different common features between the terms of a metaphor. Metaphors are interpreted and they are inter- preted differently by different readers and hearers. Conse- quently, the idea that there can be a literal paraphrase of a metaphor which preserves its sense is no longer widely held, for such a literal paraphrase would have to com- mand common agreement as expressing what the metaphor means. A powerful metaphor like Macbeth’s ‘sleep that knits up the ravell’d sleave of care’ invites us to join in an exploration of points of similarity and difference. Black, in a later paper, speaks of metaphors as ‘inciting the hearer’ and likens the process to game-playing. Since this also characterizes the understanding of similes, few writers would now make a sharp distinction between metaphor and simile. Black argued that when we read a metaphor like ‘Achilles is a lion’, we read it armed with a number of commonplace beliefs about ‘lion’; these, metaphorically applied to Achilles, we draw on as we construe it. In Claudius’ line in Shakespeare’s Hamlet, ‘When sorrows come, they come not single spies, but in battalions’, we may reflect that spies threaten and undermine, carrying the fear of worse to follow; battalions, on the other hand, embody open aggression. One interesting recent contribution in a philosophical debate that goes back to Aristotle is Donald Davidson’s rejection of the idea that there is a special sort of *meaning which metaphors have, over and above the literal mean- ing. Taken literally metaphors seem nonsensical or false or only trivially true. For Davidson, it is the use of metaphor which is crucial, in making us aware of some likeness, often surprising, between apparently disparate things but without asserting that likeness. Metaphors are the growing-points of *language. A cur- sory glance shows just how much of the language of mind is metaphorical in origin. These metaphors die, of course, and lose their metaphorical force though their origins may be still visible. In recent decades, philosophers have, as well, become more aware of the role played by metaphor in science and religion. r.a.s. D. E. Cooper, Metaphor (Oxford, 1986). Andrew Ortony (ed.), Metaphor and Thought, 2nd edn. (Cam- bridge, 1993). metaphysics, history of. Metaphysics is the most abstract and in some views ‘high-falutin’ part of philosophy, having to do with the features of ultimate reality, what really exists and what it is that distinguishes that and makes it possible. Nevertheless, the exact nature of the subject has been constantly disputed, as indeed has its validity and usefulness. Philosophy at its very beginnings with the Pre-Socratics was metaphysical in character, although it was initially presented in a dress which made it sound more like physics, as witness Thales’ claim that everything is made of water. Subsequent Pre-Socratics were concerned with other attempts to understand nature and the possibility of change within it, although Parmenides argued (for the first time by means of a formal argument, even if that was given a poetical dress) that coming to be, ceasing to be, and change in general were impossible, so that his succes- sors had to counter his claim, even if they did not fully understand his arguments. By the time of Plato, with his theory that the true realities were Forms (or Ideas), abstract exemplars or paradigms, of which sensible things were only imperfect copies, the distinction of metaphysics from physics became clear, since these realities were quite distinct from the world with which physics has its con- cern. Since the Forms were also universal in character his theory also initiated metaphysical arguments concerning the status of *universals, something that has gone on ever since. The term ‘metaphysics’ originated, however, as a title given to some of Aristotle’s works in the catalogue of the edition of them produced by Andronicus of Rhodes in the second half of the first century bc (although it may have come from an earlier library classification). It meant sim- ply the works which followed those on physics in the cata- logue. But those works, which were concerned with being, both as such and in respect of various *categories of it, especially *substance, contain discussions concerning matters which have an obvious continuity with later metaphysical theories. Hence it is reasonable to see Aris- totle’s Metaphysics, untidy though it is in the form in which it has come down to us, as the first systematic treatise in metaphysics, containing not only discussions of the notion of being and what has the best claim to that title but also criticisms of earlier thought on the subject, particu- larly Plato’s Theory of Forms. Those Forms are soundly rejected. Aristotle believes in universals, certainly, but they are features of the world itself, which is made up of things with *essences, belonging to a system organized in terms of genera and species. The notion of species corres- ponds to that of *form as Aristotle construes that, but material things have not only form but *matter too. Among beings, which Aristotle thought were classifiable in terms of a system of categories, things in the category of substance have the greatest claim to that title, and among them those which are nearest to being pure form. God, whose nature is, in Aristotle’s view, pure form, is the high- est kind of substance, and thus the highest kind of being, so that what it is for something to be is best seen in God, who comprises the end or goal to which other things tend, and who, as the *prime mover, is also the so-called *final cause of the movements of the heavenly bodies. Post-Aristotelian philosophy saw the world as organ- ized under different principles, though the influence of Aristotle was strong. Epicurus thought that everything, including ourselves, was composed of atoms moving in a void, and was to be explained in those terms. The Stoics, by contrast, thought of matter as forming a continuum, but subject to rational or so-called ‘seminal’ principles due to pneuma (breath or spirit), which gives everything life. Platonism went though many vicissitudes, and at the end of the period of Greek philosophy took a somewhat mys- tical form in Neoplatonism, led by Plotinus, according to 590 metaphor which the Forms are organized under a unitary principle, the One. At the opposite extreme from this is the world of matter, responsible by its negativity for evil. The mystical goal is an identification with the One, but it is a goal to be reached through philosophy, not by any religious process. Nevertheless, Neoplatonic ideas had a considerable influ- ence on religious thinking, including that of Augustine. Plotinus’ main disciple, Porphyry, wrote on, amongst other things, the Aristotelian doctrine of categories, say- ing that the ontological status of species and genera was uncertain, and Boethius, commenting on that, thereby transmitted to later medieval thought the problem of the status of universals, which loomed large throughout that period. There was considerable argument between schools of thinking about universals in the early Middle Ages, between realists (e.g. William of Champeaux), nominal- ists (e.g. Roscelin of Compiègne), and conceptualists (e.g. perhaps Abelard, although his position on the issue is not entirely clear), who respectively claimed that what was general was to be found in nature, words only, or thoughts only. With the rediscovery of Aristotle in the thirteenth century, after a period of ignorance of his philosophy in the West, realism about universals became the accepted view, until a revival of nominalism, particularly with William of Ockham, in the fourteenth century. There was, however, a connection between the issues over uni- versals and theological issues, particularly the doctrine of the Trinity. The other main metaphysical