Russell’s The Problems of Philosophy John Perry March, 1997 [In Bertrand Russell, The Problems of Philosophy, New York: Oxford University Press, 1997, vii–xxvi.] There are many good introductions to philosophy, and many important philosophy books, but only a handful that are both; the book you have in your hands, The Problems of Philosophy by Bertrand Russell, is one of these, and one of the best. Russell is an important and fascinating figure, no doubt the most read, most honored, and most reviled English-speaking philosopher of the twentieth century. He was born in Wales in 1872 and died there in 1970. A member of a famous British family, Russell graduated from Trinity College Cambridge with first class honors in mathematics and moral sciences. His work in logic and philosophy was widely admired and honored; he was a Fellow and then a lecturer in Philosophy at Trinity. He was elected a Fellow of the Royal Society in 1908. He was awarded the Order of Merit in 1949 and the Nobel Prize for literature in 1950. Russell was not an ivory-tower academic; he was very involved in social policy and politics, especially after the outbreak of the First World War and during the cold war. Russell became a figure who was hated by millions of peoplewho knew little or nothing of his philosophical ideas—and admired by many others. Over the course of his 98 years he vigorously opposed, among other things, British involvement in WorldWar I, the Soviet Union under Lenin and Stalin, nuclear arms, and American involvement in the Viet-NamWar. In 1916 he was fired from Trinity College, Cambridge, and in 1918 he was put in jail, both in connection with his protests against the First World War. In 1940, as a result of public outrage and a taxpayer lawsuit, he lost a job at City College in New York before he teaching his first class. In 1961 he was jailed for his activities against nuclear arms. In the storm of protest surrounding his appointment to City College, Russell was untruthfully portrayed as a communist. But the most important objections to him seem to have been based on his views on marriage and morals. Russell defended marriage as “the best and most important relation that can exist between human beings” (Marriage and Morals, 115). But he saw nothing wrong in sexual relations before marriage, or in temporary childless liaisons between college students. (The widespread adoption Thanks to Andrew Irvine, David Israel, ElizabethMacken and Ed Zalta for comments on the penultimate draft. 1 2 of these practices in our own age should probably not be wholly credited to Russell’s influence.) Russell published this little book in 1912. At the time he was a Fellow of Trinity College at Cambridge University in England. With one of his Cambridge colleagues, G.E. Moore, he was leading a revolt against the philosophical movement in which he was trained, British idealism, inaugurating the movement known as “analytical philosophy.” If you are a student in a British or American university, many of your teachers would probably describe themselves as analytical philosophers; the movement Russell and Moore founded is alive and well after almost a century. In the years just before he published this book, Russell was just finishing Principia Mathematica, a three-volume work that, together with the works of Gottlob Frege, set the stage for modern logic. Russell wrote Principia with another Cambridge colleague, Alfred North Whitehead. If you take a logic course, whether in a philosophy, mathematics or computer science department, you will be will be studying some of Russell’s ideas, and using a notation only slightly evolved from that which he and Whitehead adopted. Russell’s logic and Russell’s philosophy were closely connected. Although The Problems of Philosophy contains no logical notation, it is built around some of the central ideas of Russell’s logic. By studying this book, then, you will not only receive a good introduction to many of the basic problems of philosophy, you will be introduced to themby the very words that helped inaugurate two of the most importantmovements in twentieth century philosophy. In this book Russell focuses on our knowledge of the physical world. He begins with some beliefs he has about the table in his room: that there is such a table, that is oblong brown and shiny, that the table he sees is the same he feels his arm resting on. Russell wants to know if he really has this knowledge and what sort of thing the table is. The answer he gives is that he does have this knowledge, and that the table is a material object, independent of his or anyone else’s mind. Russell’s main target is idealism, the view that “whatever exists, or at any rate whatever can be known to exist, must be in some sense mental” (36). A radical form of idealismis solipsism; all that exists ismy mind and its ideas; theworld is justmy dream. If one is a solipsist, one doesn’t have much motivation to share one’s view with others, so we seldom meet real solipsists in the literature of philosophy, but only philosophers like Russell who seek principled reasons for not being solipsists. We are more likely to meet advocates of another form of idealism, which had almost the status of common sense in the early parts of the century, at least in philosophical circles. According to this sort of idealism, space and time and tables and chairs and other physical objects are not just the figments of one person’s imagination, but they are nevertheless basically