The notion of consensus is something that comes up a lot, the more time you spend investigating and debunking pseudoscience, alternative medicine, conspiracy theory and claims of the paranormal. As skeptics and scientists, we are often fond of pointing out the consensus position is our best attempt at getting to what is most likely to be true. Meanwhile, among conspiracy theorists and anti-vaxxers, the term is often derided as nothing but an example of mainstream “sheeple” groupthink.
“Consensus” is an important concept, but it’s also one that isn’t as straightforward as we skeptics might make it out to be, at least to the everyday person who doesn’t spend a lot of their time thinking about how we know what we know, and how we know what to trust.
What do we mean when we talk about consensus? Generally speaking, we mean “what most of the people who know what they’re talking about would say is most likely to be true”, but that’s a hard thing for a layperson to judge, especially if they have little to no expertise in that particular subject. How does the everyday person judge which people know what they’re talking about? What should be the right threshold for “most of the people”, and how do we estimate it? How does that judgement of consensus differ from an argument from popularity, or an argument from authority, or both? It is distinctly different, but the difference perhaps isn’t wholly easy to articulate in a clear, black and white, low-ambiguity way. It starts to get tricky, and it’s easy to see why people can be swayed by rhetoric that picks at the fuzzy edges of what consensus means, because consensus can be illusory.
In a case I wrote about last year, a man named John Lawler visited a chiropractor and received a spinal manipulation treatment which fractured his spine and ultimately led to his death. The General Chiropractic Council, the regulator of chiropractors, held a hearing to determine whether the chiropractor in question, Arlene Scholten, had acted unprofessionally. One important question during the hearing was whether it could have been foreseen that the treatment she gave Mr Lawler would be dangerous for him, given the weakened condition of his spine. The committee found that Scholten had not acted unprofessionally, because expert testimony at the hearing explained that the consensus among chiropractors was that such a spinal fracture was unforeseeable. The consensus among chiropractors was that chiropractic is safe – and, reasoned the regulators, and who could be better placed to offer an expert opinion on what a chiropractor ought to have foreseen than another chiropractor? Arguably, the hearing simply followed the consensus position of the experts – but, crucially, were they the right experts?
In a similar vein, there have famously been long lists of engineers putting their names to arguments that the fall of the twin towers on September 11th must have been a controlled demolition. Seeing such a list, comprised of people with some degree of expertise, hits the parts of our brains that have been trained to take consensus as indicative of correctness – it takes a bit more examination to ask important questions, like “how many of those engineers were structural engineers with experience of controlled demolition, rather than, say, software engineers?” or “how many engineers are there who agree the official narrative is more likely to be true?” or “of the group with the most expertise in building structure and demolition, what do they make of these claims?” It can be difficult, without having trained oneself to perform these kinds of sense-checks, to recognise what level or flavour of expertise should be trusted, and which experts count as consensus.
Who counts as consensus on which subjects can get surprisingly granular, to a level that isn’t necessarily immediately obvious or accessible, especially to a layperson. Take, for example, the medical consensus in 2004, which was that peptic ulcers were a chronic condition, often brought about by things like stress. Even experts in gut health at the time, who had spent their lives and academic careers building up credible expertise in this very specific subject, would have signed on to that consensus. Yet, in 2005, Barry Marshall and Robin Warren proved that ulcers were actually caused by H. pylori bacteria, winning a Nobel Prize and proving the other experts wrong. Their discovery means that, in 2004, the consensus was clearly incorrect, and instead two chaps in Australia were right.
Equally, heliocentricity as a model for the universe had been proposed in the 5th century BCE, but it was almost 2000 years later that it was accepted by the consensus. Anyone who proposed a heliocentric model in the early 1600s would have been considered by the expert consensus to be barmy, until Copernicus, Kepler and Galilei eventually challenged and changed that consensus.
This may all seem like a quite compelling argument against trusting the consensus. In reality, what I’ve actually been engaging in here is a sleight of hand about what we mean when we talk about consensus: the heuristic isn’t that the consensus is always correct and cannot be wrong; it’s that, by and large, all things being equal, we’ll be right far more often than we’re wrong if we follow what the consensus of experts conclude, knowing that any conclusion is provisional, pending further evidence. Effectively, all swans are white right up until you see a black swan, but seeing one black swan doesn’t then prove that all swans are black, or that people were stupid to have ever believed in white swans.
Importantly, appreciating all of this takes close examination, reflection and pause for thought, and as human beings going about our lives in society, we’re not built for that kind of constant sense-checking; we rely on heuristics, mental shortcuts, and things that allow us to take actions without having to weigh up every single thought and possibility. The alternative would be exhausting, and unworkable.
This isn’t therefore, an argument against relying on the consensus for helping us decide what is likely to be true, and what isn’t. But we do need to be careful about how we talk about the scientific consensus and what we mean by it – that we don’t mean it as argument from popularity meets argument from authority, or that we think scientists are infallible, or that we think it’s impossible for the consensus to be wrong. Instead, we should remember to make clear that what we mean by consensus is simply that it is the current conclusion of most experts in their field, and that following that provisional conclusion gives us a better shot at finding truth than listening to outlier voices, inexpert voices, or just whichever voices shout the loudest.
