Casteñeda's purpose is to investigate the notion of production, which is central to the idea of causality. In doing so, he touches upon themes we have mentioned before and pushes the discussion forward in an interesting way which we have the resources to assess and criticize.

I. Three Principles Concerning Causality (from pp. 17-19)

    (C.1*) We are often more certain that an item c has caused an item e than of any causal law under which c and e fall.

    (C.2*) We establish causal generalizations on the basis of singular causal instances.

    (C.3*) If an item c causes an item e, then there are properties F-ness and G-ness and circumstances Z such that c is F, e is G, c is in Z, and if any item of the same category as c is and in Z, then it causes an item that is F and is G of the same category as e.

Note: The last principle, the so-called principle of the piggybackness of causality, seems too strong and indeed seems to entail a strong form of determinism. However, there may be some other more plausible principle that tries to tie causity (and causal power) to various structures or forms.

II. The counterexample to Hume's explication of the paradigm instance of causality

Casteñeda produces a counterexample which preserves all of Hume's official criteria for causality but in which the production of the effect clearly comes from a cause other than the one which satisfies Hume's criteria. He then goes on to claim that the essence of causality consists in the communication of a sort of "world orderliness" that he calls causity. Let us ponder the following two paragraphs.

    "There is something perplexing about a transfer of motion. We seem to be hypostasizing motion. On the other hand, the motion of an object is rather a string of successive states of the object in question, each state lasting only an instant or moment. Thus, the view that the motion of ball A is destroyed ... and that the motion of ball B is created at the appropriate subsequent instant looks better. But at the surface ontological level, the level at which we deal with causes either in daily experience or in scientific research, it really does not matter whether we suppose that something passes, so to speak, bodily across time from the cause to the effect, or we merely suppose that something in the cause is destroyed at the moment of causation while a counterpart, even replica, of it is created in the effect. Whatever the profound metaphysics of time and causality may be, the crucial thing is that on the surface of it causation is communication and transmission of something in the cause to the effect, whether by replication or by actual carrying over across time. This transfer of something, including the transfer of certain orderliness, is the substance of causation.

    "In our example, nothing leaves ball A to go to ball B. Mechanism M absorbs the motion of ball A fully. But it is not required that anything in mechanism M receives that motion--although something may do so. In any case, we know that there are causes that do not, at least in appearance, involve motion. But perhaps every case of causation is at bottom the transfer of motion from the cause to the effect, or from the objects involved in the event that is the cause to the objects involved in the event that is the effect. This is an empirical matter that we cannot decide by means of philosophical reflection on the general nature of causality. The fundamental point is that in causation there is a transfer (or metaphysical replication) of something in the setup containing the cause to the setup containing the effect. For convenience, neutrally, so as to beg no empirical issues, let us dub causity that, in general, which is characteristically transferred from the causal network to the effectal network. Thus, with the same neutrality we adopted in the preceding paragraph regarding the transfer of motion, we shall speak of the transfer of causity. And as before, we allow that there may not be a literal transfer across time of some selfsame item, but only an appropriate matching of a vanishing causity at a given spatiotemporal position and a contiguous creation of causity--perhaps of a different species--at an adjacent, or the same, space-time location" (pp. 22-23).

Some questions: Is Casteñeda operating at a high enough level of abstraction? Why does something have to be destroyed in the cause? We can conceive of a perpetual heat source that by nature maintains a steady level of heating capability without consuming anything internally. Would such a heat source fail to be a cause of heating simply by virtue of the fact that nothing is destroyed in it? Again, why talk of creation here? Does he mean creation ex nihilo? That seems far-fetched.

Compare what Casteñeda says with the following passage from Charles Hart's Thomistic Metaphysics (Prentice-Hall: New York, 1959):

    "On the finite level, from the standpoint of the subject, [efficiency] involves change and the actuality of a passive power, which is the accident of passion. From the standpoint of the effect, it involves a certain becoming or production of something. From the standpoint of the agent, it implies the actuality of its active power, what we called its `second act', which in turn involves a previous transition of the agent from the inaction of mere power-to-act to actuality. Now while all three elements are required in the efficiency of which we have experience, we can immediately say that the factor of subject with the change in it (accident of passion) is not required by efficiency as such, as witness the creative efficiency of God by whose action the whole existence of the effect is produced without subject or change.

    "It is limited efficiency that makes this element necessary, not efficiency as such. Nor will the agent's transition from power to actuality be a necessary preliminary if the agent is entirely in act. In any case the change in the agent's active power is not part of its efficiency as such, but a preliminary in finite efficiency. Therefore in efficiency as such, we are left with a being called agent in a state of perfection or actuality, radiating a corresponding production in some other being or some other part of the same being considered as distinct. Without this production of effect, the action (actuality) is not properly efficient nor is the agent truly agent. It is precisely this production of the effect that principally and directly denominates the efficient cause as an actual agent, while at the same time referring terminatively and exercitively to the being effected. No other elements would seem to be required in the Thomistic view for the essence of efficiency. As thus described it may be applied analogously, of course, to every being, whether its efficiency, always measured by the extent of its act of to be [esse], is limited or unlimited. Hence St. Thomas can say: `A being is an efficient cause to the extent that it acts' [In V Metaphysics, 2, n. 765]. Thus efficiency is the production of the effect viewed actively as coming from a being and designating it as a cause."

III. Causity, Existence, and Etiological Properties

Causity, Casteñeda tells us, cannot be existence, since existence does not admit of degrees. Contrast this with the Thomistic notion of esse.

Causity must be measureable, he claims. (Perhaps this is alright for certain kinds of physical causation, but it is not obvious that it constitutes the core of the notion of causity.) On p. 24 he talks loosely of causity "migrating" from object to object, "riding on the backs of the objects involved in the causal reaction."

Casteñeda also talks of the etiological properties that ground causity and causal power. On an Aristotelian theory such "properties" constitute, in the first instance, the structure or form of the object involved. Some of what Casteñeda says in this section (pp. 24-25) seems tied too much to a materialistic and perhaps deterministic account of causality, although if we think just of causal interactions within the realm of pure nature (say in physics or chemistry), what he says makes sense.

I will simply ignore here the various principles that Casteñeda proposes for causity, though I think they bear looking into.