Part I: Natures and the Limits of Individuality

1. Comment on the Humean assumption

H & M characterize the Humean Assumption as follows: What is not logically necessary is absolutely contingent, so that in a given well specified physical setup, anything whatever may be assumed to be possible.

Note that theistic Aristotelians cannot reject this assumption as it stands; they must add "given only God's general concurrence with the natural causes operative in the setup. This is a minor point, but one which leads H & M astray in their discussion of the fiery furnace in Causal Powers.

2. The distinction between nature and substance

I would prefer that H & M not draw this distinction, though again the point is a minor one. Their claim is that a rational account of the changes in a thing's powers is possible given that there is a material basis which allows us to see how changes in powers are correlated with changes in the arrangement of the perduring entities through which the thing is structured. This saves us from having to appeal to a wholly mysterious replacement of one set of powers by another.

Note that this explication of powers and changes in powers via structure (or form) cannot eliminate powers, since presumably the elements in the microstructure themselves have powers that can change. Nonetheless, it may well be that we get a deeper understanding of higher-level powers by associating them with deeper-level structures and properties. Some seventeenth-century Aristotelians seem not to have seen this point. (Note McMullin's reference to Scotus.)

3. Change into vs. replacement by

According to H & M, Humeans try to assimilate all change to replacement, so that at its core all change is mysterious and can be given no rational explanation in terms of the continuity underlying the natures of the things involved. H & M give several plausible explanations of how alterations in a thing's microstructure can explain alterations in its causal powers, where the same microentities are preserved throughout. So we can't have a fountain pen changing into, as opposed to being replaced by, a pig.

Note: If one recognizes the possibility of orderly substantial change, then one can distinguish substantial change from transubstantiation, the latter consisting in the complete replacement of a substance, including its material basis, with a new substance, having a wholly different material basis. This distinction plays the same role H & M want to use the change/replacement distinction for and is, I believe, preferable, since, given that transubstantiation can be described by the words 'change into', we can more easily see why Humeans slip into 'change into' language. The distinction we want is a distinction between two senses of 'change into', one of which presupposes and the other of which does not presuppose an enduring material basis.

Still, a problem arises at the ultimate level when we can no longer appeal to microstructure. What are we to say about the case in which an electron collides with a positron and a gamma ray results. Isn't this just like the fountain pen/pig paradigm pushed by the Humeans? Where is the something that perdures through this change in the way that microentities perdure through macrochanges?

The answer here is energy conceived by scientists as that which plays just this role of "material basis" in such cases. Matter perdures and takes on different forms (as we say). [Notice that the Aristotelian conception of matter is the conception of that which underlies change and is not tied to the notion of stuff; thus energy might, as a principle of potentiality, play the role of Aristotelian matter.]

In section IV below H & M identify the concept of a field (of potential) as the physical concept of something fundamental, since a field is the most plausible candidate for an entity whose powers just are its nature.

Part II: Natural Kinds and the Defects of Event-Ontology

     I have little to add to this discussion. The main point is that the discrete and atomistic nature of events taken as ontologically basic makes it impossible to give a rational account of changes in powers and substances, since there are no intrinsic connections between events; that is, there is no intrinsic metaphysical basis for the connection between any two events if events are taken to be ontologically fundamental. This induces the problem of what connects various properties together and concomitantly generates a serious problem of induction, one which threatens to undermine the basis for inductive practices. We can go into this in class.

Part III: The Restoration of the Concept of a Natural Agent

     Though I have a few quibbles of a mainly terminological sort here (is it power that is observed as over against action? Why equate power and agency?), I basically agree with the metaphysical and epistemological points made. I am not sure that H & M have captured adequately the distinction between natural and rational agents, but I will not worry about that here. The notion of a threshold for action is an important one in a theory of agency (see p. 204 l).

     Look at the four points on pp. 203. (I do not, by the way accept the business about the direction of time being derivable merely from the direction of agency.) Also, think about the ball/glass example on p. 204 r.

