Jump to ContentJump to Main Navigation
Powers$

George Molnar and Stephen Mumford

Print publication date: 2006

Print ISBN-13: 9780199204175

Published to Oxford Scholarship Online: October 2011

DOI: 10.1093/acprof:oso/9780199204175.001.0001

Introduction: George Molnar and Powers

Chapter:
(p.1) Introduction: George Molnar and Powers
Source:
Powers
Author(s):

Stephen Mumford

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780199204175.003.0001

Abstract and Keywords

This introductory chapter discusses the purposes for the creation of this book and briefly tells about the nature of metaphysics and the questions it tries to answer. It sheds light on the debate to which the late George Molnar was contributing and trying to answer, and shows in detail the background to the works of Molnar and Powers. The chapter also gives an account of the most controversial claims of the book. Lastly, the chapter explains the history of the unfinished manuscript and indicates the principles used to reach its current state.

Keywords:   metaphysics, Molnar, Powers

This is a book of analytic metaphysics by the late George Molnar. It concerns subjects that a number of contemporary metaphysicians regard as the most central and important. Its author had a theory that was virtually complete when he died. He planned to continue working on the book and apply the theory to a number of other problems in philosophy. Unfortunately, that work remained unfinished. The theory itself, however, was complete enough, and worked out enough, that it can stand alone. We can only speculate on how much better the book would have been had Molnar seen through his project to completion.

This introduction has a number of purposes. First, a context is set for the debate to which Molnar was contributing and some of the problems are established that he was trying to solve. Second, the background to Molnar's own work is detailed. This will include some biography but will lead to an account of his other contributions to philosophy, during two spells in the profession. I will then consider the argument of Powers itself, during which I will try to identify what is important and controversial in the work. I will justify a claim that Molnar's theory is a substantial contribution to the existing debate. There will be further detail on two of the most controversial claims of the book: that there is physical intentionality and that there are ungrounded powers. Finally, I will explain the history of the unfinished manuscript and indicate the editorial principles that saw it through to its current form.

The Debate

An area of metaphysics that has increasingly concentrated minds is the issue of dispositional properties. What are they? How do they (p.2) differ from other properties? Are they bona fide? How do they relate to other categories such as events, causes, and laws? Philosophers have wanted to answer these questions because the notion of a disposition has been useful in both the philosophy of mind, most notably in Ryle (1949), and the philosophy of matter. Physical dispositions are long recognized; indeed, Ryle explained mental dispositions as analogues to well-known and accepted physical dispositions such as solubility and fragility. More recently, however, physicists have invoked properties of fundamental particles that have an appearance of dispositionality. Further, some philosophers are arguing now that the laws of nature may be explained in terms of the dispositional properties characteristic of natural kinds.

Philosophers have said widely varying things on the question of what dispositions are. Some follow the empiricist line, of Humean origin, that states they are nothing at all. This view finds expression in a conditional analysis where the ascription of any disposition can be rephrased as affirming the truth of a conditional that has no dispositional elements. An ascription of solubility to x, for instance, means nothing more than ‘if x is placed in liquid, x will dissolve’. The opposite view is that dispositions are real and ineliminable properties, which can be distinguished, for instance, as being the causal powers of objects, and it is this realist line that Molnar defends. The realist line has come under constant attack from empiricist adversaries. Empiricists argue that there is just no need to invoke a separate category of powers in addition to categories such as events and their categorical (non-power) properties. If there is nothing more to the ascription of a power property than asserting the truth of a conditional, and that conditional mentions only events with their categorical properties, then power ascriptions can be reduced away into non-powers. Carnap (1936–7) had argued this line, though the precise form of his ‘reduction sentences’ needed some refinement. Ryle fell into the same category and was a defender, if anyone was a defender, of the ‘naive’ conditional analysis. In contemporary metaphysics, David Lewis (1998) has been the chief advocate of the Humean view and he has tried to show that, although the naive conditional analysis has problems, a reformed version is tenable that preserves its Humean spirit. Molnar argues against this view, primarily in Chapter 4.

