Contact: Mary Ann Meyers, Ph.D., Senior Fellow

Top-down causation refers to the effects on components of organized systems that cannot be fully analyzed in terms of component-level behavior but instead requires reference to the higher-level system itself. A sweeping and fundamental concept, it is not only a philosophical idea but also a key ingredient in the emergence and functioning of complex systems, including life and the human brain. Together with bottom-up causation, top-down causation enables genuine complexity to emerge within specific levels of the hierarchy of complexity and causation. It also links the various levels of the hierarchy in a manner that undermines any simple-minded version of reductionism. A growing literature on complexity and emergence is providing an analysis of how this happens. Nevertheless, there are some who deny that it has any significance, or even reality.

The core issue is volitional agency. The topical focus of the discussion at the Royal Society amongst sixteen scientists and scholars has the potential to link scientific research with progress in philosophical and theological scholarship because of the fundamental importance of causation in daily living and in moral life. Symposium participants will consider the following specific sets of philosophical questions:

  • What counts as “top” in top-down causation? What counts as “bottom”? Would “whole-part causation” be a better description? If so, what difference would that make?
  • Are there any top-down realities to do the causing? If so, are they properties or substances? If properties, “of what” are they properties? If substances, of what are the substances made/composed?
  • What would be required for a supervening or emergent entity to have causal powers? And what would be required for “downward causal” power in particular?
  • Where might we get some explanatory (or metaphysical) “cash value” out of employing or hypothesizing top-down causal entities? Gaia? Agents? God? Molecules? Minds?
  • How would we or could we know that the relevant causal powers are not (even in principle) reducible to constituent properties of the “bottom” level of reality? If we cannot answer this question, does this rule out top-down causation as empirically useless?
  • How could we empirically test for the existence of the relevant causal powers?
  • Is top-down causation a single category, or are there distinguishable different kinds? It has been proposed that there are five different kinds of top-down causation in the hierarchy of complexity, namely: algorithmic top-down causation; top-down causation via non-adaptive information control; top-down causation via adaptive selection; top-down causation via adaptive information control; and intelligent top-down causation. Is this a good classification? Are there other kinds that should be taken into account?
Two key issues in physics are:

  • How can there be room at the bottom for any top-down causation at all?
  • Can non-physical entities have physical effects? If so, how is this possible?
And related to these are a set of subsidiary questions:

  • What is the relation of information to causality? What is the nature of information?
  • Is quantum theory necessary for top-down causation to be possible?
  • How does the environment influence quantum states and outcomes?
From the perspective of chemistry, crucial issues include:

  • What distinguishes chemistry from physics? What is chemical identity?
  • To what extent does quantum mechanics explain chemical phenomena?
  • To what degree are properties of physical entities altered by incorporation in molecules?
Biological issues in this discussion relate to the nature of causation in hierarchically structured networks, as well as to the nature of genetic determinism and evolutionary causation. These include:

  • What distinguishes biology from physics and chemistry? What is biological identity?
  • What is the ontological status of the structure and relationships in a biological network?
  • What is the nature and ontological status of biological information?
  • What limits on genetic determinism result from the effects of biological context?
  • Is multilevel selection a defensible option?
  • Is it the way that evolution works?
  • Is “chance” a genuine causal category?
Underlying all of the above questions is the relation of macro levels to micro levels, which remains mysterious. Among the questions confronting us are:

  • How is the second law of thermodynamics (a key requirement of physics at the macro level) related to biological development and the growth of information (key biological properties), on the one hand, and the time-reversibility of micro physics, on the other?
  • Is the macro arrow of time determined by top-down causation, bottom-up, or in some other way?
A central matter at the individual level is how top-down causation relates to the way the individual mind functions and to free will. But because this particular issue was dealt with in depth at an earlier Humble Approach Initiative symposium, which led to the volume, Downward Causation and the Neurobiology of Free Will (2009) edited by Nancey Murphy, George F.R. Ellis, and Timothy O’Connor, it is not a core focus of the London gathering.

The relation of top-down causation to language is, however, a central part of the discussion, and dealing with it inevitably leads to social issues. At the social level, top-down causation from society to the individual seems rather obvious, and the fact that, as Merlin Donald wrote in A Mind So Rare (2001), the “human brain is, quite literally, specifically adapted for functioning in a complex symbolic culture” has been known for a long time. The new point is that in the light of modern neuroscience, this effect is not just from society to the individual, but spans multiple levels and leads to the understanding that individual minds cannot be understood on their own; they can only be understood in relation to the society in which they are embedded. Family life, language, peer pressure, education, and role models are key ways in which this top down causation takes place from society to the brains of individuals.

Because the computer metaphor is dominant in most current discussions of neuroscience, it is important to consider the issue of top-down causation in digital computer systems, where the causal nature of the lower physical levels is particularly transparent. Issues include:

  • What is the ontological nature of computer programs?
  • How is the hierarchical nature of software related to that of hardware?
  • What category of causation is taking place in artificial neural networks? Genetic algorithms?
These difficult and compelling questions are being posed in a venue closely intertwined with the history of science. The Royal Society was founded 350 years ago when an “invisible college” of natural philosophers met at Gresham College to hear a lecture on astronomy by the young Christopher Wren. Their decision to found an organization for promoting the still new and controversial “experimental learning” has contributed profoundly to the advancement of knowledge throughout the world.