mental; consciousness pervades the whole universe, in a most unexpected way. One version of this view, the easiest to grasp, is theistic idealism, Bishop Berkeley’s view, that the world we know, including the physical world, is a system of God’s ideas. The God Berkeley had in mind was very much the Christian God, a personal creator and sustainer of the universe. Other forms of idealism, including the idealism of F.H. Bradley against which Russell and Moore revolted, conceive of a more impersonal 3 Absolute to provide an all-encompassing consciousness. In philosophy the term “realist” is applied to some philosophers to contrast them with other philosophers who deny some category of objects. One can be a realist about all sorts of things. Philosophers have denied the existence of minds, matter, numbers, space, and time and objective moral principles, to mention just a few things, so other philosophers, who defended the reality of these things, were realists by contrast. Russell was a realist on two key issues, universals and material objects. In both cases he was opposed to much of nineteenth century idealism. Universals are what particulars have in common. Suppose you are wearing a red sweatshirt and so am I. The sweatshirts are particulars. The properties they have in common are universals. Being red is a universal, and so is being a sweatshirt. Particulars are in one place at any given time; universals are not. Particulars instantiate or exemplify universals. The same universal can be exemplified by many particulars, in different places, at the same time. Philosophers have taken different attitudes towards universals over the centuries. Some, the realists, like Plato and Russell, are very enthusiastic about them. “Nominalists” think that all that particulars really have in common are the words we apply to them, like “red” and “sweatshirt”. “Conceptualists think that all that particulars have in common are the ideas we have of them.1 You own your sweatshirt, and I own mine. Between you and your sweatshirt, and me and my sweatshirt, there is another commonality, the common relation, owning. So owning is a universal, but a universal relation rather than a quality like redness. Relations appear to be very important. Among the most important are spatial relations (being next to, above, below, etc.), temporal relations (occurring before or after) and causation. Without these relations, it seems the world wouldn’t have much structure at all. Nevertheless many philosophers, including even some who were realists about qualities, denied the reality of relations. Russell thought that the denial of relations was one of the mistakes that led to idealism and that one thing that led to this mistake was an inadequate logic. His own logic treated relations on a par with other universals. The other important sense in which Russell was a realist concerned particulars. Idealists did not deny the whole category of particulars, and in fact would agree that tables, chairs, planets, and other physical objects were (in some sense) real. But they denied that there were any objects that were neither minds nor ideas, and didn’t depend on mentality of any sort for their existence—what we usually mean by “material objects”. In opposition to the idealists, Russell thought thatmaterial objects were real—although, as we shall see, he didn’t think we were acquainted with any of them. Two other important ingredients in Russell’s ontology are facts and propositions, both of which are, in some way, complexes, with universals and particulars as constituents. Facts constitute the way things are; particulars having qualities and standing in various relations to one another. Propositions are the possibilities we grasp in thought; they represent particulars as having qualities and standing in relations. True 1Still others think that in addition to the particular sweatshirts, there are particular rednesses; (G.E. Moore was destined to get into a big argument on this issue with G.F. Stout.) 4 propositions correspond to the facts. Russell says very little about facts in The Problems of Philosophy; he says more about propositions. He uses the concept of a proposition to elucidate some of the main principles of his view, but propositions bothered him, and in Chapter XII he seems to do away with them. He considers the example of Othello’s false belief that Desdomona loves Cassio. Rather than supposing that there is a proposition which serves as the object of Othello’s belief, he supposes that the belief has three objects, Desdomona, Cassio and the relation of loving. The belief is true if there is a complex unity, a fact, involving them in the right order, i.e. with Desdomona the subject and Cassio the object; if there is no such fact, it is false. This theory would allow Russell to do without propositions, but it was not easy to spell it out in detail. The beliefs that Desdomona loves Cassio, and the belief that Cassio loves Desdomona, for example, are quite different, but involve just the same objects. Russell distinguishes these by the “sense” of the believing, but it is not clear exactly how this works. We’ll set Russell’s late chapter qualms about propositions aside, and return to his central ideas. Let’s call the table in Russell’s study “A” and the chair next to it “B”. Assume that the facts are as follows: 1. A is a table 2. B is a chair 3. A is next to B The first two facts involve two particular objects and two properties that at least at