The Read, Write, Compare technique is a great way to improve your writing and a method memory research backs. You read a page you love, write it from memory, then compare your version against the original.
The copying that teaches nothing
Re-reading a text is the most common study habit there is, and one of the weakest. When the words sit in front of you, your eye glides over them and your brain reports a warm feeling of recognition. I know this. But this feeling is a trap named the fluency illusion. The easy, smooth practice convinces you that you understand more than you do.
Copying by hand is re-reading with extra steps. You see the sentence, your hand reproduces the sentence, and nothing inside you has to do the work of building it. The prose passes through and leaves the writer unchanged.
What Stevenson and Franklin did instead
In 1887, Robert Louis Stevenson described the drill that taught him to write. Whenever a passage pleased him, he sat down at once and tried to reproduce its qualities from memory. If he failed, he tried again, and if he failed again, he tried again. He called it playing “the sedulous ape,” and the failures, he insisted, were the training. Even in those losing bouts he got practice in construction, rythm and the fit of the parts.
Stevenson did not copy the text in front of him. He closed the book and tried to “ape that quality” without looking.
A century earlier, a teenage Benjamin Franklin built almost the identical exercise. He took essays from The Spectator, jotted a short hint for the sentiment of each sentence, set the notes aside for a few days, then rebuilt each essay in his own words with the book shut. After that he compared his version against the original, found his faults, and corrected them. He even turned the prose into verse and back again to widen his stock of words.
Two self-taught writers, separated by a hundred years, landed on the same learning method: read, write from memory, compare. They were running an experiment on themselves long before anyone could explain why it worked.
Why writing from memory beats copying
In 1978, researchers ran 5 experiments comparing words people generated themselves against the same words handed to them to read. The generated words won every time, across recall and recognition. Producing something yourself fixes it deeper than receiving it. This is the generation effect.
Retrieval does similar work. Researchers had students study a passage, then either re-read it or set it aside and write down everything they could remember. On a test days later, the students who pulled the material from memory beat the re-readers, even though the re-readers had spent more time with the words in view.
The act of reaching for a memory strengthens it, while the act of looking at it again mostly produces the feeling of knowing. In learning we like to call this desirable difficulties. Closing the book to rebuild a page does the work that rereading wil never accomplish. The effort to reconstruct what you’ve half-forgotten is exactly what drives the gain. Read, Write, Compare forces all three moves at once. You generate the prose, you retrieve the text, and the comparison hands you immediate feedback.
The Read, Write, Compare technique
Here is the full sequence, built on the same logic Franklin and Stevenson used:
The comparison step does the heavy lifting. The gap between your sentence and the author’s is the lesson. Here’s where the learning happens.
Where it goes wrong
There are 3 ways to waste the effort.
You have to take the steps yourself, then check against the answer. The finished text is the solution. Read, Write, Compare puts you back in the driver’s seat, taking each turn by hand.
The bottom line
You are not trying to reach a destination another writer already reached. You are learning to drive. So the next time a paragraph stops you cold with how good it is, don’t admire it and move on, and don’t copy it out. Close the book, write it from memory, and let the gap between your version and theirs show you the road they took.
Frequently asked questions
It’s a self-teaching method for writers. You pick a page you admire, read it a few times, close the book, write the page from memory, then compare your draft against the original to see where the two diverge. The differences reveal the choices the author made. Robert Louis Stevenson and Benjamin Franklin both taught themselves to write using versions of this drill.
Less than most people hope. Hand-copying lets you re-read without retrieving or generating anything, and re-reading builds a sense of fluency more than durable skill (Roediger & Karpicke, 2006). Writing the passage from memory and then comparing it works better, because it forces the harder mental moves that build ability.
There’s no fixed timeline, and it depends on your cadence. One page a week, worked through the full compare step, will teach you more about sentence rhythm and word choice within a month or two than years of passive reading. The gains compound, because each page adds choices you’ve tested against a master’s.
A single page of prose you’d be proud to have written, by an author one step above your current level. Reach too high and the exercise turns frustrating without teaching, because you can’t make a fair attempt (Bjork & Bjork, 2011). A paragraph that stops you mid-read is usually a good signal.
No. Knowing the page by heart is the end point, not the goal. The learning happens in the act of choosing words, comparing them against the original, and noticing where you went a different way. Recite a page without ever comparing your draft to it and you’ve trained recall, not craft.
Copywork keeps the original in front of you while your hand reproduces it, so you re-read instead of building. Read, Write, Compare closes the book and makes you generate the prose from memory first, then check it. The closed book is the whole difference, and it’s where the learning lives.
Leave feedback about this