concerns of medieval philosophers were similarly theologically orien- tated—particularly the existence of God and the nature of the soul. Anselm in the eleventh century became famous or notorious for his so-called *ontological proof of the existence of God, maintaining that God’s existence fol- lowed from the fact that God is that than which no greater can be conceived. The great Thomas Aquinas in the thir- teenth century took a more Aristotelian line on the argu- ments for God’s existence, relying in the main on considerations (which owe much to Aristotle) concerning the supposed nature of the world which point to the need to assume the existence of a deity. Aquinas also took, with modifications, an Aristotelian line on the nature of the soul as the form of the body, provoking questions, not easily answered, about how this view was to be reconciled with belief in immortality. After the Renaissance, during which there was a revival of Platonism, often in other forms of mystical dress, Descartes initiated a change in the approach to philoso- phy, although preoccupation with scholastic notions such as that of substance remained. Descartes’s orientation in philosophy was mainly epistemological in character; it might indeed be said that his metaphysics was founded on epistemological considerations. For the thesis for which he has become known—the radical dualism between mind and body as distinct substances—was founded on the claim that we have a more direct access to (and thereby a clearer and more distinct idea of) our minds than to our bodies. His rationalist successors Spinoza and Leibniz were also very concerned with the *mind–body problem. Spinoza maintained that mind and body were to be construed simply as different aspects of one substance, but that was in a context of argument which was directed to the conclusion that there can be only one substance, God or nature, and that what we are and what happens to us is strictly determined because we are modifications of that one substance. Spinoza thought, nevertheless, that there was a sense to speaking of freedom which lies in an acceptance of the necessity that the determinism entails. Leibniz, by contrast, thought that there was an infinite number of substances, which were simple, though capable of reflecting an infinite number of points of view. He came to think that these substances could only be what he called *‘monads’, which were simple, like the ego in ourselves. Monads were organized in such a way as to fall under a dominant monad, which was God. Leibniz also held that everything that happened to a substance was necessary to it, but that God created the world of sub- stances according to the principle of *sufficient reason, which made this world the best of all possible worlds. Despite the Spinozistic necessity that this seemed to entail as far as human-beings are concerned, Leibniz thought, but did not convince others, that a form of freedom was still possible. One might think that the British Empiricists, Locke, Berkeley, and Hume, were, because of the empiricism, mainly epistemologically orientated. This is superficially true of Locke, whose Essay tries to set out the structure and limits of human understanding, but he is concerned with substance, for example, although in a way that owes much to Boyle and contemporary physicists. He also pre- sents a theory of *persons and *personal identity, which has provoked considerable recent interest, and, in his theory of abstract ideas, he sets out once again a concep- tualist theory of universals. Berkeley was more nominalis- tically inclined, and attacked that theory because he thought that it let in a doctrine of material substance to which he was also opposed. In place of the latter Berkeley put forward the view that ‘to be is to be perceived’, so that the only things that properly exist are ideas (the objects of perception, as he thought) and spirits (which include our- selves and particularly God). Berkeley’s theory is thus the first instance of full-blooded *idealism. Of the three British Empiricists, Hume was the most anti-metaphysical, but his doctrine of impressions and ideas in a way continues Berkeley’s thinking, and Hume admits at one point that his impressions are in one sense what deserve the title of substances. Ontologically, therefore, Hume regards real- ity as consisting merely of impressions and corresponding ideas, and he expresses a form of scepticism about both material bodies and self. In both cases we have bundles of impressions and ideas which have a certain constancy and coherence, and it is these characteristics which make us believe, but do not justify such belief, in bodies and the self. It has to be said, however, that Hume thought, with some reason, the resulting position over the self particu- larly worrying. metaphysics, history of 591 It is Kant who stands at the culmination of this, being opposed to what he regarded as the speculative meta- physics handed down by the rationalists, but concerned, as the Empiricists were, with the limits of the human understanding in such a way as to allow more than they had done. Kant accepts a form of idealism, which he called *‘transcendental’, saying that space and time were merely forms of experience and had no application to what he called *‘things-in-themselves’, the unknowable reality which he thought must be assumed to underlie and in some way be responsible for experience. Kant’s idealism was not, however, merely subjective, as Berkeley’s was, in that he sees the *understanding as bringing to bear in judgement certain principles, derived from the *cat- egories or formal concepts which it supplies, in such a way that the forms of objective judgement can be distin- guished from merely subjective ones. In particular, Kant thought, objective experience can be seen to involve causality and principles of necessary connection, despite Hume’s scepticism on this. All this, a sort of metaphysics of experience, can be regarded as a substitute for trad- itional metaphysics, which Kant thought of as concerned with God, freedom, and immortality, but as involving an attempt to use reason beyond the boundaries to which it was properly limited. Part of his Critique of Pure Reason involves an attempt to show that such improper uses of reason lead to contradictions and the like. Kant’s account of what is necessary about human understanding and the limitations, by comparison, of rea- son presented a kind of watershed for metaphysics, but philosophers were soon to try to circumvent his conclu- sions in a variety of ways. Fichte objected to the whole idea of things-in-themselves, arguing that the ego or self actively posits a non-self opposed to it, so that in effect that non-self exists only for the self, while constituting some- thing necessary, an absolute. His idealism is thus the first instance of so-called absolute *idealism. Schopenhauer, on the other hand, thought he could produce good reason for believing that there was only one thing-in-itself and that this was to be identified with will. Both philosophers, however, accepted a form of idealism. The most radical development in that respect was the philosophy of Hegel, who thought that reason could certainly do what Kant thought impossible, leading to the idea of an identification of self and object. This form of absolute idealism was worked out in terms of a system of developing categories, culminating in what Hegel called the absolute notion in which ‘Spirit knows itself as Spirit’. Hegel’s metaphysical system is both monumental and encyclopaedic in charac- ter, claiming to bring all phenomena within its terms of reference. It has been seen as either marvellous or repul- sive by different commentators. There were, of course, reactions to it. *Existentialists, beginning with Kierkegaard, objected that existence pre- cedes