It was Pamela Paul who reminded me about the great failing of books.

Pamela edits The New York Times Book Review. That makes her one of the bookiest people on earth. And in an interview last year, she ‘fessed up about where our relationship with books fails. “I remember the book itself,” she told The Atlantic, speaking about her general experience of reading. “I remember the physical object. I remember the edition … I usually remember where I bought it, or who gave it to me.”

And then she said it: “What I don’t remember – and it’s terrible – is everything else.”

Yes, this book expert forgets most of what she actually reads.

She is not alone. I’m now fairly sure that her experience is the same as most people’s. Books don’t embed their content in people’s heads.

Everyone forgets

As an example, Pamela offered her experience of Walter Isaacson’s biography of Benjamin Franklin. Two days after reading it, she said, she doubted she could even offer a general timeline of the American Revolution.

I immediately sympathised with her, because I have had the same experience. I mostly read non-fiction, and I have often been shocked by the way I forget much of a book’s content within a few days of reading it.

Brenda, my wife, reads mostly fiction, and has a pretty similar experience. In fact, almost everyone who talks with me about this has the same experience.

When you think this through, you see some fairly radical implications. We live in a culture which reveres books and book-reading. At school we are told to read the books and not just buy the Cliff’s Notes. At work, we are told that the smartest managers read books constantly. But all of that may be wrong.

Surprisingly bad at teaching

A recent essay by Andy Matuschak takes this to its logical conclusion. “As a medium, books are surprisingly bad at conveying knowledge,” Matuschak writes. And then he adds: “and readers mostly don’t realise it.”

In other words: books don’t do their job, and no-one calls them out.

Matuschak doesn’t make this damning judgement out of ignorance. Quite the opposite: he has spent several years working on instructional design for Khan Academy, that great YouTube-based educator.

His experience has taught him that people learn in a number of ways with varying degrees of effectiveness. Books work only occasionally, mostly in the hands of “active readers” who react to the text as they go, turning over and analysing ideas as they read.

These “metacognitive” skills are tough to learn. Textbooks often have exercises designed to encourage them, but almost no-one does those exercises unless they have to.

Puffed-up magazine articles

I had the opportunity last week to talk to academic economist and famed blogger Tyler Cowen, whose Marginal Revolution is one of the finest and most wide-ranging publications of the past two decades.

Cowen, a famously voracious reader, churns out pocket book reviews for the site; like Paul, he qualifies for the title of “famous reader”. He’s also an admirer of Matuschak, and has even raised money for Matuschak’s efforts to improve on the book.

Cowen argues not only that books are a bad way of teaching, but that they often have less to teach than their impressive 300 pages would suggest. “Smart people often overrate books,” he told me drily. “Too many of them are puffed- up magazine articles.”

From my own experience writing magazine articles about books, he’s right.

A foundational error

Those of us who spend our lives working with ideas believe disproportionately that if we read books, they will teach us. Yet that foundational belief appears to be wrong.

The people like Matuschak who are thinking about the limits of the book are also thinking about how to improve learning. It’s 2019, and we have a lot of possibilities for learning that didn’t exist just a quarter-century ago. More on that in a future column.