For context, we must go back to the early twentieth century and the way science was often presented. A view known as Positivism held that science was about objective factual activities that, while they didn’t span the gap between Kant’s noumena and phenomena, nevertheless was the best we could get, and the science in question was physics.[1] As Rutherford repeatedly stated, “In science there is only physics. All else is stamp collecting.” This raised more than a few hackles among social scientists and biologists, including the famous ornithologist and evolutionary biologist Ernst Mayr, who in his later years turned to philosophy and history of biology. Having stated that when he turned to philosophy to aid his scientific work,
[a]fter switching my studies from medicine to zoology (particularly birds) following the completion of my preclinical examinations, I took courses in philosophy at the University of Berlin. But to my disappointment, they built no bridges between the subject matter of the biological sciences and that of philosophy. Yet in the 1920s and 30s a discipline was developing that would eventually be designated "philosophy of science." In the 1950s, when I became acquainted with the teachings of this field, I was again bitterly disappointed. This was no philosophy of science; this was a philosophy of logic, mathematics, and the physical sciences. It had almost nothing to do with the concerns of biologists. [1997, xiv–xv]
Mayr was a bit off the mark here. Among others, Henri Bergson and Pierre Teilhard de Chardin had by then developed what they saw as philosophies of biology (to various degrees of acclaim and disapproval by scientists),[2] and a nascent philosophy of the life sciences was arising even in analytic philosophy (e.g., Morton Beckner’s 1959 book). However, Mayr wasn’t wrong overall. Rutherfordian physics primacy did indeed rule the beginnings of twentieth century philosophy of science, at least in the analytic traditions, and there were real issues of application to the biological sciences.
In particular the task of science was revised to make Theory (yes, with a capital) the main if not sole target, and relations between different sciences were expected to be resolved by the process of reduction, so that all objects, processes and causes in one field were ideally to be reduced to the terms of another, deeper, theory. This was made canonical in a widely influential text, The Structure of Science by Ernest Nagel in 1961.[3] And of course the deepest layer of science was supposed to be physics, at least in the positivistic sense. This led to what is often called the “Layer Cake” view of science and metaphysics, with physics at base (an ideal physics we don’t have quite yet), then chemistry, biochemistry, biology, psychology and sociology (or maybe sociology first and then psychology, depending on your own philosophy[4]). Emergence was the most common defence of this layering – each layer emerges out of the immediate layer below. And this is what Mayr held - not only is biology a science in its own right, it cannot be reduced to lower layers, because emergence is always more than the action of its component parts. He called his view “organicism”, a kind of modern vitalism. Mayr was a zoologist and held the standard opinion in zoology that organisms were what philosopher Elliot Sober called the “benchmark” entity in biology (botanists are not quite so inclined). Moreover, he thought this was a kind of holism, in which whole objects had properties not had by parts. This isn’t how philosophers view holism. Instead they hold the view that for any complex, all aspects of the whole contribute to the properties of parts, especially in epistemology. In short, this is “downward causation”, a term coined by the psychologist Donald T. Campbell.
However, monists cannot accept this. A metaphysical layer cake supposes “spooky action” at emergent levels. Campbell and others who adopted this approach, especially in the 1960s onwards, needed to explain how it is that “mind”, which is in their view quite physical but emergent, can cause, for instance, physiological changes or actions. They were opposing the philosophical view of mind as an excretion of the brain, a view known as epiphenomenalism. I would also oppose that view, but I do this by rejecting the existence of “mind” as an emergent phenomenon. So far as I am concerned, “mind” is just the physical actions of a certain class of organisms interacting with their bodies, environments, and if they are semantic, their cultures. And what unifies the notion of mind is, in general, similarity to what human beings do, which is about as anthropomorphic as it is possible to get.