first glance seem like qualities rather than relations: being a table and being a chair. The third fact involves two particulars and a spatial relation, being next to. A number of important philosophers, including Leibniz and nineteenth century British idealists against whom Moore and Russell were revolting, held that relational facts could not be basic. Russell’s rejection of this doctrine of “internal relations” was a very important turning point in the development of his philosophy. We are assuming each of 1-3 to be a fact. Facts for Russell have constituents, the objects, properties and relations that are involved in them. Fact 1 has A and the property of being a table as constituents; Fact 2 has B and the property of being a chair; fact 3 has A and B and the relation of being next to. Every fact or proposition must have at least one universal as a constituent. You can’t get a proposition just by piling up particulars. For example, Socrates is not a proposition. If we add Plato, we still do not have a proposition. Even ifwe add Aristotle we do not have one. But if we take Socrates and the property of being a philosopher, we have enough to make a proposition. So Russell believes in particulars, universals, including both qualities and relations, and propositions. And he believes that these are real things. By this he means that they are not to be identified with either mental or linguistic objects, although, as mentioned above, he was to retreat from this position in the case of propositions. 5 Now suppose that one had a list of all the facts. What all would it include? Surely, of all of the particulars, which qualities they had and in what relations they stood to other particulars. Would there be only positive facts (A is next to B), or would there also be negative facts (A is not on top of B)?Would there general facts (All tables have flat tops)? Or only singular facts (This table has a flat top, that table has a flat top, etc.)? Whatever answer we give to these questions about facts, the answers we must give concerning propositions are clear. Surely there must be negative and general propositions, for these are the kinds of things we can know and believe. One can believe, the proposition that every table has a flat top, for example, even if one is not acquainted with most of the tables in the world. What sort of constituents does this proposition have? It has no particulars as constituents. It has the universals being a table and having a flat top and the relations, among universals, of inclusion, signified by “every”. This relation obtains between universals A and B if everything that has A has B, the A’s are included in the B’s. The fact that one can have thoughts just involving universals is important for Russell’s account of a priori knowledge. Russell had a Platonic attitude towards universals; one could be acquainted with them, and know things about them intuitively, without being acquainted with any particulars that exemplified them. One of the big accomplishments in the developments in logic at the turn of the century, in which Russell played a large part, was the development of an account of words like “every” and “some”. Without the tools logic provided, one might think that “Every table has a flat top” expressed a proposition with an odd object, named “Every table”, as a constituent. It is very hard to say what this thing, Every table, might be, especially if one has to distinguish it from Some table, Each table, etc. By taking “Every” to stand for a higher order relation, we can avoid these ontological oddities and the problems that come with them. So there are as many propositions, or possibilities, in the world of possibilities as there are sequences of universals and objects that fall under them. But what is involved in being able to think those propositions—what is involved in having thoughts that correspond to the various possibilities? At this point Russell enunciates the fundamental principle of his epistemology: Every proposition which we can understand must be composed wholly of constituents with which we are acquainted (58). What does Russell mean by “be acquainted”? He has two things in mind, that pull in different directions. The word “acquainted” suggests the familiar distinction between people we have actually met, and those we have just heard about, and ideas based on this familiar distinction drive much of the discussion. But Russell also thinks of acquaintance as the fundamental relation that minds have to other objects. Since acquaintance is a relation between minds and other objects, those other objects must exist. This conception of acquaintance drives it away from the familiar one, as we shall see. When I see something red, assuming I am not color blind, I have the experience of redness; I am acquainted with the universal red. This allows me to think about redness, 6 even after I am no longer experiencing the red thing. And if I see one thing to the left of another, I have experienced “being to the left”, a relation, and so I can think thoughts involving that universal. What about particulars? At first pass, it seems we are acquainted with the particulars that we have experienced. So you are acquainted with your “acquaintances”; the people you have met and know; with the places and buildings you have seen; and so forth. Now we have the makings of a very simple account of thinking. We experience a variety of objects; universals of various types and particulars. We