METAPHYSICS, NATURE OF

Almost everything in metaphysics is controversial, and it is therefore not surprising that there is little agreement among those who call themselves metaphysicians about what precisely it is that they are attempting. In beginning a discussion of the nature and validation of metaphysical arguments and theories, the best course we can follow is to list some of the standing preoccupations and ambitions of metaphysicians. For this purpose we need to make the assumption that there is a distinct class of metaphysical philosophers, a class into which such thinkers as Plato, Thomas Aquinas, René Descartes, Benedict de Spinoza, and G. W. F. Hegel would fall and from which purely critical or analytic philosophers like the later G. E. Moore would be excluded. It has to be admitted, however, that the line between metaphysical and nonmetaphysical philosophy is exceedingly hard to draw, for many metaphysicians from Plato on have been expert in the supposedly nonmetaphysical pursuit of analyzing or clarifying ideas, while few self-styled analysts have contrived to stick to pure analysis without the open or covert advocacy of a metaphysical point of view.

Setting these difficulties aside, we may note three main features of metaphysics as traditionally practiced. First, metaphysicians have constantly aspired to say what there is in the world or to determine the real nature of things; they have been preoccupied, that is, with the concepts of existence and reality. Their interest in these concepts springs from a double source: from the reflection that the surface show of things often misrepresents them, with the result that we are set the task of determining their real as opposed to their apparent constitution, and from the need to specify what ultimately different kinds of things there are in the world, a need that presses itself on our attention when we wonder whether, for example, minds or numbers are independent existents. The first of these tasks might seem to belong to the scientist rather than the philosopher, for science, too, makes constant use of the distinction between the apparent and the real; we shall indicate in the next paragraph why metaphysicians have not been ready to accept this proposal for lightening their labors.

Second, metaphysics has been commonly presented as the most fundamental and also the most comprehensive of inquiries. It claims to be fundamental because questions about what there is or about the ultimate nature of things underlie all particular inquiries. If you are to assess the results of mathematical investigations, for instance, you need to determine the ontological status of mathematical objects, and according to the theory, this is a task for the metaphysician. The claim of metaphysics to be comprehensive is more difficult to justify. One possible line of support for it, followed by Aristotle, is found in the reflection that questions about existence and reality, along with those about potential and actual being and about causation that are also raised by metaphysicians, cut across the boundaries of particular sciences and arise in connection with every sort of subject matter. Thus, metaphysics is comprehensive just because of its extreme generality. But there is another way in which the claim to comprehensiveness has been advanced. It has been customary to say that whereas sciences like physics and mathematics are departmental studies each of which deals only with a part or particular aspect of reality, metaphysics, by contrast, is concerned with the world as a whole. This explains why philosophers have been unwilling to accept the suggestion that scientists might be left to determine the true nature of things. A scientific theory purports to explain, for example, the real constitution of matter or the fundamental mechanisms of the human body but not to draw the distinction between appearance and reality in an entirely general way, not to tell us, to give an instance, whether matter is the ultimate reality, as materialists suppose, or whether it is itself a manifestation of spirit, as Hegel tried to argue.

This contrast between metaphysics and the particular sciences is sometimes developed in yet another way, again, as will be apparent, to the great advantage of metaphysics. It is said that inquiries in the individual sciences are carried out under assumptions it is the business of metaphysics to make explicit and either to justify or to correct. Metaphysics, by contrast, proceeds without assumptions and is thus fully self-critical where the particular sciences are in part credulous. This line of argument goes back to Plato, who tells us that mathematicians postulate the existence of "odd and even numbers" and "three kinds of angles," and implies that these "hypotheses," taken as "starting points" or "bases" in mathematics, could find their justification and thus lose their hypothetical character in the comprehensive "synoptic" study Plato called dialectic. The dialectician is a man who leaves nothing unquestioned, and just because of this the results of all other inquiries must be seen as no more than provisional; they await ratification or correction from the dialectician. The apparently arbitrary and obviously vague character of this suggestion has not prevented its having a continued appeal to philosophers. Even today, we sometimes hear it said that we need not be unduly disturbed by, for example, the findings of physiologists and psychologists, since the proponents of these sciences work under assumptions it is the business of philosophers to uncover and correct in the light of their knowledge of the whole man (for an argument on these lines see J. S. Haldane, The Philosophy of a Biologist, Oxford, 1935).

