A Process Ontology for Biology

John Dupré

Reflection on the last hundred years of physics might naturally lead one to suppose that the ancient debate as to whether the world was ultimately composed of things or processes had been resolved in favour of the latter. Quantum physics, whatever else it may be, seems to constitute a decisive rejection of the atomism at the root of traditional thing, or substance, ontologies. My interest, however, is in larger scale entities, those that are the subject matter of biology. Within this narrower sphere, the basic parameters for most subsequent thought were laid down by Aristotle, and the Aristotelian tradition is firmly on the side of things rather than processes.

Organisms were for Aristotle the standard exemplars of what he called primary substances. A primary substance was a thing of a particular kind, for example a cat. Particular cats, such as Tabitha, are assumed to have sharp boundaries: everything is either part of Tabitha or not. Subsequent substance-based thought about the living world has seen substances as forming a hierarchy. Cats are composed of organs, organs are composed of cells, etc., and cats might themselves make up larger entities such as species.

Against the tradition of substance thinking an important minority of philosophers have defended more processual views of reality. These include such major figures as Leibniz and Hegel, but the definitive exposition of a processual metaphysics is in the work of Alfred North Whitehead, for example in his classic, Process and Reality. Nevertheless, in the metaphysics and philosophy of science this tradition has had very little impact in recent years. My current ERC-funded project, ‘A Process Ontology for Contemporary Biology‘ aims to redress this balance.

One common statement of the difference between a substance- and a process-centred metaphysics is that the former sees being as basic, and the latter prioritises becoming. For Aristotle substance was closely connected to essence: to be a cat was to possess the essence of catness, the property or properties that made that entity a cat. For a processualist, on the other hand, a cat is a pathway from zygote to kitten to mature animal to death. Any time slice of this sequence can only be properly understood as part of this longer process, and in terms of both its preceding history and its possible future trajectory. No property need be common to every stage of a cat life cycle; it is rather the relations between stages in its history that constitute it as a cat.

For a substance theorist, what require explanation are the changes that occur to an entity, and the conditions under which an entity can remain the same thing through change. The latter question is most visible in the longstanding debate about personal identity: in what sense can I be the same thing as the foetus or infant from which I originated, given that neither my matter nor my properties have remained the same? For a process theorist, on the contrary, the central questions concern rather how a combination of processes can maintain the appearance of stability and persistence in an entity that is fundamentally only a temporary eddy in a flux of change.

I noted that an Aristotelian substance metaphysics assumes that things have determinate boundaries.  It is no doubt possible to modify this assumption and allow for some fuzziness about exactly where an entity begins and ends. But—and this is one of the central reasons for preferring to think of life in processual terms—it increasingly appears that the attempt to identify even fuzzy boundaries as objective features of the living world may be fundamentally misguided. The omnipresence of symbiosis begins to suggest that isolating traditional organisms from the complex set of relations that are required for their persistence may be at best a pragmatic exercise for particular purposes. 90% of the cells in a human body, for example, are microbial symbionts, huge numbers of which are required for healthy functioning, and others of which have more or less cooperative relations with the whole system or important subsystems. And humans are probably as close as one could get to a traditional biological object with sharp boundaries. In general, it seems increasingly plausible that there is no unique way of distinguishing biological things from the multiple processes by which they are sustained, which supports the hypothesis that the ontology of biological things is less fundamental than that of living processes.

Does this very abstract issue matter to our understanding of biology? Of course part of the answer is just that we would like to know how things are. Metaphysics has value in its own right. But I also believe that adopting a process-centred approach can make a difference to quite practical questions in biology. I shall briefly sketch some examples to illustrate this claim.

My first example is rapidly becoming an historical one. It is the understanding of the nature of the genome. One of the deepest problems in biology is to explain development, the process by which organisms reliably grow into organisms of the right kind. Dogs give birth to puppies not snakes or cabbages. Preformationism, a solution to this problem often illustrated by quaint seventeenth century pictures of tiny babies squatting in the head of a sperm is an attractive solution to this problem: the structure of the adult is there from the beginning. Not long ago genomes Dupre - preformationismwere widely seen as offering a sophisticated version of such a solution. The actual structure of the adult was not there from the start, to be sure, but in the genome we could find a programme or a blueprint from which the organism could be constructed.

This interpretation of the genome sees the genome as a stable thing with remarkable properties that somehow encode or represent the future organism. But it is now clear that the genome is, on the contrary, an entity in constant flux, a process, I would say, rather than a thing. Although nucleotide sequence is a highly stable feature of genomes, even this is maintained as such by a range of editing and repair processes. Moreover, the function of the genome is constantly sensitive to interactions with a variety of molecules in the cell which, in turn, change the physical structure of the system in ways that determine what genes are expressed. The genome, in short, is a fully interactive participant in cellular processes. I said this was increasingly a historical issue, as the failure of the idea of genome as blueprint is no longer very controversial among well-informed scientists. The interpretation of this change in terms of process ontology is, no doubt, less widely agreed.

A second, more general, issue that is transformed by a processual perspective on living systems is the distinction between structure and function. It is common to think of biological objects having particular structures that enable them to perform particular functions. But if these ‘objects’ are in fact constantly fluid and evolving processes, this perspective can be misleading. Structure and function are intertwined aspects of process. A good example of a phenomenon that invites such a view is plant development. The growing meristem of a plant is typically an opportunistic growth process capable of producing a variety of structures—leaves, flowers, roots—in response to the environment it encounters. Even the attempt to distinguish sharply between these traditional structures is often problematic.

Proteins are perhaps the paradigmatic examples of biological entities for which structure is assumed to determine function. However, this simple structure/function analysis of proteins has had increasing difficulty as it has been found that many proteins serve a range of functions (‘moonlighting’ proteins); that most proteins do not have a fully determinate structure (‘intrinsically disordered’ proteins); and that the interaction between an enzyme and the molecule with which it interacts, does not fully fit the traditional lock and key molecule, but rather involves a considerable amount of mutual configuration. All of these phenomena fit better into the view of the protein molecule as a dynamic entity, the causal powers of which are constantly being reconfigured in relation to the processes in which it participates, than into the classical model of a thing with a fixed nature that determines once and for all what it is and what it can do.

As a committed pluralist, I don’t want to assert dogmatically that the world is composed of processes not things. However, I am confident that a process ontology provides a generally more illuminating view of the living world, and certainly the same seems true of contemporary physics. At the very least the question whether things or processes provide a better framework for interpreting science is one that should be a central concern for everyone interested in the metaphysics of science. And perhaps it is a question that matters beyond just getting to the truth. As Whitehead wrote in Process and Reality:

There is urgency in coming to see the world as a web of interrelated processes of which we are integral parts, so that all of our choices and actions have consequences for the world around us.

John Dupré
University of Exeter