people who build and instruct computers must know those rules before they can make machines that obey them. Computers themselves could not conclude to such rules by any process of their own mechanisms.—This is an illustration of the fact that human thought, on the one hand, and what computers do and can conceivably be made to do, on the other, are by no means the same. A machine can quite well be made to observe and record observations and generalise and abstract from what it observes. But just as it does not fall in love with other machines and produce little ones, or co‑operate with other machines in socially producing and distributing their means of sustenance, so it lacks the capacity of developing for itself the categories of thought and the rules for operating with them. It is not made of the right stuff for doing such things. In an age of automation and computation it is important to work out with great care principles concerning categories as well as a calculus of formal logic.

Secondly, even some of the most apparently trivial category rules are not so obvious as they look. Take colour, for example. Information about how things are coloured includes not only such items as "This is red" but also "This looks red". Thus it can be stated: "This is red, but it looks yellow in this light". Again, how tell the colour of, say, the sunset on Mars, when there is no one there looking at it? Indeed, what does anything look like when no one is looking at it? Such considerations show that it would be mistaken to regard information about colour as information about certain simple qualities which either do or do not belong to things irrespective of their relationships with other things. Propositions specifying colour are evidently relational propositions, though this is disguised by the verbal forms in which they are often expressed. Hence there is plenty of room for painstaking investigation of the categories of propositions even in the case of such apparently trivial examples as that of propositions stating colour, and it yields conclusions which are far from obvious at first sight.

This kind of discovery of the not so obvious in the obvious was what originally led to the use of the expression "dialectics". Thus Plato illustrated dialectics by holding up a finger and asking if it was short or long. It was short in relation to longer things, and long in relation to shorter things. From this he reached the important and by no means obvious conclusion (essentially a conclusion about category, though he did not put it in these terms) that propositions specifying length are a category of relational propositions. From this, incidentally, it already becomes evident that there is no sort of incompatibility, as has sometimes been suggested, between dialectics and formal logic. To say, as Plato said, that a finger is short and long—or short and not‑short, and long and not‑long—does not contravene the formal logical principle of non‑contradiction.

Of very great interest are categories of propositions at what may be, termed different levels of abstraction, together with the connections between them. For example, to describe and distinguish things in terms of their sensible qualities (such as colour) is to operate at a certain level of abstraction. But information about what produces these sensible differences (for instance, what makes things look one colour rather than another) belongs to another level of abstraction and is formulated in propositions of different category. And the second category of information does not replace the first, but connects with it. Together these different categories of information constitute information about (in the case of colour) how bodies reflect light and how that affects our senses. This kind of connection of propositions at different levels of abstraction and of different category is of great importance in the management of information. For on it depends the proper assembly of relatively abstract items of information into something more concrete‑filling in and generalising our conceptions of the matters that concern us.

All information is abstract (how could it not be?). And what I have just called the management of information (which is the practical business of thinking, as distinct from what Engels called "idealist fancy") consists of assembling abstractions. Information is not only negative and positive (for example, "This rose is not white" and "This rose is red"), and more or less generalised ("All roses have thorns" and "All red roses have thorns"), and more or less abstract ("This rose is red" and "This rose is red and sweet‑smelling")—all of which are differences of form and are treated in formal logic. It also exemplifies a host of what may be called different modes of abstraction, which are not distinguished in formal logic because they are not formal differences. For example, to talk about colour and to talk about shape, to talk about sensible appearance and to talk about physical‑chemical constitution, to talk about internal structure and to talk about interaction and interrelation with other things, and so on, all exemplify different modes of abstraction.

The sense in which the much‑abused word "category" is used in the context of a logical investigation of categories is the sense in which category of proposition corresponds to mode of abstraction. In general, different categories of proposition result from different modes of abstraction. To take a very simple example. "This is red" and "This became red in the process of being painted" are propositions of different category—the first exemplifying what one might (in semi‑Hegelian language) term a category of "being" or "quality", and the second of "becoming" or "causality". Their connection by way of abstraction is evident, since the first proposition is relatively more abstract than the second, and could be called an abstraction from the second. In both information is assembled, and the assembly of information in the second is such as to render a more concrete account of the same thing of which the first gives a more abstract account.

To investigate differences of category, therefore, is not simply to describe the ways words are permitted to be put together in various languages, as Ryle and other linguistic philosophers have supposed. It is to investigate modes of abstraction and their assembly, which are expressed in the permissibility and impermissibility of combinations of words.

