Analytic Philosophy

also called Linguistic philosophy, a movement, dominant in Anglo-U.S. philosophy in the 20th century, distinguished by its method, which has focused upon language and the analysis of the concepts expressed by it.

The methods that have dominated British philosophy for most of the 20th century and American philosophy since somewhat more recently have been called Linguistic and Analytic because language and the analysis of the concepts expressed by language have been a central concern. Though Australia and the Scandinavian countries have also contributed to this movement, it has won a very limited following elsewhere. Although there is a unity of outlook in the tradition, individual philosophers and movements within it have differed, often radically, about the goals and methodology of philosophy. A leading figure prior to the mid-20th century, Ludwig Wittgenstein, an Austrian-born Cambridge philosopher, for example, may have been unique in the history of philosophy in having engaged in two periods of profoundly influential philosophical productivity of which the later work was in large part a renunciation and a sustained argument against the earlier. Yet both the early Wittgenstein (represented by his Tractatus Logico-Philosophicus, 1922) and the later (represented by the Philosophical Investigations, 1953) are central examples of Analytic philosophy.

Moreover, the aims assigned to the philosophical study of language have often been different. Some philosophers, among them Bertrand Russell and the early Wittgenstein, have thought that the underlying structure of language mirrors that of the world - that from an analysis of language a philosopher can grasp important truths about reality. This so-called picture theory of language, though influential, is generally repudiated by current Analytic philosophers. Another important dispute concerns whether everyday language is defective, vague, misleading, and even, at times, contradictory. Some Analytic philosophers have thus proposed the construction of an “ideal” language: precise, free of ambiguity, and clear in structure. The general model for such a language has been symbolic logic, the growth of which in the 20th century has played a central role in Analytic philosophy. An ideal language, it was thought, would resolve many traditional philosophical disputes that have arisen from the misleading structure of natural languages. At the other pole, some philosophers have thought that many philosophic problems have come from paying too little attention to what men say in everyday language about various situations.

Despite such disagreements, Analytic philosophers have much in common. Most of them, for example, have concentrated on particular philosophical problems, such as that of induction, or have examined specific concepts, such as those of memory or of personal identity, without attempting to construct any grand metaphysical schemes - an attitude that has roots as ancient as those of the Socratic method exemplified in Plato's dialogues. Almost invariably Plato began with specific questions such as “What is knowledge?” or “What is justice?” and pursued them in a way that can be viewed, without undue strain, as philosophical analysis in the modern sense.

Ideally, a philosophical analysis illuminates some important concept and helps to answer philosophical questions involving the concept. A famous example of such analysis is contained in Bertrand Russell's theory of definite descriptions. In a simple subject–predicate statement such as “Socrates is wise,” he said, there seems to be something referred to (Socrates) and something said about it (that he is wise). If, instead of a proper name, however, a “definite description” is substituted, as in the statement “The president of the United States is wise,” there is apparently still something referred to and something said about it. But a problem arises when nothing fits the description, as in the statement “The present king of France is wise.” Though there is apparently nothing for the statement to be about, one nevertheless understands what it says. Consequently, a pre-World War I philosopher, Alexius Meinong, celebrated for his Gegenstandstheorie (“theory of objects”), felt forced by such examples to distinguish between things that have real existence and things that have some other sort of existence; for such statements could not be understood unless they were about something.

In Russell's view, philosophers such as Meinong were misled by surface grammatical form into thinking that such statements are simple subject–predicate statements. In reality they are complex; in fact, an analysis of the foregoing example shows that the definite description, “the present king of France,” is not an independent unit in the statement at all. Upon analysis, the statement is a complex conjunction of statements: (1) “There is a present king of France”; (2) “There is at most one present king of France”; and (3) “If anyone is a present king of France, he is wise.” But, more importantly, each of the three components is a general statement and is not about anything or anyone in particular. There is no phrase in the complete analysis equivalent to “the present king of France,” which shows that the phrase is not an expression, like a proper name, that refers to something as the thing that the whole statement talks about. There is no need, therefore, to make Meinong's distinction between things that have real existence and things that have some other kind of existence.

General viewpoint of Analytic philosophy

Nature, role, and method of analysis

Analytic philosophy is concerned with the close and careful examination of concepts.

Status of philosophy in the Empiricist tradition

In spirit and style Analytic philosophy has strong ties with the Empiricist tradition, which stresses the data received through the senses and which, except for brief periods, has characterized British philosophy for some centuries, distinguishing it from the more Rationalistic trends of continental European philosophy. It is not surprising, therefore, that Analytic philosophy should find its home mainly in the Anglo-Saxon countries. In fact, the beginning of modern Analytic philosophy is generally dated from the time when two of its major figures, Bertrand Russell and G.E. Moore, both Cambridge philosophers, rebelled against an anti-Empiricist Idealism that had temporarily captured the English philosophical scene. The most famous British Empiricists - John Locke, George Berkeley, David Hume, and John Stuart Mill - had many interests, doctrines, and methods in common with contemporary Analytic philosophers. Although many of their particular doctrines are favorite targets of attack by Analytic philosophers today, one feels that this is more the result of a common interest in certain problems than any difference in general philosophical outlook.

Most Empiricists, though admitting that the senses fail to yield the certainty requisite for knowledge, hold nonetheless that it is only through observation and experimentation that justified beliefs about the world can be gained; i.e., a priori reasoning from self-evident premises cannot reveal how the world is. This view has resulted in a sharp dichotomy among the sciences: between the physical sciences, which ultimately must verify their theories by observation, and the deductive or a priori sciences - e.g., mathematics and logic - the method of which is the deduction of theorems from given axioms. Thus, the deductive sciences cannot give justified beliefs, much less knowledge, of the world. This consequence was one of the cornerstones of two important movements within Analytic philosophy, logical atomism and Logical Positivism. In the Positivist's view, for example, the theorems of mathematics are merely the result of working out the consequences of the conventions that have been adopted for the use of its symbols.

The question then arises whether philosophy itself is to be assimilated to empirical or to a priori sciences. Early Empiricists assimilated philosophy to the Empirical sciences. They were less self-reflective about its methods than contemporary Analytic philosophers are. Being preoccupied with epistemology (theory of knowledge) and the philosophy of mind, and holding that fundamental facts can be learned about these subjects from individual introspection, they took their work to be a kind of introspective psychology. Analytic philosophers in the 20th century, on the other hand, have been less inclined to appeal ultimately to direct introspection. Moreover, the development of rigorous methods in formal logic seemed to promise help in solving philosophical problems - and logic is as a priori as a science can be. It seemed, then, that philosophy must be classed with mathematics and logic.

