Our Knowledge of the External World by Bertrand Russell

First published: 1914

Type of Philosophy: Epistemology, metaphysics

Context

The significance of Our Knowledge of the External World is that it proposed a new method for philosophizing. Although the method suggested was not strictly novel (it had been previously used by such mathematical logicians as George Boole, Giuseppe Peano, and Gottlob Frege), it was modified by Bertrand Russell and transferred from mathematical to philosophical subject matter. Russell called it the “logical-analytic” method or the method of “logical atomism.” Since the publication of these lectures, the method has been taken over and radically modified and broadened by a large school of philosophers who call themselves “analysts” and who constitute an important group in the modern philosophical world.

The analytic method, as Russell formulated it, is less ambitious than that of classical philosophy in that it does not claim to determine the nature of reality or the universe as a whole. What it does is both less speculative and less sweeping, but more scientific. It uses the techniques of modern logic and modern mathematics and employs such concepts as class, relation, and order for the purpose of clarifying and solving some of the perennial problems of philosophy that have not yet yielded to satisfactory solution. There are many such problems, but Russell here considers those concerned with the nature of the external world, with how the world of physics is related to the world of ordinary sense experience, and with what is meant by space, time, continuity, infinity, and causation. The book consists of illustrations of the application of the logical-analytic method to these problems for the purpose of showing its fruitfulness. Russell insists that his results are to be taken as tentative and incomplete, but he believes that if any modification in his method is found necessary, this will be discovered by the use of the very method that he is advocating.

The Faults of Traditional Logic

The philosophies of two typical representatives of the classical school—F. H. Bradley and Henri Bergson—are first examined in order to show the errors that these men made. Bradley found the world of everyday life to be full of contradictions, and he concluded that it must be a world of appearance only, not of reality. His error lay in his attempt to determine the character of the world by pure reasoning rather than by going to experience and examining what he found. Bergson believed that reality is characterized fundamentally by growth and change; he then concluded that logic, mathematics, and physics are too static to represent such a world and that a special method called “intuition” must be employed. His mistake was twofold: First, he supposed that because life is marked by change and evolution, the universe as a whole must be so described; he failed to realize that philosophy is general and does not draw conclusions on the basis of any of the special sciences. Second, his emphasis on life suggests that he believed philosophy to be concerned with problems of human destiny. However, such is not the case; philosophy is concerned with knowledge for its own sake, not with making people happier.

One reason why the logical-analytic method was not more widely employed earlier, according to Russell, is that it was new and only gradually replaced some of the earlier and erroneous conceptions of logic. Traditional (syllogistic) logic had been quite generally abandoned as inadequate. The inductive logic of Francis Bacon and John Stuart Mill had been shown to be unsatisfactory because it cannot really show why people believe in such uniformities as that the sun will rise every morning. Belief in universal causation cannot be a priori because it is very complicated when formulated precisely; nor can it be a postulate, for then it would be incapable of justifying any inference; nor can it be proved inductively without assuming the very principle that one is trying to prove. Russell claims that Georg Wilhelm Friedrich Hegel made the mistake of confusing logic with metaphysics, and that only mathematical logic provides the tool by which one may hope to solve philosophical problems.

The Logical-Analytical Method

Mathematical logic is both a branch of mathematics and a logic that is specially applicable to mathematics. Its main feature is its formal character, its independence of all specific subject matter. Looked at formally, it can be said to be concerned with propositions. The main types of propositions are atomic propositions, such as “Socrates is a man”; molecular propositions, which are atomic propositions unless connected by and, if, or unless; and general propositions, such as “All men are mortal.” In addition, there must be some knowledge of general truths that is not derived from empirical evidence. For example, if we are to know that all men are mortal, we must also know that the men we have examined are all the men there are. This we can never derive from experience because empirical evidence gives us only particular truths. Thus, there are certain general truths that, if they are known, must be either self-evident or inferred from other general truths.

It is obvious that one of the oldest problems of philosophy is the problem concerning knowledge of the external world. In applying the logical-analytical method to this problem, Russell finds that what we have to begin with (something that is always vague, complex, and inexact) is knowledge of two kinds: First, our acquaintance with the particular objects of everyday life—furniture, houses, people, and so on; and second, our knowledge of the laws of logic. These may be called “hard” data because doubt in such cases, while it could occur, normally would not, and if it did, would probably be considered pathological. Most of us would be willing to add to these data certain facts of memory, some introspective facts, spatial and temporal relations, and facts of comparison, such as the likeness or unlikeness of two shades of color.

