Causality and Levels of Describing Reality
We regard reality on different levels which reach from quanta and elementary particles at the lower end, via molecules, and macroscopic objects to structures. The behavior of each level is emergently induced by the behavior of the entities on the level below.
We describe the behavior on the different levels with specific physical laws, like quantum physics, classical physical laws for macroscopic objects, or models describing the physical properties of matter.
On structural level, dissipative structures occur, the interaction of which on structural level can not be described by any physical law. Instead we use cause-and-effect interrelations to describe their interactions. One structure causes an effect in another structure. Causality thus solely occurs on the structural level.
All these aspects only relate to our description of reality. Reality itself just behaves in a certain way and especially does not know anything about our physical laws.
Before answering the question how causality can be understood, first causality needs to be defined. Here the definition of Honderich shall be used: A cause shall be regarded as an individual trigger or a small set of triggers, which induce an effect, which would not have happened, if the triggers would not have occurred, irrespective of some consistent variation of the surrounding environment (Honderich: A Theory of Determinism: The Mind, Neuroscience, and Life-Hopes. Oxford University Press, Oxford, 1988). While the first condition, the so-called counterfactual conditional, corresponds to the common-sense understanding, the second condition that a cause is only causal, even if some consistent variations occur, restricts the generality and will become important in the following. Causality will now be analyzed on the different levels, at which we describe reality.
The first level regarded is the molecular level. As seen on the previous page, molecular action is highly interconnected in a universal network of interactions. These interactions induce random motion. The consequences become obvious, if it is testwise assumed that one could link the behavior of one molecule as effect to the behavior of another molecule as cause. Then, any variation of any of the other molecules of the universe would within pico seconds lead to a different behavior of the affected molecule. Thus, the influence of the causing molecule on the affected molecule would not be independent of the consistent variations of the surrounding environment. Thus, molecular interactions can not be regarded as being causal. Nevertheless in a deterministic world view the behavior of each molecule is fully determined by the interactions with all other molecules.
Similar considerations apply on the quantum level, which can be described based on quantum physics. In principle on that level two viewpoints can be taken. Either can the Copenhagen interpretation of quantum theory be applied, which assumes that the probability functions, which are described by quantum theory and which account for the uncertainties observed on that level, are all there is to reality. The alternative viewpoint is to assume that an exact and deterministic reality exists on that level, where the particles are described by so-called hidden variables, where this exact reality can only not be accessed experimentally due to fundamental limitations. In the Copenhagen interpretation, the elementary particles are assumed to behave randomly within the probability functions characterized in quantum theory. Thus, any causal interaction between elementary particles can be excluded, since randomness determines the behavior. If an exact behavior is assumed, the same consideration as on molecular level apply. Deterministic chaos is expected, factually also leading to a behavior, in which causality can not be found, as just described for molecules. Thus, on the quantum level, no causality can be found.
As a next level that of macroscopic objects is to be regarded. A typical example for this level are billiard balls that hit on a billiard table. Before analyzing causality on that level, it is required to clarify what is actually to be considered on the macroscopic level. Of course, the hit between two billiard balls can also be described on molecular level. At the same time the details of the hit depend on the almost perfecly spherical shape of the balls, which actually refers to the next level to be considered, namely structure. Thus on the macroscopic level, only material properties and the fundamental laws on how macroscopic objects interact for example described with Newton's laws of motion are to be regarded. The material properties like the hardness of ice or the density of the billiard ball are determined by the properties of the molecules constituting the billiard ball. This determination is no causal interrelation, because it can not be attributed to any individual molecular parameter, like the interaction force between two molecules at a specific orientation and distance. If any of these molecular parameters would be changed, identical macroscopic properties could nevertheless be realized, if some of the other molecular parameters are appropriately adjusted. Also physical laws for macroscopic objects do not constitute causality. In the equation force equals mass times acceleration, neither the force is cause for the acceleration nor is the mass cause for the force. The physical law rather constitutes a relation between observed variables, which are expected as outcomes of a physical experiment. Thus, also on macroscopic level no cause-and-effect interrelation can be found.