Conceptual, linguistic, and scientific analysis

The question remained, however, what philosophy's function and methodology are. For a great many Analytic philosophers who do philosophy in the minute and meticulous manner of G.E. Moore and, in particular, for those who have made Oxford the centre of Analytic philosophy (see below, Later trends in England: Oxford philosophers), its business is the analysis of concepts. For them, philosophy is an a priori discipline because the philosopher in some sense already possesses the concept in which he is interested and needs no observations in order to analyze it.

Philosophy can be seen either as conceptual or as linguistic analysis. In the analysis of the concept of seeing, for example, the philosopher is not expressing purely linguistic concerns - with, say, the English verb “to see” - though an investigation of what can be said using that verb may be relevant to his conclusions. For a concept is independent of any particular languages; a concept is something that all languages, insofar as they are capable of expressing the concept, have in common. Thus, philosophers who stress that it is concepts that they analyze attempt to rebut the charge that their problems and solutions are merely verbal.

In contrast, other Analytic philosophers have been concerned with how expressions are used in a particular, nontechnical, everyday language. Thus, the term ordinary language philosophy has been applied by critics as a term of opprobrium to such philosophers. An influential study, The Concept of Mind (1949), by Gilbert Ryle, a prominent Oxford Analyst, is an example of a work that some critics took to depend in large part on a trivial appeal to how English speakers talk; but many of Ryle's arguments could equally well have been given by Analytic philosophers who would look upon the term ordinary language with horror.

The problem of perception illustrates how Analytic philosophers who do conceptual analysis think of the goal of philosophy as both different from and complementary to science. Physiologists, psychologists, and physicists - through experiments, observations, and testable theories - have also contributed to man's understanding of perception. There is in the sciences, however, a strong tendency to advance beyond earlier positions, which seems to be absent from philosophy. In philosophy, for example, the account of perception given by such 20th-century Analytic philosophers as G.E. Moore and the Positivist A.J. Ayer has a close connection with that of Locke in the 17th century.

The difference between philosophy and science is that, whereas the scientist investigates an actual occurrence, such as seeing, the philosopher investigates a concept that he already possesses quite independently of what he might discover through the occurrence. Whereas the scientist begins by supposing that he can recognize examples of seeing and is already exercising the concept, the philosopher wants to know what is involved in seeing in the sense of what conditions one can use to classify cases as examples of seeing. He may want to know, for example, whether certain conditions are necessary or sufficient. In testing the philosophical theory that, for an observer to see an object, the object must cause a visual experience in him (the causal theory of perception), one does not set up a scientific experiment. It would be of no use to set up situations in which various physical objects are not causing any visual experiences in order to see whether they still can be seen. For if the theory is correct, no such experimental situation will be an instance of seeing; and if it is wrong, merely describing a hypothetical situation would suffice. The question is one about how situations are classified, and for that purpose hypothetical situations are as good as real ones.

Therapeutic function of analysis

For some philosophers in the Analytic tradition, especially those influenced by Wittgenstein, the analysis of concepts has therapeutic value beyond the intrinsic enjoyment of doing it. Even scientists and laymen in their philosophical moments generate problems by not understanding the proper analyses of the concepts that they employ. They are then tempted to formulate theories to explain these difficulties, when instead they should be sorting out the roles of the concepts, which would show them that there was no problem to begin with. Thus, the failure to see how psychological concepts - sensations, emotions, and desires - are employed has led philosophers to such problems as how one can know what is going on in another's mind or how desires and emotions can produce physical changes in the body, and vice versa. Analysis of the concepts involved would, in this way of looking at philosophy, “dissolve” rather than solve the problems, for philosophers would come to see that their formulations of the problem rest on mistakes about the concepts involved.

This way of looking at philosophy has often been criticized as making it merely a clearing up of the confusions of other philosophers and therefore a sterile enterprise. The confusions, however, need not be only those of other philosophers. Scientists, for example, can also generate philosophical theories that affect how they design their experiments, which may, thus, be subjects for philosophical therapeutics. Behaviorism in psychology - which views emotions, desires, and attitudes as being dispositions to behave in certain ways - seems to be a philosophical theory and perhaps to be based on a confusion about the analysis of psychological concepts. Yet Behaviorism has influenced psychologists in their approach to the science. Thus, in this view, philosophy can have a therapeutic value beyond the sphere of philosophical games.

Philosophy, in spite of its abstractness, has traditionally been concerned with human needs, and the therapeutic model may even fulfill this ideal. Laymen, as well as philosophers, for example, are bothered by the thought that their actions are determined not by themselves but by prior conditions. This is a problem that, if the therapeutic view is correct, rests on the misunderstanding of such concepts as causation, responsibility, and action, which need clarification.

Formal versus ordinary language

The role of language as a central concern of Analytic philosophers is the dimension most involved in disputes about the methodology employed. Philosophers outside the Analytic movement tend to think that its preoccupation with language is a departure from philosophy as classically conceived. Yet Plato and Aristotle, medieval philosophers, the Empiricists - and, in fact, most of the philosophers whose works have been considered important - have found it essential to talk about language. There are serious differences, however, about what role language should play. One such difference concerns the importance of formal languages (in the sense employed in symbolic logic) for philosophical problems.

Development of mathematical logic

Since the time of Aristotle, logic has been allied to philosophy. Until the late 19th century, however, logic was largely confined to formulating elaborate rules for one fairly simple form of argument - the syllogism; and there was a lack of systematic development of the subject along lines that had been taken in mathematics since early times.

Almost from the beginning, mathematicians had rigorously exploited two important techniques: (1) the use of the axiomatic method (as in Euclid's geometry) in developing the subject; and (2) the use of schematic letters or variables for stating general truths in the subject (thus, one can write “A + B = B + A”, in which any names or numbers whatsoever can be substituted for A and B, and the result will still be true).

It is surprising that logicians through the ages failed to grasp the power of the use of schematic letters. When they finally began to employ these and other mathematical techniques, they made great contributions to man's understanding of the subject.

