The Problem of Fit
The problem of fit is the puzzle of why powers only seem to team up with particular other powers and not with every power. Traditionally, it appeared that the fit was always between an active power with a passive power. However, I think we need to seriously consider whether this is how things really work or whether it is merely a feature of our ordinary conceptual scheme, generated by our fairly superficial perception of the world of ordinary middle-sized dry goods. Because, in the disparity between our everyday conceptual scheme and the scientific image, it is not just that the apparently asymmetric interactions between different kinds of powers are explained by science to be symmetric interactions, but arguably also, to a great extent at least, interactions between the same kind of powers. The water and salt molecules do not interact in terms of different types of powers, but in terms of interactions that are all electrostatic. Billiard balls exert forces on each other in virtue of their respective momenta. Particles repel or attract in virtue of charge. In the scientific image, it appears that like power responds to like. If that is the case, the answer to the problem of fit is that objects interact in so far as they possess the same powers. However, again we have come to a point in the development of this view that is in its infancy, and I leave it at that.
What Is Doing the Work: Powers or Particulars?
As I have argued, it is the powerful particular that exercises influence and which is influenced, rather than the events in which the particular is involved. I could therefore be interpreted as taking side with substance causation over event causation (but see qualification below). However, contemporary powers-based accounts are typically either silent on the role of the particular bearing the powers, or explicitly argue that it is the powers and not the particulars that are doing all the work (Mumford & Anjum 2011; Buckareff 2017). My position is the reverse; that it is the powerful particular, rather than the powers of the particular that do the work, even though of course the particulars exert different kinds of influences depending on what kinds of powers they have. The difference may seem minuscule but matters for our understanding of the inner structure of a persistent entity, and for any attempt to tie together our understanding of causation, persistence, and change.
In my view, a power—like any property—is just the way an object is, and powers aren’t the kind of entity that can be independently engaged in an activity, nor are they entities that can change. To say that an object’s momentum is doing something to another object’s momentum, while the object that has the momentum isn’t doing anything, seems like a category mistake. In saying this, I don’t think I differ much from other writers that have argued in favour of substance causation, i.e. the position that particulars can be causes, such as Lowe (2008), Helen Steward (2012), and Ann Whittle (2016). We agree the particulars and not their powers are doing the work. However, my view is that interactions—not particulars— are causes although I view interactions as mutual exertion of influence between particulars rather than between powers. To clarify, Lowe, Steward, and Whittle understand causation more along the standard view and therefore understand interactions as unidirectional actions, and
Powers 145 consequently think that one of the interacting objects is doing the causing, while the other changes.
The argument that contemporary writers appeal to for the primacy of powers is based on the same observation that Hobbes once made, namely that it appears that particulars exert an influence not merely because they are particulars but because they have certain powers. In Hobbes’ words: ‘the agent hath its effect precisely such, not because it is a body, but because such a body, or so moved’ (1656: Ch. IX, 3). However, we would do well to observe that Hobbes doesn’t really draw the conclusion that we must therefore regard powers as entities that can in themselves exert an influence. He consistently continues to write that it is the Agent that acts on the Patient and not that a power acts, even though they act and are acted upon in virtue of possessing certain powers. To insist that powers are somehow separate and independent from their bearers seems to rely on the acceptance of the modern analytic conception of substance, and the deficits of that view are spelled out in Chapter 7. On the Aristotelian view of substance, properties are the way an object is, rather than a separate entity that is somehow coupled to the object. So, even if an object acts in a certain way not merely in virtue of being a bearer of powers, but because it has this or that power, it doesn’t automatically mean that the power is the right kind of entity to act.
Another problem with powers-based accounts that ignore the role of the particulars is that they fail to connect explicitly to questions about the proper subject of change and to issues about persistence. This is admirably put into focus by Williams (2019: Ch. 8), although I do not agree with Williams’ favourable views on perdurantism and even creation ex nihilo. But I do agree with him that, to properly connect powers-based accounts to traditional disputes about change and persistence, we have to address the role played by particulars.
The fact that the role of the particular is largely ignored in the literature has serious repercussions. For instance, it leads C.B. Martin to use a flawed analogy—the two triangles example—to illustrate his view, which in turn allows Mumford & Anjum to interpret that analogy as proof that Martin is not presenting us with an alternative view of causation but as an alternative to causation, i.e. not a view of causation at all (Mumford & Anjum 2015b). Martin’s example of two paper triangles that together ‘mutually manifest’ a square when put side by side is meant to illustrate the contrast between his and other views on mutual manifestations:
You should not think of disposition partners jointly causing the manifestation. Instead, the coming together of the disposition partners is the mutual manifestation: the partnering and the manifestation are identical. This partneringmanifestation identity is seen most clearlywith cases such as the following. You have two triangle-shaped slips of paper that, when placed together appropriately, form a square. It is not that the partnering of the triangles causes the manifestation of the square, but rather that the partnering is the manifestation.