The opposition to powers has not taken this form only, however. In addition to the conditional analysis, there has been a line of argument (p.3) based on a principle of microphysical reduction. David Armstrong (1968), (1973) was a chief proponent of this line. The central idea of the account is that to each disposition of a particular there corresponds a categorical property of that particular such that the disposition is reducible to that property. When a glass is fragile, for instance, its fragility may be entirely explained by the substructure of the glass, such as the bonding between molecules. The persistence of such a substructure may explain what it is for a disposition to be possessed by an object between manifestations. Such properties would explain the truth of counterfactual conditionals, therefore, which disposition ascriptions seem to entail. Molnar has arguments against this position, mostly presented in Chapter 8.

Realism about powers is a view that has gathered momentum in the contemporary debate. There have been a number of landmark contributions, such as Mellor (1974) and Martin (1984, 1993c, 1994). My own Dispositions (1998) was intended to uphold the view. Since then, Brian Ellis (2001, 2002) has done a fine job in defence of realism about dispositions. Molnar worked on the present book before Ellis's were published. Ellis uses a realism about dispositions in an attack on the whole Human metaphysic. Only in Molnar's Powers, however, do we get a detailed defence of the ontological status of power properties. Within realism and the anti-Humean movement, this book ought, therefore, to be considered one of the key texts.

George Molnar: The Man and his Work

George Molnar was a multifaceted man. He was born on 14 May 1934 into a Jewish Budapest family. George, together with his whole family, faced Nazi persecution but were saved from the concentration camps by a Swedish diplomat, Raoul Wallenberg, who used bogus documents and Sweden's neutrality to keep thousands of Hungarian Jews from the gas chambers. After the War, George arrived in Australia, where his father had already fled.

In 1953 he started at Sydney University, reading Economics, but switched to Philosophy and was taught by the influential John Anderson until 1956. However, he dropped out of formal education in the hope of making a living as a professional gambler. This was not a success but he got by until returning to complete his degree and (p.4) graduating in 1964. His ability was rewarded with a tenured position at Sydney and he was highly regarded for his lecturing, which was mainly in political philosophy. During this time he produced noteworthy work such as ‘Defeasible Propositions’ (1967) and the respected and anthologized ‘Kneales' Argument Revisited’ (1969).

Molnar then became gripped by the spirit of the times. He became a leading light in the bohemian and anarchist movement known as the Sydney Push. He was part of the Libertarians, the intellectual wing of the Push, who recommended anti-authoritarianism and sexual freedom. His political principles led him to believe that the position of an academic philosopher was morally untenable in current society and in 1976 he resigned his position. He decamped to England, settling in Leeds, and took up the causes that had become his passion. These included nuclear disarmament, far-Left revolution, women's rights, children's rights, gay rights. At the time, he worked at a crèche he had set up and took part in many protest movements. He moved in with his long-term partner Carlotta McIntosh and both returned to Australia in 1982. He took up what appeared to be a respectable position at the Department of Veterans’ Affairs but this was still in the interest of one of his causes, namely the plight of the Vietnam veterans. During this time of public service he worked on a number of publications, some for the DVA and some philatelic.

He rose to a senior position as assistant director of the Disability Assessment Unit, but in 1996 he was able to return to academic philosophy. He produced a number of papers and worked on the current book, Powers. In addition to the papers associated with the Powers project, George had a paper accepted by the Australasian Journal of Philosophy on ‘Truthmakers for Negative Truth’ (2000). In 1998 he was appointed the Senior Research Fellow at Sydney University to edit John Anderson's papers. According to Carlotta McIntosh, this was the happiest time of his life. He was pursuing his work with a renewed vigour, was working at the cutting edge of contemporary metaphysics, and was fulfilling the promise in philosophy he had shown earlier in his career. Then, in August 1999, he had a heart attack on the steps of the University's Fisher Library and died in hospital the following week.

Just four published papers in metaphysics may not seem a lot for a lifetime and might not qualify their author as one of the great metaphysicians of our time. An examination of them reveals a (p.5) notable philosophical intellect, however: meticulous, incisive, and elegant.

