Epilogue on Berkeley ^[*]

Roy Wood Sellars

MY DICTIONARY defines an epilogue as a concluding part to a literary work; and it can be here so taken. What I shall be concerned to do is to comment very briefly on the course philosophy took in the period after Berkeley. I shall even comment on the present scene. I was early convinced that the rise of human cognition in the matrix of perceiving was a very basic philosophical problem. There was quite a flurry of energy devoted to the question in the first three decades of this century. Philosophers have differed as to what was accomplished. There is the negative attitude exemplified by John Dewey. Professor Schneider quotes Dewey as saying, "We did not solve the problem. We got over it." But, if I remember events correctly, Dewey came late into the field of epistemology after the New Realists and the Critical Realists had had their debate. He criticized both and kept to his logic of inquiry. I have argued elsewhere that both Dewey and Woodbridge were puzzled by the status of sensations, which seemed to be located in the brain. If these are the first objects of knowledge, how could one get to the outside world? Their gambit was to reject such entities and start with the outside world.

It seemed to me that this was a sort of begging the question. What was the status of consciousness? Dewey was always very vague on this topic. In fact, he moved to a modified kind of behaviorism.

I, on the other hand, persisted in thinking about it. It seemed to me that a new approach was needed. This would stress the use made of sensory factors in a frame of directed reference. These factors [55/56] would operate in the directed act as informative. In animal life, they would have a guidance role, as even Berkeley noted. But, at the human level, there would be added the use of them as informative of the stimulating object which would also be the object cognized. From this situation would emerge the formulation of empirical facts about the object cognized. The base of these facts would be in the informational appearings found in the sensory data. This means that sensations would now function as disclosures, or manifestations, of the object. They should no longer be regarded as terminal entities, as they were for Berkeley and Hume. Instead, they function in a larger context. Concepts and language would now enter to aid in this cognizing effort.

Now I do think that human cognition is a wonderful achievement. But, surely, it had humble beginnings. As I see it, it developed in the matrix of guided response. Here we come to animal life and note the function of the sense‑organs. The working unit is sensorimotor. Man has emerged with delayed responses and the use of his sensations for the deciphering of his environment. They clearly lend themselves to this employment. Mental ability is, of course, added. It is in this fashion that human cognition has emerged. The very word, fact, is a sign of this arrival.

Let me indicate some changes in analysis indicated by this approach. It was customary to speak of the content of perception as against the object of perception. In the older view the content could only be sensations and images. Now, as I see it, this content is changed into the operative use of this material into appearings of the object. Added to this transformation is the development of concepts about the object being perceived. The content ceases to be a tertium quid intervening between the knower and the object and becomes the instrument of knowing.

I am now going back to Berkeley, Hume, Reid, Kant and others to show why and how they "missed the boat." In all their ability, they did not understand the “from­and‑to” mechanism of perceiving.

Berkeley, we saw, reflected the then current view of matter as inert, something pushed around like a billiard ball. Berkeley also accepted the view of sensations as terminal effects. I was told recently, during a visit to Trinity College, Dublin, the locus of Berkeley's early work, by Professor Furlong that Berkeley was as much influenced by Malebranche as by Locke. And I think one can see this in his theory of how ideas are inserted into the human mind. [56/57]

The result of this view of matter as inert was that he could see no role for it in sense‑perception. Spirit was introduced as causally active with volition as its paradigm. It is quite otherwise with us with our stress on energy. A weakness of Locke upon which he jumped was Locke's substratum view of the support and inherence of properties. I think we move from facts about things to our categorical notions of them. The object I look at may be cognized as red, round and so far away. There is no necessary finality about this cognition. It is a point of empirical departure. And we may decide that redness is correlated with light reflected from the object to our eyes. Properly interpreted, these cognitions do not conflict.