3. DIALECTICS

In his paper on Categories Ryle said that "category propositions" describe and distinguish "the logical type" of different expressions used in statements, and so concern "only collocations of symbols". But he added that this does not equate them with "the propositions of philologists, grammarians or lexicographers . . . Nor does it imply that they can say nothing about the 'nature of things'. If a child's perplexity why the Equator can be crossed but not seen, or why the Cheshire Cat could not leave its grin behind it is perplexity about the 'nature of things', then certain category‑propositions will give the required information about the nature of things. And the same will hold good of less frivolous type‑perplexities."

Because he confined philosophy to describing "collocations of symbols", Ryle may be justly likened to Horatio. At the same time, because he set philosophy the task of formulating category‑propositions, and considered that this (especially in its "less frivolous" applications) would throw some light on the "nature of things", it may be said of him that there is more in his philosophy than he dreamed of. For the yield of this investigation is nothing more nor less than the principles of materialist dialectics.

Dialectics, as an investigation and discipline distinct from formal logic, deals with categories—and so studies and distinguishes modes of abstraction and assembly of abstractions. Evidently, for the purposes of managing information there is required not only the observance of formal consistency, as enjoined by formal logic, but also the observance of the rules of abstraction and assembly of abstractions, as enjoined by dialectics. Dialectics comprises the procedures for arriving at less abstract and more generalised working conceptions of ourselves and our environment by the proper conduct and assembly of the different modes of abstraction.

Thus, for example, it is generally agreed among Marxists that dialectics has to do with understanding things in their inter‑relations and changes, as opposed to the "metaphysical" way of considering things separately, out of relationship and in abstraction from their changes. Evidently, to get a concrete picture of any phenomenon we must assemble the available information in a way that adequately reflects the actual interconnections and motion of things, and thus understand the separate properties of things, and their temporary states, as products of processes of interaction and change. The dialectical approach consists in doing this—its "laws" are the laws for doing it. As Lenin observed, the essence of dialectics is "the concrete analysis of concrete conditions".

This makes the word "dialectics" a many‑purpose one—as it is, indeed, in common use. Thus one may speak of "dialectics" in general, as the universal method of procedure of right thinking, and also of "the dialectics" of any topic one chooses to mention. One can speak of "the dialectics" of the thought‑process in general, as management of information, and also of "the dialectics" inherent in the subject matter of thought. But there emerges as of the very greatest interest and importance the investigation of the most universal modes of abstraction and assembly which are variously exemplified and applied in dealing with any subject matter on which information may be abstracted and assembled.

Hegel has the credit of being the first to open up this department of investigation, in his Science of Logic. True, he obfuscated the whole issue by working with a merely mystical conception of category. But as Marx said of him (in the Afterword to the second German edition of Capital): "The mystification which dialectic suffers in Hegel's hands by no means prevents him from being the first to present its general form of working in a comprehensive and conscious manner. With him it is standing on its head. It must be turned right side up again, if you would discover the rational kernel within the mystical shell." If only linguistic philosophers could so far overcome their prejudices as to ignore the mystifying misuses of language in Hegel, they might be able to extract the "rational kernel" for themselves, and break out of their own shell. As Engels remarked in the Dialectics of Nature: "If we turn the thing round, then everything becomes simple, and the dialectical laws that look so extremely mysterious in idealist philosophy at once become simple and clear as noonday."

To this department of investigation belongs, for example, the investigation of the rules of qualitative and quantitative determination and their connection—the so‑called "law of the transformation of quantity into quality". What is subject to qualitative determination is subject to quantitative determination, and vice versa; and these are so connected that alteration of the one is not independent of alteration of the other. Again, there belongs too the investigation of the so‑called "unity and interpenetration of opposites" and of "dialectical contradiction". To describe processes, motions and operations always involves the deployment of polar oppositions, such as "attraction and repulsion", "increase and decrease", "growth and decay", "addition and subtraction", "forwards and backwards", and so on.

But investigation of universal laws of dialectics remains an open field. It is something that has been projected but not yet systematically done. And the laws that have been written down, following Hegel, still lack both the precision of formulation and the systematic derivation to be expected of anything that can rank as science. The laws of dialectics should be, as Engels claimed, "as simple and clear as noonday". If they are not, and if their interconnection is not evident, that is because not enough work has been done on their formulation. (A case in point is the so‑called "law of the negation of negation".)

