On Evidence

Over the years years, I’ve seen countless assertions that take this form:

“There is no evidence to support the claim that …”

It is so common to encounter this assertion that I almost find it cliché. And misleading. In many ways.

Here’s why.

The absence of evidence

Much of this essay is going to be a bit controversial, so let’s hopefully start by something that’s relatively tame. I hope it’s uncontroversial to say that the absence of evidence is not evidence of absence, and in particular, it is not an argument to support the negation of the claim. What should happen when there is no evidence is that everybody “falls back” to their priors (as in Bayesian probability).

Note that priors are mostly subjective. However, in actual discourse, very often the one who asserts “There is no evidence…” expects the other side to adopt their priors. In such cases, it is demonstrably a mistake. Everyone has uses their own priors, you can’t force priors on other people.

One would think that learned people don’t make such mistakes, but it happens quite often in similar but slightly different forms. After asserting the lack of evidence to support a claim, they expect people to fall back on priors within bounds of the mainstream narrative. The problem is that historically speaking, the mainstream consensus can shift dramatically, so sticking to it is not necessarily a fallback to a more “probable truth”, but rather, in fact, a fallback to political correctness. Occasionally I see people using the term “burden of proof” in such discourse, satisfied that the “establishment” has their back and they really only have to discredit the other side to make a point (without presenting evidence of their own). What is really happening here is not an inquiry into truth, but rather a power play dynamic: “I have the power to assert that there is no evidence, and if you can’t rebut the assertion to our satisfaction, you have no choice but to accept our (the establishment’s) priors.”

It is of course natural that mainstream views are backed by more resources and more people, and thus has an inherent advantage. But is exploiting this advantage in discourse really a good way to find truth?

Extraordinary evidence

It is often said that “extraordinary claims require extraordinary evidence”. In theory, this statement is fine and should be observed. The devil is of course in the details.

Let us consider what extraordinary evidence actually is with an example. Let’s say last night you and a couple friends saw an alien spaceship land in your backyard. The next morning it was gone. You might think this obviously counts as extraordinary evidence that aliens exist. But if you try to tell anyone about this “fact”, they’d just tell you that you’re making one hell of an extraordinary claim (with nothing to back it up).

In fact if you think deeply what “extraordinary evidence” is, you might come to the conclusion that it isn’t really a real thing. Truly extraordinary “evidence”, no matter how air tight you think you had it, could be re-interpretted as a “claim” because they are not readily believed. So you’d get into an endless recursive loop of trying to prove your evidence using more evidence. Of course, extraordinary claims have been accepted by the mainstream in the past, but they never merely rested on “extraordinary evidence” per se – rather, they often involved bitter political struggles, sometimes deaths[1], and a lot of waiting time for this “extraordinary” status to wear off, so that when the time is ready, “ordinary” evidence sufficed to convince the majority that the claim is true.

That said, extraordinary evidence does exist, in a somewhat surprising way. I quote Mark Xu’s observation that “Strong Evidence is Common“[2]:

One time, someone asked me what my name was. I said, “Mark Xu.” Afterward, they probably believed my name was “Mark Xu.” I’m guessing they would have happily accepted a bet at 20:1 odds that my driver’s license would say “Mark Xu” on it.

The prior odds that someone’s name is “Mark Xu” are generously 1:1,000,000. Posterior odds of 20:1 implies that the odds ratio of me saying “Mark Xu” is 20,000,000:1, or roughly 24 bits of evidence. That’s a lot of evidence.

While I think his numbers are probably a bit off, his main thesis is pretty insightful. I dispute the numbers, in particular the 20:1 odds bet, because while nobody doubts that a person knows their own name and what is on their driver’s license, the odds of a random stranger lying about their name is probably higher than 20:1.

And thus we arrive at the crux of what “extraordinary evidence” means – do we believe in the person making the claim? For some reason, we believe such extraordinary claims as “My name is Mark Xu”, but not “I saw an alien spaceship in my backyard.” I think there are various ways to justify this discriminatory approach, but I suspect it all boils down to whether the claim fit inside the mainstream scientific narrative. While on the surface the claim that “My name is Mark Xu” is highly improbable, it is easily believed because it does not challenge our scientific understanding. Whereas the alien claim seems to require a re-evaluation of our understanding of physics at the very least.

And thus if we rephrase this approach in a more cynical language, we can say that the social force that people call “science” readily assumes people are either incompetent or lying for challenging the consensus narrative. Perhaps they are right to do so, since if the mainstream scientific understanding is mostly correct, then presumably most challengers are in fact incompetent or lying. But in that case, is it really that hard to understand why some people are bitter and hateful towards that attitude? Especially when what they honestly believe as “extraordinary evidence” is said to be “extraordinary claims”? And more generally, do we really have to assume people making extraordinary claims are incompetent or lying in order to maintain a rational society? I think this is something people should think about.

Extraordinary subjective evidence

As we have seen, it takes a lot to change the minds of many people. It is almost impossible to challenge the mainstream consensus on any given topic with just a couple pieces of evidence (or claims).

But changing one person’s mind is actually not as hard. It is still hard, but not quite impossible. Because nobody presumes themselves lying or incompetent. So, for example, a person who does not believe in aliens would be very shaken to see a spaceship landing on their backyard, and it would be difficult for them to completely deny their own subjective experience. They might not fully believe that aliens exist even after the experience, but most likely they would be much less certain about their convictions. In particular, even if they thought what they saw was not real, they’d have much more sympathy with other people reporting UFO sightings for example.