Among the developments that occurred in the 19th century, primarily through the work of mathematicians, those of the Englishman George Boole, creator of Boolean algebra, and of Georg Cantor, the Russian-born creator of set theory, are especially important inasmuch as they gave promise of bringing logic and mathematics closer together. The one figure who was both a mathematician and a philosopher and so might be credited with the marriage of logic as a philosophical subject with the techniques of mathematics was Gottlob Frege (died 1925), of the University of Jena in Germany. Historically, Frege, whose works are now appreciated in their own right, was important principally for his influence on Bertrand Russell, whose monumental work, Principia Mathematica (1910–13), written in collaboration with Alfred North Whitehead, together with Russell's earlier Principles of Mathematics (1903), awakened philosophers to the fact that the use of mathematical techniques in logic might prove to be of great importance for philosophy. Its symbolism had the advantage of being closely connected with ordinary language, whereas its rules can be precisely formulated. Moreover, work in symbolic logic has produced many distinctions and techniques that can be applied to ordinary language.

Divergence of ordinary language from formal logic

Ordinary language, however, seems to differ from the artificial language of symbolic logic in more respects than its lack of precisely stated rules. On the surface, it often appears to violate the rules of symbolic logic. In the English statement “If this is gold [symbolized by p], then this will dissolve in aqua regia [symbolized by q],” for example, which in symbolic logic is expressed in a form known as the material conditional, p ⊃ q (in which ⊃ means “If . . . then . . . ”), one of the rules is that the statement is true whenever “This is gold” is false. In ordinary language, on the contrary, one would not count the statement as true merely on formal logical grounds but only if there were some real connection in the world of chemical reactions between being gold and dissolving in aqua regia - a connection that plays no role in symbolic logic.

Among Analytic philosophers the existence of many such apparent divergences between symbolic logic and ordinary language has generated attitudes ranging from complete mistrust of symbolic logic as relevant to non-artificial languages to the position that ordinary language is not a proper vehicle for the rigorous statement of scientific truths.

Interpretations of the relation of logic to language

Symbolic logic has been viewed by many Analytic philosophers as providing the framework for an ideal or perfect language. This statement can be taken in two ways:

1. Russell and the early Wittgenstein thought of logic as revealing, in a precise fashion, the real structure of any language. Any seeming departure from this structure in ordinary language must therefore be attributed to the fact that its surface grammar fails to reveal its real structure and is apt to be misleading. As a corollary, philosophers who have held this view have often explained philosophical problems as arising from being taken in by the surface features of the language. Because of the similarity of sentences such as “Tigers bite” and “Tigers exist,” for example, the verb “to exist” may seem to function, as other verbs do, to predicate something of the subject. It may seem, then, that existence is a property of tigers just as their biting is. In symbolic logic, however, the symbolic equivalent of the two sentences would be quite different; existence would not be represented by a symbol for a predicate but by what is called the existential quantifier, (∃ x), which means “There exists at least one x such that...”

2. The other sense in which symbolic logic has been seen as the framework of an ideal language is exemplified in the work of Rudolf Carnap, a 20th-century semanticist, who was concerned with what the best language - especially the best for the purposes of science - is.

One distinctive feature of the formal language of Principia Mathematica is that it becomes, when interpreted, a language of true-or-false statements. In ordinary language, on the contrary, one is not restricted to statements of truths; in it one can also issue commands, ask questions, make promises, express beliefs, give permission, and assert necessities and possibilities. Consequently, many philosophers have developed nonstandard logics that incorporate the non-assertoric features of language. Thus, various systems of logic have been formulated and studied (see logic).

On the other side of the coin, many philosophers - most notably the later Wittgenstein and those influenced by him - have thought that attempting to put language into the straitjacket of a formal system is to falsify the way that language works. Language performs a multitude of tasks, and even among expressions that seem to be alike in the way they function - those sentences, for example, that one might think are used simply for expressing facts - examination of their actual use reveals many differences: differences, for instance, in what is counted as showing them to be true or false and in their relationships to other parts of language. Formal systems, according to this view, at best oversimplify and at worst can lead to philosophical problems generated by supposing that all language operates strictly according to a simple set of rules. Accordingly, far from settling philosophical disputes by getting underneath the misleading exterior of ordinary language, formal systems add their own share of confusion.

Early History of Analytic Philosophy

Reaction against Idealism

During the last decades of the 19th century, English philosophy was dominated by an absolute Idealism that stemmed from the German philosopher G.W.F. Hegel. For English philosophy this represented a break in an almost solid tradition of Empiricism. The seeds of modern Analytic philosophy were sown when two of the most important figures in its history, Bertrand Russell and G.E. Moore, broke with Idealism at the turn of the 20th century.

Absolute Idealism was avowedly metaphysical in the sense that its adherents thought of themselves as describing, in a way not open to scientists, certain very fundamental truths about the world. Indeed, what pass for truths in the sciences, were, in their view, not really truths at all; for the scientist must, perforce, treat the world as composed of distinct objects and can only describe and state the relationships supposedly holding among them. But the Idealists held that to talk about reality as if it were a multiplicity of objects is to falsify it; in the end only the whole, the absolute, has reality.

In their conclusions and, most importantly, in their methodology, the Idealists were decidedly not on the side of commonsense intuition. Thus, a Cambridge philosopher, J.M.E. McTaggart, argued that the concept of time is inconsistent and cannot therefore be exemplified in reality. British Empiricism, on the other hand, had always thought of common sense as an ally and science as the model of the way in which to find out about the world. Even when their views might seem out of step with common sense, the Empiricists were generally concerned to reconcile the two.

One can hardly claim that Analytic philosophers have universally thought of themselves as on the side of common sense and much less that metaphysical conclusions (on the ultimate nature of reality) are absent from their writings. But there is in the history of the Analytic movement a strong anti-metaphysical strain, and its exponents have generally assumed that the methods of science and of everyday life are the authentic ways of finding out the truth.

Founding fathers: Moore and Russell

The first break from the Idealist view that the physical world is really only a world of appearances occurred when Moore, in a paper, “The Nature of Judgment” (1899), argued for a theory of truth that implies that the physical world has the independent existence that, apart from philosophical theories, it is naively supposed to have. Though the theory was soon abandoned, it did represent a return to common sense.

The influences on Russell and Moore - and thus their methods of dealing with problems - soon diverged, and their different approaches became the roots of two broadly different methodologies in the Analytic tradition.