essence, and that Hegel’s thought left out individu- ality. This was an objection to the idea that reality could be seen as such only in terms of an all-comprehensive system. Marx, using, at any rate initially, somewhat Hegelian terms, tried to turn the system on its head by insisting on the materialist and social basis of all thought and thereby of reality. Hegelian thought had a late influence in Eng- land towards the end of the nineteenth century, particu- larly in F. H. Bradley, although he objected to the more systematic aspects of Hegel’s thought. To his *monism (the belief that reality was one) there was in turn a reaction in the *logical atomism of Russell and perhaps the early Wittgenstein, according to which reality involves a plurality of *sense-data, which, like Leibniz’s monads, constituted absolute simples. Subsequently, the anti- metaphysical theory of the Logical Positivists, such as Ayer, who, on the basis of the principle that the meaning of a statement is to be found in its method of verification, argued that metaphysical statements were nonsensical, put metaphysics out of fashion, where on many popular views it remains. In fact, however, it continues, and Strawson’s so-called ‘descriptive metaphysics’, according to which ontological distinctions between individuals or objects of identifica- tion are made relative to speaker–hearer discourse, is something of a return to Kant, though without the ideal- ism. Elsewhere in Europe there has been, for example, Heidegger’s anti-scientific concern with the nature of being and with Dasein or presence in the world. This pres- ence is of a kind that only human individuals have, and Heidegger saw it as having an intimate connection with time, on a view of time which sees it as having fundamen- tally to do with the ideas of past, present, and future, and not simply temporal relations between events. These alternative conceptions of time have been a central issue in Anglo-Saxon metaphysics too, ever since the Cam- bridge Hegelian McTaggart argued, early in the twentieth century, that time must, essentially, have to do with past, present, and future (or, as it is sometimes put, ‘tense’) and that, because every event is all three and thus in possession of incompatible attributes, time is unreal. To the objec- tion that events have all three attributes at different times, McTaggart argued that this only produced an infinite regress. Different philosophers have drawn different morals from these claims, including the moral that time must fundamentally have nothing to do with ‘tense’. It is equally arguable that the correct conclusion is that the ‘tensed’ point of view is indispensable to an account of reality and that it is the attempt to do without it in charac- terizing reality that causes the trouble. Heidegger has his own and different reasons for emphasizing a ‘tensed’ con- ception of time, in that he is concerned to bring out what presence in the world, in his sense, entails—in particular that it must have its end in death, when time ends for us. In other quarters again, particularly in the USA and Australia, an emphasis on science has produced its own scientistic metaphysics, according to which only sup- posedly scientific characterizations of reality will do. Such views not only reject the kind of ‘tensed’ conceptions of time which I have noted above, but also underplay the kinds of point of view that are arguably involved in self- hood and thereby in any reality which involves selves. 592 metaphysics, history of Such a metaphysics tends inevitably to be materialist, though not necessarily in the kind of way in which Marx- ian thought is materialist. It is simply assumed that all that exists in the end is particular incidences of matter in motion and that what seems at first not to be that is in fact identical with some form of it. Nevertheless, although Cartesian *dualism is widely rejected as a great mistake as well as a great obstacle to the successful development of philosophy, the pressures deriving from what led to that dualism in the first place—the first-person point of view— remain, and are emphasized by some philosophers, e.g. Thomas Nagel. And so it goes on. d.w.h. *descriptive metaphysics; materialism; modality and metaphysics; opposition to metaphysics; revisionary metaphysics. D. W. Hamlyn, Metaphysics (Cambridge, 1984). —— The Penguin History of Western Philosophy (London, 1987). A. J. P. Kenny, Ancient Philosophy (Oxford, 2004), ch. 6. ——Medieval Philosophy (Oxford, 2005), ch. 6. Jonathan Lear, Aristotle: The Desire to Understand (Cambridge, 1988). T. L. S. Sprigge, Theories of Existence (Harmondsworth, 1984). Ralph C. S. Walker, Kant (London, 1978). Margaret D. Wilson, Descartes (London, 1978). metaphysics, opposition to. Opposition to metaphysics has come from both within philosophy and outside it. *Logical Positivism, though now defunct, was particu- larly hostile to what its adherents saw as the meaningless, because unverifiable, claims of metaphysics. These objec- tions foundered on the impossibility of providing an acceptable criterion of *verifiability. But the deference to empirical science displayed by the Logical Positivists is still a feature of much Anglo-American analytic philoso- phy, creating an intellectual climate inimical to the pursuit of speculative metaphysics. This hostility is paralleled in the popular writings of many scientists, who seem to think that any legitimate issues once embraced by metaphysics now belong exclusively to the province of empirical sci- ence—issues such as the nature of *space and *time, and the mind–body problem. Such writers are often blithely unaware of the uncritical metaphysical assumptions per- vading their works and the philosophical naïvety of many of their arguments. But it is ironic that the deference shown by many philosophers to the latest scientific theor- ies is not reciprocated by the popularizing scientists, who do not conceal their contempt for philosophy in general as well as metaphysics in particular. More recent hostility to metaphysics comes from the post-modernists and deconstructionists, who wish to pro- claim that philosophy—and certainly metaphysics—is dead. These writers represent metaphysics as a temporary aberration of the Western intellect, denying the notion that it is a pursuit of perennial questions for which timeless answers may legitimately be sought. Of course, these crit- ics of metaphysics, in repudiating any objective concep- tion of truth in favour of a fashionable cultural relativism, can make no common cause with the scientific critics, whose quite contrary assumption is that science provides the royal road to objective truth and ultimately to a final ‘Theory of Everything’. With enemies so divided amongst themselves, metaphysics may comfort itself with the thought that so many people can’t be right. The very fact of such widespread disagreement over fundamentals demonstrates the need for critical and reflective meta- physical inquiry, pursued not dogmatically but in the spirit of Kant. Despite all this hostility, metaphysics and *ontology are currently enjoying a modest revival amongst professional philosophers, who are no longer embarrassed to discuss such issues as the nature of substance and to advance real- ist theories of *universals. But much of this work is highly technical, involving sophisticated applications of *modal logic, and consequently it is difficult to convey its results to a lay public. There is thus a danger that such work will be dismissed as a revival of scholasticism without rele- vance to everyday concerns. That would be a pity, and so it is not only the duty but also in the interest of metaphys- icians to make their work more accessible, with a view to countering the relativistic and scientistic dogmas of our time. Perhaps the most serious intellectual threat to