This may perhaps be thought a great failing. But to work out reliable theory it is not necessary first to work out all the principles which reliable theory exemplifies. On the contrary, no more than you have to work out all the laws of locomotion before you can walk do you have to work out all the laws of thought before you can think. It has been reported that Marx was always meaning to write a systematic treatise on logic and dialectics. But he never did. And even if he had tried, he might not have succeeded. For such an enterprise calls for a lot of highly specialised work, and most of this was done afterwards, and some of it by linguistic philosophers. To deal with dialectics requires a grounding in formal logic. Engels did make a few remarks about formal logic in his notes for the Dialectics of Nature, as well as about mathematics in Anti‑Dühring; but today these date badly. The well known practicality of Marx: prevented him from writing his treatise on logic and dialectics, not only because as a practical man he had other things to do, but also because it is never practical to undertake tasks for the fulfilment of which the conditions are not ripe. For the same reason he never tried to work out the detailed principles of socialist planning, which his followers now have to wrestle with. But what he did work out provides a secure starting point for socialist planning and socialist philosophy alike. He did establish the dialectical materialist approach, leaving it to his successors to work out the abstract principles of its logic as that becomes necessary for purposes of understanding, criticism and further development.

4. THE LAWS OF DIALECTICS

I have already remarked that the findings of an investigation of categories are distinguished from the propositions of formal logic in that they are not formal, and so are not to be symbolically constructed and tested in the way logical formulas are. But they are not empirical generalisations, and so are not comparable with even the most fundamental laws formulated by empirical sciences. For instance, the laws of thermodynamics deal with transformations of energy from one form into another. But "the law of the transformation of quantity into quality" is not a comparable law, just as "transformation of quantity into quality" is not comparable with "transformation of mechanical motion into heat". These exemplify different uses of the word "transformation".

In the Dialectics of Nature Engels called the laws of dialectics "the most general laws", and said that they were "abstracted" from "nature and human society". But if he meant, or is interpreted as meaning, that laws of dialectics are comparable with, say, laws of motion as formulated by Newton or Einstein, differing only in being even more general, then that simply exemplifies confusion in the use of the words "law" and "general" (admittedly very confusing words), since the latter are empirical laws and the former are not. There is not a shred of evidence to suggest that Engels was actually guilty of any such confusion—that he really thought that, for example, the "law of transformation of quantity into quality" was a transformation law of the same logical type as, say, the first law of thermodynamics. Some of his interpreters did afterwards perpetrate such confusion—but Engels himself simply did not deal with such logical questions, which had not yet been raised at the time he was writing. The point can, if one likes, be put like this—that the difference between these laws is not only a quantitative difference, of degree of generality, but a qualitative difference; they are different kinds of law, or exemplify different uses of the word "law".

Perhaps the difference can be most perspicaciously brought out in terms of Dr Popper's criterion for a "scientific" (or empirical) law. Such a law, he said, must be falsifiable. Sure enough, the laws of thermodynamics, for example, are falsifiable, even though never actually falsified. But principles or laws concerning categories, including "the most general" ones, or universal "laws of dialectics", are not falsifiable—or if they are, they are not correctly formulated. For instance, certain scientific or empirical laws stating and forecasting how particular changes happen state and forecast how certain quantitative changes bring about certain qualitative ones. Such laws are subject to empirical test (falsifiability), as to whether the connection of qualitative with quantitative change is always in fact as stated. But that qualitative and quantitative changes are always connected is not similarly subject to test. That they are connected in a particular way in a particular case has to be empirically established or disestablished—but not that they are connected. Thus "the law of the transformation of quantity into quality" is, as Engels said, exemplified in scientific laws, and in that sense it may be said to be "abstracted" from the same factual data as are generalised in scientific laws. But at the same time it is not a more general empirical law, comparable to scientific laws, and the method of arriving at its formulation and of testing its correctness is entirely different.

In his inaugural lecture at Oxford (Philosophical Arguments, 1945) Professor Ryle said that "a pattern of argument which is proper and even proprietary to philosophy is the reductio ad absurdum". He was evidently referring to those arguments in philosophy which concern, in his phraseology, the logical investigation of categories. Correctly formulated principles concerning category, or "category propositions", are such that their breach results in "absurdity"—and this is the test of such principles. That (I believe) is in principle correct. But he had earlier ended his paper for the Aristotelian,, Society on Categories by asking: "But what are the tests of, absurdity?" And this vital question was still not answered. In criticising, above, the linguistic interpretation of category, I suggested the way of answering this question.

