I attended a couple of very interesting talks in philosophy of biology yesterday. One was my friend Y discussing his research in the history of evolutionary biology, and the other was a critique of Carl Craver's conception of levels of explanation in biological phenomena. If any insight was common to both, it was that intractable problems arise when you develop and employ concepts for scientific interpretation for the sake of simplifying phenomena.
A short version of what I took away from the account of Craver's idea. Organic and ecological systems are ridiculously complex, so it's a good idea to figure out how to simplify and systematize your understanding of them. In this case, Craver described a series of obstacles and difficulties in making sense of causation between micro, meso, and macro levels of biological phenomena. So at one level of analysis, bodies and processes affect each other through causes. But when processes at one level of analysis affect processes at another level, they don't cause these other processes. Instead, the notion is that these processes constitute each other.
Consider the relationships among our cells. These are incredibly complicated exchanges among chemical processes, which, taken as a whole, constitute a whole organism. The processes between the cells are in relationships of mutual causation. All these processes working in tandem constitute the organism called Adam Riggio.
However, levels of analysis aren't absolute. Think about an army who loses a battle because, in the process of assembling itself for battle, suffers a contagious viral infection among its troops, allowing the opposing army to devastate them. The viral infection, despite existing on a micro level relative to the macro level phenomenon of this multi-nation war, has literally caused war casualties. When such viruses are biological weapons purposely released into the opposing army, then the relationship of causation is even more clear.
If you divide causation from constitution by level, then you won't adequately understand the entire event. Even though conceptually, taking multi-level causation into account is very difficult to handle, you have to do it. If the phenomenon you're dealing with is complex, you won't adequately understand it if you simplify the concepts you use to make sense of it.
It's a common platitude in philosophy, and quite often in ordinary life as well, that if you can't make your explanation of some phenomenon remarkably simple to understand, then you haven't done a very good job of explaining. But when the phenomenon in question is as complicated as a meteorological or climatological phenomenon, an organism's internal processes, or a growing viral epidemic in society, a simple explanation will not do. Simplicity is, in many ways, the worst type of error in thought, because it is precisely so easy to understand. When understanding comes to easily to you, it's incredibly difficult to convince yourself, let alone convincing other people, that your concepts are inadequate. That you have cut away too many relevant aspects of the world from your analysis in the name of simplification.
Correcting that kind of mistake isn't just difficult because you can so easily become comfortable in your simple understanding of events. It also requires admitting that you need to do a lot more work. And no one likes making their lives more complicated. Even if it's necessarily complicated.