Bet on Existence
(short)
There’s a concept I hadn’t quite crystallized until a recent (friendly, if frustrated) disagreement:
It’s basically always correct, when there’s a disagreement about how diverse or varied a given population is, to bet on the “more diverse” claim.
Like, if Person A is saying that all dogs are [one way] and Person B is pushing back with “what? No, no, dogs have a wide spread on that axis,” then you should pretty much always bet on Person B being right.
In part, this is because I think there are many, many more ways for Person A to have missed the existing diversity—
(Social dark matter/conformity pressures/status games making people want to appear more normal or aspire to a particular standard, double illusions of transparency, typical mind fallacy, confirmation bias, reality distortion fields, being in a bubble)
—than there are ways for Person B to have hallucinated the diversity.
We keep seeing this, over and over again, in practice—people in the same town live in meaningfully different worlds, not all students learn the same way, not all mothers share the same experience of motherhood, not all gay men want the same things out of sex, turns out people interpret the same words and phrases as meaning wildly different things, turns out people have all sorts of non-overlapping reasons for going to the gym, turns out autism is a spectrum like a color palette rather than a spectrum like a line stretching from zero to one hundred, etc. etc. etc.
We find out that something we thought was homogenous and monolithic isn’t far more often than we find out the reverse; I actually can’t think of any examples where it turned out that a category was less diverse than originally believed oh right Brontosaurus and Apatosaurus turned out to be the same species I guess.
This concept hooked up with another one in my brain recently, when my husband pointed out [something that I interpreted as meaning] that claims of value being present in a given piece of art vastly outweigh claims of that art being valueless.
(Seems pretty straightforward on reflection, no? If Person A is claiming “this song/painting/poem/sculpture/movie/performance piece has no value” and Person B is claiming “uh, well, I found value in it, actually,” pretty much the only way for Person A to win is if Person B is just straight-up lying.)
The two concepts did not use protection, and their offspring now lives in my brain as one of those occasionally-wrong-but-usually-right rules, like Occam’s Razor or Chesterton’s Fence:
Claims of existence beat claims of non-existence.
Now, you gotta be careful what you allow under the heading “claims of existence,” because in fact some people are schizophrenic and some people are liars and a lot of people are wrong in the conclusions they draw about what exactly they observed and also by the way there’s a vast conspiracy to indoctrinate literally billions of children into believing in a monotheistic god that we have zero actual reason to believe in.
But. While I don’t buy that e.g. the people who believe in ghosts saw actual ghosts, I do default to believing that most of them did in fact see something.
(There are enough otherwise-reasonable people with no particular incentive to lie about it making the claim even in the face of mockery or ridicule that, like … come on?)
Logicians will find this whole concept unimpressive because it’s a very well-worn tenet of logic that you only need one counterexample to disprove an absolute; if someone is claiming that X definitely never happens then you only have to find X once to conclusively win the argument.
But while many people understand this principle, I find that in practice people forget to apply it. They frequently react to anything outside of their direct, lived experience with bewilderment and denial, insisting that it’s just not possible for, I dunno, a stable polyamorous triad to exist, or for someone to genuinely dislike the taste of chocolate, or for a person to have two X chromosomes but be phenotypically male.
These days, when I encounter a direct, personal testimonial that amounts to “it’s possible for someone to have this experience or preference that you’ve never thought of” or “I’ve witnessed this interpersonal dynamic that you seem to believe isn’t possible” or “I’ve observed this weird phenomenon out in nature,” I basically just immediately update to “okay, sure, I guess the world includes that thing that I previously thought it did not include.”
I keep the update tentative, to account for people being confused or deceptive—
(Vaccines that cause autism really don’t seem to exist, we looked for them pretty hard)
—but what I don’t do is assume that, just because I’ve failed to notice this particular thing before—
(In my mere thirty-nine years of life, during which I’ve managed to meet like 0.0001% of all people and have seen approximately 0.0001% of the surface of just this one single planet, at a generous overestimate in both categories)
—that it therefore does not exist, and the person telling me otherwise is wrong.
Postscript: in addition to my usual begging-for-subscriptions, I’ll add that this past year has been unusually rough, financially, and if you’re one of those people who’s like, man, what am I even gonna do with all of this money, I don’t know how to efficiently convert it into happiness and More Cool Things Happening In The World, you can also make one-time donations (or buy some hours of my time and expertise) and it will in fact make a difference and I will in fact be genuinely grateful.



Latest (2015) word from paleontologists is that Brontosaurus and Apatosaurus are different species (and different genera!) after all. There's a good summary here: https://svpow.com/2015/04/07/welcome-back-brontosaurus-and-other-first-thoughts-on-tschopp-et-al-2015/
It's definitely worth bearing in mind how fraught non-existence claims are (because, as you point out, only one counterexample is needed to disprove the claim; any universal claim about current humans has 8.2 billion chances to be wrong), but I don't think the "always bet on more diversity" heuristic holds up very well if one is thinking quantitatively about distributions, rather than about existence vs. non-existence (and thinking about distributions is much more useful).
This is easier to see concerning traits where we have hard data. The average height of German women is 5'5" with a standard deviation of 2.6 inches. There are a lot of people in Germany and the distribution is fatter-tailed than an idealized normal distribution, so it's definitely not trivial to pick a particular height and confidently claim that literally no woman is shorter or taller than that. (Dwarfism or gigantism come from specific disorders rather than being extreme samples from the usual sources of variation. The world's tallest woman at 7'0" is 7.3 standard deviations out, and almost certainly wouldn't exist if the distribution were truly normal.)
But "someone vs. literally no one" isn't a particularly decision-relevant piece of information for most applications? You write about betting about "how diverse or varied a given population is", but such claims are usually construed as being about the distribution, not the most extreme outlier fortunate enough to actually exist. It does make sense to be quantitatively more skeptical of any particular account of a 6'0.6" (+3σ) or 6'3.4" (+4σ) or 6'6" (+5σ) woman in proportion to how likely people are to lie or exaggerate about such things, and I think that's what the "Person A" side in these disputes is usually trying to get at (even if most people can't articulate it as carefully as I'm doing here).