(Martin 2008: 51)
To be sure, it is an ill-chosen example. The triangles can form a square without exerting any influence on each other at all, and without producing any intrinsic changes at all in each other, and thus without becoming a unitary entity. It is an example of a mere Cambridge change. If Martin is not guilty of simply choosing a bad analogy, Mumford & Anjum are right to understand him as advancing what they call a ‘compositional view’, in which powers come together to constitute a mutual manifestation rather than producing genuine changes in anything (Mumford & Anjum 2015b).
Now, I find it difficult to reconcile this ‘compositional’ understanding with Martin’s overall view but admit that it is not easy to decisively refute it. But it does at least seem reasonable to say that if the bearers of powers had a more protruding role in the discussion about powers, it is unlikely that Martin would have been tempted to use the example. At the very least, he would have realised that his readers would be wondering whether he actually thinks causation is a mere Cambridge change, and explicitly address that worry. The fact that he uses the example without addressing the worry leads me to think that the triangle example only signifies a bad day at work. All he wants to say, I think, is that powerful particulars come together to jointly influence each other and that as they influence each other they change to give rise to the effect; that is one sense in which you can say that the coming together is the mutual manifestation. The following passage can support that interpretation:
In coming to a new account of cause-effect, one should put the matter in the active voice in terms of the manifestings of the manifestations of the disposition in its reciprocatings with its reciprocal disposition partners as active partnerings with mutual manifestings. That is cause-effect redesigned and improved.
(Martin 2008: 4)
The passage is a mouthful, I’ll give you that, but it is not easily reconciled with Mumford &c Anjum’s interpretation of the two triangles passively merging to manifest a square.
Indeed, I find that Mumford & Anjum are also guilty of not taking into account the role of the particular, wherefore they too become guilty of using a flawed analogy—the vector model—to illustrate their view of causation. They suggest we can illustrate powers-based causation using vector arrows situated in what they call a ‘quality space’ (2011: Ch. 2), building on a suggestion by Lombard (1986). The term ‘quality space’ is meant to capture the idea that a change in an object is always with respect to some or other property, and that an object always changes from one determinate qualitative state to another determinate state by passing continuously through the intermediary states. An object which accelerates from an initial velocity vt to a final velocity vn will accelerate through all the intermediary velocities v„ vy v4 ... before reaching vn. We can thus think of any specific qualitative state as occupying a position on a continuum of adjacent determinate states, that continuum being thought of as the dimension of change for that quality (how it can potentially change). The vector model thus represents the dimensions of change of the sum total of (causally relevant) properties located at a given spacetime point during a causal interaction.
The quality space can be either one, two, or three-dimensional; each dimension representing a spatial dimension. In a one-dimensional quality space, the locus of the powers—their position in spacetime—is represented with a vertical line, and vector arrows pointing either left or right represent the potential dimension of change of each power, towards some or other manifestation property in the given situation. In a three-dimensional quality space, the locus of the powers in spacetime is represented with a dot in a 3D coordinate system, and the power-vectors point outwards in all directions from the dot.
The flaw of the one-dimensional model is that it cannot be used to illustrate that in a real-world interaction the powers belong to at least two different objects. Consequently, it depicts causal interactions merely as a nexus of co-present powers that can change along a given dimension regardless of how the powers are distributed between the objects involved. Really, like the parallelogram used in classical mechanics, the one-dimensional vector model can only accurately represent the forces operating on one object at a time. It is therefore not really a model of an interaction, but only of one side of an interaction.
The three-dimensional model could possibly be used, because it allows the introduction of multiple loci of powers from which arrows emanate representing the dimensions of potential change. Mumford & Anjum never do that, and explicitly express that they think it makes no practical difference to use the one-dimensional model for the sake of simplicity (2011: 45). Since they only use the one-dimensional model, my point about the flaw in the analogy is still valid.
The moral is: don’t forget about the particular in your account of causal interactions. After all, any account of causation that aims to explain how causes produce effects in a manner that accommodates the moral of the genetic principle—that production is never coming into being out of nothing or complete annihilation—must clarify the role of the continuants, because only they are able to persist through an interaction to provide the raw material out of which a new state of affairs comes into being.