experience the particulars as having various properties and standing in various relations to one another. But once we have experienced these things, we can pull our ideas of them apart. If you have met Fred who is bald and Gertrude who wears a dress, you can think of Gertrude as bald and Fred as wearing a dress. If you have seen Fred sitting beside Elmer and Gertrude racing with Frances, then you can think of Fred sitting beside Frances, or Elmer racing with Fred, or any combination you want. So thinking is taking apart and putting together ideas that one has on the basis of experience. Let’s consider again what we are acquainted with. Suppose you have never met Fred, but you have talked to him on the phone. Are you acquainted with him? Of course in ordinary parlance we might say that you were, but are you in Russell’s sense? Or suppose you haven’t even talked to Fred, but you have read about him in the newspaper? Or suppose you have read about Fred not in the newspaper but in the history books, since he was dead long before you were born? Are you acquainted with him in those cases? Here the two things Russell has told us about acquaintance tug in somewhat different ways. I don’t feel, for example, that I am acquainted with Russell himself. Although our lives overlapped, I never laid eyes on him, never heard him speak in person. I have read a great deal about him, read a great deal by him, and seen photographs of him. But I never saw or talked to him in person. So it seems that I am not really acquainted with Russell. But I can think about Russell; I’d better be able to, since I am writing an essay about him! But on Russell’s principle of acquaintance I must be acquainted with Russell if I am able to think about him. So I must be acquainted with Russell after all. Russell solves this dilemma with his distinction between knowledge by acquaintance and knowledge by description. This distinction rests on an important idea from his logic, his theory of descriptions. In Russell’s logic, there are two kinds of terms that we use for particular objects, names and definite descriptions. “Bertrand Russell” is a name, “The author of The Problems of Philosophy” is a definite description. Let us say that names and descriptions both designate objects, so both “Bertrand Russell” and “The author of The Problems of Philosophy” designate the same person. But they do it in quite different ways. Names name, while descriptions denote. By using descriptions, we can think and say a lot of things about objects with which we are not acquainted, and even about objects that don’t exist. Someone might think, “the highest prime number must contain at least one zero,” even though there is no 7 highest prime number. Someone might think, “The PrimeMinister of the United States is probably a man,” even though the United States has no Prime Minister. Definite descriptions, then, pose a challenge to the principle of acquaintance. We saw earlier that by treating “all” and “some” as involving higher order relations among universals, Russell was able to avoid ontological oddities like the object Every man. Russell’s famous “theory of descriptions,” is the discovery that these same ideas allow one to treat definite descriptions in a very satisfying way that allows us to have thought about all sorts of things, without having to give up the principle of acquaintance, or countenance all sorts of weird things to have thoughts about. On Russell’s view, “The author of The Problems of Philosophy lived in Cambridge” says exactly the same thing as:  Someone authored The Problems of Philosophy,  Every person who authored The Problems of Philosophy is that person, and  that person lived in Cambridge. Now if we look at the whole, three-part, proposition, we see that Russell himself is not a constituent of it. We have the relation of authoring, the book The Problems of Philosophy, the relation of living in, the town of Cambridge. Then we have the “logical apparatus”; the relation of identity and two relation between universals, every and some. Russell isn’t a constituent at all.2 So here is a proposition that doesn’t have Russell as a constituent, but still is about him in some sense. Since Russell wrote The Problems of Philosophy, and is the only person that did (i.e., everyone who did is identical with him), then it is where he lives that will determine the truth of the proposition. As I think about Russell in writing this essay, and as you think about him in reading it, we know him by description, not by acquaintance. But note that in this essay I do not restrict myself to descriptions of Russell, I also use the name “Russell”. How can I do this, if to understand its meaning I have to be acquainted with Russell himself? Russell held the doctrine that often what appear to be names are really descriptions (54). He would maintain that, because I am not acquainted with him, when I use the name “Russell” in this essay, it is really functioning for me as a description—probably a very complex one involving many items I believe about Russell. Now notice one important feature of the theory of descriptions. Sometimes we have thought about things that don’t exist at all. For example, many people believe in Nessie, a monster that is supposed to live in Loch Ness in Scotland. So here is a proposition that we might want to discuss: Nessie is slimy. 2Cambridge might not be either, for “Cambridge” might be a hidden description; see below. 