In ‘Defeasible Propositions’ Molnar considers this troublesome class of proposition and shows how no simple and reductive treatment of them is easily found. Defeasible propositions have the unusual feature of being both general but also permissive of counter-examples. Power terms fall into this category. Hence, it is true that water freezes below 32°F even though there are circumstances in which some particular sample of water does not freeze below 32°F. Defeasible propositions may remain true despite exceptions. We cannot reduce such a proposition to a universal statement with an ‘all else being equal’ clause because such clauses cannot be explicated in a non-trivially true way. But Molnar does have a positive proposal:

The moral to be drawn from these considerations is that the analysis of a defeasible proposition must include some reference to a principle of relevance which functions as a principle of exclusion ranging over all simple property predicates, known or unknown. To say that the standard F is G is to say that any F is G if it has those properties which, according to the appropriate principle of relevance, suffice to distinguish it from all exceptional cases. (1967: 189–90)

Two years later ‘Kneale's Argument Revisited’ appeared. This paper concerned William Kneale's argument concerning laws of nature (1950, 1961), which Molnar thought had not been given due consideration or a satisfactory response. The argument was that, on a certain broadly Humean account of laws, we cannot say that propositions express unrealized empirical possibilities, though that is what we would ordinarily take them to express. The problem arises when laws of nature are taken to be adequately expressed in propositions that are universally quantified, omnitemporally and omnispatially true, contingent, and containing no local predicates (such as ‘in Smith's garden’). In this Humean account, if nothing is F, anywhere or anywhen, then it is a law of nature (or statement of a law of nature) that ‘Nothing is F’. But this entails that ‘Something is F’ is inconsistent with a law of nature and thus not a possibility. Hence, if there is never, anywhere, a river of lemonade, the statement ‘there exists a river of lemonade’ is deemed not to express a possibility. Kneale's argument forces us to decree such statements as either true or impossible, as ‘there exists a river of lemonade’ will be possible only if it is at some time true. This conclusion is counter-intuitive, as we would ordinarily take such (p.6) statements to be false but possible. The argument sets up a puzzle, therefore. It shows that there are problems for this account of laws when coupled with certain accounts of truth and possibility.

Molnar went on to discuss four possible ways of resisting Kneale's conclusion, for example, placing extra empirical requirements on laws such that, if ‘something is F’ is false, it follows that ‘nothing is F’ is true but not necessarily that it is a law. Molnar favoured a different way of resisting the conclusion, however, which Kneale also seemed to prefer. This was the strengthening modality strategy that rejects the contingency of laws of nature. Molnar was suggesting, though admittedly not proving, a necessitarian view of laws of nature. He saw it as the best answer to Kneale's argument. It deems that from the falsity of ‘something is F’ one cannot infer that ‘nothing is F’ states a law of nature. More importantly, the strategy, once endorsed, resists Kneale's argument without any undesirable consequences. Molnar says very little about the necessitarian view of laws, however, other than that such a view is able to answer Kneale's argument. The paper shows him to be one of the first modern proponents of a view that currently has a growing popularity (see Bird 2001, Ellis 2001, and Lombardo 2002).

David Armstrong acknowledged the clarity and importance of this paper when he quotes it as an example of the problem of the regularity theory and refers back to it throughout his study of laws in What is a Law of Nature? (1983). This is all the more impressive when one considers that Armstrong cites relatively few sources and that ‘Kneale's Argument Revisited’ is still held as an exemplar fourteen years after publication. In the passing of those years, others had attempted more illuminating discussions of the issues but failed to improve upon the work of Molnar, who had then retired from professional philosophy.

From a philosophical perspective, 1976–96 appeared to be Molnar's wilderness years, but Carlotta McIntosh, who was with him throughout, is able to tell us that he retained his interest in metaphysics. This is further evinced by the discovery of an earlier version of Powers found among his papers that probably dates from around 1983. The interest in powers predates the existing book by some years, therefore.