If metaphysics is to make good its claim to be uniquely self-critical, its propositions must be shown to be exempt from intellectual challenge as those of no other study are. Descartes, in fact, tried to offer such a demonstration. He argued first that such commonsense assertions as "There is a table under the window" were in every case open to theoretical doubt: However much I seemed to perceive a table, it might be that I was under perceptual illusion or was dreaming. Next, he maintained that even propositions whose truth appeared to be evident, such as those of mathematics, could not be accepted as necessarily in order. An evil demon could be deceiving me into thinking them clear and distinct when they did not really deserve this description. But matters were different when we came to the fundamental metaphysical truth "I think, therefore I am." This truth was such that in the very act of doubting it, one reaffirms it. To doubt is to think, and in thinking that I might not exist, I make clear that I do. Hence, there is at least one truth about whose correctness I could not be in error, and this is a truth of metaphysics. But Descartes was not content to stop at this point. He went on to argue that if I, a being with obvious limitations, certainly exist, then just as certainly there exists a perfect being whose nature is such that he would never deceive me into thinking that true which is not in fact so, once I have satisfied myself that it is by the test of clear and distinct perception. The effect of this move was to provide a guarantee for the findings of the sciences, which were otherwise open to "hyperbolical" doubt. We could henceforth be assured on metaphysical grounds that whatever was clearly and distinctly perceived was true. As for the propositions of metaphysics itself, their truth was guaranteed by their connection with the cogito, which, as we have seen, could not be intelligibly questioned.

The interest of these arguments for our present purpose lies not in their details but in the basic claims they involve. The propositions of metaphysics, according to Descartes, are intellectually impregnable, and in this respect they contrast not only with the beliefs of common sense but also with the pronouncements of the sciences, at least when these are considered apart from their metaphysical guarantee. But from where can they derive their unique certainty? The only possible answer is from their being the products of reason when that faculty is put to work in the fullest and freest way. The result will be that metaphysics is not only the most fundamental of studies; it is also one that relies for its results on the efforts of reason alone.

Metaphysics and the Supersensible

Thus far, we have observed three main features in the projected science of metaphysics. It claims to tell us what really exists or what the real nature of things is, it claims to be fundamental and comprehensive in a way in which no individual science is, and it claims to reach conclusions that are intellectually impregnable and thus possess a unique kind of certainty. Now, many critics of metaphysics have suggested that these claims could be justified only if metaphysics were a factual science providing us, on the strength of rational insight, with knowledge of things or aspects of reality that lie beyond the range of the senses. Nor is this view without support from practicing metaphysicians. Plato drew a contrast between "things seen" and "things unseen" and argued that only things unseen were proper objects of knowledge. From his time on there was a standing tendency to identify the province of the metaphysician with what was vaguely called the supersensible, or the realm of the intellect. Aristotle, for example, distinguished between sensible and insensible substance and assigned the investigation of insensible substance to "first philosophy," or metaphysics. Medieval and early modern philosophers thought of God, the "being of beings," as an entity without bodily extension or shape and for that reason considered him outside the province of the empirical sciences. More generally, it was widely believed that behind the phenomena that present themselves in everyday experience, there lie realities whose existence and properties can be established only by use of the intellect and that can hence be described as noumena, or intelligible objects. In this view, the proper concern of metaphysics was to give us news about noumena.

From the eighteenth century on much ingenuity has been displayed in showing the untenability of this position. The idea that there might be a science that was at once factual and purely intellectual drew its firmest support from the example of mathematics. David Hume suggested, however, that the concern of the mathematician was not with matters of fact and existence but solely with "relations of ideas": His aim was only to make explicit what was already implicit in the premises from which he started. The propositions of mathematics were indeed necessary truths, but by the same token they gave no information about the world. If an inquiry was to pronounce on matters of fact, its method must be empirical, not conceptual, and this meant that its results could not possibly claim to be intellectually impregnable, for anything established on the strength of experience might need to be amended or even withdrawn in the light of further experience. There were no final empirical truths.

