Alan: Hi everyone! Welcome to In Limbo Conversations. Today, we have with Stephan Guttinger. He is a philosopher of biology with a background in biochemistry. He is a research fellow at the Centre for Philosophy of Natural and Social Science (CPNSS) at the London School of Economics. His research focuses on the replication crisis in science and how ongoing paradigm shifts in biology affect (or should affect) our thinking about health policy.
In the first part, Stephan, you explain the lure of the thing-view as being the “well-defined” boundaries that are set up as part of methodology that is spawned by this perspective, which seem to help clean up the messiness of context/environment thus enabling us to study the entity in isolation. Are there any existing processual-methodologies that can be called upon as reliably as the essentialist methodology? Or does the dynamic nature of entities in the processual perspective mean that the methodologies have to be constructed from scratch for each case?
We do things in a certain way, we shape the world we encounter through specific practices, and this then instils in us a belief that the part of the world we are looking at is shaped in a certain way.
Stephan: That is a good question. To answer it, I probably have to further qualify what I think about the lure of a substance-view, and in particular about the relation between ontological background assumptions and actual practice. One factor here is the question of directionality. As you put it in the question, it looks like a one-way relation, meaning there is path from ontology to practice: researchers have a certain picture of what the world is like, and this then guides and shapes their practice. I think that will be the case for some people or research contexts. But I think the relation is often also the other way around. We do things in a certain way, we shape the world we encounter through specific practices, and this then instils in us a belief that the part of the world we are looking at is shaped in a certain way. This leads to some interesting tensions. For instance, the one thing a protein biologist always encounters in practice is constant change. Proteins unfold, or they switch fold, they fall apart, and so on. The one thing you constantly do as a protein biologist is to try and somehow keep that darn thing stable. This means that you keep all your samples on ice, you use buffers that keep your proteins at a certain pH, you don’t want to go too high or too low with your salt concentration, and you run your experiments for only a couple of hours, not days. These practices are so wide-spread and deeply ingrained that they almost become invisible. Once you stop seeing them, all you see are stable proteins with distinct structures and well-defined functions. You have moved yourself into a thing-like view of the protein world, but you have done so exactly because you were so painfully aware of the deeply dynamic nature of the things you are dealing with! When new technologies and practices emerge, then a different mode of ‘seeing’ proteins can open up (no necessity here). This has happened, slowly, in protein biology through the emergence of NMR and new X-ray crystallography methods. And it is also happening in virology, through new sequencing technologies. This can then change our picture of what we have in front of us. The dynamics of the entities that are investigated are now more in line with the technologies we have. The new methods and instruments work whilst also accommodating dynamics. We don’t need to stabilize and rigidify in order to trace and observe ‘things’. But this does not mean that these new methods ‘belong’ to a process perspective, whereas there is another set of practices that belongs to a substance view. What I am trying to say here: the connection between ontological assumptions and practice is certainly much more complex than I might have made it sound in the blog posts. But one thing that certainly stands out is that settling on a thing-view provides the researcher with a much easier narrative and way forward. You are justified in studying something in isolation, you don’t have to engage in a whole range of furher investigations. From a process perspective, more is needed. So, it is maybe not so much that there is a set of ‘process-related methodologies’ that always have to be built from scratch because of the dynamic nature of the world. Researchers will always build on a similar set of evolving practices. But if you explicitly adopt a process-view, you have to construct a more complex body of experiments, and checks, and further data, whereas with a thing-view you could more easily claim that you are done for now and that you have shown what protein X (or virus Y) is and how it behaves.
Alan: So- how does the thing-view of viruses, the understanding of them as these substances with essential properties have methodological implications? How do such implications show up in, say, the limited effectiveness of antiviral drugs (which are usually produced in-vitro but are executed or implemented in-vivo)? Could you show/tell us about the chain of thought here? If possible, could you also tell us some other pandemic instances where the thing-view of viruses could methodologically affect decision making in scientific practice?
