The Atlas of Reality
Page 20
B = {the power, when combined with A, of producing a change from D to C; the power,when not combined with property A, of producing a change from C to D}
C = {the power of producing a change from B to A; the power of producing a change fromA to B}
D = {the power of producing a change from B to A; the power of producing a change fromA to B}
The picture is this: when a body has both properties A and B, it has the power to change another body from D to C, but not vice versa. When a body has either property A or property B, but not both, it has the power to change another body from C to D, but not vice versa. The properties C and D have powers that treat A and B equivalently. In such a world, there must be a difference between the two properties (or supposed power-bundles) A and B, since there is a real causal difference between having just one of the properties and having both of them simultaneously. However, if we try to identify the two properties with their power-bundles, we face a kind of vicious circularity. The only difference between the two bundles depends on there being a difference between A and B. If we substitute A for B and B for A, we transform the A-bundle into the B-bundle, and vice versa. Figure 6.1 illustrates the symmetry of the situation. The horizontal arrows between A and B, and between C and D, represent possible changes, and the diagonal and vertical arrows represent the causal powers of the five relevant combinations of properties: C, D, (A&B), (A&∼B), and (∼A&B).
Figure 6.1 A Symmetrical Law-Book
The roles of A and B in this law-book are perfectly symmetrical. Each can be described as one of two properties that, by itself, results in the transition from D to C, and, in combination with the other, results in the opposite transition from C to D. The two properties A and B are analogous to the two indistinguishable steel spheres in Max Black's famous thought-experiment (Black 1952) (see Section 9.3.2.1 for details). Black used his example to argue that two objects can be distinct although indiscernible. This is turn could be used in an argument for individual haecceities. In the same way, the symmetrical law-book in Figure 6.1 can be used to argue for sicceities, since the causal profiles of A and B are not sufficient to individuate one from the other.
One complication: we have to distinguish between two types of causal profile: property-relative profile and systemic profile. A property-relative profile is a set of powers (active and passive) that are defined in terms of the other properties with which the given property interacts. A property-relative profile assumes that we can identify the other properties, distinguishing them from each other and from the property in question. A systemic profile is a set of powers that are defined abstractly, in terms of positions or nodes within a network of nomic connections. The difference between the two kinds of profiles emerges in cases of symmetrical law-books, as in Figure 6.1. The property-relative profile of property A is that of causing the D-to-C transition (when by itself), and causing the C-to-D transition when jointly instantiated with B. Property B has a different property-relative profile: it has the profile of causing the D-to-C transition (when by itself) and causing the C-to-D transition when jointly instantiated with A (not with B).
In contrast, the two properties have exactly the same systemic profile in this case. Both have the profile of being the first of four properties P1, P2, P3, and P4, of causing the transition from P4 to P3 when instantiated by itself, and of causing the opposite transition from P4 to P3 when instantiated with another of the four, P2. In the case of the profile of property A, the role of P2 is played by B, P3 by D, and P4 by D, while in the case of property B, the role of P2 is played by A. Nonetheless, the systemic profiles are themselves exactly the same.
Thus, we can distinguish two forms of Causal Structuralism: a strong form in which properties are individuated by their systemic causal profiles, and a modest form in which properties are individuated by their property-relative causal profiles. The modest form is really a version of Dual-Aspect Theory (discussed in the next section), since we need something like sicceities to distinguish between two property-relative causal profiles (see Hawthorne 2001: 376). If A and B do not have different sicceities, what distinguishes the property-relative profile of A in this case from that of B if not the sicceities of A and B? If we cannot account for the real distinctness of the two property-relative profiles, we cannot account for the real distinctness of A and B.
Therefore, it seems that there must be something in addition to the powers themselves (a difference in qualitative characters or sicceities) that accounts (in a non-circular way) for the difference between A and B. Only by accounting for the difference between A and B in this way can we account for the difference (in such a world) between the power of instantiating both properties and the power of instantiating just one of them.
6.2.2 Dual-aspect vs. mixed two-category theories
C.B. Martin has articulated and defended what the Dual-Aspect Theory of properties (Martin 1994, Armstrong, Martin, and Place 1996). On this view, every property has two aspects, its causal profile or powers, and its sicceity. Since properties have sicceities, they cannot always be individuated by their powers. It is possible for two distinct properties to have the same causal profile. At the same time, on Martin's view, every property has at least one power and has that power (and all of its other fundamental powers) essentially.
To make this idea concrete, return to the two thought-experiments from the last section: swapping phenomenal red for red*, swapping one type of electron for two. In each case, Dual-Aspect Theorists will conclude that there is a real difference. Even if all the red-lookings became red*-lookings, and had every other causal connection to things that red-lookings actually have, they would still be phenomenally red and not red*, by virtue of having the sicceity of phenomenal redness and not that of red*-ness. Similarly, there is something about the property of being an alpha-electron that makes it intrinsically and essentially different from the property of being a beta-electron, even if the causal profiles of the two properties are indistinguishable. If, in the next moment, alpha- and beta-electrons were to swap places, the world would change dramatically, even if we couldn't tell the difference.