A natural reply to this is to argue that even if every factual inquiry must begin from experience, it need not necessarily terminate there. Why should not the metaphysician argue from the characteristics of things sensible to the existence and the nature of things supersensible, as, for instance, Thomas Aquinas and John Locke thought they could? Immanuel Kant was much concerned about the proper answer to this question. He allowed—and here he showed more sympathy with metaphysicians than empiricists then or now—that such concepts as cause and substance, which figure prominently in supposed inferences from the phenomenal to the noumenal, have a necessary character; in Kant's terminology they are a priori, as opposed to empirical, concepts. But he denied as stoutly as Hume that they can therefore be used to carry us beyond the range of possible experience. The question "What brought that about?" is a necessary question, one we cannot rationally refuse to ask, but the answer to such questions must always be sought within experience. If we try, as, for example, Descartes did, to maintain that there must be a First Cause, a necessary being entirely different from the contingent things with which we are familiar, we cease to attach any clear meaning to the concept of cause, for, as Hume saw, it is an essential part of the idea of cause that a cause precede its effect. We can talk about causes as long as we remain within the sphere of the temporal; once we step outside it, the concept loses its determinate character. And what is true of cause here is also true of substance and other metaphysical notions. We can give sense to the concept of substance if we understand it as the permanent that persists through change, but if we eliminate the reference to time, we are left with no more than the logical notion of that which is always a subject and never a predicate, an idea that in its pure form is too indeterminate to be put to metaphysical or, indeed, any other use.

Another attack on metaphysics as the supposed science of intelligible reality was made by the logical positivists. It is a mark of those propositions that belong to accredited sciences like mechanics or genetics, they argued, that we know in principle how to test them; we can see what difference it makes that they are true rather than false. But if a metaphysician comes along and tells us that what really exists is not trees or tables but, say, monads, what tests can we apply to determine the truth of his statement, and what difference does it make if it is true? By definition monads are entities that could never be encountered within experience, nor is their presence supposed to have particular empirical consequences like that of electrons and similar unobservables postulated by natural scientists. Thus, a metaphysical thesis will be compatible with any state of affairs whatsoever, just as the propositions of logic and mathematics are. But if this is so, how can it possibly be maintained that metaphysics gives us information about the world, even the unseen world? The news it purports to bring can only be news from nowhere.

These highly general refutations of a particular conception of metaphysics have seldom been found convincing by metaphysicians. One reason for this is that they fail to come to grips with individual metaphysical arguments, for example, with the cogito. Another is that they appear to prejudge the case against this sort of metaphysics. Why, for example, should it be supposed that a metaphysical thesis must make an empirical difference? Another cause of their failure to carry conviction, however, may be found in the fact that many metaphysicians have worked with a different concept of their subject, one that does not involve it in the claim that it provides information or rivals the empirical sciences. This conception will be considered below.

Metaphysics without Ontology

We have already seen that metaphysicians have wanted to say both that their propositions possess a peculiar certainty and that they are significant as a purely analytic proposition is not. In Kantian terminology they pretend to the status of synthetic a priori truths. Now, many critics of metaphysics have made the assumption that a proposition could be synthetic a priori only if it at once stated a truth of fact and was established by conceptual means alone, a combination they regard as impossible. Facts must be established empirically; pure thinking can lead to the knowledge only of analytic truths. But if we look at Kant's alleged synthetic a priori judgments, particularly those he called principles of the understanding, we see that they make no claim to state facts, even very general facts. A principle like the principle of causality is not a very wide empirical truth, mysteriously known in a nonempirical way; it is, on the contrary, the expression of a rule of procedure that serves to tell us not what properties things have but how to interpret them. Kant supposed that principles of this sort had a special sort of necessity, though they did not logically compel; they owed this, he thought, to the fact that they are prescribed by the human mind as principles specifying what is to count as objective in our experience. Thus, we take it to be a feature of what is objectively there that no quality is present except in a determinate degree, that nothing ever goes entirely out of existence (all change is transformation), that nothing happens except for a reason, and so on.

Kant himself intended this doctrine to have limited application. He thought of the principles of the understanding as prescribing the form of the phenomenal world that we know by means of the senses and investigate in the natural sciences. In his view there were other aspects of experience, in particular the activities of the moral agent, in regard to which they had no legislative force. But it is possible to think of an extension of Kant's doctrine and imagine a set of principles that would prescribe the form not just of one department of experience, but of experience as a whole. A set of principles of this kind would tell us how to organize the data of our experience in such a way that we could give a unitary account of them; it would thus help us make sense of the scheme of things entire. Possessed of concepts of this sort, we could hope to resolve the apparent inconsistencies of science and common sense, together with the more serious conflicts between science and religion and science and morality. We should then be masters of an overall point of view enabling us to see things synoptically or have a set of ideas that would allow us to differentiate the real nature of the universe from its merely superficial aspects. We should, in short, be in possession of a metaphysics.

