Part I
Metaphysics Chapter 1
The Power of Power
STEPHEN MUMFORD
1. Philosophy
In some ways it was a disadvantage while in others an advantage that I did not study mainstream analytic philosophy as an undergraduate. My degree was in history of ideas (major) and politics (minor) in a small provincial town that knew little of the mainstream of received orthodoxy. The philosophy I took was Marxism and Existentialism though with some of the great works of political philosophy thrown in. I thus studied a miscellany of deep and troubled thinkers such as Kierkegaard, Nietzsche and Heidegger, and knew Hobbes, Locke and Mill only as political philosophers. I read no Hume at all, nor Aristotle, and only Kantâs moral philosophy. The only logic we were taught was syllogistic.
At times I am aware that I lack some of the training of my peers. I had to teach myself analytic philosophy on the hoof, often encouraged by the necessity of teaching it, but all the while conscious of huge gaps in my knowledge. At times I felt an imposter. But one thing I did bring from my undergraduate studies was a sense that philosophy was an important matter that should deal with the biggest questions. If you take Marx, Hobbes or Nietzsche as your model, you have a sense of the scope and ambitions of philosophy. It is about very serious tasks: building an all-encompassing system; explaining the human condition; showing what is to be done, facing unwelcome truths. I have never felt that my proper business was tinkering at the edges: for instance, developing minor amendments to other peopleâs theories that make little difference to the world. I had grand designs. If I was going to bother with philosophy at all, it would be so that I could offer a systematic view of the world. It should be radical, different: I should challenge rather than defend the paradigm. And any theory I developed should explain everything. To this day I worry whether this ambition is a pretentious vanity.
My initial work was piecemeal and I didnât know how it would all hang together. But I have started to think of late that the hope of a single, all-encompassing theory might be realised. In this paper I outline for the first time how I think these ideas might form a systematic metaphysics and possibly realise the impetuous ambition of my youth.
2. Mind
Temporarily casting aside my deep and troubling books, I landed by chance at the University of Leeds, commencing masterâs studies in philosophy of mind. I lived almost on fresh air, having to take part-time work in a comic-book shop and then at the University library. On my entry, Leeds seemed a bastion of late-Wittgensteinianism. Advancing a positive thesis in the Department did not seem the done thing when there was instead the business of finding problems in what philosophers said. Some may have thought of me as the fly voluntarily entering the bottle. But change was coming. My arrival coincided with that of Robin LePoidevin, a PhD supervisee of Hugh Mellorâs, who taught that we should take metaphysics seriously. This was the start of a major transformation for Leeds and also for me.
One of the first books I set to reading, before meeting Robin, was Gilbert Ryleâs (1949) The Concept of Mind, as it was suggested reading on the course. It was between those covers that I first came across the notion of a disposition and it was an encounter that would have a profound effect on my intellectual development.
Ryle argued against the myth of the ghost in the machine. There was no Cartesian inner theatre of thoughts and sense data, hidden from everyone elseâs view. Instead, the mind could be understood as a complex bundle of dispositions. To believe that it is raining is just to be disposed to behave in a certain way: to be disposed to take in the washing, pick up an umbrella on oneâs way out, or not go out at all, and so on. And Ryle says the same of emotions, desires and even sensations. To be in love is just to have a complex bundle of dispositions: to be disposed to buy flowers, write bad poetry and a host of other things. To be in pain is just to be disposed to behave in a certain way: to moan, to protect an injured body part, to seek medication.
What interested me was not just the account of mind, which had the potential to be qualia-free (see Dennettâs 1991 development of the view), but what Ryle had to say about dispositions. Ryle didnât quite go so far, but his is often the view thought of when we speak of the simple conditional analysis. He compared mental dispositions with ordinary, non-mental ones, such as solubility and fragility. When we say that a thing or substance is soluble we mean something more or less along the lines of: if it is in liquid, then it dissolves. The ascription of a disposition gives us an inference ticket. The disposition neednât show itself all the time but I will nevertheless be able to infer that in certain circumstances it would. To say that someone speaks French, for instance, doesnât mean that they are speaking French all the time, or now, but means that under certain circumstancesâon holiday in Provence buying bread, for exampleâone can infer that they would be speaking French.
I loved reading Ryle. But when I got to the end, I thought that something was lacking. Why were these conditionals true? What underwrote an inference ticket? What sustained them when they were not being put to the test? What kind of being did they have? And if there was nothing making them true, why should we accept these conditionals? Ryle resisted any temptation to answer these questions, for if there were something that grounded these conditionals during the times when their antecedents were false, this thing could be precisely the sort of inner state that Ryle was denying. It could be exactly a thought or sensation in the Cartesian inner theatre. Such a worry prevented Ryle from taking disposition ascriptions at face value: as the ascriptions of properties.
3. Dispositions
The problem I had discovered was already known. Indeed it is precisely the worry that supposedly led C. B. Martin to propose the truthmaker principle. (It is Armstrong 2004: 1â3 who attributes this principle to Martin.) In his case, the concern arose from Berkeleyâs phenomenalism. According to Berkeley (1710), to claim that there is a chair in your room is to claim that if someone enters the room they will have chair-like sensations. But the chair is not an independently existing object that causes these sensations, so he says. We can see that the view and associated problem is entirely parallel to Ryleâs view of mind. Both material objects and then mind are analysed away in terms of the truth of a conditional. The conditional describes the possibility of a sensation or the possibility of behaviour, respectively. But Martin asked what the truthmaker was of such conditionals. They are contingent truths, not true merely analytically, for instance, but then it seems that if they are true, they have to be true for a reason. For a conditional to be true, there has to be some feature of the world that grounds the possibility of the consequent upon realisation of the antecedent. So what made it true that someone in love would write poetry when they had a spare moment?