Russell was a major influence on those who approached philosophical problems armed with the technical equipment of formal logic, who saw the physical sciences as the only means of gaining knowledge of the world, and who regarded philosophy - if a science at all - as a deductive and a priori enterprise on a par with mathematics. Russell's contributions to this side of the Analytic tradition have been important and, in great part, lasting.

Moore, on the other hand, never found much need to employ technical tools nor to turn philosophy into a science. His dominant themes were (1) the defense of commonsensical views about the nature of the world against esoteric, skeptical, or grandly metaphysical views and (2) the conviction that the right way to approach philosophical puzzles is to ask exactly what the question is that generated the puzzle before trying to solve it. Philosophical problems, he thought, are often intractable because philosophers have not stopped to formulate precisely what is at issue.

G.E. Moore

Because of these two themes, Moore enlisted much more sympathy among Analytic philosophers from the 1930s and onward who were followers of Wittgenstein's later writings, of Gilbert Ryle's postwar The Concept of Mind, and of John Austin's work (see below Oxford philosophers). These philosophers, like Moore, saw little hope in advanced formal logic as a means of solving traditional philosophical problems and believed that philosophical skepticism about the existence of an independent external world or of other minds - or, in general, about what men label as common sense - must be wrong. The followers of Wittgenstein also shared with Moore the belief that it is often more important to look at the questions that philosophers pose than at their proposed answers. Thus, unlike Russell, who was important for his solutions in formal logic and ideal models of language, it was more the spirit of Moore's conception of philosophy than its lasting contributions that makes him a seminal influence.

The Idealists were given to arguing for what, in Moore's eyes, were outrageous positions. Thus, in his essay “A Defence of Common Sense” (1925), as in others, his defense was not only against such Idealist doctrines as the unreality of time but also against any of the forms of skepticism - about the existence of other minds or of a material world - that philosophers have espoused. The skeptic, he pointed out, usually has some argument for his conclusion. Instead of examining such arguments, however, Moore pitted against the skeptic's premises quite everyday beliefs, such as, for example, that he had breakfast that morning (thus time cannot be unreal) or that he does in fact have a pencil in his hand (thus there must be a material world). His challenge to the skeptic is to show that the premises of the skeptic's argument are more certain than the everyday beliefs that form Moore's premises.

Although some commentators have seen Moore as an early practitioner of the appeal to “ordinary language,” his appeal was really not to what it is proper to say but rather to the beliefs of common sense. His rejection of anything that offends against common sense, however, was influential not only in the release that it afforded from the metaphysical excesses of absolute Idealism but also in its impact on the continuing attitudes of most Analytic philosophers - even though they may have given it a linguistic turn.

Moore was also important for his vision of the proper business of philosophy - analysis. He was puzzled about what is the proper analysis of “X sees Y,” in which Y designates a physical object (e.g., a pencil). There must be a special sense of “see,” in which one does not see the pencil but only part of its surface. And finally - and most importantly - there is also a sense in which what is directly perceived is not even the surface of the pencil but, rather, what Moore called “sense data” and which earlier Empiricists had called “visual sensations” or “sense impressions.” Moore's problem was to discern the relationships among these various elements in perception and, in particular, to discover how a person can be justified, as Moore fully believed he is, in his claims to see physical objects when what he immediately perceives are really sense data. The idea that sense impressions form the immediate objects of perception has played a large role in Analytic philosophy, showing once again its Empiricist roots. It later became an important source of division, however, among the Logical Positivists. Most post-World War II Oxford philosophers, however, together with those closely influenced by Wittgenstein's later work, have found sense data to be as unpalatable and unwarranted as Moore had found McTaggert's doctrine of the unreality of time to be.

Bertrand Russell

One of the recurring themes in philosophy is the idea that the subject needs to be given a new methodology. Among Empiricists this has often meant making it more scientific. From an early date, Russell enunciated this viewpoint (which was not shared by Moore), finding in the techniques of symbolic logic a measure of reassurance that philosophy might be put on a new basis. Russell did not see the philosopher, however, as merely a logician. Symbolic logic might provide the framework for a perfect language, but the content of that language is something else. The job of the philosopher is - for Russell, as it was for Moore - analysis. But the purpose is somewhat different. In most of Russell's work, analysis has the task of uncovering the necessary assumptions - especially about the kinds of things that exist - for a description of the world as it is. For the most part this description is the one that science gives and is therefore realistic. Thus, Russell's use of analysis was openly metaphysical.

The question then arises of how philosophical analysis, which is concerned with how men talk about the world, can presume to give any answers about how the world is. The search for an answer begins with the above-mentioned theory of descriptions - a theory that seems to be closely tied to linguistic concerns. It will be recalled that Russell considered that such definite descriptions as “the author of ‘On Denoting’” are not really expressions used to refer to things in the world but that, instead, they make the statements in which they occur into quite general propositions about the world, to the effect that one and only one thing of a certain sort exists and that it has a certain property. Because there must be some way, however, of directly speaking of the things in the world, Russell turned his attention to proper names. The name Aristotle, for example, does not seem to carry any descriptive content. But Russell argues, on the contrary, that ordinary names are really concealed definite descriptions (“Aristotle” may simply mean “The student of Plato who taught Alexander, wrote the Metaphysics, etc.”). If a name had no descriptive content, one could not sensibly ask about the existence of its bearer, for one could then not understand what is expressed by a statement involving it. If “Bosco” were a name in this sense (without any descriptive content), then merely to understand the statement that Bosco exists or the statement that Bosco does not exist presupposes that one already knows what the name Bosco refers to. But then there cannot be any genuine question about Bosco's existence, for just to understand the question one must know the thing to which the name refers. Ordinary proper names, however - Russell, Homer, Aristotle, and Santa Claus - as Russell pointed out, are such that it makes sense to question the existence of their bearers. Thus, ordinary names must be concealed descriptions and cannot be the means of directly referring to the particular things in the world.

Names in the strict logical sense, then, are very rare; Russell, in fact, suggests that in English the only possible candidates are the demonstrative pronouns, this and that. Yet, if men are ever to talk about the actual things in the world directly, there must be the possibility of such demonstrative expressions underlying their language - in their private thoughts about the world if not in their public language.