The kind of absurdity that results from breach of category principles is not the same thing as violation of the rules of the actual or significant use of given languages. For example, it is no more a violation of those rules to say that quantitative changes can go on and on indefinitely without involving qualitative changes (which Hegel and Marx said should not be said), than it is to assert the independence and immortality of ghosts which temporarily inhabit the machines of our bodies (which Ryle and Marx said should not be said). The "absurdity" of such statements consists rather in what I have called their uninformativeness.

Those very simple and obvious examples of "category mistakes" which linguistic philosophers quote from Lewis Carroll are of this kind. For example, when the White King misunderstood Alice's statement that she "saw nobody on the road" he was expressing the curious view that the person Nobody is the person who is there when no visible or tangible person is there. This raises a laugh; but its "absurdity" is of the same kind as that perpetrated by primitive peoples and theologians who say that your immortal soul is that part of you which thinks and feels and acts when you yourself have ceased to think or feel or act. Of the same order of absurdity is the statement that the Cheshire Cat's grin remained after the Cat had disappeared. Some such absurdities are very easy to recognise, and are funny; and others are not so easy to recognise, and are serious. To the latter class belong those absurdities which result from breach of the general principles of materialism and dialectics. The point of these general principles, and what makes their formulation worthy of being called "the science of thought and its laws", and their observance as essential for right thinking as is observance of the laws of formal logic, is that they are the principles for maintaining informativeness and removing illusion and fantasy.

We can very well speak, if we want to, in violation of laws of dialectics, of things which are not affected by other things and are eternal, indestructible and changeless. Language permits us so to speak, and it involves no grammatical or syntactical error, no violation of principles of formal logic and no logical contradiction. But if we try to describe a world consisting of such things, then we are describing something unrecognisable and unverifiable. For how could we recognise anything or verify its existence unless it affected us in some way? And how can anything which is fixed and changeless and does not interact with anything else possibly so affect us? To speak of such things is to speak of things in principle unverifiable—like, for example, Leibniz's monads. Statements about them are uninformative or, as Engels put it, "fantastic", or as Ryle puts it, “absurd".

Similarly with quantitative specifications unconnected with qualitative ones, or with processes free of opposition or dialectical contradiction. For example, we can very well describe a physical system in which the law of gravity did not hold in the form we know it. Thus bodies might attract each other with a force not inversely proportional to the square of their distance; or in the formulation made by general relativity theory, quite different geometries might apply. But what of a physical system in which forces acted between bodies but none of these forces were attractive? It is not very hard to see that such a system is an "absurdity". One can talk of such a system, but it is mere "fantasy". So in Dialectics of Nature Engels wrote about the necessary correlation of attraction and repulsion in physical systems, and showed that this was a case of the more general principle of "unity of opposites". Again, one can speak of a system in which quantitative changes did not involve qualitative ones. But how could we recognise or measure any quantitative changes unless there were recognisable qualities connected with them? And how could we formulate generalisations or "laws" governing such changes except in terms of such connection? As Engels put it in Dialectics of Nature, "the dialectical laws are really laws of development of nature, and are valid for theoretical natural science". They are "laws of development of nature", and theoretically "valid", because no material system and its laws could be described in concrete terms except in conformity with the dialectical principles. Hence they are always and necessarily found exemplified in natural as well as social science, whenever we undertake "the concrete analysis of concrete conditions". Their formulation is first suggested as a result of observing their occurrences and exemplification (somewhat as mathematics is first arrived at from observation and practical application, and not produced ready‑made a priori out of people's heads, the concepts of number and measure, just like those of quantity, quality, opposition, and so on, being derived from experience). But at the same time, once they are formulated, the necessity, of dialectical laws, and consequently their logical distinction from empirical laws, can be demonstrated, and their formulation be then systematically corrected and made more exact in the procedure of demonstration.

5. DIALECTICS AND MATERIALISM

An important consequence of this account of dialectics is that it renders invalid the separation so often made in expounding Marxist philosophy between materialism and dialectics.