meta- physics as traditionally conceived comes from the move- ment towards *naturalism in contemporary philosophy, taking its lead from W. V. Quine’s advocacy of ‘natural- ized epistemology’. With the theory of knowledge recon- ceived as, in effect, a branch of empirical psychology and the concomitant rejection of the traditional distinction between *a priori and a posteriori truth, the claim of meta- physics to have a distinctive subject-matter and method has been put under some pressure. However, just as the cruder scientistic and relativistic enemies of metaphysics may be accused of promoting a particular metaphysical dogma under the guise of an onslaught on metaphysics in general, so too may this charge be levelled at its naturalis- tic critics. The normative categories of reason and truth transcend naturalistic reduction and cannot, without prag- matic incoherence, be argued out of existence. e.j.l. H. Kornblith (ed.), Naturalizing Epistemology (Cambridge, Mass., 1985). R. Rorty, Philosophy and the Mirror of Nature (Oxford, 1980). S. Stich, The Fragmentation of Reason (Cambridge, Mass., 1990). metaphysics, problems of. In contemporary philosophy, problems of metaphysics often take the form of a trilemma concerning some large and important feature of our lives or discourse, a trilemma whose terms are: illu- sion, well-founded appearance, and fundamental reality. In recent decades these problems most often tend to arise against the backdrop of a broad *naturalism and, often, scientific *realism. The problems themselves may be viewed as demands for possibility explanations: How are values and norms possible in a world of facts? How are minds and mental phenomena possible in a world of matter in motion? How is freedom of action or will pos- sible in a world of scientific law? How can there be abstract metaphysics, problems of 593 entities in a world of events and other contingent particu- lars? In each case the same troublesome trilemma presents itself. Concerning any of these realms or dimensions—e.g. values, the mental, freedom, and abstract—there is the view that it is all a big illusion, that there is really nothing in that realm or dimension. Alternatively, it may be held that there are real denizens of the realm in question, with a reality as fundamental as that of any particle or field pos- tulated by our physics. And according to a third, irenic view, though there are real enough entities or phenomena in the target realm, none is fundamental, all deriving rather from more basic entities or phenomena. All indeed are said to resemble ordinary bodies—tables, quantities of water, cats and dogs, etc.—in being real enough, though derived from the existence and organization of more basic entities: from cells or molecules, etc. For example, it might be held that values and the normative are a complete illusion. Thus, for the non- cognitivist, normative and evaluative concepts do not represent any mind- or language-independent constituent or aspect of reality. Rather, their significance is only func- tional: like that of the imperative mood or the exclamation mark. It is an illusion to suppose that the goodness of a juicy, sweet apple attaches to it just as do its redness and its roundness. For others, however, the goodness of an apple is just as objective a property of it as its roundness, or, certainly, its redness: just as objective a property, and just as real, and fundamental. But there is a third, irenic, option, according to which the goodness of the apple and the rightness of biting into it are real and objective enough, but ‘well founded’ on more fundamental properties: e.g. on the apple’s disposition to cause, or on the biting into it actually causing, a sufficient balance of pleasure over pain (especially when compared with the alterna- tives open to the agent at the time). This third option comes with two interesting suboptions: first, adding further that the evaluative and normative phenomena in question are not only well founded (bene fundata) but also actually reducible by definition or analysis to the underlying realities that give rise to them; and, alternatively, remain- ing deliberately non-committal on that issue, claiming only that the phenomena in question do supervene on underlying realities, whether or not they are reducible to them or definable or analysable in their terms. Similar issues arise with regard to the realm of *minds and the mental. Let us assume that reality is constituted of particulars (whether substances or events) with the prop- erties that characterize them and the relations that inter- relate them. Just what is included among these particulars, properties, and relations has been a matter of considerable controversy in the history of Western philosophy. Idealists view reality as ultimately spiritual or mental. For them the basic particulars are subjects of thought or experience, souls or spirits or monads, the world of matter in motion being nothing more than a stable appearance to our minds. If we say there are snowballs, for the idealist we are right at best in the sense that in certain circum- stances our minds are disposed systematically to experi- ence combinations of whiteness, roundness, and coldness. The foundation of the existence of such supposed objects therefore lies in the contents of our minds. For the idealist, physical bodies are rather like images in a rich and stable dream. And we are essentially minds, subjects of thought and consciousness. Leibniz and Berkeley were idealists. Materialists and physicalists view material or physical objects or events as more fundamental than minds or egos or their modes of thought or experience. Accordingly, they would reduce mind to matter rather than matter to mind. For the materialist there are no fundamental sub- jects of consciousness, no souls or spirits. We have minds simply because we think, sense, feel, etc., and we do all this as rational animals with properly functioning brains and nervous systems. Hobbes was a materialist, as have been most contemporary philosophers who write in the analytic tradition. The token physicalism (and *anomalous monism) of Donald Davidson is also a kind of physicalism of particulars, since it accepts events as basic particulars and regards these as without exception physical. Finally, dualists admit both souls and bodies as funda- mental entities. Neither mind nor matter is reducible to the other and there is no problem of reducing either to the other. For the dualist the problem lies rather in under- standing how mind and matter can possibly interact. Descartes was a dualist. So far we have considered metaphysical options on the nature of basic particulars. There are also similar options on the nature of fundamental *properties or states of affairs. Thus one can be a property phenomenalist, for whom the only fundamental states are mental, e.g. sen- sory experiences, all other states being ‘reducible’ to or at least derivative from these. For a physicalist with regard to states or properties, on the other hand, only physical properties are fundamental; hence any state constituted by a particular having a prop- erty, or by a number of particulars related by a certain rela- tion, is a fundamental state only if the particulars are all physical and the properties and relations are all physical. The type–type identity theory is an option open to such a physicalist, and believers in type–type identity might hold the identity to be necessary, as did logical behaviourists, and as do functionalists. Alternatively, a believer in type–type identity might opt for ‘contingent identity’, as with the functional-specification view of David Lewis. Finally, property dualists admit both physical proper- ties irreducible to the mental, and mental properties irre- ducible to the physical. Recent debates over qualia, over the