In our own day, science has developed sensors to gather information beyond the reach of our own sense‑organs. The recent excursions to Venus illustrate this. The sensors are instruments which gather information and radio them to the earth. There the messages are decoded. And the decoding enables us to use the information about temperature on Venus. Now, as I see it, evolution developed our sense‑organs for guidance and, ultimately, for cognition. Science assumes the fact of cognition. It has been one of the tasks of philosophy to try to understand it. I took the task seriously. I was recently told at Oxford that Professor Austin was inclined, after he had undercut phenomenalism linguistically, to leave the rest to science. I certainly wish to cooperate with science which has recently been doing such remarkable work on nervous receptors. But I still think that philosophy can supplement this work by undercutting some traditions in epistemology.

I pass now to Hume. He may be said to have continued the outlook of Berkeley and Locke in making sense‑impressions and images terminal rather than, as I hold, informational. I shall argue that even Kant did not make this transition. His things‑in‑themselves did not manifest themselves in the sense‑manifold.

Many hold that there was a touch of agnosticism in Hume. The mind and the material world may have "ultimate, original qualities" but these are beyond human knowledge. If so, his view may have resembled that of Gassendi, the father of modern atomism. What Hume did was to bring out into sharp relief the implications of his sensationalism. He belonged to the period of the Edinburgh Enlightenment and had no religious axe to grind. He even examined religious cosmic beliefs quite objectively. He was puzzled by the self for he could find no sensory basis for it. And with the Self went Berkeley's [57/58] Spirit and volition as causal. Causality became a case of routine expectation.

The critical realist cannot accept this reduction of the categories. To him, they need analysis but they are, essentially, well‑based cognitions of the constitution of the world. Like space and time, they need refinement and clarification.

G. E. Moore was puzzled by Hume but he, himself, was unable to relate his sense‑data to things. It is this that I have done in a functional way. Russell hardly made the effort. He swung to logical construction. I had conversations with him and he seemed to stress percepts rather than perceiving. My referential approach was quite alien to him. It has led me to wonder about the competence of logicians in epistemology. I shall mention no names.

I turn now to Reid. I have a certain sympathy with his rejection of ideas as primary objects. But I do not see that he reanalyzed perceiving in my fashion. He rightly stressed language and common sense but fell back on intuition too readily. He was convinced, however, that mental operations involve three factors: the mind which performs the operation, the object of the operation, which be seems to have thought was intuited or apprehended without any tertium quid intervening, and the operation itself. My reanalysis of the content of perception into sensory material and their cognizing use in the context of objective reference is, so far as I can see, not present. He appealed to language in a way analogous to the present "ordinary language" movement but, as I see it, he did not realize how ingenious nature had been. I, myself, moved between the introspective tradition and Watsonian behaviorism. Neither seemed to me to have an adequate epistemology. The introspective tradition stressed, like Russell, percepts rather than perceiving. How could one move from introspection to commerce with the world? Idealism beckoned. And Watsonian behaviorism just observed the behavior of rats and other mammals but had no curiosity as to the nature of observing itself. Why should it? Did not all the sciences make observations? How could one break down the barriers between introspection and behaviorism? My reply is that a good epistemology will help. But, even without that, I find that I agree with Gestalt psychology that traditional introspection did not do justice to sensory organization and the influence of the environment. Thus it tended to be reductive. As a philosopher, I would stress perceptual experience and its categories but I am, of course, interested in development. [58/59]

John Stuart Mill struck out at Reid and Hamilton. But he, himself, as I look at it, could not get to the external world. Matter was reduced to the possibility of sensations. There are a lot of blind alleys in philosophy. I suppose that proves its difficulties. It may well be that the attainment of an adequate epistemology may simplify many things. Many technical analyses may then fall into line. Reid's successors got into disputes. As James Ward pointed out in his Naturalism and Agnosticism, one dispute was over the question whether we see the sun itself or an appearing of the sun. I contend that we see the sun, itself in a referential way, in using its appearing cognitively. This, of course, is just a good start. But the right start is tremendously important. I think of the energy and ability that has gone into logical positivism, into existentialism, even into pragmatism. Many points are well taken, and yet are the foundations sound?