Stephan: I think the thing-view comes to the fore most clearly in some drug-design studies. The reasoning underlying many of these studies is that there is a given set of proteins that the virus contains, each of which has a well-defined structure (or a structure that switches between two well-defined states) and a function. There is a single cell entry mechanism that the virus relies on, and for which one of the viral proteins is responsible (e.g., the spike protein of coronavirus). To block viral entry all the researcher has to do is to design a drug that sits on the surface of the spike protein in a manner that blocks the interactions needed for viral entry. The drug would then shut down viral reproduction, because the virus can no longer gain access to cells.
The dynamics of the entities that are investigated are now more in line with the technologies we have. The new methods and instruments work whilst also accommodating dynamics. We don’t need to stabilize and rigidify in order to trace and observe ‘things’. But this does not mean that these new methods ‘belong’ to a process perspective, whereas there is another set of practices that belongs to a substance view.
Now, this approach can of course be successful. Under certain conditions, and for a certain amount of time, viral processes are stable enough to be affected by such interventions. But these stabilities are of course not some sort of intrinsic or eternal feature of viruses. The system that is targeted is more fluid and diverse. Researchers might be able to ignore these dynamics when they work in silico or in vitro. The computational models of proteins (or the isolated protein in a test tube) don’t evolve, and interventions that look promising in such a research setting might not be as effective in the non-laboratory context. A molecular biologist doing docking studies, combined perhaps with some in vitro tests, will be able to find a compound that does the job of blocking viral replication. You can find plenty claims to that effect in the literature. But those working with the actual process, the living virus, encounter a different reality. This is also reflected in the fact that very few compounds are effective, even though they might have displayed high efficacy in in vitro studies.
Regarding the second part of the question: I think a thing-view of viruses can also affect other aspects of pandemic-related reasoning or practices, not just drug design. An example here is the approach to avoiding the next pandemic. In this case, a thing-view might promote a picture of viruses as simply sitting out there, in their ‘natural habitat’ so to say. I quickly touched on this in the second blog post, where I look at authors such as David Quammen, a science writer who has written a lot on the One Health approach.
Alan: Something you said really struck me. You had said- What (such) a mutant cloud is depends on what it does. Could you tell us how we could explain the novel coronavirus in terms of mutant cloud dynamics? Especially taking into account the larger systemic changes like shifts in practices of animal husbandry, the flow of capital, the dynamics of viral populations? If we understand coronavirus as a mutant cloud, then what kind of network are we looking at, through the process-view, which could help us to understand how we reached here today?
Stephan: That is a very interesting question. There are some papers out there on the cloud nature of coronaviruses, and also SARS-CoV-2 more specifically (for instance here or here). But a lot of this work reflects on the cloud as a molecular phenomenon, analysed in a relatively narrow context (a single patient, for instance). This is of course shaped by how this ‘cloud’ can be seen, namely through genome sequencing. Sampling and sequencing are bottlenecks here. It is relatively easy to take samples from one patient, or a well-defined group of patients, and to monitor them over the course of a few days. What you point to in your question is a much broader analysis of the ‘cloud’ nature and its implications. Such analyses, I think, are still largely missing. To say something about these larger networks we would need a much broader and more diverse sampling process. We would need more environmental sampling, comparing rural populations and urban populations, including animals, both wild and in a farming context. And it would have to be a global sampling approach to understand how the virus as a more abstract cloud has moved and is moving. Another issue is also how the mutant cloud changes once you are in a pandemic – pre-pandemic and in-pandemic sampling might have to look slightly different. Because, as has become painfully clear over the last six months, once the pandemic is here the world moves differently. Interventions such as social distancing, face masks, or vaccinations will re-shape how the virus behaves. So, the dynamics that you might observe in pre-pandemic times might not apply to the in-pandemic context. But the point of course remains the same: if we understand the virus as a cloud, either as a more specific cloud that moves within a single body, or as a more abstract cloud that starts to engulf the planet, we think differently about ‘the’ virus and ‘its’ behaviour when we use the cloud metaphor. Time, place, and interactions become part of what makes the viral process the process it is (for a certain amount of time). The investigation and analysis of this cloud will not reach an end-point. I guess that goes back to the first question, the openness and pervasiveness of investigation that a process view demands, much more, perhaps, than a thing-view might.