Molnar returned to professional philosophy in the late 1990s and this was marked with his return to print with ‘Are Dispositions Reducible?’ (1999). As this was an integral part of the Powers project, I will not pass comment on its arguments until later. The period also produced a free-standing piece of philosophy, however, with (p.7) ‘Truthmakers for Negative Truth’ (2000). Molnar argues in this paper, which he saw accepted but did not see in print, for the wholly negative conclusion that there are, as yet, no discovered positive truthmakers for negative truths. The finding brought him no joy, however, as we need such truthmakers for the negative truths which clearly are true. He finished the paper by candidly admitting his own lack of success:

I confess, with much gnashing of the teeth, that the Holy Grail of positive truthmakers for negative truths remains undiscovered. We need positive truthmakers for negative truths but we have no good theory of what these might be. That is the sad conclusion from the arguments of this paper. I have criticised proposals by other philosophers for solving the problem of negative truths, but that criticism must be tempered by the acknowledgement that where they have failed, so have I. It is an impasse and at present I cannot see the way out. (2000: 85)

Molnar impresses the urgency of the problem on us by setting it up in the following way. He offers a realist metaphysics that holds:
  1. (i) The world is everything that exists,

  2. (ii) Everything that exists is positive,

  3. (iii) Some negative claims about the world are true,

  4. (iv) Every true claim about the world is made true by something that exists.

Claims (i)–(iv) jointly imply that negative truths have positive truth-makers. But Molnar proceeds to show how all accounts so far offered, which attempt to provide positive truthmakers for negative truth, fail. Thus, ruled out is the exclusion of the negative truth by a positive truth-maker, as already dismissed by Bertrand Russell (1918). Absences of truthmakers will not work, as they would have to postulate negative facts. Molnar shows that there are good reasons to think that there are no negative facts: they would be mysterious, they would fail the Eleatic Stranger's reality test (Plato's Sophist 247e) by being acausal, and they cannot be directly perceived, contrary to the claim of Richard Taylor (1952: 444–5). Totality facts which, together with positive facts, could serve as truthmakers for negative truths, are rejected; not least because they are not positive facts. They are ‘no more’ facts, which look negative.

Need one really find these truthmakers for negative truths? The obligation can only be avoided if one rejects one of (i) to (iv), above. (p.8) But for any realist, the denial of any of (i) to (iv) is difficult. A fifth escape is to deny that the truthmaker has to be something that exists, but Molnar sees this as a desperate move also. The obligation remains and that is why Molnar accepts the failure to find such truthmakers as his own failure as much as those who have preceded him.

The Argument of Powers

There are four distinct sections to Powers, which would have corresponded to separate parts had the book been completed. The first section (Chapters 1 and 2) sets out a general metaphysical background against which the theory of powers is to be developed. This is not as detailed, or as introductory, as intended. There was to be a different first chapter that, like many first chapters, the author was to write last. It would have eased the reader into metaphysics and the issue of powers, but almost nothing of it has survived. We do, however, have discussion of the substantial and important commitments necessary for an understanding of Molnar's theory. Molnar argues that properties are tropes: non-repeatable particulars as opposed to universals. Both realism and nominalism are in part right and in part wrong, necessitating a move to tropes, which retain the best features of realism and nominalism. Next, Molnar argues for selective realism about properties. Properties and predicates are not isomorphic, in agreement with Armstrong's rejection of the ‘argument from meaning’ (1978: ch. 13). This leaves us with a ‘sparse’ theory of properties, where best science, not philosophy, tells us which properties there are. A number of distinctions are then stated and clarified with a view to their deployment later in the book.

Chapters 3 to 7 offer the main theory of powers. This is presented in the form of a fivefold characterization of powers with each chapter describing and defending one of the features. These are directedness, independence, actuality, intrinsicality, and objectivity. By directedness, Molnar is claiming that there is such a thing as physical intentionality on a par with the mental intentionality discussed by Brentano and others who have followed him. This claim will be one of the most controversial of the book. By independence, Molnar means that the existence of a power is independent of the existence of its manifestation. Hence, a fragility trope can exist without its manifestation (p.9) (in breaking) ever existing. Powers exist whether manifested or not. It is here that Molnar dismisses the famous (or infamous) conditional analysis of power ascriptions. In Chapter 5, Molnar only briefly defends the actuality of powers. He thinks it absurd to defend in depth something so obvious. Chapter 6 defends the intrinsicality of powers. Powers are intrinsic properties of their bearers, so having a power is independent of the existence of any other object and this is contrary to, so requires a rejection of, Popper's account of propensities as properties of the entire experimental set-up. The final characterizing feature of powers is objectivity. Physical powers do not depend on how we cognize them. This is a rejection of the Humean view that all necessary connections are in some sense mind-dependent.

