The Problem of the Levels and Kinds of Causation in Biological Sciences: Some Remarks
Complex living systems consist of several organizational levels, which often are interdependent in different ways. This multi-layered organization poses the problem of causation, which is scientifically and philosophically deep. This is especially true for the metabolic, cellular and physiological systems, as well as for the nervous and cognitive systems. In all these systems, upward and downward causation are causally interrelated. This important fact has led the heart physiologist Denis Noble to argue that there is no privileged level of causality in biological systems. Moreover, “higher levels in biological systems exert their influence over the lower levels. Each level provides the boundary conditions under which the processes at lower levels operate. Without boundary conditions, biological functions would not exist” (Noble 2012).
Studying the causal pathways in brain dynamics, the Swedish biologist Hans Liljenstrom remarks that downward causation from larger to smaller scales could be regarded as evidence that multi-level “both-way” causation occurs (Liljenstrom 2016). He investigated, on the one hand, how cortical neurodynamics may depend on structural properties, such as connectivity and neuronal types, and on intrinsic and external signals and fluctuations; on the other, to what extent the complex neurodynamics of cortical networks can influence the neural activity of single neurons. More precisely, Liljenstrom attempted to show that the neural activity at the microscopic level of single neurons is the basis for the neurodynamics at the mesoscopic network level, and fluctuations may sometimes trigger coherent spatio-temporal patterns of activity at this higher level. Irregular chaotic-like behavior can be generated by the interplay of neural excitatory and inhibitory activity at the network level. This complex network dynamics, in turn, may influence the activity of single neurons, causing them to fire coherently or synchronously. Thus, Liljenstrom concludes: “this downward causation is complementary to the upward causation” (Liljenstrom 2016: 189).
From simulation results, applying both to bottom-up mechanisms like noise-induced state transitions and to top-down processes like network modulation of neural activity, Liljenstrom is led to stress that events and processes at the microscopic level of single neurons can influence the mesoscopic neurodynamics of cortical networks, which in turn are associated with cognitive functions at the macroscopic level.
It is apparent that internal noise can cause various phase transitions in the network dynamics, that may have effects on higher level functions. For example, an increased noise level in just a few network nodes can induce global synchronous oscillations in cortical networks and shift the system dynamics from one dynamical state to another. This in turn can change the efficiency in the information processing of the system.
(Liljenstrom 2016: 185)
This kind of situation, however, needs to be related (or can be correctly understood only in relation) to another important aspect of the neurodynamics of cortical networks. In fact,
neuromodulation, whether related to the level of arousal or as a consequence of attention, can regulate the cortical neurodynamics, and hence the activity of its constituent neurons. The firing patterns of single neurons are thus, to a certain degree, determined by the activity at the network level (and above). For example, neurons in visual cortex may fire synchronously and in phase, as a result of cholinergic modulation during attention.
(Liljenstrom 2016: 186)
These arguments clearly show that
the intricate web of interrelationships between different levels of neural organization, with inhibitory and excitatory feed-forward and feedback loops, with nonlinearities and thresholds, noise and chaos, makes any attempt to trace the causality of events and processes useless. In line with the ideas of Noble, it seems obvious that there is, in general, both upward and downward causation in biological systems, including the nervous system. This also makes it impossible to say that mental processes are simply caused by neural processes, without any influence from the mental on the neural.
(Liljenstrom 2016: 186)
R. W. Sperry already stressed this crucial point when he wrote:
A traditional working hypothesis in neuroscience holds that a complete account of brain function is possible, in principle, in strictly neurophysiological terms without invoking conscious or mental agents; the neural correlates of subjective experience are conceived to exert causal influence but no mental qualities per se. This long established materialist-behaviorist principle has been challenged in recent years by the introduction of a modified concept of the mind-brain relation in which consciousness is conceived to be emergent and causal. Psychophysical interaction is explained in terms of the emergence in nesting brain hierarchies of high order, functionally derived, mental properties that interact by laws and principles different from, and not reducible to those of neurophysiology. Reciprocal upward and downward, interlevel determination of the mental and neural action is accounted for on these terms without violating the principles of scientific explanation and without reducing the qualities of inner experience to those of physiology. Interaction of mind and brain becomes not only conceivable and scientifically tenable, but more plausible in some respects that were the older parallelist and identity views of the materialist position.
(Sperry 1980: 195, see also Eccles 1986)
In the light of the last remark, the debate on the philosophical distinction between the “functionalist” version and the “monist” version of “nonreductive” “physicalism” might appear meaningless. While, in the first version, one maintains that mental phenomena are realized in physical properties and processes, in the monist version, one holds that every event that can be provided with a mental description can also be provided with a physical description. In either version, even though there are no scientific laws by which mental phenomena could be “reduced” to physical phenomena, the underlying causality of the world remains entirely physical.
In life sciences, we need to rethink of the concept of biological causality in new, deeper terms. One key point is that higher-level phenomena cannot be understood by simply analyzing the lower levels. The importance of systems biology is connected to the limitations of molecule-centered approaches. Systems biology has shifted the focus from the identification and characterization of molecular components towards an understanding of networks and functional activity. However, a further significant shift remains to be made: refocusing our attention away from pathway-centered approaches to an understanding of complex multi-level systems. In other words, our understanding of cellular functions must be integrated across multiple levels of structural and functional organization: from cell tissues and organs to the whole organism, and from cell functions (growth, proliferation, differentiation and apoptosis) to the physiology of organs or the human body. To quote H. Kacser, “to understand the whole, one must study the whole” (Kacser 1986). The idea is that, if you want to understand a tissue, you need to study it as a whole. Now, organs and tissues are multi-level systems manifesting both “bottom-up” determination and “top-down” determination: the whole (organ or tissue) is the product of the parts (tissues or cells, respectively), but the parts in turn depend upon the whole for their own functioning and maintenance. In more philosophical terms, this means that higher-level systems in biological phenomena may change in very significant ways properties of lower-level systems or entities. In other words, these entities behave at lower levels in novel and irreducible ways.
Following O. Wolkenhauer and A. Muir (2011), we stress that living systems, from organisms to organs, tissues and cells, are phenomena of organized complexity whose relationships and properties are largely determined by their function as a whole. The tissues of our human body are self-organizing systems: every cell owes its role to the action of all its surrounding cells, and it also exists for the sake of the others. The whole (tissue) and its parts (cells) reciprocally determine the functioning of each other. For instance, the pacemaker rhythm of the heart is not only caused by the activity of the ions channels at the molecular level, but it is also dependent on the functioning of the organ, and even the body, as a whole. The systems biologist Denis Noble convincingly demonstrated the importance of such downward causation in simulations of the heart rhythm, where feedback from cell voltage was removed and fluctuations in ion current ceased. To understand such phenomena in multi-level systems, it is not only important to understand molecular mechanisms but also to understand the organizational maintenance of the system at higher levels.