There can be no doubt that many of the classical metaphysical systems can be thought of as conforming to this schema. In the system of Aristotle, for instance, the key concepts are teleological, and their articulation is to be found in the doctrine of the four causes. It is axiomatic in Aristotle's thought that everything serves a purpose; Aristotle's ambition is to find the point of each phenomenon and thus specify its place in the articulation of the whole. He attempted to carry through his program not only at the biological level, the most obvious source of the concepts involved, but also above and below it—in moral, political, and social life, on the one hand, and in physical science, on the other. His success in these spheres is unequal, but that does not affect the general character of the enterprise.

The popular philosophy of materialism, again, can be seen as an attempt to make sense of the world as a whole on the basis of a distinctive set of first principles. The primary thought of the materialist might be expressed in the axiom that there is nothing that cannot be satisfactorily explained in natural terms; belief not merely in the competence, but also in the omnicompetence, of natural science is a prominent item in his credo. The materialist sees the world as a vast mechanism; whatever happens is the result of natural causes, and all other phenomena must be assessed and understood on this basis. Thus, the phenomena that characterize religious and moral life can be taken in psychological and social terms as things whose causes are ultimately natural, though scarcely in the terms favored by those who engage in them. Religion, as Sigmund Freud said, is an illusion but not an unintelligible illusion; science can account for it, as it can account for everything else.

Finally, Hegelianism made a conscious attempt to produce a metaphysics that constitutes an overall reading of experience. The central concept here is the concept of spirit; it is alleged that everything can be understood in terms of this concept once we take account of the fact that spirit cannot fulfill its potentialities except by working on and against something not itself—in Hegel's peculiar language, "its own other." Thus, we can make sense of the existence of a world of nature in this system; it is there to subserve the purposes of spirit. We can make sense of the social world, too, for many of the characteristics of mind are intelligible only when people are aware of one another and know that others are aware of them. Self-respect and self-contempt would be cases in point.

Each of the systems mentioned could be said to rest on a basic idea or intuition, an idea articulated in a series of concepts taken as definitions of reality and applied, with greater or less success, to the whole range of experience. To appreciate the force of such a system, we need to grasp the basic idea as well as understand the articulated concepts; we have to see the world as the metaphysician in question saw it. The deviser of a metaphysical theory thus becomes a man with a vision of the scheme of things entire. It is important to add, however, that he is not merely a man with a vision, in which case he would be indistinguishable from a philosophical poet. He needs to work his vision out in a theory; he needs to argue his case both by adducing those facts that immediately support it and by explaining those that on the face of things do not.

It seems clear that most of the standard claims for metaphysics can be understood with this account of the matter. Since the first principles of a metaphysical system have prescriptive force, exactly as Kant's principles of the understanding had in regard to the world of nature, they can be properly thought to compel every rational thinker. Their certainty is not the certainty of logic, and yet it exceeds that of any individual statement of fact, for facts are descried only within a framework that these principles provide. Again, even if a system of this kind does not tell us precisely what there is, it nevertheless pronounces on the real character of the world as opposed to the surface show. According to the materialist, for instance, there seem to be features of experience that transcend the natural realm, but in the end it turns out that this is not so. Everything, including men's thoughts and actions, can be accounted for satisfactorily in natural terms. That a scheme of this kind is comprehensive, wider than that of any particular science, goes without saying; that it is fundamental because it is concerned with the coordination of ways of thinking in widely differing spheres is also obvious. True, there is no straightforward counterpart in this type of theory for the criticism by metaphysics of the assumptions of the particular sciences: Metaphysics not being a source of knowledge in itself, it cannot be claimed that other studies are dependent on it as, say, chemistry is dependent on physics. But this circumstance will not prevent this type of metaphysician from putting his own construction on the results of the sciences, as the example of Hegelianism shows. He may have no warrant to question such results, but all the same he may insist on interpreting them in his own way when he offers his reading of experience as a whole. Hegel was doubtless too brusque in his treatment of Isaac Newton and John Dalton, but it does not follow that the whole project for a philosophical treatment of natural phenomena is a mistake.

