Health is one of our major personal concerns. Despite all the scientific progress, it seems that we are still fighting an uphill battle against the complexities in biological systems, leading to rising costs and limiting our progress. To understand if digitization might help us to cope with complexity, we talked to Sandra Mitchell, professor for History and Philosophy of Science at the University of Pitsburgh. In her book “Unsimple Truths: Science, Complexity, and Policy,” Mitchell proposes a multiple conceptualism to better understand complexity.
KG: We have based our scientific progress mostly on the method of deconstruction, by fragmentizing complex systems into those parts that we could analyze and understand and then reassembling them to get to the whole. Is that approach dead?
SM: I would not say so. It was and is very successful but mostly for very stable properties. It still works great, as you can see with the Human Genome project for example. But truly and deeply understanding complex systems with unstable or changing properties is beyond its limits. I think that for complex systems we need more analytic tools than deconstruction, understanding the parts in isolation, and then trying to synthesize to understand the whole system.”
KG: In our understanding of nature, and let us focus on medicine and biology for now, do we miss something very profound so that we can’t get to the complete picture?http://digitalistmag.wpengine.netdna-cdn.com/files/2015/11/sandra_mitchell.jpg
SM: I would say that we miss what is novel in complexity. In complex systems – and almost all biological systems are complex – a deconstruction will give us partial answers, only. We usually choose the approach according to the answer we want to get or according what we can achieve now. That is perfectly okay, as we made huge progress in science. But those insights may be constraining and will give us only partial pictures. So to get to a richer picture, we need a different approach.
KG: Completeness is a great catchword. I just read “Beyond Words: What Animals Think and Feel,” a book by Carl Safina. In it he makes the statement that scientists often neglect the emotions of animals, as it cannot be verified by standard scientific methods, despite the obvious, that some are obviously emotional, smart, and self-aware beings. He uses the example of native Indian cultures that seem to have a better grasp of the emotional and cognitive capabilities of animals, e.g. wolves. Of course they came to their insights by rather intuitive ways. Would intuition still be in your line of “integrative pluralism”? Is there any line that you would not cross?
SM: Now I have a catchword with wolves. Did you know that the reintroduction of wolves into the Yellowstone Park in 1995 changed the flow of rivers? We clearly knew what wolves would do and how they behave but nevertheless we got a surprising answer on a systemic level. Coming back to your question, we definitely have to agree on shared standards when we add more approaches to our scientific methods. I certainly could name some weird approaches but I was thinking more about aesthetical and spiritual approaches and non-linguistic interactions, such as body language. I don’t know if there is any fixed line we should not cross. I think it is more a matter of what would work and what would not to give us reliable and repeatable results.
KG: We currently witness a digitization of biology and medicine, getting us more data than ever before. Many scientists have the hope that we can finally solve many problems, like treating cancer. Do you share that view?
SM: Yes, I certainly think that Big Data can help fight cancer, as cancer is a disruptive process that should be analyzed at multiple levels. But Big Data would probably not get much beyond ”gene x does y,” which would not be sufficient for the bigger goal. So for epigenetics and the interaction of all co-factors involved, my view is that we need indeed multiple concepts as well as more data.
KG: So you see there is a potential for science in the ongoing digitization?
SM: The question is if we have the tools to model the coupled interactions, and I think that the advances in data analytics and modeling are promising. The question is if we will be able to integrate data from multiple concepts.
KG: In business, complexity is regarded to be something bad and in addition something that should and can be simplified. Is complexity an inbuilt feature of all systems? And how far can it be simplified?
SM: The goal is not to remove complexity but to manage it. Complexity can express many outcomes and respond fast to changes. That is the evolutionary advantage that those systems have. So simply reducing complexity might mean getting fewer answers and possibly failing.
KG: Are we able to think complexly or is that simply too much for our primate brain?
SM: We certainly can. The multisensory integration in our brain that happens in every moment and produces our adaptive behavior shows that we have been capable of coping with complexity for thousands of years already. I think that we have to shift our expectations when dealing with complexity, away from thinking of knowledge as a fixed state but more as a flux of insights.
Read more on the digitization of medicine and do not miss the forthcoming article by Sandra Mitchell and Angela Gronenborn “Why after 50 years are Protein X-ray Crystallographers Still in Business?” in the British Journal for the Philosophy of Science.
The article originally appeared on the Digitalist