Having characterized powers so, Molnar enters a third section in which he answers some of the further questions that must be addressed before we have a completed theory of powers. Chapters 8 is on the relationship between powers and their grounds in a so-called causal base. Molnar rejects the claim that all powers must be grounded. Although many powers do appear to have such a causal base, the powers of the subatomic particles appear to have no sub-structure so cannot be causally based. The groundedness claim is not borne out empirically, therefore, providing philosophers with the problem of the missing reduction base. Molnar categorizes and dismisses each of the resisting responses that have been offered to the problem, from the claim that the missing base is there but unknown to the claim that such powers are ‘ultra-grounded’ (see 8.4.2) in relatively macroscopic properties. The best response, therefore, is acceptance: there are ungrounded powers. But this leaves further work to be done. We will have to explain the difference between a grounded and ungrounded power. We will have to give up causal analyses of powers in general because we have accepted that, for some, there is no causal base. Chapter 9 develops further the ontology of powers. The groundedness of those powers that are grounded is explained in terms of derivability—one of the concepts explained in Chapter 1 (1.4.1). A taxonomy of theories of the ontology of powers is introduced. The taxonomy differs from that in Dispositions (Mumford 1998: 1.5) in some key respects. We agree that the division between dualists and monists is the most important division but Molnar divides the monists into pan-dispositionalists, categoricalists, and neutral monists, whereas I had used the less transparent terms dispositional monists and categorical (p.10) monists for the first two of those subdivisions. An initial evaluation is given of these positions but it is only in Chapter 10, after considering whether there are any non-powers, that Molnar states his own preferred position. Because he thinks there are non-powers, he opts for a property dualism: there are both powers and non-powers. What are the non-powers? In brief, they are the S-properties, which include spatial location, temporal location, spatial orientation, and so on. These, basically positional, properties fail the test for powers. They are not directed, independent or intrinsic properties, as described in Chapters 3, 4, and 6. Nevertheless, the S-properties have causal relevance, so pass the Eleatic Stranger's reality test. Where objects are located makes a difference to what effects they have on each other (10.3). How can something be a non-power yet have causal relevance? The locations of objects affect the outcomes of the workings of the powers (10.4). Alternative theories of what count as non-dispositional properties are then shown not to match this account (10.5).

Chapter 11 is a consideration of some objections to the general theory of powers that Molnar has offered. He defends his theory against two main charges that pull in opposite directions. He summarizes the two objections thus: ‘According to one, ontological seriousness about irreducible powers empties the world of something that it contains. According to the other, it imports into the world something that does not exist’ (11.1). Against the first objection, Molnar shows that his theory is not subject to a vicious regress. Against the second objection, he defends the necessary connections denied by the thesis of Humean distinctness. This completes the theory of powers.

There was to have been a lengthier final part, ‘Powers at Work’, in which the completed theory of powers was applied to various other areas of metaphysics in an attempt to show the connections with, and centrality of, powers. The book's subtitle, ‘A Study in Metaphysics’, indicates that Molnar did not see powers as some peripheral and specialist sub-area of metaphysics. Rather, it is one of the most important parts and could be the key with which we might unlock many other philosophical problems. His task was to show how powers, understood in the way he has described, relate to various problems in an enlightening way. Unfortunately, just two problems were addressed in a substantial form: those of causation and modality. There is every indication that Molnar was hoping to offer similar treatments to a host of other issues.

(p.11) Molnar's Theory of Physical Intentionality

Something more should be said on the key, controversial claims of the theory.