Like Dual-Aspect Theory, the Mixed Two-Category Theory posits the existence of dual-aspect properties. In addition, it proposes that there are pure non-powers, properties which have their causal profiles contingently. The most plausible examples of pure non-powers are spatial and temporal properties and relations, such as size, shape, distance, and duration. If there are other pure non-powers, we can call them ‘pure qualities’.
Def D6.4 Pure Quality. A pure quality is a fundamental, non-spatiotemporal property that has its fundamental powers contingently.
Mixed Two-Category Theory requires either that spatial and temporal relations have their fundamental powers contingently or that there are pure qualities or both.
It seems we can imagine a perfectly inert world, in which material bodies occupy places at times, even though no material body has any active or passive power. Let's suppose that each body has as its only immanent power the power of inertia—the power to keep moving in a constant velocity. We can imagine that the bodies are made of a ghostly matter, so they are able to pass through each other without disturbance. In such a world, it would seem that the spatial and temporal relations would exist, even though they had no power of any kind, whether active, passive or immanent. If so, then all spatial and temporal relations must have all of their fundamental powers, if any, contingently.
Do spatial and temporal properties have any powers at all, even in the actual world? They seem merely to provide occasions for the exercise of the powers of other things, by providing instances of contact and proximity needed to trigger the active and passive powers of body or instances of temporal duration needed for the exercise of immanent powers. Space and time in and of themselves seem to be causally inert, never causing or preventing anything by themselves. If this is right, then the Mixed Two-Category Theory would seem to be correct.
6.3 Objections to Strong Powerism
<
br /> The arguments for Powerism have already been presented, in the form of objections to its competitors. Let's turn to some objections to Powerism itself.
1. Implausibility of necessary laws of nature. Both Dual-Aspect Theory and the Mixed Two-Category version of Powerism entail that most of the laws of nature—all of those laws that describe the fundamental powers of fundamental kinds of things—are metaphysically necessary, since they are entailed by the essences of the kind-properties involved. This entails further that all of the derived laws that are logical and mathematical consequences of the fundamental laws are also metaphysically necessary. For example, it would turn out to be necessary that water freezes at 0 °C.
We could certainly imagine a world in which water freezes at some other temperature, a world, for example, in which water freezes at the same temperature and pressure as mercury. Since imagination is a guide to possibility (PEpist 1), this gives us good reason to believe that such a world is possible. The best versions of Powerism must treat these reasons as misleading in this case. This is at least a theoretical cost.
2. Determinism and recombination (Patchwork Principles). Jonathan Schaffer (2005) has argued against the thesis that the laws of nature are metaphysically necessary, on the grounds that necessary deterministic laws would drastically undermine our knowledge of alternative possibilities. If the laws of nature are both metaphysically necessary and deterministic, then there is only one possible world that begins as our world does, namely, the actual world. If the laws of nature are deterministic, then there is no possible world that agrees with our world up to some point of time and then diverges thereafter, since this would involve a violation of the world's laws, but those laws are supposed to hold in every possible world.
Worlds in which things go differently from the actual world would have to be different from the very beginning. Worlds that look like ours at the beginning but diverge would have to be populated partly or wholly by things with alien properties, properties that are never instantiated at all in the actual world, because only such alien properties could obey laws of nature other than the ones we are familiar with.
Such a picture of possibility would deprive us of the method of recombination: the ability to construct alternative possibilities by using mental “scissors and paste” to take apart pieces of the actual world and recombine them in new ways. Yet, this seems to be exactly how we go about forming our opinions about what is really possible. This way of forming our modal opinions can be captured by various Patchwork Principles (Principle of Metaphysics 5), which say (roughly) that recombined possibilities correspond to further possibilities. Without recombination and the associated Patchwork Principles, we would never be in a position to know anything about how things might have gone but didn't.3 We examine Patchwork Principles in much more detail in Chapter 16.
The best response for Powerists is to deny that the laws of nature are, or even could be, deterministic. Every law of nature must ascribe certain tendencies and propensities to things, but it must never prescribe exactly one possible outcome for each interaction. This is another cost to the theory, since we seem to be able to conceive of worlds with deterministic laws, and some philosophers and scientists believe that the actual world is deterministic.
3. The mystery of “natural intentionality.” As George Molnar has pointed out (2003), the thesis that powers are fundamental properties means that a kind of intentionality, similar to that displayed by human beliefs and intentions, is built into the very fabric of reality. A property is intentional if it essentially involves some relation to another thing or to a situation, without entailing that that other thing or situation be actual. For example, I can imagine a golden mountain or think about a round square, even though no such thing exists. Similarly, a power to attract negatively charged particles could exist, even if there were in fact no negatively charged particles to attract. Opponents of Powerism object to such properties as fundamental, insisting that intentionality must somehow be reducible to non-intentional facts.