Argument and Truth in Metaphysics

If metaphysics answers the description given above, a description that would fit many if not quite all of the best-known metaphysical systems, two questions immediately arise. First, we may be asked what sort of a study metaphysics is in this account. Is it a priori or empirical, and to what sorts of argument does it appeal? Second, there is the question what criteria to use in choosing among metaphysical systems. Seeing that many systems are possible, are there any objective ways of deciding that one system embodies the true or the proper way to look at the world?

argument

The answer to the first query is that metaphysics, according to this account, is neither a priori nor empirical, though it makes constant use of both deductive and probable reasoning. A metaphysician is concerned to advocate, articulate, and apply a set of basic interpretative principles, categorical principles we might call them, and principles of this kind cannot be grounded in either conceptual considerations or an appeal to empirical fact. They cannot be supported conceptually since no contradiction is involved in disputing them; they cannot be deduced from facts since they claim to apply with unrestricted validity, no matter what data turn up in experience. They may indeed be suggested by experience and commonly are, but that is not to say that they can be shown to be acceptable or unacceptable by simple empirical methods. Apart from anything else there are no absolutely neutral data to which we can appeal when supporting or attacking a metaphysical theory. For though it is the case that every metaphysician has the duty of explaining all the facts as he sees them, he also has the privilege of being able to decide what really is to count as fact. To see the importance of this we have only to reflect on the different views of religious phenomena taken by materialists and their opponents.

However, though it is true that a metaphysical theory on this account can be established neither deductively nor inductively, deductive and inductive argument both bulk large in metaphysical discussion. Like any other thinker the metaphysician is much concerned with consequences and consistency. He often wants to make the point that since p is true and p implies q ; which in turn implies r, we are logically committed to r or to contending that since q is false and p implies q, p must also be false. The very fact that a metaphysician has a theory to put forward means that he must be preoccupied with the logical connections between the concepts that constitute his system. To say this, however, is not to deny his preoccupation with fact or with probable arguments. Unlike an empirical scientist he establishes no new facts, but all the same he has a double interest in fact. First, he is concerned, more than any specialized inquirer, to see similarities in widely different areas of fact, a process that is relevant to both the formulation and the application of his theory and that involves him in much reasoning by analogy. Second, he needs to pay constant attention to the state of factual knowledge in working out and pressing home his central insight. He promises, after all, to make sense of all the data of experience, and he must consequently take continuous account of these data. The legend that metaphysicians are indifferent to fact has no foundation; on the contrary, they have a primary interest in facts of all sorts even though they do not originate any factual propositions. The extent to which advances in cybernetics have been discussed in recent years by philosophers interested in the truth of materialism affords an apt and striking illustration of this point.

truth

We saw that one charge made against metaphysics as a doctrine of what there is was that no decisive considerations can be adduced either for or against such a theory; the monads of Gottfried Wilhelm Leibniz and the Forms of Plato make no empirical difference. In this respect are things any better in our revised form of metaphysics? It must be confessed that the initial appearance is not favorable. We have emphasized that the first principles of such a system are neither analytic nor empirical; the temptation to conclude that they must accordingly be no more than arbitrary prescriptions, representing a point of view taken up for no good reasons, is strong. And though we have also urged that metaphysicians of this sort have a special interest in fact, the force of that contention is considerably weakened by the admission that they claim the right to decide for themselves what really is fact. If we arm them with this veto—and it is hard to see how they could be refused it—the question of metaphysical truth seems wholly intractable.

It could be, however, that we are setting an impossible standard for metaphysics in requiring it to possess a decision procedure as clear-cut as those of mathematics and the natural sciences. One reason that we can get a straight answer about the acceptability of a theory in physics is that physics works on principles that it does not question (such as that every natural happening will have a sufficient natural explanation). In metaphysics, by contrast, we are concerned with the comparison and assessment of precisely this type of principle. As the widest and most general of all forms of thinking, metaphysics can appeal to no fixed criteria beyond itself except to the requirements of internal consistency that any theory must satisfy. Nor is it true that every reputable branch of knowledge possesses obvious and easily applicable decision procedures. If, for example, we compare metaphysics with history instead of physics, we may begin to see that there are areas of study where dispute and disagreement play a prominent part and that still can claim to proffer understanding and enlightenment. Once we pass beyond the mere ascertaining of fact, there are many histories written from many points of view and resting on many judgments about what is historically important; it is not really possible to hope for a final decision about which, if any, is correct or even about the relative merits of any two equally sophisticated interpretations. However, we do not conclude from this that history is a pointless pursuit rational men would do well to avoid. We realize that a study like history can enlarge the mind and educate the understanding even when it does not add to the sum of public knowledge.

