Scientific Vs. Metaphysical Realism?
This is a brief post to take issue with something Graham recently said on his blog (here). Graham has said similar things before, but I think this is one of the clearest examples of his opinion about the relation between science and realism. Graham first notes that the increasing fashionability of the term ‘realism’ in continental circles means that it can be misleadingly appropriated by those who aren’t genuine realists. He repeats the idea, stated elsewhere, that it is not enough to consider the real as something which constrains consciousness/thought, but that one must be able to allow it some structure independently of a relation to consciousness/thought. One must be able to say something about the interaction of fire and cotton in order to count as a genuine realist. Now, I agree with this, for the most part. My problem comes when he draws a dichotomy between two ways in which one can be genuinely realist:-
And even if you are, then there are still at least two different possible camps. The first would be the “scientific realist” camp, which would nod approvingly about the fire example, since for them human consciousness is simply a physical reality just like cotton and fire are; science can explain it all. The second would be, for lack of a better name, the “metaphysical realist” camp, for which the fire and cotton are no more reducible to calculable physical interaction than consciousness is. Sign me up for the second camp. The first tries to turn philosophy into something it is not.
Philosophy is the handmaid of nothing. Most will be quick to agree that it’s not the handmaid of theology; no longer a controversy about that, at least not in suitably chic circles. But neither should it be the handmaid of science, and neither can it be the handmaid of politics. It can be of assistance in all of these areas, but it should no more be their slave than it should seek to enslave them. Too many examples of present-day philosophy are really just examples of attempts to enslave the discipline.
I think this is a false dichotomy. Now, we must bare in mind that Graham does qualify this by saying that there are ‘at least’ these two, with the implication that there could be more options. However, it strikes me that Graham has presented these two opposing options quite a few times, and that he generally doesn’t demonstrate what, if any, ground lies between them. This means that a lot of positions that have a nuanced conception of the relation between philosophy and science that differs from Graham’s own tend to get lumped into the former ‘scientific realist’ category, in a way that makes for some easily dispatched straw men. Given this, I think it’d be helpful to briefly explore what space lies between these opposing poles.
Firstly, we should parse the two positions offered a bit. Both are realist, insofar as they make claims about the real structure of the interactions between non-human entities. Given this, there are two differences implied between these positions. The first, and most important, difference is that, whereas the ‘scientific realist’ adopts its claims about the structure of these interactions wholesale from the scientific description of these interactions, the ‘metaphysical realist’ posits a structure which is in some way in excess of these interactions. The second difference is present more by implication, and concerns a difference in their respective conceptions of consciousness. It is implied that the former position takes consciousness to be exhausted by some scientific account of the structure of a given kind of entity (e.g., neurobiology), whereas the latter takes consciousness to be some kind of paradigm case of irreducibility, and thus to potentially tell us something interesting about the more general excess indicated in the first difference.
Taking the second difference first, it essentially comprises a difference in the strategy of explanation. The former strategy is to explain consciousness in precisely the same terms used to describe other entities, and the second reverses this priority, trying to use consciousness to provide a description of all other entities. It’s important to note that the latter strategy does not negate the realist condition laid down earlier. It’s not a matter of claiming that all interactions between non-conscious entities are relative to consciousness, but rather that there are some characteristics that all such interactions share with the interactions between consciousness and other entities. It should be fairly obvious that this latter strategy is exactly the strategy Graham himself pursues to some extent (which I discussed with him a while back here, here and here). The interesting point is that although we can definitely say that the latter strategy must side with the latter half of the first distinction we drew (as it makes no sense to generalise the conclusions of neurobiology to all other sciences), its not clear that the former strategy is precluded from holding that there is some sense in which the structure of entities (and their interactions) is some sense in excess of the corresponding scientific descriptions. It’s possible both to claim that there is some important structure common to all entities that science does not speak of, and to claim that we must understand consciousness in terms of it, rather than the other way around.
As such, I think that the second distinction just confuses the first. Although it draws a very interesting contrast between Grahams approach and that of others’, it doesn’t seem to be directly related to the real issue at hand.
Moving on to the first distinction then, the crucial issue is what this ‘excess’ of the real structure of entities over and above science’s description of them consists in. The problem is that there are at least two ways of conceiving of this excess, which I will call formal excess and material excess. Graham’s own position is very good for demonstrating this distinction, because it contains both kinds of excess. So, Graham holds that an object is split between its sensuous and real sides, and that there is a corresponding distinction between its sensuous and real properties. If I understand him correctly, he thinks that science can only ever describe the sensuous properties of the thing, but can never penetrate into its real nature. Graham is very insistent that the way in which objects elude scientific description is not a matter of quantity, i.e., it is not that we will never be able to know them completely, but a matter of quality, i.e., we can never know any of an object’s real properties because they are qualitatively different from the sensuous properties we encounter. This is a matter of material excess, because there is something about the matter of each given object which eludes scientific description. Its important to note that I’m not using the concept ‘matter’ in any materialist sense here. It has nothing to do with instantiation in a substratum, but with what distinguishes various objects (or subject matters) from one another. Ursa Minor, my left foot, and the number 6 have different real properties, and thus there is a good sense in which what is eluding scientific description in each case is different between them.