To this point, Russell had concluded that things in the world can be talked about only through the medium of a special kind of name; in particular, one about which no question can arise whether it names something or not. At this point there was a transition from questions about the nature of language to results about the nature of the world. Russell asked what sort of thing it is that can be named in the strict logical sense, that can be known and talked about, and that can tell a man something about the world. The important restriction is that no question can arise about whether it exists or not. Ordinary physical objects and other people seem not to fit this requirement.

In his search for something whose existence cannot be questioned, Russell hit upon present experience and, in particular, upon sense data: one can question whether he is really seeing some physical object - whether, for example, there is a desk before him - but a person cannot question that he has had visual impressions or sense data; thus, what a man can name in the strict logical sense and what things he can actually talk about turn out to be the elements of his present experience. Russell therefore made a distinction between what can be known by acquaintance and what can be known only by description; i.e., between those things the existence of which cannot be doubted and those about which, at least theoretically, doubt can be raised. What is novel about Russell's conclusion is that it was arrived at from a fairly technical analysis of language: to be directly acquainted with something is to be in a position to give it a name in the strict logical sense, and to know something only by description is to know only that something uniquely fits the description.

Russell was not constant in his view about physical objects. At one point he thought that the observer must infer their existence as the best hypothesis to explain his experience. Later he argued that they could be taken as logical constructions out of sense data.

Logical atomism: Russell and the early Wittgenstein

The next important development in Analytic philosophy was initiated when Russell published a series of articles entitled “Philosophy of Logical Atomism” (1918–19), in which he acknowledged a debt to Wittgenstein, who had studied with Russell before the war. Wittgenstein's own work, the Tractatus Logico-Philosophicus (1922), which can also justly be said to present a logical atomism, turned out to be not only tremendously influential on developments in Analytic philosophy but also such a deep and difficult text that it has generated a growing body of scholarly interpretation.

Russell's choice of the words logical atomism to describe this viewpoint was, in fact, particularly apt. By using the word logical Russell meant to sustain the position, described earlier, that through analysis - particularly with the aid of the ideal structure provided by symbolic logic - the fundamental truths about how any language functions can be revealed and that this disclosure, in turn, would show the fundamental structure of that which the language is used to describe. And by using the word atomism Russell highlighted the particulate nature of the results that his analyses and those of Wittgenstein seemed to yield.

On the linguistic level, the atoms in question are atomic propositions, the simplest statements that it is possible to make about the world; and on the level of what language talks about, the atoms are the simplest atomic facts, those expressible by atomic propositions. More complex propositions, called molecular propositions, can then be built up out of atomic propositions via logical connectives such as “either... or...,” “both... and...,” and “not...” - the truth-value of the molecular proposition being in each case a function of the truth values of its component atomic propositions.

Language, then, must break down, upon analysis, into ultimate elements that cannot be analyzed into any other component propositions; and, insofar as language mirrors reality, the world must then be composed of facts that are utterly simple. Atomic propositions are composed, however, of strings of names understood, as Russell had explained it, in the strict logical sense; and atomic facts are composed of simple objects, the things that could be thus named.

The details of the Russell–Wittgenstein view have fascinated philosophers by the way in which they not only formed a coherent view but also seemed to follow inexorably from the central assumptions. There are close connections between this period, which was perhaps the most metaphysical in contemporary Analytic philosophy, and traditional Empiricism. The breakdown of language and the world into atomic elements had been one of the prominent features in the classical Empiricists, John Locke, George Berkeley, and David Hume. There was also a view of the connection between language and the world - adumbrated in Russell but fully evident in the Tractatus - which has been important and influential, viz., the picture theory, which holds that the structure of language mirrors that of the world. Analysis is important because ordinary language does not show immediately, for example, that it is founded on the atomic-molecular proposition model. Another theme is that the deductive sciences - mathematics and logic - are based solely on the way that language operates and cannot reveal any truths about the world, not even about a world of entities called numbers. Finally, logical atomism, in Wittgenstein's thought as opposed to Russell's, was at one and the same time metaphysical - in the sense of conveying via pure reasoning something about how the world is - and anti-metaphysical. Wittgenstein's Tractatus is unique in the history of Empiricism in its acceptance of the fact that it is itself a metaphysic and that part of its metaphysics is that metaphysics is impossible: the Tractatus says of itself that what it says cannot be coherently said. Only empirical science can tell a man anything about the world as it is. Yet the Tractatus apparently tells him, for example, about the relationship between language and the facts of the world. For Wittgenstein, the solution of this seeming paradox lies in his distinction between what can be said and what can only be shown. There are certain things that can somehow be seen to be so - in particular, the ways in which language is connected with the world. The Tractatus could not straightforwardly tell its readers about these matters - metaphysics cannot be a body of facts expressible in any language - but the attempt to say these things, done in the right way, can show them what it cannot coherently express.

Logical Positivism: Carnap and Schlick

Wittgenstein's Tractatus was both a landmark in the history of contemporary Analytic philosophy and perhaps its most aberrant example. It not only contained the most highly sophisticated metaphysics but also was an important influence on the most anti-metaphysical of the positions taken by Analytic philosophers, viz., that of Logical Positivism, which was mainly developed by a group of philosophers, scientists, and logicians who were centred in Vienna and came to be known as the Vienna Circle. Among these, Rudolf Carnap and Moritz Schlick have perhaps had the most influence on Anglo-American philosophy, although it was an English philosopher, A.J. Ayer - whose Language, Truth and Logic (1936) is still the most widely read work of the movement in America and England - who introduced the ideas of Logical Positivism to English philosophy. Its main tenets have struck sympathetic chords in the Analytic philosophers and are still important today, even if in repudiation.

Above all else, Logical Positivism was anti-metaphysical; nothing can be learned about the world, it held, except through the methods of the empirical sciences. The Positivists sought a method for showing both (1) when a theory that seemed to be about the world was really metaphysical and (2) that such a theory was, in fact, meaningless, and this they found in the principle of verification. In its positive form, the principle said that the meaning of any statement that is really about the world is given by the methods employed for verifying its truth or falsity - the only allowable methods being, ultimately, those of observation and experiment. In its negative form, the principle said that no statement could both be a statement about the world and have no method of verification attached to it. Its negative form was the weapon used against metaphysics and for the vindication of science as the only possible source of knowledge about the world. The principle would, thus, class as meaningless many philosophical and religious theories that purport to say something about the world but provide no way of testing the truth of the statements; for example, in religion it would render suspect the statement that God exists, which, being metaphysical, would be, strictly speaking, meaningless.