In recent times, some philosophers have argued that there is such a thing as physical intentionality and that this is the mark of the dispositional rather than the mark of the mental. These philosophers are Ullin Place, who died in January 2000, Charlie Martin, and John Heil. In Powers, Molnar offers the most developed defence of the view that intentionality is the mark of the dispositional. Additionally, he argues that physical intentionality demonstrates the irreducibility of the dispositional but also that there are non-powers. This leaves him with a dualism of properties: there are two distinct kinds. These appear to be controversial claims. How can they be defended? Is Molnar's defence of physical intentionality any advance on the previous versions, which have been attacked, for instance, by Crane (1998) and myself (1999)?

The notion of intentionality comes from Brentano (1874). Famously, he suggested it as the mark of the mental. The key notion in intentionality appears to be directedness, though there are others associated with it. A thought (belief, desire, emotion, perception, etc.) always seems to be directed at, or be about, something. To believe is to believe something, that is, to believe that p. To fear is to fear something. To see is to see something. One thing that seemed to mark off such mental intentionality from any directedness in the physical world, such as the directedness of an arrow towards a target, was the possibility of intentional inexistence. I can fear an intruder who is not really there but exists only in my imagination. Despite the lack of any intruder, my fear is real and it is indeed directed towards an object that, outside my mind, does not exist. The view that intentionality is the mark of the mental still persists. But not all follow the line. Some see consciousness as the best way to distinguish the mental from the physical.

Martin and Pfeifer (1986) argued that the typifying features of intentionality are applicable to physical causal dispositions as much as to mental phenomena. A disposition is characterized, for instance, in terms of that to which it is directed. Dissolving, for example, can be understood as that towards which solubility is directed. As such an event need be possible only, because something soluble need never dissolve, there appears to be the possibility of intentional inexistence. Martin and Pfeifer's argument does not automatically support the conclusion (p.12) that there is physical intentionality, however. The notion may be bolstered to avoid the comparison. But Martin and Pfeifer ended the argument there.

It was Ullin Place (1996a, 1996b) who carried the argument forward. He argued that Martin and Pfeifer's argument could be used to support the claim that intentionality was the mark of the dispositional— physical and mental dispositions—rather than uniquely the mark of the mental. There was, therefore, physical intentionality. But Place's argument did not appear absolutely conclusive. He argued that certain features of dispositions were best accounted for by them being intentional states. But what if these same features could be explained another way? Such an explanation was my goal at one time (Mumford 1998).

Powers advances the debate in a numbers of ways. Molnar shows in detail that the directedness of physical causal powers meets all the traditional marks of intentionality that have developed in the post-Brentano literature. This includes the serious, non-linguistic, criteria of intentionality; hence Molnar's account is not clearly vulnerable to Crane's attack on non-mental intentionality (Crane 1998: 248). Molnar accepts the following characteristics of intentionality:

  1. (i) An intentional state is directed to something beyond itself, the intentional object.

  2. (ii) The intentional object can be existent or non-existent.

  3. (iii) There is indeterminacy of the intentional object, which depends on ‘partial consideration’.

  4. (iv) There is referential opacity and non-truth-functionality.

Characteristics (i) and (ii) cannot be dismissed as merely linguistic. They are seriously ontological. Nevertheless, Molnar argues that there is physical intentionality because:

  1. (i) Physical powers, such as solubility, are directed at something beyond themselves, their manifestations.

  2. (ii) These manifestations need not exist/be actual.

  3. (iii) There is indeterminacy with respect to dispositions, for example, a manifestation can be indeterminate as to timing.

  4. (iv) Statements of a capacity to F are not truth-functions of F. And the intentional object cannot always be replaced with a co-referring expression in an account of that power.

(p.13) After first considering some other objections, Molnar concentrates on providing a naturalistic account of physical intentionality. The problem arises from traditional accounts that suggest that the directedness of an intentional state requires some representation of its intentional object. Molnar answers such a concern by arguing that there are states or properties that are (a) mental, (b) not semantic or representational, but (c) intentional. Showing that there are such states provides a model of non-semantic directedness, creating the conceptual space for physical intentionality. The example brought forth is pain, which is undeniably mental, but is it non-semantically directed? Molnar thinks so. Pains meet the marks of intentionality as they are felt somewhere, in a location that may be non-existent or with a fuzzy boundary. Further, pain exhibits what Grice called natural meaning. Pains may naturally mean their locations but they do not non-naturally mean them so do not require representations of them. We have, thus, a model of non-representational directedness available to us into which fits physical directedness.