I leave Reid, and pass along. In France, Condillac identified knowledge with complexes of sensations. He was very ingenious and tried to show how we come to have the idea of an external world. Hume had worked on the same problem. We saw that Berkeley was aware of it. How do we two come to believe that we see the same thing? Since I start from a framework in which perceiving is externally oriented, I do not regard the problem as very difficult. It is solved, in part, behaviorally and, in part by communication. James gave the case of blowing out a candle. When I blow it out, your light also goes. But all this is made easier by the referential view of perceiving. Berkeley made it more difficult since he held the various kinds of sensations the primary objects of perceiving. I do, not think we need Aristotle's common sense. The point is that the context of perceiving is referential and sensations are used in this context and subordinated to it. Pointing goes naturally with perceiving. So do demonstratives. Since G. E. Moore made sense‑data terminal, it is not surprising that he had heavy going with his proof of an external world.

When we pass to the United States we find John Dewey an interesting transitional thinker. Kant and Hegel had long come on the scene and the result had been what is usually called Anglo‑American idealism. Kant was alive to the science of his day, which was essentially Newtonian. He was, however, not a physical realist. While recognizing a sense‑manifold, he did not appreciate its disclosing function. Rather he amalgamated it with a priori forms, such as space, time, causality, substance, and arrived at a phenomenal construction [59/60] of the world. In this outlook he was motivated, I believe, both by his inadequate view of perceiving and his assumption that the physical world involves self‑contradiction. This he thought he had shown in his Transcendental Dialectic. Modern logic and mathematics have undermined his arguments. But they played an important part in the ensuing rise of romantic idealism. Fichte took his departure from the Self while Hegel developed a logico‑historical dialectic which paid small attention to the material world. One must remember that chemistry was just beginning to get a framework and that biology was still in the future. Technically speaking, the Hegelians rejected Kant's things‑in‑themselves as meaningless, as outside thought. These had been Kant's gesture to realism. One was now left to such terms as experience and thought. Not unsurprisingly, epistemology was downgraded. One heard, at most, of the subject‑object relation.

There were protests outside official philosophy, as by Diderot and the scientists who began to flirt with the idea of a revival of materialism. But this was disregarded by official philosophy. Feuerbach, however, protested and influenced the rise of dialectical materialism, as expounded by Marx and Engels. It is only of late that Feuerbach has been carefully studied. And that is largely due to the rise of Marxism. I return to John Dewey and the United States.

Dewey had been introduced to Reid and Hamilton by Torrey of the University of Vermont. The stress was now on ontological dualism and intuition as a bridge. This seemed the sole alternative to the sensationalistic empiricism of Hume. After this, Dewey was brought to Kant and Hegel by G. S. Morris. Later he was affected by Darwin and William James and developed a logic of inquiry in which the stress was on the reflective solution of problems. William James, on the whole kept, with modifications, to the empirical tradition of Mill with, however, more stress on purpose and teleology.

Dewey moved to his experimentalism, instrumentalism and behaviorism in a quite logical way. As I have indicated, the idealistic tradition downgraded epistemology except as a reflection of its outlook. It is not surprising that the initiative here was taken by a group of younger men, the new realists and the critical realists. Dewey came into the situation from the outside, so to speak, and pointed out weaknesses in both. As regards critical realism, be devoted his attention to the "essence" doctrine of Santayana and Drake. So far as I know, he never carefully studied my analysis of [60/61] perceiving. Hence I can appreciate his remark to Professor Schneider and also discount it. Schneider, himself, it would seem never understood my effort to get a new kind of direct, referential realism in which sensations are given an informational role. This, of course, undercuts both presentationalism and representationalism. Schneider decided that nothing new was offered. The next generation of American philosophers turned largely to Europe. There are some signs of a renewed interest in critical realism.

If I were asked why I turned to epistemology, my answer would be that I was interested in science and that I could not see how idealism connected up with it. Perceiving offered a linkage.