The principle of verification ran almost immediately into difficulties, most of which were first raised by the Positivists themselves. The attempt to work out these difficulties belongs to a more detailed study of the movement. It is sufficient to note here that these problems were sufficient to make most subsequent Analytic philosophers wary of appealing directly to the principle. It has, however, influenced philosophical work in more subtle ways.

With the principle of verification in hand, the Positivists thought that they could show a great many theories to be nonsense. There were several areas of discourse, however, which failed the test of the principle but which were simply impossible to rule out as concealed nonsense. Foremost among these disciplines were mathematics and ethics. Mathematics (and logic) could hardly be written off as nonsense. Yet their theorems are not verifiable by observation and experiment; they are known, in fact, by pure a priori reasoning alone. The answer seemed to be provided in Wittgenstein's Tractatus, which held that the propositions of mathematics and logic are, in Kantian terms, analytic; i.e., true-like the statement “All bachelors are unmarried” - in virtue of the conventions that lie behind the use of the symbols involved.

About ethics or, more precisely, about any statements involving value judgments, the Positivist view was different, yet still of lasting importance. In this view, value judgments are not, like mathematical truths, necessary adjuncts to science. But they cannot be put off as nonsense; nor, obviously, are they true by definition or linguistic convention. The usual view of the Positivists, called emotivism, is that what look like statements of fact (e.g., that one should not tell lies) are really expressions of one's feelings toward a certain action; thus, value judgments are not really true or false. The Positivist's position was that neither mathematical nor ethical statements could be dismissed, as were metaphysical propositions. Both had then to be exempted from the principle of verification; and this was done by arguing that their statements are not really about the world: mathematical truths are conventions, and ethical statements are merely expressions of feelings. The divorce of ethics from science, once again, reflects an old Empiricist theme, to be seen, for example, in David Hume's dictum that from matters of fact one cannot derive a conclusion about what ought to be nor vice versa.

Later history of the movement

“Philosophical Investigations”: the later Wittgenstein

A crucial turn that initiated developments that were destined to have a lasting and profound effect on much of contemporary Analytic philosophy occurred in 1929, when Wittgenstein, after some years in Austria during which he was not philosophically very active, returned to England and established his residence at Cambridge. There, the direction of his thought soon shifted radically away from his Tractatus, and his views became in many ways diametrically opposed to those of logical atomism. Because he published none of the materials of this period, his influence on other English philosophers - and ultimately on those in all of the countries associated with Analytic philosophy - spread by way of his students and those who heard him in the small groups to whom he spoke at Cambridge. His style, too, changed from the semi-rigorous and formally organized propositions of the Tractatus to sets of loosely connected paragraphs and remarks in which the ideas are often conveyed more by suggestion and example than discursively. The result has been that one of the major splits within the ranks of Analytic philosophy is that between those who derive their methods from the later Wittgenstein and those who have followed the Tractatus.

Although Wittgenstein's thoughts ranged over almost the entire field of philosophy, from the philosophy of mathematics to ethics and aesthetics, their impact has been felt most, perhaps, where it has concerned the nature of language and the relationship between the mental and the physical.

Language and following rules

In logical atomism, as shown above, language was conceived as having a certain necessary and fairly simple underlying structure that it was the job of philosophy to expose. Wittgenstein began to tear away at this assumption. Language, he now thought, is like an instrument that can be used for an indefinite number of purposes. Hence, any effort to codify how it must operate by giving some small set of rules would be like supposing that there is some rigid necessity that a screwdriver (for instance) can be used only to drive screws and forgetting that screwdrivers are also, quite successfully, used to open jars and to jimmy windows. Language is a human institution that is not bound by an outside set of rules - only by what men consider to be correct and incorrect. And that, in turn, is not really a matter for a priori theories to consider.

The notion of a rule and what it means to follow a rule was especially prominent in his writings. Several concerns made this point of particular interest to Wittgenstein. In mathematics and logic, emphasis was being placed on the rules for manipulating the symbolism. As has been seen, symbolic logic has also been a model for the underlying structure of language. If this fact is coupled with the fact that Russell and the Wittgenstein of the Tractatus saw language as reflecting these rules and with the general Empiricist tradition that explains how language operates by each person following internal rules and standards for the use of his words, the picture of the system that Wittgenstein thought mistaken then emerges, and it becomes clear why he placed the notion of a rule so centrally.

Natural languages, however, are significantly different in that one does not first learn the rules and then use the language; indeed, prior to learning the language, one would not know what to do with rules. Mathematics and logic are, in this sense, bad models for language because they aim at setting out beforehand the rules and principles that are subsequently to be used. They encourage the belief that language must have a rigid structure and that, without rules, no language would be possible. The “rules” that one might plausibly discern in the language that one speaks are not, as rules, already there, in a ghostly way, guiding what one says; they are either generalizations from the finite data of what is counted as correct or incorrect, or they are rules that, as Wittgenstein metaphorically expressed it, one puts away in the archives - one adopts the rule but only after the fact.

Following a rule, however, was a concept that Wittgenstein saw as wrongly analyzed in many classical views about language. Thus, he cast irrevocable doubt on the prevalent theory - typified best, perhaps, in John Locke's Essay Concerning Human Understanding (1690) - that to use an expression meaningfully is to have in one's mind a standard or a rule for applying it correctly. Against this theme, Wittgenstein's point was that a rule by itself is dead - it is like a ruler in the hands of someone who has never learned to use it, a mere stick of wood. Rules cannot compel nor even guide a person unless he knows how to use them; and the same is true about mental images, which have often been thought to provide the standard for using linguistic expressions. But if rules themselves do not give life to words but require a similar explanation for what gives them life, then there is a useless regress and no (philosophical) explanatory value in the whole apparatus of internal rules and standards.

