Science as Process Versus Science as Behemoth

I live on a small island off the southwest coast of Ireland. When I want to go out, I get a boat to the mainland. It’s a 10-minute crossing. I usually spend the time hanging out in the wheelhouse, catching up with the skipper and deckhand on duty.

One morning a couple of months ago, our conversation drifted on to global warming. In response to my concerns, the skipper casually commented that there was nothing to worry about, because there was no such thing as global warming. I spluttered. I cited NASA, global scientific consensus, the United Nations’ reports, melting glaciers, rising sea levels. As we pulled in to the pier, he countered, “I’ve been working on the sea all my life. If the sea levels had risen, I would’ve noticed.” Infuriated, I muttered something insulting about arrogance and stormed out.

How had we reached such conflicting conclusions? Most of the research trying to figure out why people reject scientific explanations focuses on the people themselves and their cognitive deficits, social marginalization, lack of education, fantasy-proneness, or irrationality. I’d read and written about such theories before, but there was a problem: I knew the skipper, and I didn’t have any doubts about his ability to problem-solve or understand evidence. He used explanatory virtues extolled in philosophy of science—simplicity, unification, coherence, and empirical adequacy—to infer the best explanations for engineering and navigational problems in the course of a day’s work. If he possessed explanatory virtues and used them every day, why was he ending up with such a disastrously wrong conclusion?

The conclusion marked the battle line: I accepted the scientific account of global warming, and he rejected it. Our conversation had ended when those conclusions clashed, and I stormed out of the wheelhouse on the back of a seemingly intractable disagreement. I fumed and fretted for weeks before apologizing for storming out. Some tentative further comments were exchanged; I argued for scientific consensus, technology, the gradual accumulation of masses of data. I was told I shouldn’t believe everything I read in books. I fumed again.

Waving data around and calling it evidence hadn’t changed anything. Gradually, the source of the intractability became clearer. It lay much further back up the chain from the evidence or the explanation to more over-arching questions: Who decides what counts as evidence? Why should I believe a theory that fails to describe the world in front of me? I was further dismayed to find that a theory lurked beyond these questions: the pessimistic meta-induction from past falsity. There it sat, like a seagull perched on the gunwale, squawking: “Why should we believe the claims of science when they’ve been wrong so many times in the past?” If your electrician was wrong about a problem with the radio system more than once, you’d find another electrician. You can’t trust the explanation if you don’t trust the source. Psychological theories about false beliefs tend to blame individuals for engaging in bad epistemic practices when they reject scientific explanation, but that didn’t seem to be the issue at hand. The skipper’s epistemic standing seemed sound. He had simply gone meta.

This was why my attempts to pummel him with evidence had been so resoundingly unsuccessful. The intractability in our positions had formed prior to any evidence about global warming. The pessimistic meta-induction from past falsity gave the skipper reason to doubt science in general as a source of true claims about the world. I acknowledged past falsity, but clung to scientific method like a life-buoy, relying on it to justify my trust in scientific consensus as a source of reliable (if contingent) beliefs. Our conclusions hadn’t changed; I insisted that science was reliable and the evidence about global warming was conclusive. He insisted that science can’t be trusted to make those kinds of claims. We were deadlocked.

In his 2018 New York Times op-ed “Knowledge, Ignorance and Climate Change”, N. Ángel Pinillos suggested that in conflicts like these, we might shift from knowledge to probabilities. Maybe pointing out that I agreed with 97 percent of climate scientists that it was extremely probable that global warming was man-made, rather than insisting that I knew it was, would loosen up our standstill. But the intractability didn’t seem to be about evidence at all. It was about who was in charge of the evidence. My fuming had abated, mollified by a better understanding of the skipper’s reasoning, but the fretting continued. It wasn’t enough to shrug off the intractability and resign us to disagreement when the dispute we were playing out was so common, and the consequences of global warming so severe. There has to be a better response to intractable disputes than just throwing up our hands and turning our backs on discussion.

A chance for traction came when I read Daniel Wilkenfeld and Tania Lombrozo’s 2015 paper “Inference to the Best Explanation and Explaining for the Best Inference,” In it, they discuss shifting the focus in explanations from the product (the explanation) to the process (the explaining). They argue that the process of explaining has cognitive benefits, even where the conclusion reached is untrue. I was persuaded, but beyond that, I thought that concentrating on the process of explaining, rather than the final claim of the explanation, might unstick deadlocked disputes like that between me and the skipper. Rather than bashing each other over the head with rival conclusions, focusing on the explanatory process might keep the conversation moving in a way that promoted clarity and defused the conflict. We might find things to agree on within the process even if we disagreed on the end product. It could draw out the stages of reasoning we each went through on our way to choosing the best explanation we could find, and clarify the precise loci of disagreement.

Still, we can’t pretend that the conclusions don’t matter, especially with something as critical as global warming. Conclusions direct actions. Focusing on the process of explaining seems like a promising move when we want to gain traction in deadlocked disputes and counter polarization, but it’s worrying that we can apply proper epistemic processes in providing an explanation or choosing between competing theories and still end up with a conclusion that contradicts the evidence. If people don’t trust science as a source of reliable information about the world, the core intractability remains. We faced a deeper problem: Down at the root of the intractability, the skipper and I didn’t even agree on what we meant by “science.”

Refocusing on explaining rather than explanation has more to offer here because it lines up with the two different perceptions of science that seem to be operating in our dispute and in others like it. As Wilkenfeld and Lombrozo point out, science itself is more like explaining than explanation. It’s a process, not a product. That’s my understanding of science, too; it’s a dynamic, poly-vocal system of gathering data and evaluating theories according to stress-tested established methods. Those who share the anti-science suspicion see it less like a process and more like a product—like an anonymous univocal behemoth that spits out claims. On this view, the falsity or failure of previous scientific claims shows that the behemoth is unreliable and can’t decide for us what to believe. Science-as-behemoth is capricious and monolithic, bullying the non-science world into believing what it tells them to believe, and then changing its mind later anyway. When we disagree about whether or not science can be trusted to provide reliable explanations, we’re using two very different pictures of science. No wonder we disagree.

Here’s a plea for agreement: Science explains things. That’s the general idea. Science explains how things in the world work, but it’s got a bubbling credibility problem. The problem is stoked by a perception of science as a monolithic machine that churns out opaque explanations, rather than a massive conglomeration of moving parts all involved in an ongoing process of explaining. If science is understood as an explaining process rather than a random generator of a product called “explanation,” the failure or revision of past theories is easier to understand. By focusing on what we’re doing when we explain things, we can then draw correlations with how science works through a similar process. It’s a big shift, and a big ask when you’re getting deadlocked on fundamental beliefs with potentially catastrophic consequences. It’s a lot easier to just throw your hands up and storm out of the wheelhouse. My local motivation to shift my focus lies in not wanting to fall out with the boat crew. The global motivation, for all of us, is a much more serious concern.

Editor’s note: This article received honorable mention in the APA Blog’s first Public Philosophy Award for Undergraduates contest.

Rhona J. Flynn

Rhona J. Flynnjust completed her undergraduate degree at University College Cork (Ireland), taking joint honors in philosophy and study of religions. She is an artist and writer, and has received several academic awards.