The Main Aspects of the Issue
One motivation behind the general intuition of the layered view is to find a place for mental properties in the physical order. Given the emergence relation, mental properties are thought to emerge from underlying physical properties, in present terms from neurophysiological properties. So, a property of the mental kind emerges from a property of the physical (neural) kind in that the former depends on the latter but cannot be reduced to or identified with it. The supposed advantage of the emergence relation is twofold: from one side, it guarantees autonomy; on the other side, it preserves dependence.
Marc Bedau has stressed that:
(1) Emergent properties are somehow constituted by, and generated from, underlying processes;
- (2) Emergent properties are somehow autonomous from underlying processes
- (Bedau (1997: 376)).
From one side, the concept of emergence has been invoked to give higher- level properties a causal status, that is, placing them in the causal order. This is an ultimate consequence of Bedau’s first point, the fact that emergent properties are constituted by underlying processes. These underlying processes must somehow interact both with other emergent and with nonemergent properties, thus sharing, at least in part, their causal structure. It is in this latter respect that emergentism is different from epiphenomenal- ism. Indeed, having properties emerging and not playing any causal role is tantamount to proposing a form of epiphenomenalism, the view that some properties are effects but not causes of other properties. From the other side, the concept of emergence has been invoked to give some properties an autonomous status, to recognize them as properties of their own. This is the anti-reductionist thread, the second point in Bedau’s quote, the view that emergent properties are somehow “more” or “different” with respect to the base properties from which they emerge.
It is the conjunction of these two features of emergent properties that is at the origin of many discussions and difficulties arising from this topic. Emergent properties are different from their base properties, and they somehow take part in their causal order nevertheless.
In particular, the problem of having higher-level properties exercising causal influence on lower-level properties, the so-called downward causation, is one of the thorniest issue for those who want to defend the very idea of emergence. The appeal to downward causation is needed to avoid a host of difficulties. Suppose, in fact, that mental properties emerge from physical properties, are not reducible to these but are causally related solely to other mental properties. In what sense, then, is my raising my arm voluntary, when I claim it is, if my will is severed from, or causally inefficacious with respect to, my physical movements? Similarly, how can we say that it is in virtue of what I have perceived that I have the belief I declare myself to have? In order to have mental properties playing a causal role, downward causation comes in handy, with causal influence cutting across levels.
The resulting picture is the following. Mental properties emerge from physical (neural) properties. They are not identical to physical properties, so these are autonomous from them, and do form a level of their own; at the same time, they can exercise causal influence on physical properties. So, there must be a form of downward causation, or causation from above. Given this general picture, I wish to raise a problem for it.
If the causal efficacy of emergent properties is defended, downward causation has to be accepted. Both emergence and downward causation presuppose the presence of levels of reality. However, emergence and downward causation pull in opposite directions with respect to my best reconstruction of what levels are. The upshot is that emergence stresses the autonomy among levels, while downward causation puts the distinction between levels at risk of a reductio ad absurdum, with the further consequence of blurring the very notion of downward. Thus, emergence and downward causation are not a match for each other vis-a-vis the layered view of reality licensed by the concept of level. In order to argue for this, first I need to be clear on what emergence is and what the desiderata are for it.