existence of irreducibly qualitative and experiential aspects of one’s experience, have divided mentalists on the affirmative (e.g. Ned Block and Jerry Fodor with their ‘absent qualia’ argument, and Thomas Nagel with his appeal to subjectivity), from physicalists, especially func- tionalists, on the negative (such as David Armstrong and Daniel Dennett, with their attempt to reduce conscious- ness in general to propositional attitudes). Some (e.g. Syd- 594 metaphysics, problems of ney Shoemaker) have tried to reconcile *functionalism with acceptance of qualia, but for our purposes the accep- tance of qualia is the important move, if such qualitative aspects of experience are supposed a fundamental sort of mental property not reducible to the physical, nor super- venient upon it. With regard to *freedom of action or of the will, one faces the same set of options. One might hold freedom to be a complete illusion, since we are natural beings caught in the web of physical law from cradle to grave. Or one might alternatively hold that freedom is a basic fact of life: one might rather deny that a human life could ever be wholly caught within a web of natural laws. A third, more irenic, position is possible, moreover, according to which we do enjoy freedom, but a freedom that is after all com- patible with the sway of physical law over every detail of human life. There is important support for this alter- native, quite apart from the implausibility of any meta- physics that tries to set a priori limits to how much science might achieve in understanding human behaviour. For consider the postulation of libertarian action not pro- duced by antecedent conditions in accordance with physical law. This is what the libertarian believer in funda- mental freedom accepts. But it is puzzling how that can help secure the kind of freedom desired: namely, the kind that would support the attribution of responsibility to the agent, and the assignment of praise or blame, reward or punishment. The reconcilist—or compatibilist—camp for its part still owes a large and challenging debt: reconcilists must still explain how such freedom can be reconciled with the fact that, on the assumption of *determinism, every physical detail of one’s life is already determined prior to one’s birth. They must introduce, in the teeth of that impressive fact, some crucial distinction among one’s actions, between those that are nevertheless not ‘com- pelled’ in some appropriate sense, and for which we can remain responsible, and those that are thus compelled, which relieves the agent of responsibility. Metaphysical world-views have derived from episte- mological constraints. Thus one might be impressed by the difference between one’s own consciousness, to which one enjoys introspective access, on one side, and, on the other, the supposed world of physical fact beyond. Philosophers have long puzzled over how such a funda- mental chasm could ever be bridged by reason. How could one ever know about the reality beyond on the basis of what one knows immediately about one’s own con- sciousness? One cannot deduce how it is beyond one’s consciousness simply from how it is within it: illusions, hallucinations, dreams, sceptical scenarios like that of the *brain in a vat and the Cartesian evil demon, establish that impossibility clearly enough. (*Malin génie.) So it would seem that at best one must argue one’s way to the external world through some inductive form of reasoning. But to many this has seemed hopeless if the world beyond is con- stituted by phenomena of a wholly different order and inaccessible to our experience. For how could one so much as understand such ‘phenomena’? And, besides, even if one could somehow understand them, how could one know about them? Presumably one would have to establish inductive correlations from which one could then generalize and on the basis of which one could argue from the character of one’s experience to what lies out- side. Considerations such as these, deriving from the needs of epistemology, have led philosophers in the empiricist tradition to one or another form of idealist *phenomenalism, to the view that reality is through and through constituted by experience, in the form of impressions and ideas (and, for some, subjects of such experience). Rationalists for their part have equally reasoned from assumptions concerning knowledge and understanding to metaphysical conclusions of great scope. For the rational- ist mind, reality must be comprehensible through and through. It must be possible for a mind powerful enough and well enough stocked with information to attain a complete understanding of the universe. Here the follow- ing assumptions are in play: (a) the universe is the totality of facts; (b) to understand a fact is to understand why it is a fact, why it obtains; (c) if a fact cannot be understood in its own terms, if it is not self-explanatory, then in order to understand why it obtains one needs an explanation of it, an explanation of why it obtains; (d) a complete under- standing of the universe would be an understanding of all facts; (e) X is an explanation of the fact that p (of why it is the case that p) only if X is a set of true assumptions (facts) that jointly logically imply the fact that p via some prin- ciple of lawful regularity. Fundamental *laws would be (by definition) unex- plained, there being no more fundamental laws to explain them. But that would be an obstacle to complete under- standing only if the laws in question would require explan- ation in order to be understood. Could any laws or facts be (in a sense) self-explaining? Consider the fact (F) that nothing is diverse from itself. What could possibly explain such a fact? It is not easy to think of anything else which might explain anything so fundamental. Even if it turns out that in fact there is nothing external that can supply such explanation, would that show a lack in our under- standing of F? Don’t we understand F as well as we ever understand anything, even without need of external explanation? If so, we have then in F a fundamental fact that requires no (external) explanation in order to be understood. The two relevant features of F are evidently, first, its necessity, i.e. the fact that things could not possibly have been otherwise than F says they are, and, second, the obviousness of that necessity, it being obvious that things could not possibly have been otherwise than F says they are. Any such fact will be perfectly well understood in its own terms and will need no further, external explanation. Suppose the natural order to consist of material part- icles in various configurations, moving and reconfiguring in accordance with physical law. Even if we assume that the series never had a beginning in time, why is there such a series when there might have been a different one or even a changeless void instead? In answer to this it would metaphysics, problems of 595 not do to spin an infinite series of explanations of particu- lar contingencies within the series by appeal in each case to other antecedent contingencies in that same series, and expect to have answered thus the legitimate question why there is such a series at all. Leibniz understands our predicament in roughly these terms and takes it as his own. Here is a very brief sketch of his resolution: (a) the best possible world is necessarily best; (b) *God, being necessarily omniscient, necessarily sees it to be best; (c) God, being necessarily infinitely good, necessarily wills that world to be; (d) since God is neces- sarily omnipotent, that world necessarily comes to be, and it is hence, of course, our world. If a world (or universe) is a totality of facts, then if a world W differs from a world W′ in any detail no matter how small, then W and W′ must be two different worlds. Hence if by Leibniz’s account