A comparison with metaphysics that is in some respects even closer is provided if we consider the interpretation of a literary text. The data the literary critic confronts—I am thinking of someone who offers a reading of a controversial literary work like Hamlet or Faust —are "harder" than in the case of metaphysics, but this does not prevent the appearance of a wide variety of conflicting theories. And it happens that there are no accepted criteria for deciding among the various theories; all that each critic can do, in the last resort, is explain his way of looking at the text, marshal the points in its favor, and invite the reader to test the matter for himself. But we need not conclude from this that it will be a matter of luck or, perhaps, of psychology which theory will win the reader's approval. At the end of the day, he can be entirely convinced of the authenticity of one particular reading, and he can be persuaded that it offers more enlightenment, covers the central points more impressively, and does better justice to the evidence than its rivals. He may not be able to produce knockdown grounds in favor of his choice, but that is not to say that he has made it for no reason at all.

Metaphysical argument is like literary argument in that it reaches no apparent end; it is like it again in terminating, insofar as it ever does terminate, in an insight that is more personal than public. The old dream of a demonstrated metaphysics whose propositions were even more certain than those of mathematics could scarcely be further from realization. But it would be wrong on that account to think that the concepts of truth and falsity have no application in metaphysics. At the lowest estimate we can describe one system of metaphysics as more illuminating than another. We must, however, decide for ourselves what is really illuminating and what is not. As in the case of the humanities in general, we cannot just learn the truth from another.

Contemporary Antimetaphysics

Theories that profess to deal with "the world as a whole," however they are meant to be taken, are today more often objects of suspicion than of interest, thanks to the influence of G. E. Moore and the later work of Ludwig Wittgenstein. Moore himself never attacked metaphysics explicitly, and indeed his early work, both in logic and in moral philosophy, showed pronounced metaphysical leanings of a generally Platonic kind. But the "Defence of Common Sense" with which he came to be most prominently associated was evolved as a counterblast to views put forward by contemporary metaphysical philosophers, views that, as Moore saw them, could be maintained only by someone prepared to disregard what he evidently knew to be true. When F. H. Bradley, for instance, argued that time is not real, Moore thought this an absurd paradox since the reality of time is taken for granted in any statement containing a temporal expression. If time is not real, it cannot be true that yesterday was Friday or that I had my breakfast before leaving for work. Moore's procedure here, which is to call the metaphysician's bluff by reminding him of what in an off-duty moment he will himself acknowledge that he knows, was generalized by some of his followers into an all-round exposé of metaphysics, which they represented as necessarily consisting of paradoxes and evident falsehoods. For this purpose the thesis that everything is material did not differ from its rival that everything is spirit; both were, when taken seriously, obviously false. There might be a point in maintaining such a thesis (it could be a revealing paradox, according to John Wisdom, or serve a deep-seated psychological purpose, according to Morris Lazerowitz), but in no sense could it express what was really the case.

Moore and his followers assume here that there can be only one correct description of a situation and that in matters like dating or temporal precedence it is known to all of us. It is not obvious that this view is correct, for it could be, as Bradley thought, that a description that was valid and serviceable at the commonsense level would need to be superseded when wider considerations were taken into account. One way of putting Bradley's view is to say that metaphysics claims to offer a conceptual scheme in terms of which we can give a description of the world that is ultimate and comprehensive, but that it also recognizes the existence of many subordinate and more limited schemes, each of which has its point in the characterization of appearances. The Bradleian doctrine of degrees of truth and reality is obviously relevant here, and it cannot be said that Moore gives it very serious consideration. But even if this point had to be granted, the respectability of metaphysics might still be in doubt, for the whole notion of an ultimate description of the world is itself suspect thanks to the work of Wittgenstein.

According to Wittgenstein, a principal source of philosophical error has been the idea that the primary function of language is to describe. The truth is, rather, that we engage in many different "language games," each of which serves its own purpose and each of which is authentic at its own level. There can be no question of ruling any such game out of court; the fact that it is played is sufficient evidence that it is appropriate. Nor are different sets of language users rivals; it could not be said, for instance, that physics gives a truer picture of the world than common sense or that the naïvetés of everyday moral language are corrected by the psychologist. If we keep these diverse languages apart, we see that each has its own point and utility. The idea of a finally correct language that would embrace and replace them all is clearly the height of absurdity, and, hence, metaphysics in its revised form is no more acceptable than was metaphysics in the shape of news from nowhere.

