Stroud on Hume

David Hume, as I often argue, is overrated.  But that’s not his fault.  It’s the fault of those who do the overrating.  So, rather than beat up on him (as I have done recently), let’s beat up on them for a change.  Or rather, let’s watch Barry Stroud do it, in a way that is far more genteel than I’m inclined to.

The problem, as I’ve often pointed out, is this.  Hume’s most famous conclusions – his skepticism about induction, his treatment of causality, his subjectivism about value, and so forth – are widely celebrated by contemporary philosophers.  Even those philosophers who don’t agree with these conclusions tend to regard them as formidable.  Yet few if any of these philosophers would accept the basic philosophical presuppositions on which Hume’s conclusions all rest.  For example (and as I often complain), Hume conflates concepts with mental imagery.  That is a very crude philosophical error, it has been known to be a crude error since at least Plato and Aristotle, and it has been known to contemporary philosophers to be a crude error since at least Wittgenstein.  There is very little if any “punch” left to Hume’s philosophy once this error is exposed.  Yet his conclusions continue to be taken seriously long after such underpinnings have collapsed.  They are like a ghost that continues to walk the earth long after the death of the body. 

In his book Engagement and Metaphysical Dissatisfaction: Modality and Value, Stroud raises a similar complaint.  Contemporary philosophers take Hume’s doubts about causality as a feature of mind-independent reality very seriously.  Yet they do not accept either the account of perception that these doubts rest on, or some of the other conclusions Hume draws from that account.  And it is not clear how they can consistently take the one without the others.  Stroud writes:

Many philosophers of more recent times remain in a broad sense followers of Hume on the status of causation without accepting such a severely restricted conception of the scope of perception.  They appear to hold that we can perceive and thereby have a conception of physical objects and other enduring things and states of affairs even though the idea of causal dependence between such things in the independent world remains problematic or metaphysically dubious.  The source of their doubts is not easy to determine.  One possible source is the assumption that we never perceive instances of causal connection or dependence.  A different but related possibility is that causal dependence is thought to be unperceivable because of the doubtful intelligibility of the idea of such a connection.  In any case, it certainly is still widely believed that we never perceive causal connections between things.  By now the view is hardly ever argued for.  The most that is usually offered in its support is a reverential bow in the direction of Hume, but with no acknowledgment of the restrictive theory of perception that Hume's own denial rests on. (p. 23)

Hume thinks of perception as the passive reception of “impressions” such as a sensation of color, a sharp pain, or a twinge of fear.  “Ideas” in turn, as he uses the term, are faint copies of such impressions – essentially mental imagery of a visual, auditory, tactile, olfactory, or gustatory sort, along with memories of emotions and the like.  Few philosophers today would endorse such a crude model of perceptual experience or concept formation.  Yet that model underlies Hume’s doubts about causation as an objective feature of reality.  We have a set of impressions (a succession of visual experiences of a round whitish patch, say) that we take to be the motion of a cue ball, followed by another set of impressions (a “knocking” sound followed by a succession of visual experiences of a round black patch) that we take to be the motion of an 8 ball.  But we have no impression of a causal power or force by which the first generates the second.  Hence we have no idea of such a force or causal power.  We just find that impressions of the former sort are constantly conjoined in our experience with impressions of the latter sort.  This leads us to expect the latter on the occasion of the former, and we project this subjective expectation onto the world.  But we have, Hume claims, no reason to think it really corresponds to any objective feature of the world.

(On the “skeptical realist” reading of Hume, he does not intend to undermine our commonsense belief that there really are such causal features in external reality, but merely denies that we can have any cognitive grasp of them.  But it’s hard to see how one could consistently push the latter point without ending up in essentially the more radical skeptical position traditionally attributed to Hume.  I cannot myself help wondering whether the recent heavy going about “skeptical realism” among Hume interpreters might be much ado about not a whole helluva lot.  But that’s neither here nor there for present purposes.) 

Since contemporary philosophers wouldn’t buy this story about our perceptual and cognitive faculties, it’s hard to see why they remain impressed by the conclusions about causation Hume draws from them.  That’s part of Stroud’s point.  The other part of his point is that Hume draws other lessons from the same account of perception and cognition, lessons that contemporary philosophers are not so impressed by.  In particular, Hume concludes that we have no idea of mind-independent physical objects either, because he thinks we have no impression of such things.  We have, for example, only this fleeting impression of a round whitish patch, that other fleeting impression of a round whitish patch, a fleeting impression of a “knocking” sound, etc., but no impression of any substance that underlies and ties together these different impressions.  Here again we are in his view really just projecting onto the world something that is merely subjective, namely the brief relative stability of some of our impressions (where an impression is something essentially mental rather than mind-independent).  This is the origin of our belief in mind-independent objects, and it has no more sound a basis than our belief in objective causal connections.

Now, contrary to what non-philosophers sometimes think, few contemporary philosophers really take seriously the idea that there are no mind-independent physical objects.  They may regard it as an interesting puzzle, but not as a live option.  The idea that objective causal power and necessity might not really exist is taken to be a live option, though.  And what Stroud is puzzled by is why that should be the case given that the other Humean skeptical conclusion is not taken seriously.

Nor in Stroud’s view is it just the common Humean foundation of these two kinds of skepticism that makes this combination of attitudes problematic.  That is to say, the problem is not just that Hume himself based his skepticism about causation and his skepticism about physical objects on the same flawed account of perception and cognition.  It’s also that, even apart from that, it is hard to see how one could consistently believe in mind-independent physical objects without also attributing to them real causal powers.  Stroud writes:

This raises a general question about how or whether a person could think about and understand the objects this view admits that we do see.  Could we have a conception of a world of visible, enduring objects at all if we could never see what any of those objects do, or see them doing it?  Hume's actual view does not face this difficulty.  He thinks not only that we never see a stone break a window, but that we never see a stone or a window either.  Hume acknowledges the need to explain how we get even so much as the idea of an enduring object from the fleeting perceptions we receive, and how we come to think of such things as perceivable.  But for those who think we can see an object and know what it is and where it is and what will happen if certain other things happen, but that we never see the object doing or undergoing any of the things it does, there is a special problem. (p. 24)

What Stroud is appealing to here is the thesis – common to (though spelled out in very different ways by) both Kant-inspired writers like P. F. Strawson and contemporary neo-Aristotelians and Thomists – that we cannot make sense of the notion of a world of independently existing substances except as causally related in various ways.  (Think of the Scholastic thesis agere sequitur esse or “action follows being” – that is to say, that how a thing acts reflects what it is.  If a thing does nothing, then it cannot be said to have being at all; and if it does have being, then it must be capable of doing something, which entails causal power.)

If that’s correct – and obviously it’s a claim requiring elaboration and defense – then skepticism about mind-independent objects and skepticism about causation stand or fall together.  As Stroud notes, Hume is at least consistent on this score, since he opts for skepticism in both cases.  It’s the selective skepticism (and selective Humeanism) of some contemporary philosophers that Stroud thinks dubiously coherent.

But it’s not a universal tendency.  Where causation is concerned, the Humean ghost is at long last being exorcised in some quarters, as evidenced by books like Mumford and Anjum’s Getting Causes from Powers and the neo-Aristotelian literature.  (Naturally, I’ve tried to do my part as well.)