By contrast, what is formally excessive about objects in Graham’s system is something that does not vary between objects – it is the very structure of the object qua object. That objects have a fourfold structure, and that they engage in relations of vicarious causation, precipitated by allure, is not something that science can discover. However, this is not because this structure is essentially inaccessible, but rather that it is just not within the province and methodology of science to address these kind of metaphysical questions. Science is not metaphysics. Fair enough. It is the fact that Graham claims that we can know this structure, and that he purports to give an account of it, which distinguishes Graham’s position as realist on the condition laid out above. Graham is not providing an account of the specific (material) details about the real structure of interactions between fire and cotton, but he is instead giving us an account of the formal structure of interaction between objects, which holds for fire and cotton in the same way as it would for any other two objects. This is indeed something that a correlationist can never do.
However, once we recognise these distinct types of excess, it becomes unclear as to what ‘scientific realism’ is. The general implication of the opposition between ‘scientific realism’ and ‘metaphysical realism’ is that the former has no place for metaphysical conceptions of the general structure of entities (e.g., of objects qua objects). This would indeed be the case if ‘scientific realism’ denies the kind of formal excess of the structure of entities in relation to science discussed above. This would be tantamount to some form of positivism. However, while it seems sensible that if one makes some claim to a kind of material excess of entities in relation to scientific description then one must also hold to some formal excess, insofar as the very claim of material excess is not something science itself is in a position to articulate, it nonetheless seems entirely possible to hold that there is some form of formal excess without a corresponding claim of material excess. One could hold that there is a good sense in which science describes entities as they are, but that it takes for granted some implicit conception of the Being of entities that it is no position to explicitly articulate or question. This kind of explicit concern with the Being of entities would then be a matter for philosophy (conceived as either metaphysics or ontology, depending on one’s preference). Exactly what the upshot of this kind explicit concern with Being entails is open to debate, but I think we can recognise that one can have it, and thus not be a positivist, without thereby denying that science can fully describe its objects. One might want to say that science can (though not necessarily does) provide an entirely adequate knowledge of entities, just that there is a different form of knowledge of entities that it is not directly concerned with (even if there is some important relation between them).
Broadly speaking, it thus seems as if we have three different positions: anti-metaphysical realism (positivism), and two forms of metaphysical realism, one which denies any form of material excess and one which accepts it. Graham’s position (and its affiliates) falls into the latter of the three categories.
Problem solved? Not quite. The issue is that Graham is strongly suggesting that any position which gives some form of primacy to science falls into ‘scientific realism’, as opposed to ‘metaphysical realism’. Although we haven’t discussed the various forms that the relation between science and philosophy can take, it seems fair to say that there are weaker ways of formulating the primacy of science than straight-up positivism, and thus one can be both ‘metaphysical’ and ‘scientific’ to some degree. Indeed, in the same post Graham picks out Manuel DeLanda as the other major proponent of unabashed continental realism, and DeLanda has just this kind of position. DeLanda holds that there is an important role for metaphysics as something distinct from science, but nonetheless that this metaphysics must essentially be sensitive to science. Moreover, he doesn’t obviously make any claims to the kind of material excess that Graham does. I must admit that I tend to be close to DeLanda on these issues.
It seems as if the implied criticisms of ‘scientific realism’ that Graham makes in this post are intended not just to hit genuine positivists, but anyone who singles out some kind of privileged role for science, and even more so, anyone who thinks that science can genuinely know the objects it describes. This seems to be the only way to read it, given that there just aren’t many avowed positivists out there these days (excluding the strand of neo-positivism that seems to come with Dawkin’s style atheism), at least not within the continental field, and thus it would be redundant if this was its only target. However, if it does indeed aim at a position broader than positivism, it doesn’t do enough to distinguish it from positivism. This is what I meant when I said that Graham lumps a bunch of positions together, and in the process makes some of them into straw men.
To conclude, if there’s going to be a genuine debate had about the relation (or non-relation) between philosophy and science, it needs to be put on much firmer footing than this kind of opposition. There are more options than blind slavery and radical independence. The first step here is to rigorously distinguish between the general structure of entities qua entities (or objects qua objects), with which science is unconcerned (if not indifferent to), and the particular structure of given objects, with which science does concern itself (even if it is an open question as to whether it succeeds). In talking about the ‘real’ nature of objects, it is often very easy to confuse the two.
This entry was posted on March 17, 2010 at 1:39 pm and is filed under Discussion with tags Graham Harman, Metaphysics, Positivism, Realism, Science. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.