In his little book on Dialectical Materialism (which was extracted from the larger History of the Communist Party of the Soviet Union, 1938) J. V. Stalin separated materialism from dialectics by saying that the former was "the theory" and the latter "the method". Mechanistic or pre‑dialectical materialism is, perhaps, "a theory"—that is to say, a "metaphysical theory" which attempts to state the ultimate constitution and structure of the universe. But dialectical materialism is not "a theory" in that sense, or in any sense of the word in which "theory" may be contrasted with "method". The fact is that the derivation and substantiation of the principles of dialectics is at the same time that of the principles of materialism—they are inseparable and are principles of exactly the same sort, being demonstrated in terms of the correct marshalling of categories in informative discourse.

Thus, for example, Ryle was quite right in calling the idealist theory of "the ghost in the machine" a "category mistake", even if one may disagree with his account of category. And so is it a category mistake to adopt the more general idealist approach that "thinking is prior to being", as it is to adopt such typically "undialectical" ideas as that things are what they are independent of their relations with other things and of their modes of coming into being and ceasing to be, or that qualitative and quantitative changes take place independently, or that processes proceed free of contradiction. To think in a consistently materialist way is to think in a dialectical way and to think in a consistently dialectical way is to think in a materialist way. These ways are the laws of thought, because they are the ways of assembling information, keeping thought within the bounds of informativeness and excluding illusion and fantasy. Materialism is not "a theory" that requires experimental verification and could be falsified, any more than it is a "metaphysical theory" which is somehow established a priori; nor are the laws of dialectics laws that require experimental verification and could be falsified; for any verifiable theory is materialist and dialectical.

The "mystification which dialectics suffers in Hegel's hands" was due to the way he separated dialectics from materialism, and tried to base formulation of laws of dialectics on idealist theory. The way Marx began to put this right was by getting rid of "preconceived idealist fancy". Hegel started with a preconceived and absurd idealist theory, to the effect that Thought or The Idea exists timelessly and independently of anyone's thinking, and that the world is created to be as Thought thinks it. Then (supposing himself to have what Ryle and others would now call "privileged access" to The Idea) he tried to write down in a book the sequence of Categories (in this context capital letters are used as a mark of respect for timelessness) in accordance with which the world is created. The result, naturally enough, was "mystification". Marx, on the other hand, did not start with any "preconceived idealist fancy". He started from "real premises from which abstraction can only be made in the imagination . . . the real individuals, their activity and the material conditions under which they live". He started to work out how these real individuals must think and assemble their information in order to inform their practice. That is what he meant by turning dialectics "right side up again".

To work out the laws of thought is to work out the principles in accordance with which we must think in order to inform our practice. It is not, therefore, an a priori inquiry of the sort Kant imagined, but it starts, as Marx and Engels said, from "real premises" which are "verified in a purely empirical way".

At the same time, it is by no means the same as stating the laws in accordance with which subjective processes of thought proceed, as contrasted with the laws (studied by empirical sciences) in accordance with which objective processes proceed. Indeed, such a misunderstanding is an absurd one—for thought‑processes are just as "objective" as any other processes, and to find out how they proceed requires, as for any other processes, examining the evidences of their occurrence. The generalised principles of materialism and dialectics no more provide information, additional to that obtainable by empirical means, about thought‑processes and how they proceed than they do about any other processes and how they proceed. And they apply equally to both. They are not statements of information but principles of its assembly. But as such they do serve to correct erroneous views (those that take the form of illusion and fantasy), and consequently to specify the category‑formation of truthful views.

It is in this sense (and a very important sense it is) that the working out of the principles of materialism and dialectics may be said, in Ryle's words, to inform us about the "nature of things", and of their reflection in human thought; or, as Engels put it (Ludwig Feuerbach, 4), materialist dialectics becomes "the science of the general laws of motion both of the external world and of human thought". How things are and move (the forms of interconnection, the transformations of quantity into quality, the contradictions), and how this movement is to be reflected in our thought, is the discovery of materialist dialectics.