In short, a single event could change a person’s entire worldview and put them at odds with mainstream science.

In the wild parts of the Internet that have no moderation, it is common to see acolytes of the mainstream narrative “bully” people for so-called fringe beliefs. It is curious that, to them, those unfounded beliefs are caused by ignorance and lack of knowledge, while it seems that here, I think, we have made the case that subjective experiences could lead one to conclusions that are different from the mainstream.

On a related note, it’s interesting to watch people who strictly believed in the scientific worldview be hit by some mystical experience. They can get utterly shattered. (A famous example is Eben Alexander who wrote the book “Proof of Heaven” after his near-death experiences. It became a New York Times bestseller.)

I’ve personally gone through a similar process, though at a much milder level. Without going into details, since I was a child I’ve witnessed things that had no scientific explanations, and I’ve had already dabbled a bit in spiritual practices myself. But I found that witnessing something and experiencing it first hand can be totally different – Previously I was able to shunt the mystical phenomena into the back of my mind and pretend I forgot about it – but now that it metaphorically hit me right in the face, I had to deal with it, and I was still kind of shocked and had to spend a bit of time reconstructing my worldview. (This essay is part of that exercise.)

For those who vehemently and categorically reject any phenomena that are not explained by science, take heed: you might be in for a rude awakening. You are basically one crazy experience away from having your worldview shattered (and again, I’ve seen that happen to people). Rationally speaking, it’s much easier to assert that you don’t know whether something is possible, than to assert (without evidence!) that you know something is impossible.

A side note about fragility – the logic systems that modern science is built upon are also inherently fragile in that they break down upon any minor inconsistency. Robust systems should not exhibit this behavior, a couple so called “false” or “fraudulent” claims should not bring down the entire system. Robust systems aren’t scared of being fooled by a lie or two. I don’t fully know how to build such systems (but see [3]), but it’s something to aspire to.

Ignored and Ignorant

A couple months ago, I was reading Alan Turing’s paper “Computing Machinery and Intelligence” (the one which proposed the Turing Test). It’s probably a popular read these days given the renewed debate on AI and machine intelligence (this time with more practical consequences). But what caught my eye was an “off remark” about “Extra-Sensory Perception”, from which I quote:

I assume that the reader is familiar with the idea of extra-sensory perception, and the meaning of the four items of it, viz. telepathy, clairvoyance, precognition and psycho-kinesis. These disturbing phenomena seem to deny all our usual scientific ideas. How we should like to discredit them! Unfortunately the statistical evidence, at least for telepathy, is overwhelming.

We probably won’t know what specific evidence Alan Turing was referring to, but apparently he thought there was something there to the extent that he had to address the issue of telepathy in the Turing Test. Even though Turing’s paper is famous and probably has a recent resurgence due to AI hype, and presumably a significant number of people have read/skimmed through it, and most of you probably have heard of the Turing Test, I doubt many of you realized that Turing believed in some kind of telepathy.

If Alan Turing talks about telepathy and gets ignored, while people go about claiming there’s absolutely no evidence whatsoever for the purported phenomenon, it does make me wonder how many other claims from reputable scholars get ignored. I mention Turing only because he is so hard to discredit, and if I quote anyone below his caliber, it’s obvious they will be shot down as some eccentric fellow with a side hobby in weird woo-woo stuff.

To be clear, I’m not claiming Turing is an authority on ESP matters, but if some random scientist claims there is no evidence for telepathy, and Turing says the evidence is overwhelming, it’s not very obvious to me that I am supposed to conclude that Turing is mistaken. The logical next step to resolve the dilemma would naturally be to consult a trustworthy expert in ESP – unfortunately, since polite society rejects ESP as a serious subject of study, such an expert does not exist. The circle is thus complete, and the status quo is preserved.

So, is there really “no evidence” as people claim, or are they simply blissfully unaware of it? To be honest, I never thought I’d be accusing mainstream science of weaponizing ignorance in the Information Age, but here we are.

Parable of the happy fish

Let us conclude with a parable from Zhuangzi:

Zhuangzi and Huizi walked by the Hao River when Zhuang Zhou exclaimed, "See how happy the fish are swimming!"

Huizi challenged, "You're not a fish, how do you know they're happy?"

Z: "You're not me, so how do you know I don't know?"

H: "True, I'm not you, but you're not a fish either, so you can't know."

Z: "Let's step back. You asked how I know the fish are happy, implying I already know. I know by being here with them by the river."

Now, I don’t know what the moral of the story is, and also I don’t know whether the fish in the story are truly happy. Putting aside the obvious sophism, the story does make me think whether we should be more “open-minded” to accept things for which we have no “explanation”. A person who claims to understand fish feelings makes a truly extraordinary claim[4], but should we dismiss them on principle because we can’t imagine how it is possible?

In my personal experience I’ve come to realize there are types of knowledge that are true but not explainable (see discussion on subjective experiences above). In many cases, even the person professing to know something cannot explain why or how they acquired the information. Thus, in my personal thirst and greed for knowledge, I’ve realized that “trust by default” gives me access to a vastly superior repository of knowledge compared with a cynical and skeptical attitude. When “extraordinary claims” are instead treated as “extraordinary evidence”, what’s left is to explain the evidence. The exercise of coming up with possible hypotheses is engaging and fruitful. This is the proactive approach that I think would yield better results than assuming that everyone is incompetent or lying.