The argument means that we now have two candidate ways of characterizing the dispositional: the conditional entailment criterion and the physical intentionality criterion. A compromise is suggested by Place (1999): that the difference between the two might be purely verbal. The compromise could be developed into the view that intentionality provides an ontological distinction between powers and non-powers, while conditional entailment is the same distinction, at the level of concepts, between power and non-power ascriptions. However, Molnar's position stands also against any such compromise. He argues that no conditional entailment criterion can work. Some disposition ascriptions entail no conditional, for instance, such as those that are manifested at random or continuously. It remains to be seen whether any argument will be produced against this crucial point.

Ungrounded Powers

A second key claim of the book is that the simple particulars of our world are, or may be, entities with nothing but ungrounded dispositional properties. Molnar states that ‘According to all indications, some subatomic particles are absolutely simple.’ He interprets current (p.14) subatomic theory as holding that such particles have only dispositional properties and the usual reductive explanation of those properties, in terms of microstructural components, is not available because a simple entity is defined as one without components.

This is one of the key battlegrounds in the debate between Humeans and anti-Humeans. Blackburn (1990) suggests that we ought to consider the acceptance of something like ungrounded dispositions but he concedes that no satisfactory ontology has been developed for simple particulars that have only dispositional properties. Molnar, more boldly, thinks we should openly accept the existence of ungrounded dispositions.

The Humean is likely to object to Molnar's account that there is no credible account of the Being, existence or actuality of such simple particulars when their dispositions are not manifested. Simple particulars cannot consist only of ungrounded dispositional properties because there would be nothing manifest—or actual—about them. Simple particulars must manifest their properties constantly, it seems. If not, they cease to exist. Further, a property that must be manifested constantly is not dispositional at all but, rather, occurrent (some would say categorical). This suggests that a category of ungrounded dispositions cannot be inferred from simple particulars. Such properties would require a bearer between their manifestations and one is noticeably absent in this case.

Molnar's claim that simple particulars have ungrounded dispositional properties is strongly realist about dispositions in that it does not have available the standard, microphysically reductive, explanation for their presence. Most frequently, this standard explanation invokes some basis for each disposition. The basis is usually understood to be categorical but not always. Some dispositions might ground other dispositions (Mellor 1974). Given that this basis is standardly taken to be at a microlevel, relative to the disposition it grounds, by definition there will be no microbasis for the dispositions of simple particulars. That is why they are putatively dispositions that are ungrounded.

Let us consider, again, the two candidate characterizations of dispositions. The conditional option is that a disposition ascription has a special relation to a conditional statement. The question might naturally arise of what makes any such conditional true. What, in the world, is the truthmaker, where a truthmaker is whatever in the world (p.15) makes a truth true? (Armstrong 1997: 2). Ordinarily, an answer is readily available. There is a microstructural base, certain other properties of the disposition bearer, which cause such-and-such manifestations upon such-and-such stimulation. In the case of ungrounded dispositions, there is no such ground and so no such truthmaker for any such conditional. As there is no reason to assert any such conditional there is, therefore, no ground to assert the existence or presence of an ungrounded disposition.

The second characterization of dispositions is the intentionality option. The same question of Being arises. To be a disposition is just to be directed towards some possible manifestation. To be an un-grounded disposition is to be so directed and nothing else. In particular, it is for there to be no microstructural basis to this directedness (what Molnar calls, and accepts, the missing reduction base). But if such a property is unbased, what in the world is it that is directed towards some possible manifestation? Such a property looks like no property at all. It is nothing more than the possibility of some future property, when there is a manifestation. An ungrounded disposition has no Being between its manifestations and such manifestations need never be actualized.