8 Let’s suppose that in fact the Loch Ness monster doesn’t exist. Still, some people might think that Nessie exists and is slimy. So there needs to be a proposition for them to think. But if there is going to be a proposition, it seems like there must be a constituent, Nessie, who in the proposition, if not in reality, is linked with the quality of being slimy. But then Nessie must exist after all. It is generally believed that earlier in his career, Russell accepted the view that we must give Nessie and other figments of myth and imagination some sort of quasiexistence, some sort of shadow reality. On this interpretation his views were not so different from Meinong, the German philosopher famous for his doctrine of non-existent objects. After he discovered his theory of descriptions, Russell eschewed such doctrines and thought Meinongians had an inadequately robust sense of reality. All we need to get the proposition one thinks when one thinks “Nessie is slimy” are universals and particulars that really do exist, such as sliminess and Loch Ness itself. “Nessie” is really a description, something like “The monster who lives in Loch Ness”, and our proposition comes to “Some monster lives in Loch Ness, only one monster lives in Loch Ness, and it is slimy”. Here then is a rule for when to apply the theory of descriptions. Whenever a person x is thinking a propositionP about an object a, and x’s experiences don’t guarantee that a exists, then x must be thinking about a descriptively. That is, the proposition P doesn’t have a as a constituent. We see that Russell adheres to this principle throughoutThe Problems of Philosophy.3 With this in mind let us turn to the Cartesian doubts that Russell considers in the beginning of the book. These doubts show that various of our experiences do not really guarantee the existence of the objects we think about. Thus when Russell concludes that The real table, if there is one, is not immediately known to us at all, but must be an inference from what is immediately known” (11). part of what he is concluding is that we don’t really entertain propositions that have the table as a constituent at all; we can only know it descriptively. Let us suppose then that Russell says, “A is a table,” or perhaps pointing to it, say “That is a table”. This is something that Russell knows. But just what is it, that he knows? Russell analyzes this situation as follows. He has a certain sensation. A sensation, according to Russell, is always an awareness of something. When he sees the table, he has a sensation of a brown oval patch. The brown oval patch is not the sensation, but what the sensation is of. This patch Russell calls a sense-datum. The distinction between sensations and sense-data is of paramount importance for Russell, for it is 3It is not so clear that he holds that we are acquainted with everything our experiences guarantee the existence of. The existence of Russell’s self himself seems clearly guaranteed by his experiences, but at times he isn’t completely sure that he is acquainted with his self. 9 right at this point that he think idealists like Berkeley make a mistake. Berkeley does not distinguish between sensation and sense-data, and so in effect takes sense-data to be mental. But this, Russell says, is a mistake. In describing the mistake (pp. 41ff) he uses the terminology “act” (for the sensation) and “object” (for the sense-data), which may be a bit confusing since sensations don’t seem very much like acts. The terminology fits better with cases of thinking, where the mind is more active. Suppose I am thinking about London. One clearly wants to distinguish between my thinking, which goes on insidemy head, and London, which is thousands of miles a way and much too large to fit in my head. They couldn’t be more different. It is much easier to confuse my sensation of a brownish patch, and the brownish patch, but on Russell’s view the mistake would be the same. The brownish patch, the sense-datum, is not my sensation, but what my sensation is a sensation of. The distinction between act and object, and its application to the case of sensation, was crucial to the philosophies of Russell and Moore at this time. Russell seems on the whole pretty confident about sense-data, while Moore’s discussion are more agonized. Do sense-data have backsides that we can’t see? Are they the surfaces of things? Sensedata have played an important role in philosophy in the twentieth century, inspiring, among other works, J.L. Austin’s Sense and Sensibilia, an unremitting attack on the arguments for sense-data as an abuse of ordinary language. For Russell, however, sense-data and some universals are beyond doubt, because we are acquainted with them (he is inclined to include the self, too, but isn’t quite sure). We know of other objects via these objects. The table is known by description, as the cause of the sense-data with which Russell is acquainted. That is, we take the sensedata to be signs of physical objects. And we take the features of our sense-data, to be signs of the features of the physical object. It is sense-data and their properties that are the constituents of all of our thinking about the physical world, for material objects, and, at least most of their properties, are known only by description. In asking which things he can and cannot question the existence of, Russell is using Descartes’s method of doubt. But the use he puts it to is quite different from the use Descartes puts it to. Descartes distinguished between