However, Russell argues that certain very common beliefs probably should be excluded from our data. One is the belief that objects persist when we are not perceiving them. Another is the belief in the existence of other minds and of things outside our experience that are revealed by history and geography. These should be called “soft” data because they can be doubted and are actually derived from our belief in hard data. The problem, then, is to determine whether the existence of anything other than our own hard data can be inferred from these data.

In formulating the question by asking whether we can know of the existence of anything independent of ourselves, Russell finds that the terms “independent” and “ourselves” are too vague to determine whatever solution we happen to be seeking. What we can show, however, is that the appearances of an object—its color, shape, and size—change as we move toward it or away from it, or around it. Russell calls these appearances “sense-data,” and he then defines a “private world” as an object seen from a certain perspective. If we move around an object we shall discover that the sense-data change and that correlated with these changes are the bodily sensations associated with our movements. Then, if we assume the existence of other minds, we can correlate their private worlds and perspectives with ours and define an object as the class of all of its perspectives. This is much safer than inferring a thing that exhibits itself in these perspectives. The perspectives are certainly real and constitute hard data; the thing is an inference and may be mistaken. A class is merely a logical construction, and we have thus substituted a logical construction for an inferred entity. This explanation fits the facts, has no empirical evidence against it, and is free from logical contradictions. Russell thus regarded himself as having solved the problem of how we know an external world.

Sense-Data and Physics

Another philosophical problem that yields to Rusell when his logical-analytic method is applied to it is that of the relation between the world of physics and the world of sense-data. According to physics, the world consists of material bodies, located in space and time and having a high degree of rigidity and permanence. Although the bodies themselves may change, they are made up of particles that are themselves unchanging and indestructible. On the other hand, the world of sense-data for any one of us comes and goes; even such permanent things as mountains become data for us only when we see them; they are not known with certainty to exist at other moments. The problem of “getting these worlds together” involves the attempt to construct things, a single space and a single time out of the fleeting data of experience.

Permanent things can be constructed if we can find some way of connecting what we commonly call “appearances of the same thing.” Although some sort of continuity among appearances is necessary, it is not sufficient to define a thing. What is needed in addition is that the appearances of a single thing obey certain physical laws that the appearances of different things do not.

In much the same way, the dimensionless points (no one has ever seen a perfect point) of physics must be constructed out of the surfaces and volumes of our sensory experience. Using the fact that spaces can be observed to enclose other spaces (like Chinese boxes), Russell constructs “enclosure-series,” which can be called “point-producers” because points can be defined by means of them. Russell does not, however, define a point as the lower limit of an enclosure-series, for there may be no such limit. Instead he defines it as the series itself; thus, a point is a logical entity constructed out of the immediate data of experience.

Time and Continuity

It can now be seen how Russell derives the concept of durationless instants of time. Events of our experience are not instantaneous; they occupy a finite time. Furthermore, different events may overlap in time because one event may begin before the other, but there may be a common time during which they both occur. If we restrict ourselves to the time that is common to more and more overlapping events, we get durations that are shorter and shorter. Then we can define an instant of time as the class of all events that are simultaneous with one another. To state accurately when an event happens, we need only to specify the class of events that defines the instant of its happening.

Russell recalls that since the days of Zeno, philosophers have wondered about the problem of permanence and change, and especially about the apparent paradox of motion. Some motions, such as that of the second hand on a watch, seem to be continuous because we can actually see the movement; other motions, such as that of the hour hand, seem to be discontinuous because we can observe only a broken series of positions. Russell’s solution to this apparent contradiction in the nature of motion lies in the mathematical theory of continuity.

Continuity is regarded as, first, a property of a series of elements, of terms arranged in an order, such as numbers arranged in order of magnitude. Second, a continuous series must in every case be “compact,” which means that no two terms are consecutive and that between any two terms there is always another. If space is thought of as being made up of a continuous series of points, and time as a similar series of instants, then motion will consist of a correlation between a series of different points in space and a series of different instants in time. However, Russell cautions that we must not say that a body moves from one point in space to the next, for if the series is truly compact, there is no next point in space, no next instant in time. This takes care of the difficulty we experience in thinking of a continuous motion as consisting of a very rapid series of jumps from one position to another.