But this analysis, too, is built on questionable assumptions. First, is it really clear that language games or areas of linguistic activity are as distinct as Wittgenstein says they are? The point is by no means clear as far as the language games of science and common sense are concerned, for most scientists and many plain men think that the scientific account of the physical world gives a truer picture of it than that embodied in the ordinary man's everyday beliefs. Nor can we agree without further argument with the thesis that sufficient authentication is found for a language game when we note that it is played. There are, after all, games and games. In a form of game much played in the ancient world, elaborate formulas to appease the god of the sea were devised by those about to embark. As a result, a certain way of talking commanded a wide use and approval. But could that fact alone be invoked to show that it was legitimate? Surely, we should want to object that however much such language was used, its use could not be legitimate if in fact there was no god of the sea or if he exercised no influence on whether seafarers reached their destinations safely. To do this, however, is to make the propriety of a language game subject to the tenability of the factual assumptions on which it rests. Although this is not to maintain that the only use of language is to describe (which would be absurd), it is to claim a certain priority for the language game in which we say how things are.

Metaphysics as we have expounded it is concerned with resolving conceptual conflicts by finding a way of speaking that will enable us to express the true nature of the world. If we possess such a way of speaking, we have a yardstick by which to measure the ultimate tenability, as opposed to the immediate use, of particular language games—the languages of religion, science, law, and so on. It is not self-evident that each of these is in order as it is, and though the fact that they are constantly used and understood is enough to show that they serve some purpose, it does not in itself show that they are suited for the purposes those who use them have in mind. These games are indeed played, but they could, for all that, be played on false pretenses. To decide whether they are, we must have recourse to metaphysics.

Metaphysics as Analysis

Even if the foregoing account of the nature of metaphysics were accepted as generally unobjectionable, there are many philosophers who would deny that it covers everything metaphysicians have attempted or are attempting to do. In particular, it fails to accommodate an activity pursued by many contemporary analytic and linguistic philosophers that has a clear affinity with the work of some of the classical metaphysicians. The classical metaphysicians were led to ask what there is partly because of puzzles about the status of numbers and qualities. Plato had produced arguments to show that these must be independently real, and Aristotle elaborated the doctrine of categories as an answer to them. Now, there are plainly parallels to this controversy in contemporary philosophy, both in the discussions among logicians about names and descriptions (which revive the ancient dispute about the relative priority of universals and particulars) and in the arguments about the relation of the mind and body that have recently been so prominent in British and American philosophy. What is notable about these issues, as opposed to those mentioned above, is that matters of fact appear to have no relevance to their solution. If we can solve them at all, we can solve them only by thinking.

This contrast is both genuine and important; there certainly are philosophical activities that are traditionally connected with metaphysics and that cannot be subsumed either under the schema given above or under that which it was meant to replace. These activities are in essence logical or analytic, and insofar as it is confined to them, metaphysics is indistinguishable from analysis. But there is no reason to confine metaphysics to such inquiries. That metaphysicians have been speculative theorists as well as ontologists in the restricted modern sense is almost too obvious to need mention; to decide, as some commentators do, that the speculation can be set aside as regrettable and the ontology played up is at best arbitrary. Nor is it true that we can make an entirely clear-cut distinction between the two. If we look at recent work on the mind-body problem, for instance, we see that much of it is indeed logical in a wide sense of that word but that considerations of substance also come in, for example, when we discuss the nature of consciousness or of thought bearing in mind the properties and possibilities of thinking machines. An all-important motive that impels men to persist with these questions is the need to take account once more of the claims of materialism against a background in which new scientific and technical discoveries seem to lend increased support to those claims. However fascinating logical problems may be, interest in them cannot be long sustained without some external stimulus. It is such a stimulus that metaphysics of the broad kind argued for above may be expected to provide.

See also Appearance and Reality; Aristotle; Being; Bradley, Francis Herbert; Categories; Descartes, René; Dialectic; Existence; Freud, Sigmund; Hegel, Georg Wilhelm Friedrich; Hegelianism; Hume, David; Kant, Immanuel; Language, Philosophy of; Leibniz, Gottfried Wilhelm; Locke, John; Logical Positivism; Materialism; Metaphysics, History of; Monad and Monadology; Moore, George Edward; Ontology; Plato; Spinoza, Benedict (Baruch) de; Thomas Aquinas, St.; Time; Wisdom, (Arthur) John Terence Dibben; Wittgenstein, Ludwig Josef Johann.