But the point of course remains the same: if we understand the virus as a cloud, either as a more specific cloud that moves within a single body, or as a more abstract cloud that starts to engulf the planet, we think differently about ‘the’ virus and ‘its’ behaviour when we use the cloud metaphor.
Time, place, and interactions become part of what makes the viral process the process it is.
Alan: You had mentioned that it is possible that we could find ways to interfere with the dynamics of mutant cloud. We could twist the system in ways that could result in extinction of a virus. I think you had used the term ‘lethal mutagenesis’. If we could imagine something like this happening in case of coronavirus- could you tell us what it would be like? What would lethal mutagenesis of novel coronavirus mean?
Stephan: The answer here depends on being clear about the system we can or should interfere with. In the previous answer I talked about a larger system within which the more abstract ‘cloud’ moves and develops, and I contrasted this with the narrower system of the individual body. The idea of lethal mutagenesis is usually applied to the latter. The idea here is to unbalance what is already a fragile process. Viruses are racing ahead, creating new mutations at a high rate when they multiply in a single body. The same applies to SARS-Cov-2. A recent study has shown that the virus variants found in a patient who was monitored over several days changed from one day to the next. Viral processes behave in that way for a reason. Some of the new versions might come in handy when conditions change, as they might contain that one little change which means that a drug or an antibody no longer interferes efficiently with the viral process. Scientists refer to these varianst as “escape mutants”. But the whole thing comes at a cost, as most of the new variants created will not be beneficial, or even detrimental, to the virus. It is a fine balance that the viral process has to strike. The point of lethal mutagenesis is that outside intervention can tip this balance by tweaking the mutation rate of the viral process. This works because the mutation rate is not some sort of intrinsic or fixed feature of the virus. How the virus behaves is shaped and defined, in part at least, by the larger system. This includes the resources that are used by the virus-cell system to replicate the viral genome. We can replace some of these resources (for instance nucleosides, i.e. the building blocks for DNA or RNA) with analogs that make the replication process more error-prone. Many antiviral drugs work with this principle and it has just been reported in a paper in Nature that this approach also seems promising for SARS-CoV-2. So yes, the approach can work for the current virus as well. But the problem of the mutant cloud, in particular the issue of escape mutants, and the related question of resistance to these new drugs also comes up again.
Alan: There is a certain perspective which recommends that we should just box territories so that we don’t even create opportunities of contact which could turn unfortunate- like the idea of wild being out there which should be, as Quammen had said, “left the hell alone” and our urban cities as the place where we are supposed to live and this is rationalized usually by utilizing the idea of sacredness of nature- how we should not exploit or violate nature for our purposes. The sentiment- that we should not exploit the resources- seems to be well-meaning and this idea of boxing territories seems to be what comes when we apply this sentiment from a things-view. Do you feel that from a process view point, we could still talk of exploitation of resources, of maltreatment of animals? What would you suggest, in concrete measures, to respond effectively, given the process-view?
I think a process view still has a place for values, and for the idea of protection. The main difference is that these values will not be justified by reference to something ‘sacred’ or inherently valuable.
What has to be protected is seen as worthy of protection not because of what it is, but because of what it does.