Like the Yellow River, philosophy frequently changes its channel. This is sometimes in the way of modification of emphasis, sometimes by means of a new approach. It is all very complicated with side-currents arising. In my own lifetime . . . I have had to take cognizance of pragmatism, Bergsonianism, new forms of realism, the idea of emergence, analytic philosophy, logical positivism, Wittgenstein, Whitehead, existentialism, ordinary language and Marxism. It follows that one has to have balance and a strong sense of priorities.

In this Epilogue, I am charting my course as I look back at it. Thus far, I have emphasized epistemology. But I want to add ontology. After all, Berkeley was an immaterialist and was fighting deism and materialism. One strand of his argument was to the effect that to be is inseparable from being perceived. I have argued against that. The other strand stressed ineptitudes in Locke's formulae. What is this unknowable substratum which supports primary qualities? And can an inert matter explain ideas and volitions?

This belief in the inertness of matter is, of course, dated. The stress now is on dynamism and energy. The sun's energy, or power to do work, keeps life going. And Einstein's famous formula is generally accepted, that is, that energy equals mass times C squared, C being the velocity of light.

But let us consider for a moment what and how we know about material things and their constituents. I have argued that we never intuit, or literally inspect, material things. That is an illusion nourished in us by presentational, or naive, realism. Instead, we can locate things and learn about them by using the data of observation. This operation gives us facts about them. What happens in science is a technical probing of things, particles and processes to get knowledge about them. Largely, it is an affair of asking questions [61/ 62] and getting relevant answers. But these answers turn out to be mensurational reports about structure and behavior. It is in this fashion that we learn about atoms, molecules, electrons, positrons, etc. Huge machines are made for smashing atoms. These constitute the equivalent of a microscope. But, in all this, we learn only facts about the particles involved. We can never bypass information and arrive at some sort of inspective intuition of either molar things or particles. Matter is that which behaves in this sort of way and has this kind of structure. When chemists synthesize complex molecules, they work out relational patterns for the constituent atoms and get the same behavioral properties as the original substance had. I remember that Professor Sidney Hook once challenged me to tell him what matter was. It seemed to him a conceptual construct in what he called experience. Whitehead seems to have had similar ideas. That our tested thought about matter is conceptual I do not deny. But I hold these are concepts about, responsible to, garnered information.

The history of atomism is a long one, at first largely speculative, by noting states of affairs like the wearing away of stones and the drying of clothes. But Boyle developed his corpuscular hypothesis with considerable evidence. And this evidence increased after Dalton's work in chemistry. In these days of radioactivity and the atomic bomb few doubt the existence of atoms. Ernst Mach was one of the last to be skeptical but that was, I think, largely because of his epistemology, which was built of sensory elements.

I may here mention the point that when Lenin was trying to defend materialism against its critics he sought to show that science begins with material things and passes thence to reflected sensations. But his opponents to this day told him that cognitively we begin with sensations. Berkeley was still a hard nut to crack. But the reader will remember that I regard the material thing as the direct object of perceiving and the sensations aroused in the percipient are sources of information and so used. That is, the sensations are not terminal but functional. The significant use of the term, dialectical, is another question.

Locke's substratum idiom, which Berkeley rightly attacked, is no longer a natural idiom. It is the material thing we know in terms of achieved knowledge about it. As I have argued, we get no nearer to the object cognitively than facts about it, though we may handle it and tear it apart. But, in terms of facts and tested theories, we know a great deal about matter and energy. [62/63]

Berkeley's was a mind of first quality and that is why I have concerned myself with him. I attacked Cartesian dualism from another angle. In my first book Critical Realism, I argued that consciousness is extended in the way events and functions are. I was pleased that Woodbridge Riley noted this gambit in his history of American philosophy. So far as I know, no one else did. This, of course, led to my double‑knowledge approach to the identity of mind and brain. I argue that one participates in brain activity. I am not an epiphenomenalist.