Principles of materialist dialectics, then, are expressible in less negative and "frivolous" terms than Ryle supposed, when he observed that the conclusion that "the Cheshire Cat could not leave its grin behind it" enlightened us "about the nature of things". For what we learn from these principles is of considerably greater and more general practical importance. Thus, for example, the conclusion that "it is not the consciousness of men that determines their being, but their social being that determines their consciousness", or, more generally, that "being is prior to thinking", is at once more generalised than a principle about cats and their grins, and what we learn from it is in practice more important for us—since in practice no one gets led astray by supposing that grins exist independently of faces, whereas many get led astray by supposing that thinking or consciousness exists independently of real individuals and their social being. Similarly with conclusions about the dependence of qualitative changes on quantitative ones, and other "laws of dialectics".

If we consider principles of materialist dialectics, such as those mentioned as having been first formulated by Marx and Engels, my contention in this chapter is that they evidently differ from the generalisations arrived at by the empirical sciences in that they are not established and tested by the same type of directed observation and empirical test, and that their applicability in the guidance of practice is therefore also not of the same sort. Their character and application is essentially that of "laws of thought", as Engels already emphasised when he included "dialectics" along with "formal logic" in "the science of thought and its laws", stressing that it was demonstrable "independently" of "the positive science of nature and history".

However, this does not entail, as Kant supposed in his fanciful theory of "synthetic a priori knowledge", that they are somehow spun out of the inner resources of the mind and imposed by the mind upon "the phenomenal world" which it constructs for itself as its "object". On the contrary, they are indubitably and universally true of the processes of the real world, which we get to know about, but which proceed in accordance with these "laws" independently of our knowing them. Such a contention can appear paradoxical only to those who are still under the spell of the old antitheses of a priori and a posteriori, or of "analytic" versus "synthetic" knowledge.

The distinguishing feature of principles of materialism and dialectics is that they concern categories. They lay down correct ways of assembling relatively abstract items of information in order to arrive at concrete and informative conclusions. Category principles, considered in their aspect as reflecting objective reality or "the nature of things", are characterised by their absolute generality (corresponding to Aristotle's conception of truths true of "being as being"), in distinction from generalisations of natural or social sciences which deal with particular classes of phenomena. As I have said, this "quantitative difference" of degree of generality marks a "qualitative difference" in the character of the generalisations.

Of course, as Engels often and justifiably insisted, such generalisations, far from being spun out of the mind from its own inner resources independent of all experience (the conception of the mind spinning anything out of itself independent of experience is absurd) are arrived at as a result of experience. And naturally, errors or crudities in their formulation may be detected as a result of contradictions with experience. But the test or demonstration of the correctness or otherwise of such generalisations goes beyond merely ascertaining their empirical confirmation. Their demonstration depends on demonstrating, not as with empirical laws that they are not in fact falsified, but that to imagine their falsification results in "absurdity" or "fancy" in the sense of uninformativeness.

For example, Marx's principle that "social being determines consciousness" was arrived at by studying the facts of human history, and criticising the conclusions reached by historians and sociologists who proceeded on an opposite principle. But in its generalised form, "being is prior to thinking", this principle is one which must necessarily be observed in assembling information, because not to do so is to introduce uninformative fancies in principle unverifiable. In that respect it differs from Marx's formulation of the basic law of development of human society, which dealt with the particular way in which men live by developing a mode of production.

Thus Engels, with complete consistency, said on the one hand that laws of dialectics are "laws of thought", and on the other (in the preface to Anti‑Dühring) that "there could be no question of building the laws of dialectics into nature, but of discovering them in it and evolving them from it". He went on to say that "the revolution which is being forced on natural science by the mere need to set in order the purely empirical discoveries . . . must bring the dialectical character of natural events more and more to the consciousness even of those empiricists who are most opposed to it . . . It is possible to reach this standpoint because the accumulated facts of natural science compel us to do so; but we reach it more easily if we approach the dialectical character of these facts equipped with the consciousness of the laws of dialectical thought."

SOURCE: Cornforth, Maurice. Marxism and the Linguistic Philosophy. 2nd ed. London: Lawrence & Wishart, 1967 (orig. 1965). Part III, chapter 2, pp. 285-302. See also the complete book offsite.

Marxism and the Linguistic Philosophy: Contents

Science and Evaluation by Maurice Cornforth

Logical Empiricism by Maurice Cornforth

Maurice Cornforth (1909-1980) Study Guide

Vienna Circle, Karl Popper, Frankfurt School, Marxism, McCarthyism & American Philosophy: Selected Bibliography

Positivism vs Life Philosophy (Lebensphilosophie) Study Guide

Salvaging Soviet Philosophy (1)

Marx and Marxism Web Guide