our world necessarily comes to be, then every detail in it, no matter how small, necessarily comes to be. Not one grain of sand could have been different in its qualities or location. But this conclu- sion is not idiosyncratic to Leibniz’s particular proposal for how it might be possible to attain a complete under- standing of the universe. For, as we have seen already, however we fill in assumptions and laws in order to explain fully the existence and character of the natural order of events (e.g. of matter in motion from eternity to eternity), the resulting assumptions and laws had better be necessary facts if we are not just to extend the problem to another series of contingencies in the vertical direction. Suppose we lump all the assumptions into one big assumption A and all the laws into one big law L. If such necessary A and L are to explain the existence and character of the natural order (at least to an infinite being who could grasp it all), then by our account of explanation, A and L must jointly entail that the natural order does exist and has exactly the character it does have. But anything thus entailed by what is necessary must itself be necessary. It follows that if we have such explanation, then the natural order must necessarily exist and must necessarily have exactly the character that it does have. So, again, no grain of sand could possibly have been different in its qualities or location. And this result is thus seen to derive not just from anything special in Leibniz’s particular explanation, but from the very nature of what a complete explanation would have to be. What is made plausible by our reason- ing is that if a complete understanding of the universe is to be attainable to anyone, even to a being with access to all information and with no limit to his faculty of reason, then the universe must be necessary in every detail. (As an alternative to Leibniz’s, compare Spinoza’s very different but equally rationalistic and equally necessitarian world- view.) Once again we see how an epistemic commitment drives powerful thinkers to a metaphysical view about broad and fundamental features of reality. One can, of course, deny the commitment, denying that there is any- thing deeper than the natural order of contingent events. Unless there is some mistake in our reasoning, however, this would commit one to the view that there is inevitable opacity to reason, inevitable absurdity built into the universe, something that even an omniscient being with infinite reason could not wholly eliminate. And this con- sequence contributes to a powerful intellectual move- ment alternative to both the broad empiricism and the broad rationalism already sketched, a movement that cul- minates in works such as the Nausea of Jean-Paul Sartre, obsessed with the contingency of the world, and deriving its existentialist consequences about human life and society. In the twentieth century linguistic philosophy rejected traditional metaphysics as a pseudo-inquiry whose real point is or should be linguistic. This has taken the form of a linguistic *relativism (LR) according to which: When we say ‘There are three objects here, not eight’ we are really saying ‘The following is assertible as true in our language L: “There are three objects here, not eight”.’ This is, for example, in the spirit of Carnap’s Logical Syntax of Lan- guage, where Carnap defends the following theses: (i) Phil- osophy, when cognitive at all, amounts to the logical syntax of scientific language. (ii) But there can be alterna- tive such languages and we are to choose between them on grounds of convenience. (iii) A language is completely characterized by its formation and transformation rules. In that same book Carnap also distinguishes between: (s1) object sentences: e.g. ‘Five is a prime number’, ‘Babylon was a big town’; (s2) pseudo-object sentences: e.g. ‘Five is not a thing but a num- ber’, ‘Babylon was treated in yesterday’s lecture’; (s3) syntactical sentences: e.g. ‘“Five” is not a thing-word but a number-word’, ‘“Babylon” occurred in yesterday’s lecture’. And he defends the thesis that although s2 sentences seem deceptively like s1 sentences, actually they are really s3 sentences in ‘material mode’ disguise. Quine agrees that a kind of ‘semantic ascent’ is possible, as when we shift from talk of miles to talk of ‘mile’, but he thinks this kind of semantic ascent is always trivially available, not just in phil- osophy but in science generally and even beyond. Thus we can paraphrase ‘There are wombats in Tasmania’ as ‘“Wombat” is true of some creatures in Tasmania’. Quine does grant that semantic ascent tends to be especially use- ful in philosophy. But he explains why as follows (Word and Object, 272): The strategy of semantic ascent is that it carries the discussion into a domain where both parties are better agreed on the objects (viz., words) and on the main terms concerning them. Words, or their inscriptions, unlike points, miles, classes, and the rest, are tangible objects of the size so popular in the marketplace, where men of unlike conceptual schemes communicate at their best . . . No wonder it helps in philosophy. However, the use of that strategy is clearly limited to dis- course about recondite entities of controversial status. No relevant gain is to be expected from semantic ascent when the subject-matter is the inventory of the market-place itself. Tables and chairs, headaches and beliefs, and even good apples are no more controversial than words: in fact, 596 metaphysics, problems of some at least of these seem less so, by a good margin. No general conceptual or linguistic relativity, no avoidance of metaphysical discourse, can be plausibly supported by the semantic-ascent strategy offered by Quine. In addition, questions of coherence arise concerning LR. When we say something of the form ‘The following is assertible in our language L: . . . ’ can we rest with a literal interpretation that does not require ascent and relativization? If not, where does ascent stop? Are we then really saying ‘The fol- lowing is assertible in our language L: “The following is assertible in our language L: . . . ”’. This way lies vicious regress. But if we can stop the regress with metalinguistic reference to our sentences of L (and to ourselves), why can we not stop it with our references to headaches and good apples, and to tables and chairs and other medium-sized dry goods? Other ways of attacking the problems of meta- physics as mere pseudo-problems have also gained prom- inence and a wide following in recent decades, but this linguistic turn will have to serve as our example, and, as revealed with this example, metaphysics is neither destroyed nor even silenced by such attacks. This article has focused on the following general facts about metaphysical problems: (a) that many take the form of a trilemma among illusion, well-founded appearance, and fundamental reality, and arise in recent decades against a backdrop of naturalism; (b) that interrelated solutions to them—i.e. broad metaphysical positions— sometimes derive from epistemological assumptions con- cerning what is comprehensible or knowable, and the ways in which this might be so; and (c) that in contempor- ary philosophy metaphysical problems have been deni- grated by positivist and linguistic philosophers as pseudo-problems or as linguistic issues with a mask of pro- fundity. I have discussed particular metaphysical prob- lems mainly as examples, and there are many that I have not so much as mentioned: problems, for example, about *space and *time, about *substance and attribute, about *events and states, about *universals and particulars, and about *change and *identity through time. Discussion of these and other metaphysical problems may be found in this Companion under specific headings. e.s. *causation; descriptive