However, this objection suffers from the fact that no one has yet succeeded in reducing intentionality to non-intentional facts, despite many attempts to do so. Since intentionality is an undeniable reality, taking natural intentionality as fundamental seems a live option.
In addition, the mysteriousness of the intentionality of powers can be dispelled to some extent if we embrace Possibilism (12.1A.1T), the thesis that some merely possible things (in particular, merely possible future events) do not actually exist. A powerful property could be an internal relation that ties its bearer to merely possible future events. Suppose, for instance, that powerful particulars are always involved in a process of some kind and that processes involve relations among things and events, both actual and potential. Let's consider a single electron that is in the process of motion through space, giving rise to a magnetic field with the power to exert force on other moving, charged particles. The electron's magnetic power is an aspect of its being in motion, a paradigmatic case of a process. This process of motion ties the electron to certain actual past events, events involving the past location of the particle and its actual effects on other particles in its neighborhood. The process also ties the electron to future, merely potential events, events involving both certain future locations of the electron and events involving the forceful acceleration of particles in its vicinity. Some of these potential events may never actually occur. In fact, if the electron's process of motion were immediately brought to an end by encountering some new force or by the electron's annihilation, all of the process's future events might be prevented from becoming actual. For example, it can be true that an armadillo was in the process of crossing the road when it was crushed by a passing truck. The process was one of crossing the road because it related the armadillo to a future event of its reaching the other side of the road, but the passing truck ensures that that future event is never actualized (for a further discussion of processes and causation, see Chapter 28).
4. Pairing active and passive powers. Another possible objection to Powerism concerns the coincidence of finding matching pairs of active and passive powers. Why, if F's have the active power to make G's H, must G's have the passive power to H when acted on by F's? The simplest account would be one in which either active powers alone are fundamental (and all passive powers derived) or all passive powers are fundamental (and all active powers derived). But neither answer seems more plausible than the other, and it can seem implausible to take these coincidences as everywhere brute, unexplained necessities (PMeth 1.2).
The best answer to this problem involves carefully distinguishing between a causal power and our descriptions of it, or, to put it more precisely, between those descriptions that capture the true essence of the power, considered in isolation, and those descriptions that tacitly take into account the existence of other powers.
Here's a suggestion about how this disentangling might go. Let's suppose that water has the passive power of being capable of being heated, period. This fact about water doesn't entail that fire, in particular, should have the active powers of heating water, but only that there could be properties that confer such an active power. At the same time, fire might have the active causal power of heating anything with the passive power of being capable of being heated, or perhaps, of heating anything with this passive power together with certain other characteristics. Fire could have this active power, regardless of whether water, in particular, has the passive power of being capable of being heated. The two facts together entail that fire can heat water: that is, these two fundamental facts entail the two derived facts that fire has the active power of heating water, and water has the passive power of being heated by fire. No circularity or unexplained coincidence is required.
5. Negative causation. We've saved the most difficult problem for the last. We often speak about an event's being caused by an absence. In fact, as Jonathan Schaffer (2004) has pointed out, such negative causation is extremely common
. Every time one fires a gun, for example, there is a case of negative causation: pulling the trigger removes an obstacle between the firing pin and the bullet. It is the absence of the obstacle that causes the gunpowder to be ignited. Every time a nerve signal passes through the brain, a case of negative causation is involved: it is the absence of certain chemicals in the synapse between two neurons that causes an electrical impulse to be transmitted. Many cases of death (perhaps all) involve negative causation: the absence of oxygen in the bloodstream to the brain.
We can also speak of absences as effects. The causation of an absence is a case of prevention. We can prevent an explosion by cooling a chemical compound, thereby causing the absence of an explosion. One can prevent a stampede by calming a jittery steer or prevent an accident by swerving one's car to the right.
How is this a problem for Powerists? Powerism requires that we attribute causal powers to absences. But only things can have powers (whether active or passive), and absences aren't things!
Here's another way to put the point: Powerists conceive of causation as a real connection between things. If absences are nothing, then they cannot be connected to other events, either as causes or as effects. We will take up this problem again in Section 27.2, including some strategies by which Strong Powerists might meet this challenge.
6.4 Conclusion
Each of the four positions we have examined—Hypotheticalism, Powerism, Nomism, and Neo-Humeism—has its advantages and disadvantages. Neo-Humeism has the leanest metaphysical theory, since its class of fundamental truths includes no conditionals, laws of nature or attributions of power. In addition, the Ramsey/Lewis account of natural laws provides a simple explanation for our preference for simple theories about the laws. However, it faces some serious challenges, both metaphysical and epistemological. The cheif metaphysical challenge is that Neo-Humiesm denies that causal powers are intrinsic to their bearers and consequently forces us to claim that certain conceivable small worlds are impossible. The chief epistemological problem is that Neo-Humeism has to treat induction as a primitive postulate of reason, despite the apparent abundance of chaotic possibilities.