Cognitive bias and the backwards bicycle test

When it comes to patterns of behaviour in a heavily patterned society, not all feedback is created equally.

By Sherwin Arnott Media
Gears on the handlebars of the backwards bicycle

I’ve watched the video below more than ten times. I love it because it demonstrates so beautifully and thoroughly how hard it is to learn, and unlearn, some kinds of things. As Destin says, knowing and understanding are very different things.

The handle bars are rigged to turn in reverse. If you turn the handles to the right, the front wheel goes left. And it appears rideable. Many folks believe they can ride it. Destin did, when he first saw it. After all, the steering is just backward. It shouldn’t be a big deal. Simple as riding a backwards bike.

“I’m telling you right now. You cannot ride this bike.” – Destin

Destin has seen hundreds of folks try and fail. It took Destin, five minutes a day, everyday, for eight months to learn to ride the bike. Holy fuck, right? It looks easier than that.

But this is the way it goes with behavioural and cognitive patterning. They run deep and sometimes they are very very hard to rewrite.

Destin concludes that knowledge is not understanding. But we could as easily say that understanding is not behaving. Or, thinking is not doing.

This is a very cool way to understand the difficulty of challenging the habits and forces and fashions that shape our thoughts and behaviours.

There are many lessons to be learned from the backwards bicycle. One is that it’s hard to accurately estimate how hard it is to do undo certain kinds of cognitive bias. Another is that learning, and unlearning, is possible.

The lesson I want to focus on here is the significance of a good test.

See, in the case of the backwards bicycle, it comes prepackaged with its own universal test. You can’t ride the bike without riding the bike. Success and failure are obvious to all onlookers. No one can pretend to have learnt how to ride the backwards bicycle, because it’s obvious to everyone whether you can or not.

Not all deep unlearning is like this. Unfortunately.

Take for example, learning to be more anti-racist. Here in North America, we grow up, for the most part, in a dominant culture that is racist. The current of our society, if you will, is white supremacist. Oh god, I’m mixing my metaphors. So now you’re riding your bike in a river, and the current is racist and all the fish have bikes. Blrg.

My point, is that the bike many of us in Canada learn to ride is the bicycle of racism. It’s the bicycle of white supremacy.

And then one day we’re tasked with being, you know, less racist.1

And just like the actual backwards bicycle, we all look at this new set of thought and behaviour patterns and we think, yeah, okay, no problem. Well, some will first question the need to learn to ride it, because hey, “we’ve been riding it for years,” am I right? But eventually everyone will realize that maybe we haven’t, so yeah, let’s do this. Anti-racism, yay.

So you get on the backwards bicycle of white supremacy and you can’t move. And when you do, you keep tipping over. So you just sort of scoot around at impractically slow speeds, and are utterly unable to use the pedals.

But here’s the kicker. Many of your friends and supporters think you’re doing great! Everyone in your life says, hey nice work, “you’re riding the backwards bicycle of white supremacy, just like we are.”

But, ha ha ha ha ha, oh god, you’re not actually riding the bike.

And occasionally someone will try to tell you. But you won’t believe them. Because, you know, they’re haters. They don’t know what’s what. You might even accuse them of riding the wrong bicycle. Because phpht.

And, to be fair you’re doing not bad. Just committing to learn these new skills makes you more anti-racist than many.2

But learning happens best when it’s obvious to learners what works and what doesn’t.

And that’s the difference between the actual bicycle and the metaphorical one that I find so fascinating.3

The real backwards bicycle makes it dead fucking obvious to everyone that someone can’t ride it. But the successful operation of the backwards bicycle of white supremacy is only obvious to some. So when it comes to anti-racist politics, many of us privileged white folks are out there saying shit and doing shit, and there’s no universally obvious test to see whether what we’re saying or doing is actually anti-racist.

Which is why, in the absence of a universally obvious test, one heuristic is to prioritize the feedback we get from folks in the oppressed group.4 Another heuristic is to prioritize the feedback of experts.5

Destin is wiping out on his bicycle, while unlearning

  1. The Truth and Reconciliation commission had some recommendations specifically for media. ↩
  2. I’m thinking here of the writers and readers at the National Post who seem to mostly think that the Truth and Reconciliation Commission was a giant hoax and a waste of money. ↩
  3. At this point some readers may be having a problem with my analogy. After all, riding a bike is a specific neural task. As Bharat has suggested, cycling is hard coded into a neural path and eventually “delegated partially to muscle memory.” Some readers may wonder if other behaviours are like this. Do racist behaviours, for example, involve muscle memory or other kinds of deep seated cognitive patterning? The answer is yes. It does often. More than just pronunciation, much of linguistic production, for example, relies deeply on muscle memory and other kinds of procedural memory. Even word choice relies on procedural memory. Think of lyrics, passwords, slurs, and jargon, for example. But my main point is not that unlearning racism requires one to unlearn procedural memory. My point is not even that all racist behaviours are like learning to ride a backwards bicycle. My first assumption is that unlearning racism is deep unlearning, likely much more difficult than bicycles. My second point is that the backwards bicycle has a universally obvious test for success that is regrettably absent from other kinds of unlearning. ↩
  4. I’m reminded here of a post that Dan Gardner wrote way back in 2011, where he claimed, among other things, that the Muslim veil is anti-social and he actually rejected the rebuttals offered by Muslim women because they were “too close to the topic.” ↩
  5. I’m thinking here of the time that Andrew Coyne rejected out of hand the report on Canadian jails that found they were systemically racist in their approach and treatment of Indigenous people. ↩