I know that we’ve discussed this topic a lot this week, but I had some additional thoughts on scientific consensus. Perhaps this time I had a moment of lucidity thinking about my previous post about a point that I don’t think I got across very well. Thanks especially to Jonathan’s addition on the nature of probabilistic belief. He was upset that I followed up on my first comments section this week before he got to it. Ha!
It occurred to me that those who contest the validity of the scientific consensus often aren’t really promoting arriving at beliefs through scientific methods.
Again, science isn’t based off a consensus in the sense that a bunch of experts go to a conference and agree on what the truth is. This appears to be a strawman argument against those of us who encourage everyone to take into account the scientific consensus. The truth is obviously not subject to a vote, and if anyone is saying that then they are wrong.
I believe part of the major issue is that people conflate “science” with the aggregate of scientific knowledge, which is not the same thing. While in vernacular language we often use the term “science” as shorthand for the findings and data we’ve gotten as a result of scientific processes or as the community of scientists as a whole, I am going to refer to science in this post purely as the process. Science is the process of making educated guesses, simulations, replications, model building, testing, peer-reviewing results, disseminating findings, and then repeating these steps over and over. As a result of controlling experimental conditions, accounting for bias, and rigorous scrutiny, this process is useful for cutting out our preconceived notions and delivering data that gives us information on natural phenomena. When multiple studies with varying methods that look at similar problems in a variety of ways are published, the findings tend to point to similar conclusions. When all the evidence tends to point in a certain direction, this is what is meant by consensus.
When people think they shouldn’t pay attention to this consensus, they often think they are being scientific, because they are not taking into account “common knowledge” and perhaps bucking the system by creating alternate hypotheses or theories, which would be useful for any researcher in the field. After all, the biggest findings as a result of science were radical paradigm shifts that drastically overturned common knowledge at the time. What they fail to grasp is unless they are participating in the rigorous scientific process outlined above they are not participating in science in any meaningful way. Are you doing experiments, creating simulations, developing models, or aggregating studies for meta analyses or review articles? Are you getting papers through peer review? If you’re not doing this, then you aren’t really “bucking the consensus”, you are just holding an alternate, untested opinion.Let’s put it another way: the end goal of science is to eliminate as many biases as possible. It’s never going to be perfect, but over time the process is likely to self-correct and multiple orthogonal studies will able to account for flaws between methods. After the results have gone through the grueling process of peer review and the findings have been disseminated, all the most rigorous corrections and accounting for bias and human error have been done already. Once you start developing alternative opinions outside of these findings, you are reintroducing your own bias and human error back into your opinions based on the research. This is actually fairly unscientific, and taking one’s own opinion above and beyond the process designed specifically to eliminate the problems inherent with human opinion.
It’s not beyond the realm of possibility that you are actually right. No study is perfect, many are pretty terrible, and it’s possible you may be able to see a flaw in many experiments. It’s not very likely, though, that you will come across any sort of meaningful scientific truth until you actually engage in the process, you should be very cautious about believing that your stance in any way supersedes the mass aggregate of findings. At the end of the day we have no choice but to think for ourselves, but in constructing a consistent and accurate worldview we should always put more weight on evidence that seeks to overcome our human biases, and less weight on our hunches and inexperienced notions of where the science is getting it wrong.
Perhaps those who have been pushing back on me this think they can ascertain the truth by determining for themselves whether or not an individual study is good or not, but this seems like a Dunning-Kruger style mistake. When you get deeply into a research career, you start reading papers and reports with some really obvious flaws. You learn to recognize that no one study is usually sufficient to make a case for something on its own. There are famous landmark papers, of course, but those are exceptionally rare and almost always have limitations of their own. Even with good papers, it’s best to wait until multiple sources start pointing in a general direction before you put credence in any given scientific finding. It’s not elitist to say that I wouldn’t trust a layperson to distinguish between a good study and a bad study, as I struggle with that sometimes myself!
If you think you have a good “alternate” scientific opinion, the way to make sure that it holds water is to put it to the test, as many, many scientists have done before you. Until then, it’s just an idea, and it’s not scientific enough to just think for yourself and justify it by appealing to past paradigm shifts.