Bergson argued that motion is simple, indivisible, and not validly analyzable into a series of states. In replying to him, Russell argues that on grounds of physiology, psychology, and logic there is no incompatibility between the mathematical theory of continuity and the evidence of the senses. From physiology, we learn that sensations do not cease instantaneously but gradually die away. Thus when we see a rapid motion, we do not perceive it as a series of jerks because each sensation blends into its successor. From psychology, we learn that there are actually cases of sense-data that form compact series. If a blindfolded person is holding a weight and to this a small additional weight is added, he may not observe the added weight if it is small enough. If still another weight is added, he may still not notice it. However, if both weights had been added at once, he might have detected the difference. This proves that weights form a continuous series. Finally, on grounds of logic, Bergson must be mistaken. For if analysis always falsifies, there could never be two facts concerning the same thing, and this would make motion, which depends on the distinction between earlier and later, impossible.

Zeno’s Paradoxes

After a discussion of German philosopher Immanuel Kant’s conception of infinity, Russell turns to an analysis of Zeno of Elea’s famous paradoxes. Zeno had four arguments designed to show the impossibility of motion—all based on his theory of the infinite divisibility of space and time. His first argument was designed to prove that a runner cannot get to the end of a race course, for if space is infinitely divisible, he will have to cover an infinite number of points in a finite time, and this is impossible. According to Russell, Zeno’s error came from thinking that the points must be covered “one by one.” This would, apparently, require an infinitely long time, for the series has infinitely many elements. However, the series actually has a limit of 1, and this specifies the instant when the runner will complete the race.

Zeno’s second argument attempts to show that Achilles can never overtake the tortoise if he gives it a head start, for by the time Achilles reaches the point where the tortoise was at any given time, the tortoise will always be slightly ahead. Achilles must then make up this distance, but again the tortoise will be slightly in advance. Achilles gets always closer to the tortoise but never is able to pass him. Russell claims that Zeno’s error here is the same as in the previous case—an infinity of instants does not necessitate an infinitely long time.

Zeno’s third argument begins with the claim that whatever occupies a space equal to itself must be at rest. An arrow in flight always occupies a space equal to itself; hence, it cannot move. Here Zeno supposes that the arrow could not move through an instant because this would require that an instant have parts, and this cannot be. However, Russell points out that the mathematical theory of infinity has shown that, because there are no consecutive points or instants, an instant can have parts. When this is realized, Zeno’s difficulty disappears.

Zeno’s last argument attempts to prove that “half the time may be equal to double the time.” This is illustrated by a column of soldiers passing another column at rest and a third column marching in the opposite direction. In the first case, the soldiers pass only half as many soldiers as in the second case, yet these actions both take place during the same time. The error again disappears for Russell with the recognition that there is no “fastest” motion that requires that slow motion have intervals for rest interspersed.

Infinite and Finite Numbers

According to Russell’s analysis, infinite numbers differ from finite numbers in two ways: First, infinite numbers are reflexive, while finite numbers are not, and second, finite numbers are inductive, while infinite numbers are not. A number is reflexive when it is not increased by adding 1 to it. Therefore, the number of members in the class 0, 1, 2, 3 . . . n is the same as the number in the class 1, 2, 3, 4 . . . n plus 1. Even more surprising, the number of even numbers is equal to the total number of numbers (even and odd). It is therefore possible in the case of infinite numbers for a part of a class to equal the whole class.

Infinite numbers are also noninductive. Inductive numbers possess a certain property that can be called “hereditary.” This is like the property of being named “Jones,” which, if it applies to a man in many Western societies, also applies to all of his male descendants. Every property that belongs to 0 and is hereditary belongs to all of the natural numbers, and the natural numbers can be defined as those that are inductive. To all such numbers, proof by mathematical induction applies.

Philosophers have tried to define numbers, but without much success. Russell refers to the mathematician Frege, who recognized that numbers are neither physical nor mental but logical—that is, numbers are properties of general terms or general descriptions and thus are applicable to classes, not to things. The author asks: “When do two classes have the same number of members?” and answers “When a one-to-one correlation can be set up between them.” This applies both to finite classes, which can be counted, and to infinite classes, which cannot. As an example, Russell refers to monogamous countries where the number of husbands is the same as the number of wives, though we may not know how many of both there are because the relation of marriage is a correlating relation. The number of a class can then be defined as the class of all classes that are similar to it—that is, which have a one-to-one correlation with it. Note that number is not a property of a class but is a class of classes. This is a valid conception even if classes are considered to be fictions, for there is a method for translating statements in which classes occur to statements in which they do not.