As against Berkeley, I have the benefit of scientific developments; one of the greatest of these is occurring in molecular biology. This fits in with my theory of emergence and levels in nature. It would seem that DNA and RNA work by means of codes. This is a new kind of causality to be added to feedbacks and homeostasis. I earlier saw the importance of organization in nature. I find my thinking is in line with Gestalt in that I think nature organizes itself while man makes machines. This is what Köhler calls a dynamic view. He stresses sensory organization as following stimulation. As nearly as I can make out, recent work on retinal receptors, such as that of the Nobel prize winners, Dr. George Wald, and Dr. Hartline, move in that direction. Upon this base, learning advances. Epistemology, as I see it, starts in its examination from the level of perceptual claims and these concern outer things. It is this tree that I am looking at and judging about. I pointed out the preparation for this. Cognition emerges within a context. Wittgenstein stressed facts but I do not think he looked on them in this way. He seemed to me to give them more an ontological status. But I am not an expert on his perspectives.

Now all this sums up to emergent, or evolutionary, materialism. This is non­reductive in perspective. In a way, it culminates—so far as we know—in man. And the job of philosophy is to cooperate with the sciences in understanding the situation. I can quite appreciate the immaterialism and theism of Berkeley. For him, as a bishop, it was a natural stance. And he added a Platonic note in his later writings, much as did Whitehead in our own day. But I can get no nearer to Platonism than words and concepts and their cognitional application. Here we are at the level of the working of the new brain and its linguistic centers. Professor Wilfrid Sellars connects thinking with verbal expression, “The learning to use symbols in accordance with rules is a pervasive feature of concept formation.” As I see it the resultant capacity operates even when we are not [63/64] talking. Language is a social achievement whose origins are lost. But its logic and psychology are now being closely studied. A new kind of empiricism is resulting. I take this to be a continuation of critical realism. It is a compliment to a framework to seek to develop it. I believe this is being done.

I shall now move to the close of this Epilogue by means of some remarks on the past and the present.

When, in 1916, I published my first book, Critical Realism, I sent a copy to the famous British idealist, F. H. Bradley. He was kind enough to read it and to write me his reaction. It was to the effect that he could see that realism had much to be said for it but that he, personally, could not conceive how we could get beyond appearances. His effort, accordingly, was to indicate how appearances could be organized logically in terms of increasing coherence. I take it that Professor Blanshard is pursuing that objective. In a way, it goes back to Berkeley. Kant and Hegel bad, of course, been added. It has been my endeavor to show how appearings are referentially used in cognition. I take a less skeptical view of categories than does Bradley. I regard them as cognitive achievements which can be developed.

Some years ago there appeared a book in England called The Revolution in Philosophy. It largely devoted itself to showing the importance of the advances in logic to relationalism. Little attention was paid to epistemology. Because the book did not broach the question of the nature of perceiving which, obviously, played such an important part in Bradley's thinking, I could not regard it as very revolutionary. It seemed to me to be marking time. I wanted to get down to fundamentals.

I have already paid my respects to phenomenalism and logical positivism. The burden of my reproach was their neglect of epistemology. These writers were ingenious and learned within their assumptions and added to philosophy's technical equipment. Nevertheless, they wrote off the whole idea of transcendence as meaningless. I, on the other hand, had correlated transcendence with (1) reference and (2) the use of sensory information to give facts about objects. It is, of course, a worldly kind of transcendence which does not lead to a supernal realm. But it is transcendence, nevertheless. It gives a base for ontology and, if you will, metaphysics.

I have paid my respects to pragmatism in the persons of Dewey and Sidney Hook. With the former with respect to realism. With the latter with respect to the status of matter. There is much [64/65] practicality in American pragmatism, perhaps too much. It seems to me to go, at times, three quarters of the way to clarification of problems and to leave it there. Had I more space I should like to illustrate what I mean by this. But my remarks on both Dewey and Hook are samples.