Stephan: That is a good question. I think a process view still has a place for values, and for the idea of protection. The main difference is that these values will not be justified by reference to something ‘sacred’ or inherently valuable. What has to be protected is seen as worthy of protection not because of what it is, but because of what it does. Within an ecosystem of processes there can be processes that are deemed more important than others for the maintenance (or decay) of a certain state of affairs. For instance, biodiversity might be such a process. You don’t just have biodiversity, it is something that constantly grows and develops, based on a confluence of processes. And this ongoing process has effects on other processes, for instance, as some suggest, on the emergence of new infectious diseases. We don’t need fixed boundaries between well-defined things in order to have a debate about exploitation and protection, I guess. What’s more, in a process perspective there would be, I think, less of a drive to isolate particular locations or entities. Even if you identify a particular process as worthy of protection, you cannot think that process as something isolated from other processes. You are forced into a deeply interconnected reasoning. Within the box view we might end up with more simplistic solutions, such as shutting down wet markets and staying out of bat caves. These might be part of a solution, but they are unlikely to offer ‘the’ solution. This is also where the politics of a substance- or thing-view become important: a box view allows us to formulate simple recipes and solutions. There is a clear ‘us’ and ‘them’, and problem areas (or populations) can be isolated, closed, or punished. This appeals to politicians who need to present quick solutions and to demonstrate activity. In a process-view such simple solutions are not that convincing. Once more, this goes back to the very first question about methodology. A process-view not only forces a different process onto laboratory-based answer-finding, but on society more generally.
This is also where the politics of a substance- or thing-view become important: a box view allows us to formulate simple recipes and solutions. There is a clear ‘us’ and ‘them’, and problem areas (or populations) can be isolated, closed, or punished. This appeals to politicians who need to present quick solutions and to demonstrate activity. In a process-view such simple solutions are not that convincing.
Alan: If we look at the difference in process and substance ontologies, then there is the framework of a substance and its properties in the latter- could we say that for the former, for process ontology, the central categories are process and its interactions? How do the categories of relations and structure feature in process ontology?
Stephan: The question of the differences between the ontologies is an interesting one, especially as there is not just one process ontology and one substance ontology. Different authors define ‘process’ in different ways (often using quite obscure language) and the same applies to ‘substance’. On top of that, key concepts play a role in both ontological frameworks. Take the example of ‘interactions’ that you mention. Interactions are a key part of some process frameworks, but they are also a key part of substance views. The point is that they are not treated as some sort of ‘fundamental’ feature of the world in the latter. This is also picked up by Nicholas Rescher in his book Process Metaphysics. He suggests that the different views should not be seen as some sort of well-defined theories or bodies of theories, but rather as ‘points of view’ that set different priorities. Stability and change, for instance, matter for both ontologies, but for a process view change is fundamental, whereas a substance view treats it as secondary, because it sees stability as a basic feature of things. The same with relations or interactions, I guess. You cannot explain the world from a substance perspective without also factoring interactions. So, in that sense interactions or relations are central to both views. But the substance view does not see them as fundamental. With ‘structure’ the issue is similar. The point for a process view is, as I understand it, that structure is important, but it is always something changing and co-produced, rather than something that is simply given and that ‘belongs’ to one thing. A process view has a central place for the category ‘structure’, but it gives it a different priority. I hope this answers the question?
Alan: Yes, it does!
In your series, you have shown how the “substance” or “thing-view” has been the prevailing paradigm for a very long time, directing our thinking about things, with its effects evident in major policy decisions and responses. You also call for the replacement of this view with the processual view. With regards to replacing the deeply entrenched thing-view, it seems processual thinking has its work cut out for it. How then do you see process thinking eventually making a space for itself in such issues and what are the challenges it faces to this end?
