Why do I trust science?
Written on 2021-09-25
Like many people these days, I have been getting into quite a few arguments about Covid - or, more precisely, about how to deal with it. Of course, it is not in the least bit surprising that this enormous crisis has been dominating our conversations, or that there are strongly-held opinions on what we should do about it. But I have frequently been frustrated and often saddened by the supremely unscientific claims being advanced on many sides. As a biologist, this pains me particularly, and I have spent many hours trying to make the science behind the pandemic more understandable to my friends and acquaintances.
The experiences of these months have led me to think more and more about the issue of trust. Whom do I trust, and why? This may be the single most important question we're dealing with in the pandemic; an unspoken question that, in the final instance, underlies everything everyone of us does. As (mostly) rational creatures, we humans act on what we know and have recognised to be true. So how can I know what is true?
This is a question that every single one of us has answered, albeit usually subconsciously. Here, I want to make this answer explicit: I want to explore why we believe certain pieces of information and not others, and why I am so convinced that we can, despite many claims to the contrary, trust the science.
How do we know what we know?
Not surprisingly, this is a very, very old question. In fact, philosophers have been arguing about it pretty much ever since the first philosopher opened his mouth and the second philosopher begged to differ. The debate has been going on ever since in the philosophical discipline of epistemology. Thankfully, some of its core findings are still not that hard to understand.
There are three common criteria to judge whether or not a new piece of information is true, and we subconsciously apply these criteria every day. Phrasing it loosely, they are as follows:
Does this information fit in with knowledge I already have? Over the course of our lives, we all build up an internal model of how we think the world works. This model consists of all the many pieces of knowledge we have already acquired and found to be correct. So if we are presented with a new piece of information that contradicts something we already believe to be true, we tend to lend greater weight to our prior knowledge - that's why it often takes so much convincing to change somebody's mind. This is a very useful criterion, because we would be crippled if we had to rebuild our entire world view from scratch every time somebody said something to the contrary. However, it does pose the danger of confirmation bias, as we tend to be both overly critical of information that does not fit into our internal model as well as overly accepting of information that does.
Do I trust the person(s) sharing this information with me? This second criterion is another useful heuristic that allows us to quickly gauge the probability of new information being correct. Basically, we look at the bearer of the information and ask two things: “Can I trust this person to tell the truth?”, and “Can I trust this person to know what he or she is talking about?” Because we are relational beings, we often place a greater emphasis on the first, which is why we sometimes trust our good friend Johnny, whom we've known for ages, more than we do some random scientist on the telly, whom we've never met. But of course this criterion is only reliable if both of these questions can be answered with yes (we need an honest expert). Johnny might never tell a fib, but he's still not to be trusted if he's getting all his knowledge from the YouTube comments section. Conversely, the scientist might have spent his entire adult life learning about whatever topic he's talking about, but if I'm not persuaded that he doesn't have some ulterior motives I'll still be sceptical.
Can I verify this information? This last criterion, finally, is the most reliable of all - but also the most work. Basically it says: “Don't trust anything you haven't seen with your own eyes.” However, that's easy if somebody claims that the sky is currently green outside, but a lot harder if you wanted to replicate the Michelson-Morley experiment. As finite beings, it is unfortunately impossible for us to fully apply this criterion to every single thing we're ever told. But there is a decent shortcut, and that is quoting sources and methods. If an author I'm reading quotes his sources, that means I am no longer dependent on his trustworthiness, but I can (at least in theory) check things out for myself. At the same time, it improves my opinion of his trustworthiness, because he was transparent enough to open himself up to scrutiny. Likewise, if he tells me by which methods he reached his conclusions, I can now evaluate whether those methods are suitable and, if in doubt, try to replicate his work.
What makes me trust science?
So that last section was pretty abstract. But it is important, because these are some of the fundamental pillars of science. Science is all about gaining reliable knowledge about the world we live in, and scientists go to great lengths to make sure they fulfill every one of the three criteria above.