There are a number of ways in which this question could be answered and they immediately take us into the area of what a disposition is. One view is that such a conditionalâs truth just depends on the pattern of events. It may just be that within a certain period of time, a loving person does indeed write poetry for his beloved. The truth of the conditional may just supervene on this regularity. We may not be entirely sure what the regularity is, from our limited temporal perspective on it, and our belief in the truth of the conditional may be based only on inductive inference. (Quine 1960: 224 is a representative of this empiricist approach.) On the other hand, there are those who believe that there is a causal basis of the disposition. There is some persisting state, it is claimed, that is able to cause the manifestation when certain conditions are realised (Prior, Pargetter and Jackson 1982). It then becomes relatively easy to think of the having of a disposition as the having of such a causal basis that gives a certain kind of response or manifestation when it receives a certain kind of stimulus.
And what is the nature of this causal basis of the disposition? Some have used the term categorical to describe it (Armstrong 1968: 85â9). There is an alleged categorical property that grounds the dispositional behaviour of the bearer. In the case of fragility, for instance, it is alleged that the disposition resides in the crystalline structure of the glass or pottery. But what is meant by the term categorical? The way it is typically used, it just means non-dispositional, but that tells us little positive. And while metaphysicians have put dispositions under intense scrutiny, the notion of a categorical property seemed to have slipped under the radar, evading our attention.
4. Dualism
Setting aside the issue of what a disposition is, it was clear that a further question was now also raised. Are the dispositional and categorical two different types of property? There was often an assumption that dispositions were properties and they were set in opposition to categorical properties. This sets us a puzzle. Is the world populated by two entirely distinct types of property? Some have drawn this conclusion (Prior 1985), which is a view I called property dualism. But just as there is a problem of interaction for mind-body dualists, I worried about how two such distinct types of property would interrelate. Do all dispositional properties have a categorical basis, as some think? If so, does that mean that a disposition is a mere second-order property: the property of having a certain (categorical) property? If so, it would look as if it is the categorical properties that do all the real causal work: that are productive of their manifestations, for instance. As Prior, Pargetter and Jackson (1982) show, however, when the categorical properties do all the causal work, then the dispositional properties that they ground are rendered causally impotent. They were content to draw precisely that conclusion but it seemed to me wrong (Mumford 1994).
This concern was foremost in my mind when I wrote my first book, Dispositions (Mumford 1998). I wanted to make room for both the categorical and the dispositional without leaving either of them impotent and without requiring whole scale overdetermination of effects. The solution I offered was similar to the one articulated by Mellor (2000). I argued that there were just properties in the world, all of which were powerful. The dispositional/ categorical distinction was just a conceptual and epistemological one, not an ontological one (first ventured in Mumford 1995). Dispositions were properties functionally characterised, according to what they did.
I was satisfied with the way in which I set up some of the problems in Dispositions, and some people still think it a good place to start on the topic. But in time I looked for something more radical. I became involved in George Molnarâs book Powers: A Study in Metaphysics (Molnar 2003), which he left incomplete at the time of his death. I edited a finished version from the material he left behind. There was a simple criticism of my view in that book that nevertheless struck home. One is hardly a realist about powers if what makes them powers is simply how we think about them, how we characterise them, or how we know them. I certainly wanted to be a realist about dispositions or powers (and I use the two terms synonymously) but there was something not thoroughgoing enough about the realism in Dispositions.
5. Properties
Forced into a subtle but significant rethink, the connection between properties and powers seemed a crucial issue. What was a categorical property supposed to be? Nancy Cartwright once asked me this question on a Paris street, and she was genuinely mystified. I had to confess that I did not know. It seemed that all properties had to make some causal difference to the world, or at least had the possibility of making such a difference. Did we really want to allow any properties at all that were causally impotent or inert? I thought in response to the Prior, Pargetter and Jackson theory that we certainly did not want dispositions to be inert, but would the situation be any better if there were categorical properties that were causally impotent?
I looked again at Sydney Shoemakerâs (1980) early theory of properties. He argued that properties were clusters of conditional causal powers: conditional in the sense that they could do things when suitably partnered with other powers. A certain shape, plus hardness, gives a knife the power to cut, for instance. Shoemaker was claiming that properties were constituted by causal powers. He subsequently withdrew this view, stating only that properties endow powers (Shoemaker 1998: 412), but I have stated elsewhere that I think there is no good reason to back away from the original theory (Mumford 2008a).
It is interesting that Shoemaker used the example of a particular shape (being knife-shaped) as an example of a property constituted by a cluster of powers. While there are few theories of what it is for a property to be categorical we are sometimes offered examples, and shape is frequently cited. Shape is supposedly a good example of a property that is non-powerful (see Prior 1982 and Mellorâs 1982 reply). Consideration reveals this to be an implausible assertion. The way in which things behave is to a large and clear extent determined by their shape. The reason a knife can cut and an apple canât, although both objects are hard, is the differen...