Before discussing the subject that's the title of this post - how the brain makes predictions come true -- I'll shoehorn in a related personal story about my check ordering saga.
Recently Columbia Bank, which my wife and I use for a checking account, was bought by Umpqua Bank. For many years I've ordered checks for our Columbia Bank account when the supply ran low. Last month was the first time I'd ordered checks with Umpqua Bank on them.
Balancing our checkbook about a week ago, I noticed that we'd been charged on May 17 for the cost of mailing the Umpqua Bank checks I'd ordered. That got me to wondering why the checks hadn't arrived.
After that bit of wondering, every time I opened our mailbox I'd look to see if the checks had been delivered. But day after day, no checks. When May became June, I started to worry that the checks had been lost.
Yesterday I decided to contact the web site that produces and mails checks. Looking for a way to do this, I noticed a FAQ (frequently asked questions) page. One question was about what to do if a check order hasn't arrived.
The advice was to log in to your account, then click on an "order status" link. I did that. And was surprised to find that the checks had been delivered. Which I immediately remembered had happened.
So how is it that I kept looking in our mail for checks that were sitting in a drawer in our house? Here's my theory. As noted in my first two posts about Andy Clark's book, The Experience Machine, modern neuroscience says that the human brain is constantly making predictions about the world.
Thus rather than perception being an objective mirror of the outside world, the brain is using past experience to predict what is most likely to appear in a given situation. Discrepancies between that expectation and what is actually perceived are errors that, ideally, lead us to fashion a more accurate understanding of what's there.
In my case, I didn't have any experience ordering checks with Umpqua Bank on them. So even though all of my Columbia Bank orders had arrived without a problem, my brain wasn't as confident about the May check order.
When I saw the charge for the checks on our bank statement, a sense of "something is wrong" popped into my head. I didn't think that thought. It just appeared in my mind with a convincing sense of rightness.
From that point on, "something is wrong" formed the prediction in my brain when I opened our mailbox. Every day no checks were there, that strengthened the prediction. Of course, the problem was that there actually wasn't a problem.
But I needed an external push to realize that -- the FAQ about not getting a check order. For my brain was in a closed loop founded on a basic error: that the checks hadn't arrived. Sure, most people reading this are probably thinking, "You're a fool for not looking in the drawer."
No argument there. However, I never thought to look in the drawer where we keep blank checks because my brain was saying that the checks hadn't arrived with a sense of certainty that resulted in me looking for the checks every day we got mail.
Bottom line: the brain works in mysterious ways.
Regarding how the brain makes predictions come true, here's how Andy Clark describes the basic process. It shows how perception and action are much the same, since each uses prediction and error correction to do their job.
To see how this works consider a simple action such as turning my head to see the seagulls out of my office window.
...The sound of the gulls, and the fact that I'm now looking for a nice example of prediction-based action control, makes me want to look out the window and see the gulls. I do so.
In predictive processing terms, what happened is this. The sound of the gulls, and my need for a familiar example, made me strongly predict looking toward the gulls. The best way to get rid of the resulting prediction errors (which were many, since I was still actually looking at my busy computer screen) was to turn my head just so and move my eyes just the right amount.
...The deep unity (under predictive processing) of perception and action should now be apparent. There are two different, but equally effective, ways to minimize prediction errors during our encounters with the world.
The first is by using prediction errors to help us discover the best guess about how things are out there in the world. But the second is to act so as to make the world fit some of our predictions. Instead of finding the prediction that best fits the sensory evidence (perception), you now find or create the sensory evidence that best fits the prediction.
ln the case of my supposedly missing checks, my brain initially had made a prediction that the checks hadn't been delivered after I noticed that our account had been charged for them. Yes, that prediction was wrong, since the checks had been delivered and were sitting in a drawer.
But it felt right.
And every day I opened our mailbox and saw no checks, that perception made the prediction seem even more right. It was action that corrected the errors being made. Once I realized that the checks had been delivered, and that I'd forgotten putting them in the drawer, I walked to the drawer, opened it, and picked up the packet of Umpqua Bank checks.
Now the new prediction -- the checks were in my possession -- had been confirmed with no errors.
As I get deeper into The Experience Machine, I'm learning some ways we can improve our relationship with the world and ourselves. That will be the focus of my next post about the book, which I'm enjoying a lot.