Stephan: The power of the thing-view is a difficult issue to explain and address. It goes back to this tension I mentioned in my first answer. Every practicing biologist, I think, is painfully aware of the dynamic and deeply relational nature of the ‘things’ they study. At the same time, the thing-view is so powerful and appealing in biology (and elsewhere), for a range of reasons. What exactly that set of reasons is, however, I am not sure. Part of the issue here will be politics, in the sense of running a polis, or state, or research centre. A thing-view allows us to think through complex issues more easily, to propose solutions, and to get to work. On top of that, our practices, which are in place so we can actually get stuff done, make the world look more stable and individualized than it is. So, it is not only convenient to think in terms of a thing-view, but in many aspects the world presents itself to us as something stable and as composed of well-defined ‘things’! It is like with a car engine: it is a process that decays quickly, but we have all these background practices in place that keep it running smoothly and stably. Take the example of older two-stroke engines: the petrol used for these engines always had oil already mixed into it, so that there could be adequate lubrication. If it were missing, the engine would break rather quickly. But such background processes (constantly adding oil, or using refigeration in the laboratory) become standardized and part of the furniture. As a consequence, we often stop seeing them. And then the world looks much more thing-like and stable than it actually is. Everyone who owns a car, or who works in a laboratory, knows that stability does not come for free. Decay and change are basic features of biological systems and machines (they are both processes after all). But all of this actually makes it even harder to argue for a process view, I think. Because, if what I say above is true and people are aware of the pervasiveness of change, then this means they are able to forget about it, to work around it, or to ignore it (and to do so for a variety of reasons). So, making the argument for a process view is not about formulating the final and correct account of a process view. The problem is maybe more about reflecting on practices, rather than theories. Whether this can happen as part of a purely philosophical analysis is an open question. My guess would be that a much broader approach has to be taken, but what that would look like I don’t know. What I do know is that you are right when you point out that process thinking has its work cut out for it!
...making the argument for a process view is not about formulating the final and correct account of a process view. The problem is maybe more about reflecting on practices, rather than theories...
Alan: In the end: How do you feel philosophers of biology can contribute to the public practices today, in such a time of crisis? What do you feel are some things we, as philosophers, could do?
Stephan: The role of philosophy in public debate, and practice, is an interesting one. Especially in the context of a debate about something urgent like a pandemic and the question of how to deal with it. As we have seen again and again over the last six months, these debates are dominated by numbers and data, including surveys, sequencing data, models, predictions, and guidelines designed by health practitioners. Those who produce these kinds of outputs are the primary sources that people turn to for advice. I could be wrong here, but my guess is that there are no philosophers on the Scientific Advisory group for Emergencies (SAGE) in the UK? This means that philosophers are, at least on an institutional level, not really in the fold, they don’t have a seat at the table. But if that is the case, my thinking would be that not having a seat at the table can be a big advantage, because it means you are on your feet. And this means you are more mobile and you are not forced (if that is the correct word for people on SAGE) to take part in a very specific type of debate. As a philosopher we have certain liberties (even if our funding is more and more connected to ‘Impact’ and other diffuse measures). By not being at the table, and by not being part of the system that produces the primary data, philosophers can take a step back – or a step to the side, to be less negative – and run a different kind of analysis on what is happening at the moment. We of course have many pressures ourselves, especially the many of us who now have to work under abysmal temporary contracts with massive teaching loads. But at least in terms of academic freedom, we can be more nimble. We can draw lines and connections that can only be drawn based on an analysis that is less hurried, less bound up with specific goals, time pressures, or peer pressures. We don’t have a SAGE meeting or a government briefing to go to and where we need to present our numbers. I might be wrong here, but I feel we can be more experimental, and more expansive in our thinking. This allows us to draw connections and to reflect on terminology or strategies that other parts of the research landscape might neglect. Not being at the table of course also comes with risks, especially the risk that no one will listen. But every now and then, something will stick and make a small difference to how the larger issues are approached and dealt with. It might be a slow process, but as Tesco puts it so succinctly, every little helps, I guess.
I feel we can be more experimental, and more expansive in our thinking. This allows us to draw connections and to reflect on terminology or strategies that other parts of the research landscape might neglect.
Alan: Thank you so much for conversing with me Stephan!