Causation and Free Will

Russell’s final application of the logical-analytic method is to the problem of causation and free will. Causation is a complicated relation that must be described by a carefully worded formula: “Whenever things occur in certain relations to each other (among which their time-relations must be included), then a thing having a fixed relation to these things will occur at a date fixed relatively to their dates.” The evidence for causation in the past is the observation of repeated uniformities, together with knowledge of the fact that where there appears to be an exception to such uniformities, we can always find wider ones that will include both the successes and failures. The only guarantee we could have that such a causal law would continue to hold in the future would be some sort of a priori principle. The law of causality is therefore an ideal law, possibly true, but not known to be true.

Determinism can be demonstrated, Russell believes, only if we can show that human actions are theoretically predictable from a sufficient number of antecedents. This we cannot do, he claims. In the great majority of cases, volitions probably have causes, but the only evidence for this is the fact of repeated instances. Furthermore, because causes do not compel their effects, the will is free in the sense that our volitions are the results of our desires, and we are not forced to will something that we would rather not will.

Principal Ideas Advanced

•The method of logical analysis makes the resolution of philosophical problems possible by defining the limits of scientific philosophy so as to exclude speculative metaphysics.

•We can account for our knowledge of the external world by realizing, through logical analysis, that the world as we know it is a construction from the data given in sense experience; an individual’s “private world” is the class of all data within his or her perspective, and a perceived object is the class of all aspects to be found in all the perspectives that include the object.

•The conception of permanent things can be constructed by reference to appearances if points are defined by reference to an enclosure-series of spaces, and if time is defined by reference to classes of events simultaneous with each other.

•Zeno’s paradoxes of motion can be resolved by use of the mathematical theory of continuity.

Bibliography

Dejnozka, Jan. Bertrand Russell on Modality and Logical Relevance. Avebury Series on Philosophy. Aldershot: Ashgate, 1999. This work presents a criticism and interpretation of modality and logical relevance in the work of Bertrand Russell. Includes index.

Grayling, A. C. Russell. Past Masters series. Oxford: Oxford University Press, 1996. This work, part of a series on great thinkers, covers the life and accomplishments of Russell. Includes an index.

Irvine, A. D., ed. Bertrand Russell: Critical Assessments. London: Routledge, 1999. This book critically examines the life and work of Russell, including his philosophy.

Landini, Gregory. Russell’s Hidden Substitutional Theory. New York: Oxford University Press, 1998. This work takes a closer look at one of Russell’s logical theories. Includes index.

Monk, Ray. Bertrand Russell: The Spirit of Solitude. London: J. Cape, 1996. This biography of Russell examines the philosopher’s life and works. Includes bibliographical references and index.

Monk, Ray. Bertrand Russell: The Ghost of Madness 1921-1970. London: Jonathon Cape, 2001. Volume 2 of Monk’s thorough biography, covering the troubled mature years of the philosopher.

Monk, Ray. Russell. New York: Routledge, 1999. An excellent biographical introduction to the thoughts of the philosopher, clearly presented and requiring no special background. Bibliography.

Monk, Ray, and Anthony Palmer, eds. Bertrand Russell and the Origins of Analytical Philosophy. Bristol, England: Thoemmes Press, 1996. This collection of essays looks at Bertrand Russell and analytical philosophy. Includes bibliographical references.

Pampapathy Rao, A. Understanding Principia and Tractatus: Russell and Wittgenstein Revisited. San Francisco: International Scholars Publications, 1998. This work compares the philosophies and beliefs of Ludwig Wittgenstein and Russell.

Roberts, George W., ed. Bertrand Russell Memorial Volume. Atlantic Highlands, N.J.: Humanities Press, 1979. Essays from notable scholars focusing on the philosophical achievements of Russell.

Saintsbury, Richard Mark. Russell. London: Routledge and Kegan Paul, 1979. This text serves a twofold purpose as an analysis of Russell’s ideas and as a critique of their adequacy.

Tait, Katharine. My Father, Bertrand Russell. New York: Harcourt Brace Jovanovich, 1975. Tait describes her father’s flaws as a parent. This text offers a sensitive angle on the intellectual life of Russell and brings his social and educational theories under the critical lens of practice.

Vellacott, Jo. Bertrand Russell and the Pacifists in the First World War. New York: St. Martin’s Press, 1981. This book documents the origins of the pacifist movement in Great Britain, focusing on its socialist and Quaker roots.