This charge is serious. Given, as we have seen, that such ungrounded dispositions are the only properties our simple particulars are said to have, their lack of Being would suggest a lack of Being on the part of their bearers. If such dispositions were unmanifested, it would appear that the particular would have no manifest properties— nothing displayed—and any particular with no manifest properties seems like nothing at all.

Molnar's argument is that we should accept ungrounded dispositions—powers—at the basis of everything. The lowest level of existence should be taken as one of ungrounded powers for which there is no further explanation. Given what I have said above, this might seem a high ontological price to pay. But, sometimes, it is worth paying a high start-up price if the eventual benefits are considerable. The application of the ontology to other problems of metaphysics would have shown those benefits. Molnar wasn't able to demonstrate these benefits but Brian Ellis has done a worthy job (2001). One might further respond to the Humean, that while the realist ontology requires inexplicable powers, the Humean ontology requires inexplicable events, the occurrences of which must be taken as basic facts. If both (p.16) ontologies can begin only by assuming their own foundations, they are in equally strong, or weak, positions in that respect and we must look for some other basis to choose between them.

The Manuscript and its Completion

George Molnar contacted me by e-mail in the summer of 1999 and told me about the book he was working on. He told me that it was near completion and asked if I would be prepared to read it and give some comments. I agreed and said I would look forward to reading the manuscript.

We would have met that December as we were both to speak at a conference on Australian metaphysics to be held in Grenoble. We were greatly looking forward to this but the meeting never occurred owing to George's death. I did meet others in Grenoble, and heard quite a bit about George Molnar, the person. I also heard of the manuscript of Powers, which was reported to be in a good state, with a wealth of worthy material but, as yet, not quite a finished book.

That seemed to be the end of the matter until, the following spring, I got an e-mail from Tony Skillen, lecturer in philosophy at University of Kent and old friend of George's. He had access to the manuscript, via Carlotta McIntosh, and asked me what I thought. My view was that if publication was to be considered, it would be best to try to preserve the manuscript as much as possible in the form George left it but that some changes might be necessary to make it a readable book. As my idea appealed to Carlotta and to Tony, they asked if I was prepared to go ahead and do it. First I needed to see the manuscript, however. As soon as I started reading it, I realized that it was work of the highest calibre and, by the end, that it could be a significant contribution to the debate. This work deserved to be read by those working in the field. I had little hesitation in offering to help, however I could, to get the work available to a reading public.

Then followed the, at times, difficult editorial work. From what can be gathered from George's papers, he had an unusual but admirable approach to his work. The argument was planned out in his head and then written up in complete and detailed draft. This had a good side and a bad side for any budding editor. The draft chapters, early on in the book, were almost complete, polished and tightly argued. The later (p.17) chapters, in complete contrast, were non-existent and almost impossible to construct as only a few fragments appear in the remaining papers. He left a table of contents that shows how the final part of his book, ‘Powers at Work’, would have developed, but little survives that matches these proposed contents. I was reminded of what David Armstrong said to me, as we travelled down from Paris to Grenoble, when I said that George had told me the book was near finished: ‘it was near finished, in his mind’. This suggests that the book would indeed have been fully written if George had lived just another month or so.

The main theory of powers is, however, complete and it has not been necessary for me to alter much to make these main chapters finished and presentable. The reader can be confident here that the work is George's and that he said what he wanted to say. My main quandary was how to deal with the later chapters of which barely anything existed. Carlotta sent me everything that remained on his computer at the time of his death. My solution was to make a final chapter out of the relevant fragments found there. This could only be a condensed version of George's thought that drops hints of how his theory would be applied to a range of problems. We can be quite sure that George would not have been happy to present this as his finished work, and that he would have worked out his position in meticulous detail. But only if I wrote a substantial new work and tried to pass it off as George's ideas could we have anything approaching that. I had to keep my obligation to Carlotta in mind and change as little as possible. That, after all, was why she wanted me to complete the book rather than anyone else. On this last chapter, however, the reader can be assured that the ideas were all George's, even if they were not as developed as he would have wanted. This is the best we can have that accurately represents what George would have done. There was also nothing by way of introduction to Powers. The book rather plunges in at the deep end. Part of the aim of this introduction has been to make up for that. (p.18)