One proposal to explain this is that mental properties like seeing red or thinking about dogs are supervenient. Some property supervenes on physical properties when anything that has some specific kinds of physical properties must have the mental properties, but the mental categories could be instantiated[5] on many different architectures, as it were. Consider artificial general intelligence (AGI), if it can ever be achieved. An artificial system with AGI would see red and think about dogs in ways we would not, since at the very least it would be some kind of silicon-based system (if present technology is any guide). But again, this begs the question.[6] Why would “seeing red” or “thinking of a dog” be a mental category? Maybe when we do it, we simply respond to a class of stimuli that roughly approximates (because we made the categories) red and dogs. In short, maybe mental categories are functional or pragmatic classes, not physical. If that is the case, then of course they may have many different physical instantiations. Anything that matches or approximates the category instantiates those mental kinds.
In his mature reflections in holism, Mayr (2004, 76) uses the idea of a hammer, comprising the head and the handle, as an illustration of holism: the hammer has properties neither part has on its own. If that is holism, then anything at all that combines, for a use or which looks new is a “whole” that cannot be reduced. That is, if I may say so, a piss poor kind of holism. So far as I know, even the most reductionist of reductionists would have no issue with this. Nobody would cavil at claiming that two neurons have properties (i.e., the sending of signals between them) that one alone doesn’t. For holism to counter reduction, though, there needs to be “more” than just the actions of the neurons, and there isn’t. This is relational, not emergence.
In the end, emergence, holism and anti-reductionism turn on the nature of explanations. Are emergent, supervenient, or holistic properties something we discover, or something we create? What counts as “emergence” and why? That is the question. To this we turn.
Surprise!
Consider again, the novelties of evolution. Mayr writes:
The human mind seems to have been the ultimate product of a concatenation of numerous miniemergences, in both our primate and hominid ancestors. There simply was no instantaneous emergence. A product of an unbelievably complex central nervous system, the mind arose very gradually although at highly unequal rates at different stages. The period when language evolved, permitting both improved communication and the evolution of culture, was surely a period of greatly accelerated emergence of mind. [1997, 241]
A “miniemergence” is, in more usual words, a change of state. There is nothing holistic about this apart from the fact that the humans are an organism and so must, to an extent, be organised. In short, changes must not be wildly maladaptive. There needs no strange metaphysics come from emergence, to tell us that.[7] Why then is emergence such an issue? I believe it is because one thing pattern recognition systems do is identify novel (to them) patterns. And human scientists are pattern recognition systems par excellence. This gives us the clue: we see things in new ways when we do science, or learn about it, and the question is not whether these phenomena are caused by some non-physical process, but what is going on in scientific and other heads when they see things newly.
There is a term used in information theory and statistics that I think can be repurposed here: surprisal value. According to the founders of what was then “communication theory”, such as Claude Shannon (1948) and his predecessors, the information content of a signal received is the difference between a certain or uninformative message (such as “the night will lack sunshine”) which is 100% probable, and an unexpected message. This came to be known as the “surprisal value”, appropriately, as it uses the difference between the expectations of the receiver and the actual message or signal as the informativeness of that signal. In brief, it is the degree of surprise the signal evokes.
Now what a new phenomenon has in the way of information new to us is a kind of surprisal. For example, the shift from Ptolemaic astronomy to Copernican astronomy was not very surprising in terms of the dynamics of the universe: things still moved in circles in the heavens, although as the observations got more precise, circles upon circles and other deviations from the principle that all heavenly motion must be a circular motion increased. For a time, neither geocentrism nor heliocentrism held obvious benefits, although in some ways Copernicus’ view was easier to calculate.
But when Kepler demonstrated that orbits were elliptical, that was a surprise. It went against 2000 years of physics, philosophy, theology and practice. This was a novel way to think of things, and it was very informative. The prior expectation that heavenly motion was always a circle was based on the eternity of the heavens; circularity was the only eternally sustainable change. As Newton’s physics became widespread, based in part on Keplerian ellipses, though, it was clear that motion in the heavens was very far from being eternal (or at least eternally stable).
Now in evolutionary biology, it is clearly expected by many if not most biologists, or naturalists before the twentieth century and even into that age, that humans were qualitatively different in morphology and mental capacities from all other animals. The discoveries of the past century or so undercut that deeply. We know now that not tool use, nor some form of semantic communication, nor culture, nor self-awareness, nor problem-solving, nor anything else that used to mark out human capacities from the animals is in fact uniquely ours.[8] What makes it “novel”? It is that difference between what we expected (for historical and folk biological reasons) and what we have found out since Darwin and the development of comparative psychology. The emergence of the human mind is not a challenge to monism, but to dualism and human exceptionalism.
