Against Fluency

Hi friend,

In 2006, two psychologists at Washington University in St. Louis, Henry Roediger and Jeffrey Karpicke, took a group of college students and split them in two.

The first group read a short science passage four separate times.

The second group read it once, then spent the remaining sessions with a blank sheet of paper, trying to write down everything they could remember. No peeking at the text. No feedback afterward telling them what they had missed. Just the strain of trying to produce it from a mind that only had one pass at it.

Before I tell you what happened, answer honestly.

Which group would you rather have been in?

Almost nobody picks the blank sheet. The first group is obviously the sensible one. Four exposures beats one. And when the students themselves were asked to predict how much they would remember later, they answered the way you just did. The re-readers were confident. They had been over it four times, and it felt smooth and familiar and known.

Then the researchers tested everybody five minutes later, and the re-readers won. They recalled around eighty-one percent of the passage against roughly seventy-five for the group with the blank sheets.

So far, so obvious.

The researchers tested them again one week later.

This time the re-readers produced about forty percent of the passage.

The group that had spent most of its time struggling with a blank sheet produced around sixty percent.

Read that again, because the shape of it is worse than the numbers.

The group that studied the most remembered the least. The group that felt the most confident was the most wrong.

And the difference between them was not effort or intelligence or time spent on the material. It was that one group had experienced the material as smooth, and the other had experienced it as difficult.

The strain was the mechanism.

At UCLA, a memory researcher named Robert Bjork had been circling this for years and gave it the name it now carries. He calls them desirable difficulties: conditions that make learning feel worse in the moment and produce far more of it in the long run.

Spacing your practice instead of cramming it. Mixing problem types instead of drilling one. Testing yourself instead of reviewing.

Underneath sits the distinction Bjork considers the most important idea in his field, and it is worth holding onto, because it explains a great deal more than studying.

Performance and learning are not the same thing.

Performance is what you can do right now, today, under observation. Learning is the durable change that shows up weeks later, or in a situation you have never seen before.

We treat the first as evidence of the second because it is the only one we can watch. But they move independently, and sometimes in opposite directions, which means a method that looks excellent while you are doing it can be teaching you almost nothing.

The re-readers had performance. They did not have learning. And no signal was available to them from the inside that would have told them the difference.

This is where it stops being about students.

Consider what your own intake actually looks like. The book you finished and can now name in conversation. The two-hour podcast that felt genuinely profound in the car. The video where someone explained the thing so clearly that you thought, finally, I understand this.

That feeling of clarity is fluency, and fluency is a measure of how easily the material is going down. It is not a measure of whether any of it is being stored.

The smoother the delivery, the stronger the feeling, and the weaker the evidence.

Which puts you in an awkward position, because effectively everything competing for your attention is optimised for exactly that sensation. Well-edited, well-paced, frictionless. The entire modern doctrine of good content is remove the difficulty, and difficulty is the thing that was doing the work.

You are not being lied to by the people who made it. You are being lied to by the pleasant feeling their competence produces in you, and you have been treating that feeling as a receipt.

One guardrail, because this idea distorts easily into something stupid.

Bjork is explicit that not all difficulty is desirable. A difficulty helps only if you can actually overcome it and it builds real capability. Material far beyond your level, unclear teaching, noise, chaos: those are undesirable difficulties, and they impair learning without giving anything back.

Hard is good is not the lesson. The lesson is narrower and stranger, which is that the specific discomfort of trying to produce something you are not sure you can produce does something no amount of reviewing will do.

There is also a survey worth knowing about. When Karpicke and Roediger asked students how they actually study, the overwhelming majority said rereading. Self-testing came far down the list.

These were not lazy people. They were following the strategy that felt best, which is precisely the strategy that fails, and they had no way of knowing, because the signal they were using was the feeling itself.

The researchers called it the illusion of competence.

So here is a test you can run today, and it costs nothing except the discomfort of finding out.

Take something you consumed recently and were certain you understood. A book from this year, a video you sent to someone, an idea you have repeated out loud with confidence.

Close everything. Take a blank page.

Write down what it actually said.

Not the feeling of it, not the one line you remember, not your general sense of the argument. The structure. The claims. The reasoning that got from one to the next.

Most people discover something unpleasant within about ninety seconds, and the discovery is not that they have a bad memory. It is that the thing they experienced as learning was a sensation, the sensation has gone, and very little was left behind it.

What did you learn this year that you can still produce?

Until next time
Lorenc - Founder of Success Skill