The Role of Plausibility in Science

[aum]

I have been writing blog posts and engaging in science communication long enough that I have a pretty good sense how much engagement I am going to get from a particular topic. Some topics are simply more divisive than others (although there is an unpredictable element from social media networks). I wish I could say that the more scientifically interesting topics garnered more attention and comments, but that is not the case. The overall pattern is that topics which have an ideological angle or affect people’s world-view inspire more passionate criticism or defense.  Timed drug release is an important topic, with implications for potentially anyone who has to take medication at some point in their lives. But it doesn’t challenge anyone’s world view. ESP, on the other hand, is a fringe topic likely to directly affect no one, but apparently is 70 times more interesting to my readers (using comments as a measure).

 

 

I also get e-mails, and my recent article on ESP research attracted a number of angry individuals who wanted to excoriate my closed minded “scientism”.  I think people care so much about ESP and other psi and paranormal phenomena because it gets at the heart of their beliefs about reality – do we live in a purely naturalistic and mechanistic world, or do we live in a world where the supernatural exists? Further, in my experience while many people are happy to praise the virtue of faith (believing without knowing) in reality they desperately want there to be objective evidence for their beliefs. Meanwhile, I think it’s fair to say that a dedicated naturalist would find it “disturbing” (if I can paraphrase Darth Vader) if there really were convincing evidence that contradicts naturalism. Both sides have an out, as it were. Believers in a supernatural universe can always say that the supernatural by definition is not provable by science. One can only have faith. This is a rationalization that has the virtue of being true, if properly formulated and utilized. Naturalists can also say that if you have actual scientific evidence of an alleged paranormal phenomenon, then by definition it’s not paranormal. It just reflects a deeper reality and points in the direction of new science. Yeah!

 

Regardless of what you believe deep down about the ultimate nature of reality (and honestly, I couldn’t care less, as long as you don’t think you have the right to impose that view on others), the science is the science. Science follows methodological naturalism, and is agnostic toward the supernatural question. It operates within a framework of naturalism, but recognizes this is a construct, and does not require philosophical naturalism. So you can have your faith, just don’t mess with science.

 

But of course many people do “mess” with science, meaning they try to tweak or bend the rules to suit their agenda. One of the most common ways I see this happen, and it explicitly happens with ESP defenders, is the notion that any consideration of plausibility in scientific investigation representds closed-mindedness, scientism, or simply bias. But this is definitely not true, and completely misunderstands how science works. We see this attitude at work within the paradigm of so-called alternative medicine, which has embraced “evidence-based medicine”, or at least their conception of it, which explicitly eliminates prior plausibility. They do this because they don’t fare well scientifically when plausibility and prior probability are concerned.

 

Likewise, ESP and psi believers become apoplectic when you mention scientific plausibility. In my experience, they have to misunderstand what it is and how it is used, no matter how many times or how many ways I explain it to them, in order to maintain their position. For example, I recently had an exchange where the e-mailer responded:

 

Dr. Novella, your argumentation here is more reasonable, even if it still displays scientism. I’ll point to your statement, “This is especially true since they are proposing phenomena which are not consistent with the known laws of physics.” Here you’re making the known laws a determining factor for rejecting the prospect of the phenomena.  The very point of the investigations is to see if something exists beyond the known laws. Again, the evidence is the authority. The known laws don’t disallow what isn’t known. You seem to think you’re making a very sound argument, all the while exhibiting pure scientism.

 

Of course, I said nothing of the kind, and no matter how many times I corrected this fallacy, he returned to it. The idea is that since science is trying to discover new things, that already established findings or laws don’t matter and can be comfortably ignored. “The evidence is the authority” – just like with alternative medicine. He pairs this with the straw man that plausibility is “a determining factor”.

Here is how science actually works. First of all, science builds upon itself. We have to take into consideration prior knowledge, because that affects how we interpret new information. If, to give an extreme example to illustrate the point, I make an observation that seems to contradict 1000 prior observations, do we chuck out the prior observations? What is the probability that the 1 observation is wrong vs the 1000 observations?

The reality is that science does not deal with absolutes or perfect knowledge. It deals with predictive models and statistics. This is a simplified schematic, but generally we make observations about reality, we then make hypotheses to explain those observations, and we need to be able to test those hypotheses in some way (more observations or controlled experiments). When a hypothesis survives enough repeated attempts at proving it wrong, it may graduate to becoming a theory, an explanatory model. Theories are all ultimately explaining the same universe, so they also have to work together. There are entire disciplines of study that weave together multiple working hypotheses, established theories, well-established facts, and phenomena so consistently observed that we consider them laws of nature. Everything is a working model and is up for revisions, but we are building a picture of reality from the ground up. Finally, different disciplines need to work together and should not contradict each other.

One important way scientists make new discoveries is that they make observations that either cannot be accounted for with existing theories or directly contradict those theories. Scientists love this, because such anomalies point in the direction of new science. We need to modify an existing theory, or come up with a better (deeper, more complete) theory that accounts for not only all prior observations but these new ones as well. Where do I sign the grant application. To scientists, anomalies are good. But an “anomaly” implies pre-existing scientific knowledge. We are not starting from scratch with each new observation.

Here we get to a critical difference between skeptical scientists and true believers – the role of plausibility and prior probability in evaluating new evidence. Let’s say someone has a dent in their car’s fender with a bit of blood and brown fur. They say they hit a deer the night before. Is the evidence sufficient to conclude that the claim is probably true? What if they said they hit a Bigfoot? Now is the evidence enough? Would you require a higher threshold of evidence before concluding the claim is likely true? I hope so. In science when do we conclude that we have a true anomaly on our hands. A few years ago a research team said they had found faster than light neutrinos. This was the evidence. Physicists were both excited and skeptical of the claims – because that would be a significant anomaly. If true, that would be an amazing scientific discovery. But their threshold of evidence was super high. They required every possible error to be examined, and examined again, and for the results to be independently replicated. It turns out, it was an error in a cable. The anomaly vanished when the error was corrected. Physicists were right to be skeptical based upon plausibility – the known laws of physics were a reliable guide to whether or not the claims were likely to be true. That’s why they are known as laws – because they are really good at predicting reality.

So now we have people who claim that they have evidence for information going from the future to the past, or for information being transmitted from one brain to another without any detectable signal, and any known mechanism. The fact that these observations appear to contradict the known laws of physics is not “determinitive”. But it is also not just a prior bias. It affects how rock solid the evidence has to be before we conclude we have a genuine anomaly on our hands. In my opinion, this evidence (for ESP) is orders of magnitude too weak to conclude we have a genuine anomaly. The effect sizes are small, the researchers don’t have a great history for rigor, and the protocols cannot be reliably replicated. So we have relatively weak evidence being put up against rock solid laws of physics. It is not “scientism” to say that the evidence is not sufficient. And it is not scientific or logical to dismiss this dramatic lack of plausibility.

Mechanism and plausibility matter, because this is part of how we build our scientific picture of reality. That is how we close the loop – we need to be able to explain how things work, not just observe what happens. It all needs to fit together. This is partly because all human observation is flawed, we are really good at deceiving ourselves, and we may be looking at one phenomenon when we think we are looking at another. We naturally look for evidence to prove our beliefs. But science requires we look for evidence to disprove our theories, to falsify them. ESP is flawed research until proven otherwise, and proponents have not proven otherwise, not by a long shot. And yes, the threshold of evidence is higher when prior plausibility is lower – that is how science logically and statistically has to work. Otherwise we are building our house on top of a bridge made of gossamer string.

 

Source