I have a strong intuition that there is a fundamental level of reality which ultimately grounds the phenomena of the world. This intuition forms a basis for preferring certain philosophical arguments over others. For instance, take the cosmological argument. In one traditional formulation, the argument goes something like this: every effect has a cause, and if you follow the chain backwards in time, there must be a first cause. I’ve never felt this argument was very forceful – what’s wrong with an infinite chain of causes? Now, however, if you recast the argument as saying that the contingent facts of the world ultimately and necessarily depend on a fundamental fact or collection of facts, then suddenly I start nodding my head affirmatively. There can’t be an infinite chain of contingent facts depending on other contingent facts, can there? Ontological priority seems to need a starting point more urgently than temporal priority. In the famous expression invoked by Ross P Cameron in his recent paper on this topic (found via OPP): it can’t be turtles all the way down, right?
Our desire for explanations seems to drive the intuition. If an entity is shown to depend on something else, it is thought to be explained. We want this search for explanation to find an ending point in terms of ultimate constituents. In our world, we seem indisputably to encounter composite things which seem comprised of parts; this drives our search for reductionist explanations. I guess it is possible to think that perhaps the “ceiling” rather than the “floor” is fundamental; perhaps the whole of the universe is the fundamental thing and the parts ontologically depend on the whole. Now, this seems counterintuitive to me: if we start with a whole, why should there be any parts? In any case, the direction of dependence is probably less important for this discussion than the idea that there is some fundamental level.
In his paper, Cameron asks whether there is a good argument for the truth of this intuition that there cannot be an unending chain of ontological dependence. Can we, for instance, argue that if there were no fundamental level grounding other entities, then nothing would be real? Cameron concludes that this would essentially be restating the intuition, rather than providing an argument. He considers a couple of other strategies in the paper and finds no satisfactory argument. On the other hand, he doesn’t see any good arguments against the intuition either. In fact the search for a metaphysical argument for the intuition may be seen to parallel the search for a deeper and deeper ontological level: you have to start somewhere, don’t you? Why not with an intuition? He notes as an example that Leibniz never argues for the Principle of Sufficient Reason, it’s just his starting point. Now, one can’t thereby defeat a skeptic who doesn’t share the intuition, but at the end of the day I don't find that the skeptics and deflationists of the world provide very good metaphysical explanations themselves.
Cameron says we can justify the intuition against infinitely descending chains of dependence by appeal to theoretical utility. We can give better explanations for entities if we identify an ultimate ontological basis in a collection of independent entities. This may be reason enough. He notes that this won’t convince someone who thinks the search for metaphysical explanation is misguided to begin with. On the other hand he says: ”…if you believe in metaphysical explanation you should believe it bottoms out somewhere.” He ends the paper by noting that given the pragmatic way the use of the principle is being justified, we should be modest about holding forth about the necessity of its truth.
Also interesting in this context is David Chalmers' recent paper: Ontological Anti-Realism (see blog post with links). His support for anti-realism in the paper (most of which is devoted to a mapping out of the terrain of meta-ontological stances) holds out the possibility of an exception for realism about the fundamental level. Jonathan Schaffer, in his commentary on the paper argues that Chalmers’ framework actually requires realism about the fundamental level. If Schaffer’s arguments are right, it seems to help to bolster the case that if you want to pursue metaphysical explanations, you need to be a realist about the fundamental level.