Bibliography

classical works

Aristotle. Works, Vol. VIII, Metaphysics, edited by J. A. Smith and W. D. Ross. Oxford, 1928. See especially Γ 1–2; E 1; Λ 1–2, 6. Compare also W. D. Ross, Aristotle's Metaphysics, Vol. I. Oxford, 1924. See pp. lxxvii ff.

Baumgarten, A. G. Metaphysica. Halle, 1739.

Bradley, F. H. Appearance and Reality, Oxford, 1893. See especially the preface and introduction.

Descartes, René. Meditations on First Philosophy (1641). In The Philosophical Works of Descartes, translated by E. S. Haldane and G. R. T. Ross, 2 vols. New York, 1931. See especially "Reply to Second Objections."

Hegel, Georg Wilhelm Friedrich. Phenomenology of Mind (1807). Translated by J. B. Baillie, 2nd ed. New York: Macmillan, 1931.

Hume, David. Enquiry concerning Human Understanding. London, 1748. There are many modern editions.

Hume, David. A Treatise of Human Nature (1739–1740), edited by L. A. Selby-Bigge. Oxford, 1888; 1941.

Kant, Immanuel. Critique of Pure Reason (1781). Translated by N. Kemp Smith. London, 1929.

Kant, Immanuel. "Inaugural Dissertation" (1770). In Kant's Inaugural Dissertation and Early Writings on Space, translated by J. Handyside. Chicago: Open Court, 1929.

Kant, Immanuel. An Inquiry into the Distinctness of the Principles of Natural Theology and Morals (1764). In Kant's Critique of Practical Reason, edited by L. W. Beck. Chicago, 1949.

Kant, Immanuel. Prolegomena to Every Future Metaphysics (1783). Translated by P. G. Lucas. Manchester, U.K., 1953.

Plato. Republic. Translated by A. D. Lindsay. London: Deat, 1976. See especially VI–VII.

Plato. Sophist. In Plato's Theory of Knowledge, edited by F. M. Cornford. London: K. Paul, Trench, Trubner, 1935. See especially 253ff.

Wolff, Christian. Philosophia Prima Sive Ontologia. Frankfurt, 1729; edited by J. Ecole, Hildesheim, 1962.

modern works

Ayer, A. J. Language, Truth and Logic. London: Gollancz, 1936. Lively brief account of the logical positivist criticism of metaphysics. See also the introduction to the second edition (1945) for replies to objections.

Broad, C. D. "Critical and Speculative Philosophy." In Contemporary British Philosophy, edited by J. H. Muirhead, Vol. I. London: Allen and Unwin, 1924. An influential article.

Carnap, Rudolf. "The Elimination of Metaphysics through Logical Analysis of Language." In Logical Positivism, edited by A. J. Ayer. Glencoe, IL: Free Press, 1959. An important and influential article; originally published in 1932.

Carnap, Rudolf. Philosophy and Logical Syntax. London: Kegan Paul, 1935.

Collingwood, R. G. An Essay on Metaphysics. Oxford: Clarendon Press, 1940. Written in reaction to Ayer but also of independent interest.

Emmet, D. M. The Nature of Metaphysical Thinking. London: Macmillan, 1945.

Lazerowitz, Morris. The Structure of Metaphysics. London: Routledge and Paul, 1955. Mainly influenced by Moore.

Moore, G. E. "The Conception of Reality." In his Philosophical Studies. London: Routledge, 1922. Sharp criticism of Bradley.

Moore, G. E. "A Defence of Common Sense." In his Philosophical Papers. London: Allen and Unwin, 1959.

Moore, G. E. Some Main Problems of Philosophy. London: Allen and Unwin, 1953. Expounds at length the doctrine of "A Defence of Common Sense"; written in 1911.

Pears, D. F., ed. The Nature of Metaphysics. London: Macmillan, 1957. A series of modern discussions of metaphysics.

Walsh, W. H. Metaphysics. London, 1963.

Wittgenstein, Ludwig. The Blue and Brown Books: Preliminary Studies for the Philosophical Investigations. Oxford: Blackwell, 1958.

Wittgenstein, Ludwig. Philosophical Investigations. Oxford: Blackwell, 1953.

Wittgenstein, Ludwig. Tractatus Logico-Philosophicus. London: Routledge, 1922.

W. H. Walsh (1967)