The next level to be regarded is that of structure. An example of a dynamic structure is shown in Figure 1. This structure has been generated by heating a thin layer of silicon oil in a pan, where small particles of aluminum have been added. For low heating, the heat will simply be transferred though a stagnant liquid film by heat conduction. As soon as the heating is increased above a certain level, so-called roll cells as seen in Figure 1 occur as an emergent phenomenon. These structures are called dissipative structures, because they are driven by dissipation, where in this case the heat supplied from below is dissipated into the air above. This heat flux through the film induces the roll cells, where in the center of each cell the liquid is flowing from the bottom to the top of the film, then flows towards the circumference of the cell, where it flows back to the bottom of the film. If the heating is stopped, the structures remain visible, which means that such dynamic structures can in principle also lead to static structures left behind. It should also be stated that while the structures in the Bénard cells in Figure 1 of course depend on the silicon-oil molecules, the cells 'don't care', which individual molecules constitute the respective cell, the structures thus have a certain ignorance with respect to the details of the lower levels.
Figure 1: Bénard cells as seen from above in a thin layer of silicon oil heated from below and visualized with aluminum particles
The situation for the roll cells is quite different from the behavior on the levels below. If one cell is given, no physical law states, where exactly to find the other cells and for example which size they should have. Nevertheless it is clear that one roll cell directly forces the neighboring cell into existence. This can be regarded as a simple form of causal interrelation. One cell causes the neighboring cell without defining its individual detailed properties. The statement just refers to the interrelation between structures. One structure forces another structure into existence or defines the behavior of another structure. Upon inspection it becomes obvious that conventional examples for causality follow exactly the same path. For example, if Bill throws a stone which shatters a window, then it is the structure of Bill, which induces the structure of the stone to fly, which then interacts with the structure of the window, the behavior of which is then determined to shatter. Thus causality can be found only on structural level, since on lower levels no causality existed. Thus, while on lower levels quantum physics, classical physics and property models describe the behavior of reality, on the structural level we use cause-and-effect relations to describe the observed interrelations.
Finally it is important to realize that this description only relates to our description of reality. Reality itself does actually not care about these levels of our description. This is of course trivially obvious, because why should otherwise reality care about the mathematical constructs created by man in the attempt to shed some light on the interrelations of observations of reality. Presumably reality hardly cares. And of course the Bénard cells could alternatively be described by characterizing the neutrons, protons, and electrons constituting the molecules, which in turn form the roll cells. This would be a tedious task but is possible in principle. And of course the identical outcome is to be expected. Thus, it is our human choice to ignore the lower levels, when we try to describe observations on the higher levels. This interrelation becomes possibly even clearer, if a statement is regarded, which is obvious only at first sight. Isn't it obvious that the roll cells in turn determine and thus cause the molecules to move in exactly the way they do, namely following the flow regime of each cell? That this should not be so obvious becomes clear, if it is recalled that it is actually the molecules, which due to their individual structure under the given boundary conditions move such that on macroscopic level we regard the structure of motion as roll cells. Which level is now forcing which level to react? This can be solved by realizing with Ernst Mach that nature simply is (Ernst Mach: The Science of Mechanics. A Critical and Historical Account of its Development. Translated by Thomas J. McCormack. The Open Court Publishing Co., Chicago, 1919). Nature does not know any cause and effect, it also does not know or even follow any physical laws. These mental constructs are inventions of human mind to make sense of effects observed in reality. The different levels only appeal to us, because apparently it is possible to describe the observations on one level without regarding the detailed behavior of the levels below. The lower-level behavior is only entering via some averaged properties. For example the density is a result of the molecular interactions and describes for a large number of individual molecules some averaged property, in the case of density how much space all molecules together occupy on average. To ask for the density of a single molecule on the other hand does not make sense.