Coherence. Scientific results, at least in the natural sciences, are typically published in journal articles, called “papers”. Such papers generally follow a very rigid structure of introduction, methods, results, and discussion. Importantly, the first and last sections (introduction and discussion) are there to set the work done in this paper into the wider context of current scientific knowledge. In other words, the authors have to explain how their results fit in with what other scientists have found out. That definitely doesn't mean you can't publish anything that contradicts somebody else's paper, but if that happens you do have to carefully justify why you think your results are still valid. It's also important to realise that scientists are very cautious, so that a single paper rarely convinces anybody. Instead, scientific consensus and knowledge is built up from numerous studies, all approaching a common question from different directions. If a single paper comes to a certain conclusion, scientists will say “There are indications that…”. They will only say “We know that…” once many other papers have confirmed the findings of the first. Scientists do not make claims hastily, but carefully compare new evidence against existing knowledge.
Trustworthiness. Earlier this year, I had my first research paper published. As is usual nowadays, I wasn't the only author, there were three of us. It took me almost six years of study at university to get to the point where I could contribute to an actual published piece of science, and I'm quite proud of it. Most scientists who are out there writing papers have a lot more experience than I do, so any paper you happen to pick up represents the intellectual result of many decades of combined learning experience on the part of the authors. In short: these people usually know what they are talking about! And not only that - every paper in a scientific journal has to be peer reviewed. This means that other, independent scientists (usually two or three of them) have to look over it and make sure it is good enough before it can be published. Now scientists are explicitly and vigorously trained to be critical, so a peer review is quite literally a bunch of world experts in your field trying to find every mistake you made and pointing out everything that isn't quite up to scratch. If you're fortunate, they'll give their approval for publication after two rounds of criticisms (on their part) and improvements (on your part), but it is quite common for there to be many more rounds than this. So all together the competence is certainly there, but what about the honesty? There will always be one or two scientists desperate enough to increase their reputation that they won't even stop at data fabrication. Likewise, there will always be companies paying their scientists to produce the results they want. (Although note that much, and in some fields almost all, of science is conducted by independent university researchers paid by public grants, i.e. without industry involvement.) Fortunately, the structure of peer review already poses a pretty steep hurdle to anybody who tries to get bad science published in a reputable journal. And even if the reviewers miss something (they occasionally do), other scientists in the field will take note of your results. If they're not convinced, they may then write a paper of their own, explaining why they think your results are not reliable enough. (This actually happens regularly, though usually because of a quarrel about what methods to use rather than the suspicion of outright academic dishonesty.) To sum it up: scientists spend years or decades learning about their topic, and their publications are minutely scrutinised by other experts.
Transparency. But it is this last point that, to me, is the most important. Even though I have a fundamental trust and respect for scientists in general, that doesn't mean I agree with all of them. Sometimes, some scientists are just plain wrong! Indeed, one of the hallmarks of scientific research is that the scientific community is constantly debating all sorts of things. But unlike many Internet arguments, this isn't just a case of “This is what I believe and I'm right because I say so”. Rather, both sides have to show in detail how they arrived at their conclusions. And this means citing sources and methods. Most scientific paper quote several dozen other papers, often giving multiple sources for a single statement. This allows the reader to see exactly where the authors are getting their information from, and to follow that up to see where the quoted authors, in turn, got their information from. And, of course, if you're reporting the results of a scientific study, you have to explain very precisely what you did, what materials and methods you used, what data you recorded and how you analysed those data. Ideally, somebody else should be able to use your description to do the exact same thing as you and come to the same conclusion. Although that is often not feasible (scientific studies are a lot of work!), explaining exactly what you did does allow other scientists to evaluate, and if necessary critique, your methods and consequently your results. In summary: scientists don't just make a statement, they show meticulously whom they got their information from or how they conducted the study that led to this conclusion.
What makes me sceptical?
Why am I explaining all of this? Basically, this is the standard I expect when we're talking about scientific topics, such as the epidemiology of Covid-19 or the efficacy of vaccines. It is the standard I expect because as a society, we know that this standard works. Modern science is wildly successful at explaining the natural world we live in, and one of the reason we know that is because pretty much all of our day-to-day technology is based on scientific research that was conducted in just the way I describe above.
TODO
Tagged as science, philosophy, essay, covid, biology