Relation between mental and physical events

In some respects, Wittgenstein made some significant breaks with the Empiricist tradition - in his views about language and the explanation of the rigour of the deductive sciences. His treatment of the relationship between mental events and physical events also represents an important departure. Empiricists generally have started from the important assumption that what a person is immediately acquainted with is his own sensations, ideas, and volitions, and that these are mental and not physical; and, most importantly, that the things he knows immediately are essentially private and inaccessible to others. For both Moore and Russell there then arose the problem of how, in view of the privacy stressed by the sense-datum theory, the world of physical objects could be known. Wittgenstein's attack on this viewpoint, which has come to be known as “the private language” argument, has become well known, partly because it was in this area that Wittgenstein presented what could most easily be picked out as a more or less formal argument - one that could then be analyzed and criticized in an analytic manner. Even in this case, however, his style of writing was such that his precise formulation of the argument has become a main source of controversy. Wittgenstein argued that the notion of an utterly private experience would imply: (1) that what goes on in the mental life of a person could be talked about only in a language that that person alone whose mental life it was could understand; (2) that such a private language would be no language at all (this has been the main source of controversy); and (3) that the widely held doctrine that there are absolutely private mental events cannot be intelligibly stated, because to do so would be to suppose that one can publicly say something about what the doctrine itself says cannot be mentioned in a language accessible to more than one person.

The fact that Wittgenstein's argument against private language depends essentially on the question, “What is it to follow a rule?” illustrates a common characteristic of his writings, viz., that themes developed in one area of philosophy continually emerge in apparently quite divorced areas. His extraordinary ability to see a common source of difficulty in philosophical problems that seem to be unrelated helps to explain his style of writing, which seems at first sight to be a somewhat chaotic arrangement of ideas.

Analytic philosophy has also been attracted to a behaviouristic view of mental phenomena that holds that such apparently private events as the feeling of fear are not only not really private but also that they can be identified with publicly observable patterns of behaviour. The disposition toward empirical science, with observation as its foundation, united with the observation that the evidence men have of what goes on in the mental lives of other people must come from what they see of their behaviour, has often warred against the other inclination of Empiricism to regard the starting point of all knowledge of the world, for each person, as being essentially private sense experience. Wittgenstein has had tremendous influence, however, in suggesting that these two extremes are not the only alternatives. Yet attempts to state how Wittgenstein could deny the privacy of experience without espousing some form of behaviourism have not been very successful. Sympathetic interpreters have taken up the notion of “criteria,” used, but not developed in any detail, by Wittgenstein. For mental states such as fear, outward behaviour (e.g., running away, blanching, or cringing) does not constitute what it is to be in that state, as behaviourism would have it, but neither is it merely evidence of some completely private event. The problem has been to characterize the relation between behaviour and mental states so that the two are neither identical nor evidence one for the other, while still acknowledging that a knowledge of the person's characteristic behaviour is essential to understanding the notion of a certain mental state.

Later trends in England

Those philosophers who might fairly be labelled “Wittgensteinians,” who follow the methods that Wittgenstein employed in his later period, should be distinguished from those who have been influenced more indirectly by the general trends and philosophical atmosphere that arose in large part from Wittgenstein's work.

Wittgensteinians

Close students of his ideas have tended to work chiefly on particular concepts that lie at the core of traditional philosophical problems. As an example of such an investigation, a monograph entitled Intention (1957), by G.E.M. Anscombe, an editor of Wittgenstein's posthumous works, may be cited as an extended study of what it is for a person to intend to do something and of what the relationship is between his intention and the actions that he performs. This work has occupied a central place in a growing literature about human actions, which in turn has influenced views about the nature of psychology, of the social sciences, and of ethics. And, as an extension of this British influence into the United States, one of Wittgenstein's students, Norman Malcolm of Cornell University, has investigated such concepts as knowledge, certainty, memory, and dreaming. As these topics suggest, Wittgensteinians have tended to concentrate on Wittgenstein's ideas about the nature of mental concepts and to work in the area of philosophical psychology. Typically, they begin with classical philosophical theories and attack them by arguing that they employ some key concept, such as that of knowledge, in a manner incongruous with the way in which the concept would actually be employed in various situations. Their works thus abound with descriptions of hypothetical, though usually homely, situations and with questions of the form, “What would a person say if...?” or “Would one call this a case of X?” In doing so, they are following out Wittgenstein's advice that, instead of trying to capture the essence of a concept by an abstract analysis, the philosopher should look at how it is employed in a variety of situations.

Oxford philosophers

After World War II, Oxford University was the centre of extraordinary philosophical activity; and, although Wittgenstein's general outlook on philosophy - his turning away, for example, from the notion of formal methods in philosophical analysis - was an important ingredient, many of the Oxford philosophers could not be called Wittgensteinians in the strict sense. The method employed by many of these philosophers has often been characterized - especially by critics - as an “appeal to ordinary language,” and they were thus identified as belonging to the school of “ordinary language” philosophy. Exactly what this form of argument is supposed to be and what exemplifies it in the writings of these philosophers has been by no means clear. Gilbert Ryle, Moore's successor as editor of a leading journal, Mind - and especially in his The Concept of Mind - was among the most prominent of those analysts who were regarded as using ordinary language as a philosophical tool. Ryle, like Wittgenstein, pointed out the mistake of regarding the mind as what he called “a ghost in a machine” - to defeat the radical dualism of mind and body that has characterized much of philosophical thinking - by investigating how people employ a variety of concepts, such as memory, perception, and imagination, that designate “mental” properties. He tried to show that, when philosophers carry out such investigations, they find that, roughly speaking, it is the way people act and behave that leads to attributing these properties to them, and that there is no involvement of anything internally private. He also attempted to show how philosophers have come to dualistic conclusions - usually from having a wrong model in terms of which to interpret human activities. A dualistic model may be constructed, for example, by wrongly supposing that an intelligently behaving person must be continually utilizing knowledge of facts - knowledge that something is the case. Ryle contended, on the contrary, that much intelligent behaviour is a matter of knowing how to do something and that, once this fact is acknowledged, there is no temptation to explain the behaviour by looking for a private internal knowledge of facts. Though Ryle's objectives were similar to those of Wittgenstein, his results have often seemed more behavioristic than Wittgenstein's.

It is true that Ryle did ask, in pursuit of his method, some fairly detailed questions about when a person would say, for example, that someone had been imagining something; but it is by no means clear that he was appealing to ordinary language in the sense that his was an investigation into how, say, speakers of English use certain expressions. In any case, the charge, often voiced by critics, that this style of philosophizing trivializes and perverts philosophy from its traditional function would probably also have to be levelled against Aristotle, who frequently appealed to “what we would say.”

