This is the sequel to two posts, here and here, on Shepherd's theory of causation. I haven't actually reached the point of talking about Shepherd's views of the First Cause; I'm working my way thither in a roundabout way. The first of the prior posts presents some passages that I am, at least indirectly, clarifying in these posts. The second gives a brief discussion of the core idea of Shepherd's positive theory of causation, i.e., that causation is the mixture of qualities. In this post I'll briefly look at Shepherd's criticisms of David Hume's theory of causation. There are more investigative posts to come on Shepherd's causal theory, when I get around to them; in particular, I'll probably look more closely (in no particular order) at some of the arguments against Hume, at the general outlines of her positive theory of causation, at her use of experimentum crucis, at her conception of science, at her use of the design argument for the existence of God, at her speculative conception of God-as-cause, and at the causal reasoning involved in examining testimony about miracles. So there's lots left to examine; and I won't stop until I've traced through all the major parts of Shepherd's views of causation.
Shepherd is not at all impressed by Hume. She calls his theory "unstable and confused"; she says it involves "every species of illogical sophistry"; she considers it part of her work "to show that Mr. Hume's reputation for logical correctness has been overrated"; she expresses astonishment that Hume's definition of causation has "continued so long, admired, adopted, and unanswered" (Cause and Effect, pp. 129, 131, 135, 192). Her view is that Hume's position is a tissue of sophisms that are given an air of knowledge and elegance, thus misleading the casual reader. To counteract this, she attempts to strip Hume's position down to its essential elments. This is her result (I have simplified, reordered, and broken it up quite a bit to streamline her summary for this post):
a. No idea can be had except derived from an impression.
b. Experience only shows certain similar sensible qualities frequently (but not invariably) followed by other sensible qualities.
c. Thus in nature events are entirely unconnected and therefore incapable of conveying an impression of necessary connection or power.
d. In certain cases, however, there are invariable sequences of sensible qualities, on the basis of which a definition of cause and effect as invariable sequence might be built.
e. Since we are acquainted only with sensible qualities, they alone can be causes; but they have no connection with the secret powers of objects on which the effects would entirely depend.
f. Therefore like sensible qualities not being like causes might be followed by different effects.
g. It is reasonable to suppose, that an invariable sequence might be interrupted, for there is no contradiction in imagining an arbitrary change in nature's course. "Yet should a contrary imagination resist reason, and not conceive in fact this interruption as possible to take place; she may again reconsider the possibility of nature altering her course, forming no contradiction to reason" (p. 131)
h. Hence the custom of observing similar sequences of sensible qualities can alone convey the impression from which we get the idea of necessary connection.
i. Thus necessary connection is a 'fancy of the mind,' not a natural relation.
Shepherd identifies no less than seven problems with this line of thought.
1) From an examination of a particular instance, a general negative conclusion is drawn. He also deduces a general affirmative conclusion, that the future shall invariably resemble the past, from particular instances alone.
2) It claims we should deny the general relation of cause and effect on the basis of a proposition completely consistent with it, "namely, that like sensible qualities, NOT being like Causes, might be followed by DIFFERENT Effects" (p. 132).
3) A general negative conclusion is drawn from negative premises alone, "for it is concluded there is no proof for the existence of the general relation of Cause and Effect between objects;--because experience shows that like sensible causes are not like Causes; and are therefore not necessarily connected with like efects.
4) Hume shifts the discussion from examining the general relation of cause and effect to examining the criterion for ascertaining the presence of like causes.
5) The proposition that is disputed is used in making this argument: "first, in the statement that impressions are the productive Causes of ideas;--secondly, in supposing the secret powers of an object to be alone the real productive Causes of its future properties; thirdly, in conceiving Nature may alter her course for the express purpose of changing the secret powers; and that they are changed by such an alteration;--and lastly, in alleging custom to be the sole Cause (i.e. producing generating principle) of the IDEA of causation" (pp. 133-134).
6) The proposition, "The course of nature may be supposed to change," is used ambiguously; since it can mean either an uncaused alteration of the subsequent sensible qualities or of the antecedent secret powers.
7) Hume attempts to establish that custom, not reason, is the principle of causation, but allows reason to be the sole ground and necessary cause of this belief. (I presume this is part of the point of what I've listed as (g).)
Looking this over just very briefly, it looks, just at a glance, like it might be a mixed bag. I'm inclined to think that (1-3) might be fair enough, although Hume usually can be interpreted so as to overcome formal problems like these. (4) could very well be exactly right, and I'll probably look more closely at this one fairly soon. I think she's right about (5) and (6), too, but I'm not so sure as she that any of these actually are problems with the logic of the argument. She has, however, put her finger here on several aspects of Hume's formulations that make Hume very difficult to interpret on this point; a number of them are, in one form or another, still discussed in the secondary literature on Hume today. (7) is entirely too subtle for me, at least without further inquiry. There needs to be further inquiry on all of these points (my thoughts here are just first suggestions). While the summary is short enough, the overwhelming bulk of Cause and Effect is devoted to detailed arguments supporting these criticisms; and actual evaluation would have to look more closely at each of them. There is more here than meets the eye, although my claim is that what meets the eye is immensely promising.