metaphysics; idealism; materi- alism; opposition to metaphysics; pseudo-philosophy; revisionary metaphysics. R. M. Chisholm, On Metaphysics (Minneapolis, 1989). M. Dummett, The Logical Basis of Metaphysics (Cambridge, Mass., 1991). T. Honderich, A Theory of Determinism: The Mind, Neuroscience, and Life-Hopes (Oxford, 1988). J. Kim, Supervenience and Mind (Cambridge, 1993). S. Kripke, Naming and Necessity (Cambridge, Mass., 1980). T. Kuhn,The Structure of Scientific Revolutions, 2nd edn. (1st edn. 1962) (Chicago, 1970). D. Lewis, Philosophical Papers, i and ii (Oxford, 1983). T. Nagel, The View from Nowhere (Oxford, 1986). R. Nozick, Philosophical Explanations (Cambridge, Mass., 1981). H. Putnam, Reason, Truth, and History (Cambridge, 1981). W. V. Quine, Word and Object (Cambridge, Mass., 1960). N. Rescher, A System of Pragmatic Idealism, i–iii (Princeton, NJ, 1992–3). metaphysics, revisionary: see revisionary metaphysics. metempsychosis: see reincarnation. method, joint. J. S. Mill proposed to unify two of his five canons of experimental inquiry, the *method of agree- ment and the *method of difference, in a third, the ‘joint method’: namely, ‘If two or more instances in which the phenomenon occurs have only one circumstance in com- mon, while two or more instances in which it does not occur have nothing in common save the absence of that circumstance, the circumstance in which alone the two instances differ is the effect or the cause or an indispens- able part of the cause of the phenomenon’. However, this canon allows the possibility that a type of phenomenon under investigation may have more than one type of cause or that it may have a single underlying cause that is not revealed. Nor does Mill’s method show how the strength of a causal hypothesis may be only a matter of degree, though he accepts this elsewhere. l.j.c. *method of concomitant variations; method of residues. method, scientific: see scientific method. methodic doubt: see doubt. method in philosophy: see philosophical inquiry: first premisses and principles. method of agreement. J. S. Mill’s A System of Logic (1843) proposed the ‘method of agreement’ as the first of five canons of experimental inquiry. It determines that ‘If two or more instances of the phenomenon under investigation have only one circumstance in common, the circum- stance in which alone all the instances agree is the cause (or effect) of the given phenomenon’. For example, if an alkaline substance is combined with an oil in several otherwise different varieties of circumstance, and in each case a soap results, then the combination of an oil and an alkali causes the production of a soap. It is thus not an observed regularity of co-occurrence that evidences the cau- sation but an observed elimination of all but one hypothesis. However, to secure this elimination we need to test, and may not in fact know, all the eligible hypotheses. l.j.c. *method, joint; method of concomitant variations; method of difference; method of residues. method of concomitant variations. The fifth of J. S. Mill’s five canons of experimental inquiry (A System of Logic (1843)). Phenomena which vary concomitantly can be assumed to be causally related, whether one causes the other, or they are effects of a common cause. The method is useful, Mill thinks, for cases where the methods of method of concomitant variations 597 agreement and difference cannot be applied because we are facing phenomena which can be neither excluded (to see what happens in their absence) nor isolated (to exclude irrelevant factors); but it has limitations, he adds, because we often cannot tell whether all of one phenomenon relates causally to the other, and it cannot tell us what happens outside the limits of the observed variations. Like the other canons it assumes that there are causes to be found within the sphere of our present knowledge. a.r.l. J. S. Mill, A System of Logic, bk. 3, ch. 8 (London, 1843). method of difference. J. S. Mill proposed the ‘method of difference’ as the second of five canons of experimental inquiry. It determines that ‘If an instance in which the phe- nomenon under investigation occurs, and an instance in which it does not occur, have every circumstance in com- mon save one, that one occurring only in the former; the circumstance in which alone the two instances differ is the effect, or the cause, or an indispensable part of the cause of the phenomenon’. For example, when a man is shot through the heart, it is by this method we know that it was the gunshot which killed him: for he was in the fullness of life immediately before, all circumstances being the same except the wound. But in some cases it may be difficult to establish that two instances have every circumstance in common save one. l.j.c. *method, joint; method of agreement; method of con- comitant variations; method of residues. method of residues. J. S. Mill’s fourth canon of experi- mental inquiry was entitled the ‘method of residues’: namely, ‘Subduct from any phenomenon such part as is known to be the effect of certain antecedents, and the residue of the phenomenon is the effect of the remaining antecedents’. For example, said Mill, if the movements of a comet cannot be wholly accounted for by its gravitation towards the sun and the planets, the residual feature must be explained by the resistance of the medium through which it moves. But Mill recognized that in practice we may not be able to be certain that one particular factor is the only antecedent to which the residual phenomenon may be referred. So any induction by the method of residues needs to be confirmed by obtaining the residual phenomenon artificially and trying it separately, or by deriving its operation from otherwise known laws. l.j.c. *method, joint; method of agreement; method of con- comitant variations; method of difference. methodological holism and individualism. There are two large debates in *social philosophy or social ontology, both with methodological ramifications for the *philoso- phy of social science (Pettit, The Common Mind). One is concerned with how far human beings (non-causally) depend on their social relationships for the possession of the ability to think, or for the possession of some such characteristic human capacity. The atomist denies any such dependence while the non-atomist asserts that it obtains. The other debate is concerned with whether the existence of aggregate social entities—in particular, the obtaining of aggregate-level regularities—means that human beings do not conform in full to our commonplace psychological image of them as more or less autonomous, more or less rational creatures. The individualist denies that aggregate entities entail any compromise of such commonplace psychology while the non-individualist maintains that there is some more or less significant com- promise involved. There are also many other methodologically relevant debates that are loosely associated with these divisions. Some examples are the debates over whether aggregate- level social theory is reducible to psychological theory; whether individual-level explanation is preferable to aggregate-level explanation in social science; whether social scientific discovery is likely to force any revisions on our commonplace psychology; whether individual sub- jects are reciprocally influenced by the aggregate entities they constitute, as they form the concepts of such entities; whether individual agents are so constrained by the cir- cumstances of their social setting that we need only attend to those circumstances—we can ignore psychological matters of belief and desire—in predicting what they will do; and whether historically significant individual actions are generally dispensable, in the sense that had the indi- viduals involved not done what they did, there would have been others to take