In other evolutionary matters, we often see biologists asserting that some trait or another represents a complete novelty, only to have it walked back again and again by further research and better observations. Adaptive radiations are facts, but they are conditional upon expectations that, for instance, flight is rarely evolved, and so on. At some point science must revise its expectations. Actually, I think this is something true of all of us. Learning is a process of revising expectations. This is why as we move into greater and greater specialisations in our education and training, we find it necessary to unlearn what “everybody knows” so often. Even specialists must do this. This is one reason why taxonomists revise the classifications of large groups, as more information and new methods come into play.
The limits of reduction
The antipathy to reduction by holists is understandable. Mayr’s claim that biology is irreducible to physics is similar to the claim that mind is irreducible to biology, and in a sense to be explained, this is correct. With certain caveats, that is. Notice that reduction is an epistemic activity. We count it as a success when one domain is reduced to another, because it is both explanatory and parsimonious. But this does not mean all else in science is a failure if no reduction occurs. Rutherford’s dictum is at best special pleading and at worst misrepresents what science and scientists actually do. Even in physics there are irreducible properties, as witness the current disparity between quantum phenomena at the small scale and relativistic phenomena at the large, not to mention the questions of dark energy and dark matter. Much observational physics seems to be stamp collecting at the moment.
But the real problem with reduction is not that it is a virtue not often achieved, nor that it is a sine qua non of science, but that it is hard! Very hard! Consider the inferential chain from basic physics to the behaviour of an organism. It is an ideal goal, but the computational requirements are likely to be beyond any system bar the actual world’s capacities to compute. The inferential distance, the amount of data, the precision of that data, the number of variables, and the number of causal relations are so great that it is intractable. So the argument that biology, psychology, sociology and the other traditional levels of the layer cake are unique is both trivial (all unique things are unique, no matter how much they share with other things), and pragmatic. We still want to understand life and mind even if we can’t yet or even ever practically ground them in a physical description. One can reject inappropriate calls for reduction and not require some new metaphysical layer of the cake, such as monads or ghostly properties like “consciousness”. All knowledge making is undertaken in media res (in the midst of things), and we must find our anchorage where we can.
Reductionism does not imply any kind of foundational ontology. For that, one must turn to metaphysics, and I have chosen monism/physicalism on the prior grounds of parsimony, not some scientistic hegemony. And given that I hold physical causality to be the ontology of the world, there is no need to be reductionistic: all things are physical. Everything else is a matter of our epistemology and cognitive limitations. In short, not a layer cake but a pizza. Everything is arrangements of the physical order. Complexity is in the mind of the beholder, include the complexity of the mind. Awareness is a physical thing, language is a physical thing, psychology in general is a physical thing. As the meme has it, I am sitting in a park at a table with a sign saying “prove me wrong”
.
Causes and processes
Causation, which is the technical term for the theories of causes, is a well-debated domain in philosophy and especially in the philosophy of science. It has been a topic since classical times, and we know it best now from David Hume’s account of causes as the constant conjunction of sensory impressions, and the habit of associating them as causes and effects. The Humean view is one widely adopted view of causation; others are more realist, such as Wesley Salmon’s (1984) view that causal processes are signal transmissions.[9] Neither Humean associations nor signal transmissions cause anything without there being a transfer process of energy. This is sometimes called the mechanistic view of causation, but all it really denotes is that things (processes) happen in the world over time for thermodynamic reasons. Even classical mechanism involves what we would now call “entropic” increase: that is, that the available energy to move things decreases over time. So “downward” causation is basically what Salmon called the “epistemic conception” of causality: it means that we have a state description of a system (the brain) which also describes dynamic changes to parts of the system.[10] But the description doesn’t cause the internal changes.
