Kistler begins by introducing the idea that sciences often discover laws. Some of these are laws merely by virtue of being “accidental regularities” or by satisfying conditions that distinguish them from accidental universal statements (172). Kistler goes on to explain a plethora of positions that involve the existence and status of laws, but his central contention is that laws have the same metaphysical status as natural properties. Furthermore, he contends that even if natural properties are conceived in terms of “powers,” it will make no difference to the metaphysical status of laws.
Comment/Question: Do you think that this argument depends on realism of some kind? If these laws are not directly observable and may represent only useful fictions, does this seriously disrupt the argument?
11.2 The Distinction Between Powers and Dispositions
Kistler explains that dispositions are true by virtue of natural properties, conceived of as powers. This argument makes two assumptions:
- The distinction between dispositional and categorical qualities is semantic
- Powers are distinct from dispositions
Kistler then, in a similar vein to Chakravartty, describes the qualities that make dispositional properties dispositional. He summarizes this explanation as follows:
- “Object x has disposition D iff, ‘if x were in (triggering) circumstances T, then ceteris paribus, x would manifest M”
Kistler also makes the assumption that the attribution of D to x requires an intrinsic property that is the “causal basis of the disposition” (174). This causal basis is the “power” that grounds the disposition.
Comment/Question: Is this a fair assumption given our discussion on Wednesday?
So, what is a “power” exactly? Kistler wades into the categorical/dispositional debate that we had on Wednesday to evaluate this concept. In his view, this distinction can either be ontological or semantic. “Power,” in his view, is an ontological category while “disposition” is a semantic category. All powers can be conceived in a categorical or dispositional which points to the view of properties as having a “dual nature.” This ontological commitment, however, lacks an explanation as to how dispositional and qualitative aspects of an entity can coexist within the same property. Kistler thus concludes that the distinction is best understood as semantic where “dispositional and categorical predicates can make reference to the same property although they have, as predicates, incompatible properties” (175). Kistler means that dispositional predicates entail the counterfactual listed above while categorical predicates do not.
Comment/Question: Does it make sense to cast this distinction as semantic? It seems to resolve a lot of the things we were struggling with on Wednesday regarding capacities to be broken for example. Is it clever or cheap? Perhaps it takes some stance voluntarism out of the question because you do not have to choose between the two conceptions.
From our reading of Chakravartty and our previous discussion, it is not apparent why we should need to reference powers at all. It seems that dispositions can do all the work on their own. However, Kistler notes several examples that demonstrate that the power of electric charge branches into producing a number of dispositions (176). The concept of a power unites sets of dispositions.
Powers are, on Kistler’s view, theoretical properties that are justified by their utility. It is useful, he reasons, given its ability to unify the observed phenomena.
Comment/Question: If you are a dispositional realist, this unification does seem to work nicely. Do you think this escapes Professor Khalifa’s objection to Chakravartty’s use of unification? See post for Wednesday.
11.3 Laws as Parts of the Truth-Makers of Disposition Attributions
Kistler now turns his attention to the central argument of the chapter relating to laws. He argues two positions:
- Laws are required to explain the structure of the dispositions corresponding to one property, or power (179)
- Dispositional concepts and predicates are relational whereas powers are typically monadic
He focuses his attention on two of Mumford’s arguments. If either were true, then there would be no need for laws to make an attribution of a disposition. Mumford’s arguments are as follows:
- Powers are sets of dispositions, or dispositions are elements of powers (this is called the thick conception of properties)
- A power P contains all/some of the dispositions D that P gives its bearers.
- An object x’s power of being electrically charged contains x’s disposition to undergo forces in an electric field
- The electric field and force are contained in P
- If powers are internally related, P contains these internal relations to other powers
- Powers are mereological sums of dispositions
This thick conception is justified if it follows from it that laws of nature are conditionally necessary.
Nothing but the laws a property figures in can determine the identity of a property. For example, Coulomb’s law is approximately true and expresses a part of what it is to be electrically charged. This hold for other laws which feature electric charge. Kistler writes: “In all possible worlds in which the property of having mass M exists, it is such that pairs of objects having M attract each other with a force proportional to…the inverse of the square of their distance” (182). Laws are therefore conditionally necessary. There may, however, be worlds in which any given property does not exist, so laws are not absolutely necessary.
Comment/Question: Do other laws cover all the other parts? If the distinction between dispositional and categorical is semantic, how is something like extension governed by laws? Can it not exist without them?
It follows from this argument that natural properties and laws must come together and that laws are internal relations between properties. “In other words,” Kistler writes, “in every possible world in which b and c exist, they are related by R” (183).
Having established these conclusions, Kistler presents four arguments against the thick conception of properties.
- Laws explain the structure of sets of dispositions that are present in any given entity by virtue of a certain property. That is, “the attribution to electrons of the disposition to create a magnetic field if they are in motion…(Ampere’s law) is directly made true by the property of the electrons of being electrically charged” (183). This attribution is made in two steps:
- Possession of charge
- Ampere’s law that serves to connect electric charge to the existence of a magnetic field
The property of being charged also indirectly makes true the disposition to exercise forces on a wire carrying another electric current. P cannot, therefore, be constituted by a cluster of dispositions. Clusters are unstructured, and they cannot account for the fact that the disposition of electrons is linked to a varying degree to their properties. Laws, however, can explain these direct and indirect dispositions that arise from properties. Kistler views sets of dispositions as layers in an onion centered on P. The first shell are those dispositions made directly true by P and the outer ones are made true by P and multiple laws that become more and more indirect.
Comment/Question: Is this true? Couldn’t the linkage to properties be achieved without reference to the cluster?
- The thick conception cannot explain why there are no properties corresponding to subsets of these sets of dispositions. Postulating laws allows us to say that the set of laws containing property P determines the dispositions which are attributed to entities by P.
Comment/Question: I need some more clarification on the way Kistler argues for position 2.
- Without laws, we cannot make sense of the relations between different properties/clusters. If properties are clusters of dispositions, then the same disposition belongs to several clusters. For example, the cluster of speed and magnetic field contains the disposition to produce a force on moving charged objects. If properties are clusters of dispositions, one cluster should not constrain others; however, dispositions belonging to different clusters do constrain each other. If we adopt the thin conception, properties are related to other properties by laws. Dispositions in different clusters would be constrained by each other because they are determined by the same law.
- Dispositions are relational, so the truthmaker must be relational. Dispositions, as we have discussed, have relational conditions in order to be manifested. The truth maker of the attribution of a disposition cannot be monadic because the conditions of the disposition are relational. Laws allow us to conceive as the truth maker as composite insofar as it contains a property and relational insofar as the law relates the power to the condition in which the disposition is manifested.