A powerful philosophical figure among postwar Oxford philosophers was John Austin, who was White's professor of moral philosophy until his death, in 1960. Austin felt that many philosophical theories derive their plausibility from overlooking distinctions - often very fine - between different uses of expressions, and he also thought that philosophers too frequently think that any one of a number of expressions will do just as well for their purposes. (Thus, ignoring the difference between an illusion and a delusion, for example, gives credence to the view that what one immediately perceives are not physical objects, but sense data.) Austin's work was, in many respects, much closer to the ideal of philosophy as comprising the analysis of concepts than was that of Ryle or Wittgenstein. He was also much more concerned with the nature of language itself and with general theories of how it functions. This novel approach, as exemplified in How to Do Things with Words (1962), set a trend that has been followed out in a growing literature in the philosophy of language. Austin took the total speech act as the starting point of analysis, which allowed him to make distinctions based not only upon words and their place in a language but also upon such points as the speaker's intentions in making the utterance and its expected effect on the audience. There was also in Austin's approach something of the program of Russell and the early Wittgenstein for laying bare the fundamental structure of language.

Trends in the United States

Although the Oxford philosophers and the posthumous publication of Wittgenstein's writings have produced a revolution in Anglo-American philosophy, the branch of Analytic philosophy that emphasized formal analyses by means of modern logic has by no means been dormant. Since the appearance of Principia Mathematica, striking new findings have emerged in logic, many of which, though requiring for their understanding a high level of mathematical sophistication, are nevertheless important for philosophy.

Among those philosophers for whom symbolic logic occupies a central position, W.V.O. Quine, Pierce professor of philosophy at Harvard University, has been especially important. Symbolic logic represented for him, as it did for many earlier Analytic philosophers, the framework for the language of science. There were two important themes in his work, however, that represent significant departures from, say, the positions of the logical atomists and the Logical Positivists. In the first place, Quine rejected the distinction between those statements in which their truth or falsity depends upon the meaning of the terms involved and those in which their truth or falsity is a matter of empirical and observable fact - a distinction that had played an essential role in Logical Positivism and was thought by most Empiricists to be the basis for a division between the deductive and the empirical sciences. Quine, in “Two Dogmas of Empiricism” (1951) and subsequent writings, argued that the sort of distinction intended by philosophers is impossible to draw. In the course of his argument, a similar doubt was cast upon concepts traditional not only to philosophy but also to linguistics - in particular, the concept of synonymy or sameness of meaning. Quine's attack has been a threat not only to some long-held doctrines of the Analytic tradition but also to its conception of the nature of philosophy, which has generally depended upon contrasting it with the empirical sciences.

The second important departure of Quine's philosophy has been his attempt to show that science can be successfully conducted without what he calls “intentional entities.” In contrast to “extensional,” used above as an essential feature of standard symbolic logic, intentional entities include many of the common items that Analytic philosophers often assume that they can talk about without difficulty, such as the meanings of expressions, propositions, or the property of certain statements (such as those of mathematics) of being necessarily true. Quine's program - as exemplified by Word and Object (1960) - is intended in part to show that science can say everything that it needs to say without using concepts that cannot be expressed in the extensional language of standard logic. Quine's work, though by no means widely accepted, has made Analytic philosophers at least wary of uncritically accepting certain of their standard distinctions.

Since the mid-20th century, there has been an interaction between the science of linguistics and Analytic philosophy. This did not occur before because Analytic philosophers had almost always considered their study of language to be a priori and unconcerned with empirical facts about particular languages. However, a book by Noam Chomsky, a U.S. generative grammarian, entitled Syntactic Structures (1957), has produced a theory of grammar that not only has profoundly affected the course of linguistics but also bears striking resemblances to philosophical analysis. At first, some Analytic philosophers saw in Chomsky's theory a technique that could be applied to philosophy. It was subsequently considered, however, that, whereas the possibility of looking at grammar in Chomsky's way had contributed valuable concepts for philosophers, the possibility that it would become a methodology for Analytic philosophy had receded. The interchange between linguists and philosophers, however, has continued.

Analytic philosophy today

It is not possible to forecast in any detail the future trends of Analytic philosophy in Anglo-American and Scandinavian countries. It seems relatively certain, however, that the two conceptions of the subject that stem from Moore and Russell will both continue.

Analytic philosophers, mainly influenced by Oxford philosophy, and those for whom symbolic logic is a touchstone analyze many of the same problems and benefit from each other's work. Analysis in the more rigorous sense that Russell's theory of definite descriptions represents is more frequently an aim, despite the doubts of Wittgenstein and many of the Oxford philosophers. The general idea that the only ultimate explanations of the world are the scientific ones and the usual corollary that philosophy is in the service of science - which was a central idea for Russell, for the Logical Positivists, and (in recent times) for Quine - has apparently lost nothing of its vigour. The opposing tendencies, noted above, among Empiricists in general, and present also in Analytic philosophy, toward behaviourism or Materialism, on the one hand, and toward an Idealism of a phenomenalistic sort (such as that of the Irish bishop George Berkeley), on the other, are not present in the same form - mainly because of the sustained criticisms of Wittgenstein, of his followers, and of the Oxford philosophers. The battleground has shifted to a more subtle level. A substantial number of Analytic philosophers who are styled Materialists or physicalists have proposed a novel technique for reducing mental events and states to physical states. They avoid the well-exposed difficulties of older attempts in which it was held that, when one apparently talks about a separate realm of the mind - speaking of such things as thoughts, emotions, and sensations - the proper analysis of its meaning would be in terms of physical properties and events (usually observable behaviour). The novel idea, on the contrary, is that there is, in fact, an identity between so-called mental events and certain physical events, particularly those occurring in the brain, an identity that it is eventually the task of science to specify - in a way modeled after that in which science discovered that lightning is identical with an electrical discharge. The opposition against this new brand of scientific Materialism does not set up against it a view of the mind as a separate realm coexisting with the physical nor as an essentially private collection of nonphysical events and objects. Rather, the issue has been joined on the question whether the language (or perhaps the concepts) of the psychological and the physical are such as to allow for a scientifically discovered identity between items of the one and items of the other. That there still remains a division among Analytic philosophers concerning the problem of the mental and the physical (though in much altered form) shows both the continuity of the movement and the changes that have occurred.

Keith S. Donnellan *

Professor of Philosophy, University of California, Los Angeles.

Author of articles in various philosophical journals,

particularly on the theory of reference.