However, causation brings to the fore something else worth noting: all these causes are processes. A process is described by verbs – C makes E happen (mostly). In recent years there has been a philosophical revival[11] of what used to be called process philosophy, which coincidentally (or not) was devised by Alfred North Whitehead of the “the fallacy of misplaced concreteness” fame. This has the metaphysical advantage of not having to define boundaries in time or space, and the disadvantage of needing to work out what counts as a process, but I’m not going into this here. Instead, let’s reconsider what being aware is, in processual terms. Being aware is not a property but a process, a reaction and interaction between one’s cognitive and sensory systems and the rest of one’s body, conspecifics, and environment. That our central nervous system is the nexus of this process is not dualistic or supraphysical, but a simple matter of physical things making physical things happen. All the rest is a matter of convention.
References
Beckner, Morton. 1959. The Biological Way of Thought. New York: Columbia University Press.
Bergson, Henri. 1998. Creative Evolution. Mineola, NY: Dover. [Original 1907].
Blaikie, James. 1888. The Elements of Dynamics (Mechanics). New and Enlarged. Edinburgh: James Thin.
Kim, Jaegwon. 1993. Supervenience and Mind: Selected Philosophical Essays. Cambridge Studies in Philosophy. Cambridge: Cambridge University Press.
Mayr, Ernst. 1997. This Is Biology: The Science of the Living World. Cambridge MA: The Belknap Press of Harvard University Press.
———. 2004. What Makes Biology Unique? Considerations on the Autonomy of a Scientific Discipline. Cambridge, New York: Cambridge University Press.
McLaughlin, Brian, and Karen Bennett. 2023. “Supervenience.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta and Uri Nodelman, Winter 2023. Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/archives/win2023/entries/supervenience/.
Nagel, Ernest. 1961. The Structure of Science: Problems in the Logic of Scientific Explanation. London: Routledge and Kegan Paul.
Nicholson, Daniel J., and John Dupré, eds. 2018. Everything Flows: Towards a Processual Philosophy of Biology. Oxford University Press. https://doi.org/10.1093/oso/9780198779636.001.0001. [Open source]
Salmon, Wesley C. 1984. Scientific Explanation and the Causal Structure of the World. Princeton, N.J.: Princeton University Press.
Schaffner, Kenneth F. 2012. “Ernest Nagel and Reduction.” The Journal of Philosophy 109 (8/9): 534–65.
Shannon, Claude E. 1948. “A Mathematical Theory of Communication.” The Bell System Technical Journal 27: 379–423, 623–56.
Teilhard de Chardin, Pierre, 1960. The Phenomenon of Man. London: Collins. Introduced by Julian Sorell Huxley.[Originally written as an essay in 1930]
Whitehead, Alfred North. 1929. Process and Reality. Oxford, England: Macmillan.
[1] Ernst Mach in particular was a positivist promoter, rejecting atoms as insufficiently sensory and therefore mysticism.
[2] To be fair, though, French philosophy of the day probably wasn’t that popular at the University of Berlin.
[3] See Shaffner 2012 for an overview.
[4] Philosophy is either a layer above psychology or below physics, or both. Nobody knows what layer mathematics is, although a fair number like to think it grounds everything. See Reification Fallacy.
[5] Made real, or as I once said, poured into the conceptual mould with ontological cement. Supervenience is a notion developed by Jaegwon Kim. it is much more complex than I have suggested, and those who are interested should read McLaughlin and Bennett 2023.
[6] Which truly means to assume the explanandum in the explanans, or to assume the correctness of an explanatory category that needs explanation.
[7] Yes, I am paraphrasing Hamlet, badly. It’s said by Hamlet’s friend Horatio (Act 1, scene 5).
[8] I might give ground to extended recursion and displaced reference as discussed earlier, but even these are not qualitatively distinct. They are more extended or complex with human language than communication in other animals, but this is a quantitative difference.
[9] He means the physicist’s notion of a signal akin to the view that information is not lost in a black hole.
[10] In the late nineteenth century a distinction was made between kinematic and dynamic processes. The first tells us about the motion. The other gives the causal reasons. See Blaikie 1888.
[11] See Nicholson and Dupré (eds) 2018 for discussion.
Really enjoying this series, John. I hope this is heading to a book? :)