I want, in conclusion, to say a few words about Existentialism. My technical complaint is that it has no clear epistemology or ontology. I have no objection to a stress on Angst. I feel it myself sometimes. Nor do I want to treat persons like things. The human situation is very important. Call it, if you will, philosophical anthropology. It is a reflection on human life.

But my objections go deeper. So far as I can make out, the existentialists build on Kant and Husserl. They are suspicious of empiricism, naturalism and evolutionism. There is a marked subjectivistic element in their outlook. Take Husserl, for instance, one of the unwitting, founding fathers. By the use of his technique of “reduction,” he arrives at the field of personal consciousness. External objects are reduced to a sensory pole in this field. Thence he moves to a Cartesian form of idealism. The realistic analysis I have sketched is completely alien to him.

Sartre has this background. The controversy between idealism and realism is to be disregarded. Material things become an affair of appearances, much as they are for the logical positivist. Hence he rejects realistic materialism. He regards himself as left with "being‑in‑itself," which is inert and passive, and "being‑for‑itself," which is active. There is no attempt to explain this contrast. In fact, there is a marked solipsistic note in Sartre. But Sartre is an able writer and stresses freedom and commitment.

Turning to Jaspers, we find him trying to get to transcendence and God by squeezing between Kant's subject‑object relation. Somehow the world points beyond itself to an Absolute. This is an object of faith and not of knowledge. Jaspers is a brave and able man but, seemingly, out of contact with analytic forms of philosophy. There has been inbreeding here.

I cannot find much different in Heidegger and Marcel. The former makes much of ontology but seeks to reach it by a semi‑poetic intuition. He rejects the correspondence theory of truth and hopes to unveil what is by verbal efforts going back to early Greek philosophy. Heidegger strikes one as a strong man in travail with he hardly knows what. Turning to Marcel, we find him putting stress on the difference between problems and mysteries. He challenges [65/66] the sciences and the scientific view of the world. On just what grounds it is not clear.

Finally, we come to the theological existentialist, so popular in the United States, Paul Tillich. His background is, like that of Jaspers, of a Kantian type with introduction of transcendence to a God beyond any anthropomorphic God, to Being itself. I cannot, myself, find a basis for this projection. I have taken up this question in a companion book which I call American Philosophy from Within. What stands out is the separation of philosophical traditions. There is hardly genuine communication and mutual understanding. But I think this will pass. My own kind of materialism does justice to levels, the emergence of knowing, valuing and morality.

Thus ends my epilogue on Berkeley's famous Dialogue. As I look back on these sixty years, it seems to me clear that American philosophy as it came of age, deviated from the European in a certain explanatory openness. Of course Europe hardly took notice of it. And, as I see it, special circumstances, in my own case, favored a breakthrough on critical points.

* Reprinted by permission from Sellars' book, Lending a Hand to Hylas: A Restructuring of Berkeley's Famous Three Dialogues (Ann Arbor: Edwards, 1968), pp. 92‑102.

SOURCE: Sellars, Roy Wood. "Epilogue on Berkeley," in Radical Currents in Contemporary Philosophy, edited by David H. DeGrood, Dale Riepe, & John Somerville (St. Louis, W. H. Green, 1971), pp. 55-66.

American Philosophy Study Guide
includes the following:

Reflections on American Philosophy From Within by Roy Wood Sellars

Some Reflections of Roy Wood Sellars

Principles of Emergent Realism: Philosophical Essays by Roy Wood Sellars

"The New Materialism" by Roy Wood Sellars

Philosophy for the Future: The Quest of Modern Materialism

“Critique of Idealistic Naturalism: Methodological Pollution in the Main Stream of American Philosophy”
by Dale Riepe

"Ex Nihilo Nihil Fit: Philosophy's 'Starting Point'" by David H. DeGrood

Heidegger on the Essence of Truth by Marvin Farber

“Marx and Critical Scientific Thought” by Mihailo Marković

Imperialism and Irrationalism by Herbert Aptheker

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

Positivism vs Life Philosophy (Lebensphilosophie) Study Guide

Marx and Marxism Web Guide

Emergence Blog