Part IV: Powers and Potentials

     This section tries to tie the discussion of powers to that of the notion of an ultimate entity or entities:

"With the concept of the field we reach a notion in physics which through the concept of potential is most directly connected to the concepts of power and agency. In particular, a field of potentials is the most plausible candidate for an entity whose powers are its nature, so that if there are any entities which are fundamental then the field is the one most likely to be so, since it meets, through that characterization the one indispensable condition for an entity to be fundamental in science" (p. 204 l)

In explicating further this notion of a fundamental entity, they say this:

"The field ... is the field of potentials, the physicist's version of the concept of powers. Potentials are the causal powers of points, and are defined in terms of the effects they are likely to produce when a suitable test entity is at the given point. Of course we have long since abandoned the dispositionalist position that the potential at a point is only what the test body would do if it were there, in favour of the idea that a power is possessed in virtue of a nature, and thus, following in the wake of this conceptual revolution the potentials of the physicist's fields are its powers, that is the dispositions of the field in virtue of its having a certain nature at that point. No particular metaphysical significance yet lies in that characterization, since the concept of 'causal power at a point' leaves open the question of whether the nature responsible for it is independently characterizable, or whether the level is so fundamental that such powers are their natures" (205 l).

Near the end of this section, the last point is developed in more detail:

"It seems then that this metaphysical discussion can be summed up in the principle that a field is a distribution of potentials, fairly permanent causal powers ... Does the field so conceived meet the requirements that might be demanded of an entity which is to be the occupant of the ultimate level of a scientific explanation? Ordinarily the pattern of behaviour of some thing or material is explained by reference to the internal structure and properties of its components. Generally speaking we begin the progression to the ultimate by explaining the behaviour of something by reference to the powers of its component parts. More rarely patterns of phenomena are referred to the entities in which the phenonmenon is embedded. The ultimate level of scientific investigation would either be marked by reference to entities which had no components in terms of which their powers could be explained, or by reference to entities which were not themselves components of still grander entities whose influences their behaviour advertised" (p. 207 l).

And, finally:

"That we leave behind the possibility of explaining the structure and relations of the distributed potentials is just what we would expect if we are in fact at the level of an entity whose properties are its powers ... Just as corpuscularianism had two versions, the theory of the aether and the theory of action at a distance, the espousal of one or the other depending on how deep one thought one was in the investigation of the world and its processes, so the theory of individuals with powers is likely to bifurcate at the same relative point. There could be a metaphysics which recognized only point-centres of power, i.e., potentials as ultimately real and thing-like, but which made of their structured interrelations the ultimate mystery. And there could be a metaphysics which recognized the field as ultimately real and fluid-like, and the potentials as its causal powers. But we do not think that it is wise to try to construct a new monadology which would try to have it both ways" (pp. 207 r - 208 l).

     Question: What are we to make of all this? What are the limits of structural explanation? How does this discussion bear on the traditional Aristotelian distinction between form (structure) and matter? Perhaps what we have here is a delineation of the simplest forms.

    Also, compare this with the following quote, in which Ernan McMullin contrasts the sort of explanation found at the deepest levels of physics with the two main patterns of scientific explanation, viz., nomothetic and retroductive (including structural):

"It would seem that there is a third pattern, dynamic explanation, which shares features of both nomothetic and retroductive explanation but is in the end irreducible to either. Because we are trying to account for the motions of basic entities, there is nothing further we can appeal to; we cannot specify the forces or fields or energies in terms of a further type of entity. We seem to be forced back to a complex set of counterfactuals (which may be taken to express causal powers or basic tendencies or the like), specifying what would happen if a particular dynamic state of affairs were to be realized. But the causal words introduced to make the explanation conform to conventional causal patterns must be taken with great caution" ("Two Ideals of Explanation in Natural Science," p. 212).