their place. The term ‘methodological individualism’ is usually employed with a variety of connotations across the range of positions identified here. The self-described meth- odological individualist will certainly be an individualist in the sense defined above; he will probably be an atomist; and he will tend to go for the position that is thought most flattering to the status of the individual in each of the other debates. The term ‘methodological holism’ is less com- monly employed and its connotations will vary in the same way. p.p. Philip Pettit, The Common Mind: An Essay on Psychology, Society and Politics (New York, 1993). methodology. The philosophical study of *scientific method. The central question arising from this study is how to interpret methodological statements. There are three alternatives: description, convention, prescription. Under the first option, methodological statements are either interpreted as descriptions of scientific practice, or methodology is seen as a ‘science of science’ which estab- lishes correlations between practice and results. Just as sci- ence has methods which allow the successful study of electrons, so too philosophers could apply the scientific methods of, say, *cognitive science or *biology to the study of science itself. Objectors to such an approach point to the lack of a non-contentious stock of results and methods in the human sciences. Therefore, the human sciences would not be able to provide a consensus on what 598 method of concomitant variations the methods of science in fact are. An obvious reply would be to advocate the application of the methods of physical science itself. Of course, if we do not know what these methods are, we cannot apply them. Thus it seems descriptivism is either question-begging or viciously cir- cular. A common reply to be found in the writings of the *Vienna Circle and Quine is that a virtuous spiral is a better geometrical analogue than a vicious circle. Under this account, the application of a method to questions of method provides a sharpening of both the method and the questions asked. If, as Popper has argued, scientific method constitutes the rules which govern scientific behaviour, these rules may be as conventional as the rules which govern the game of chess. A problem arises if two mutually contra- dictory rule books are proposed: Which game of science should be chosen? The obvious answer is to decide which set of rules is more ‘useful’ or ‘suitable’. This assumes, of course, that we have non-conventional criteria of ‘useful- ness’ and ‘suitability’. We could appeal to the intuitions of practitioners about their activity. There are two possible sources of these intuitions. In the first case they result from the practitioner’s previous experiences of similar activities, and their association—or lack of association— with some desired outcome. This answer seems to require knowledge about what methods are correlated with what outcomes, which is the central problem of descriptivism. In the second case there is a correct answer to the question what sort of rationality motivates the rules of science. This leads us to our third position. According to normativists, methodological impera- tives are true or false much as ethical norms are under an objectivist account. In its pure form, this view is not widely held, with the exception of decision theorists like Keynes and the later Carnap. A problem arises if we ask whether a transgression of such norms could make an observable difference to the life of the transgressor. On the one hand, if it makes no difference, one might question the subject-matter: are judgements of rationality really so much hot air? On the other hand, if transgression does make a difference, the desirability of following the norms of rationality would depend on the factual differences in outcomes that result. Thus this position is also in danger of collapse into descriptivism. n.c. t.chi. r.f.h. *Mill’s methods; science, problems of the philosophy of. Rudolf Carnap, ‘Inductive Logic and Rational Decisions’, in R. Carnap and R. Jeffrey, Studies in Inductive Logic and Probabil- ity (Berkeley, Calif., 1971). K. R. Popper, The Logic of Scientific Discovery (London, 1980), ch. 2. W. V. Quine, The Pursuit of Truth (Cambridge, Mass., 1992), ch. 1, sect. 8. Meyerson, Émile (1859–1933). Born in Lublin, became a naturalized Frenchman, and worked for Jewish agencies after a brief spell as an industrial chemist. Meyerson wrote on philosophy of science and general epistemology, his main interest being the nature of thought as exemplified in its successful products. An anti-positivist, he argued, for example in Identity and Reality (1908), that scientific knowledge attempts to reach beyond mere descriptive and predictive laws to an under- standing of the nature of the reality beyond appearances. The human mind seeks the permanent behind phenom- enal change, the identity within diversity as exemplified in conservation laws, such as the law of inertia and the law of conservation of energy. And yet this identity which our reason apprehends (or perhaps constructs) cannot embrace the totality of reality, for there is also change. a.j.l. *positivism; thinking. Émile Meyerson, Identity and Reality, tr. Kate Loewenberg (London, 1930). microcosm: see macrocosm and microcosm. Mill, James (1773–1836). Scottish thinker who, after being educated at Edinburgh University, came to London and worked for a considerable time as assistant and publicist for Bentham. Most famous for the strenuously intellectual education to which he subjected his more famous son, John Stuart Mill, he wrote influential pamphlets on educa- tion and government from a utilitarian point of view, as well as a thoroughgoing associationist psychology, The Analysis of the Phenomena of the Human Mind of 1829 (which was later republished with extensive notes by his son). His most discussed philosophical work is the short pamphlet On Government, which is a rigorous a priori argument for majoritarian *democracy: since everyone acts in their own interest, only the greatest number can be relied on to protect the greatest happiness of the greatest number. r.h. *utilitarianism. Jack Lively and John Rees, Utilitarian Logic and Politics (Oxford, 1978). Mill, John Stuart (1806–73), son of James Mill. He was the greatest British philosopher of the nineteenth century, bringing Britain’s traditions of *empiricism and *liberal- ism to their Victorian apogee. The System of Logic, a product of his thirties, published in 1843, made his reputation as a philosopher. The Princ- iples of Political Economy, of 1848, was a synthesis of classical economics which defined liberal orthodoxy for at least a quarter of a century. His two best-known works of moral philosophy, On Liberty and Utilitarianism, appeared later— in 1859 and 1861. In the 1860s he was briefly a Member of Parliament, and throughout his life was involved in many radical causes. Among them was his enduring support for women’s rights—see The Subjection of Women of 1869. The leading element in Mill’s thought is his lifelong effort to weave together the insights of enlightenment and romanticism. He subscribed unwaveringly to what he Mill, John Stuart 599 . relying in the main on considerations (which owe much to Aristotle) concerning the supposed nature of the world which point to the need to assume the existence of a deity. Aquinas also took, with modifications,. phenomenon relates causally to the other, and it cannot tell us what happens outside the limits of the observed variations. Like the other canons it assumes that there are causes to be found within the sphere. in the sense defined above; he will probably be an atomist; and he will tend to go for the position that is thought most flattering to the status of the individual in each of the other debates. The