the “light of nature”— Pure Reason— and our natural inclination to believe things. In the latter category is our belief in an external world, which our ideas represent. Descartes gave little credence to what we are inclined to believe, unless it was supported by derivations from indubitable principles disclosed by the Light of Nature. So, he argues, our belief in an external world is philosophically acceptable only when bolstered by an argument from indubitable principles for a benevolent, all-powerfulGod who doesn’t fool careful people. Supplying these derivations was the job of philosophy. Russell is more modest about philosophy, expects less from the Light of Nature, and is more generous to our natural inclinations. He thinks we can accept what we are inclined to believe if it can be shown to be consistent and coherent; the job of philosophy is to find and deal with apparent problems with what we are inclined to believe. In this spirit, Russell in The Problems of Philosophy accepts a world of material objects that we know only by description, as the best explanation we have of the data of our 10 senses. Russell is not by any means of slave to Common Sense, however. As we noted, the material world of The Problems of Philosophy involves particulars and universals that we know only by description. The temporal and spatial relations we are acquainted with, as holding amongst our sense-data, for example, are but signs of the spatial and temporal relations that obtain in the material world, and science discloses that there are a number of important differences. This view is not idealism, but it is not Common Sense either. One can divide Russell’s career as a realist into various phases; extreme realism, moderate realism, and constructive realism. Russell the extreme realist gave some sort of reality to everything we could think about. After 1905, armed with his theory of descriptions, Russell could moderate his realism; he could explain how we can think about Nessie without there being a real Nessie, for example. The theory of descriptions also gave him the equipment to distinguish between knowledge by acquaintance and knowledge by description, and arrive at the picture of the structure of our knowledge expressed in The Problems of Philosophy. In his later works, however, Russell would try to dispense with physical or material objects as basic bits of the furniture of the world. He would attempt to construct them out of sense-data, rather than take them to be inferred as the best explanations of sense data. Russell pursued this philosophy of logical construction in his next book on epistemology, Our Knowledge of the External World. Russell’s philosophy of logical constructions was very influential, inspiring such twentieth century classics as Rudolf Carnap’s The Logical Construction of the World and Nelson Goodman’s The Structure of Appearance. On the whole, however, the picture of The Problems of Philosophy fits better with the temper of much of analytical philosophy as the century draws to a close. For philosophers of a moderately realistic temper, worried by the resurgence of holism and even forms of idealism in our own age, the main doctrines of The Problems of Philosophy are well worth re-examining. Suggested Reading Works by Russell Our Knowledge of the External World (Chicago and London: Open Court, 1914). The next phase in Russell’s epistemology. Principles of Social Reconstruction (London: George Allen and Unwin, 1916). This was reprinted in America under the title Why Men Fight. It contains some of the ideas that got Russell in trouble during the First WorldWar. Introduction to Mathematical Philosophy (London: George Allen and Unwin, 1919). Russell’s own introduction to his logical and mathematical ideas, written while he was in jail. 11 Logic and Knowledge, edited by Robert Marsh (London: George Allen and Unwin, 1956). This includes many of Russell’s important essays, including “On Denoting”. Marriage and Morals (London: George Allen and Unwin, 1929). Some of the ideas that cost Russell his job at New York City College. A History of Western Philosophy (New York: Simon and Schuster, 1945). Russell wrote this book to show the influence of history on political and social ideas and movements. While it is not a very dependable source of information on the philosophers Russell doesn’t like, this readable book contains many interesting interpretations of past figures and insights about Russell’s own philosophical development. My Philosophical Development (London: George Allen and Unwin, 1959). The Autobiography of Bertrand Russell; 3 vols. (London: George Allen and Unwin, 1967, 1968, 1969). Why I am Not A Christian and Other Essays on Religion and Related Subjects (London: George Allen and Unwin, 1957). A collection of some of Russell’s non-technical essays. Essays in Analysis (London: George Allen and Unwin, 1973). A collection of some of Russell’s most important technical essays. Works about Russell There are many books and articles about Russell. These two excellent Encyclopedia articles are a good place to begin: Paul Edwards, William P. Alston and A.N. Prior, “Russell, Bertrand Arthur William”. In Paul Edwards (ed.) The Encyclopedia of Philosophy (New York: Macmillan & The Free Press, 1967). Andrew Irvine, “Russell” in The Stanford Encyclopedia of Philosophy. (This new Encyclopedia is online: http://plato.stanford.edu. Irvine’s article contains links to a number of other web sites about Russell).