You’re neck-deep in IKEA assembly instructions. Furniture parts lie strewn across the floor. Your new purchase sits half-complete in front of you, mocking your fruitless hours. As an uninterested partner walks in, you let the frustration out:
“I’ve done everything correctly! Look:
- connect A with B using M1 screws
- connect B with C with the M3 bolt using the key
- join BC with D using… wait.”
You suddenly realize you haven’t joined BC with D. It all starts to click into place (literally), et voilà, you’re finished.
It’s a universal experience: the moment you try to explain a problem out loud, it all begins to make sense.
Software engineers call it “rubber duck debugging.” So where did this term come from, and why is it so effective?
Explaining aloud
This well-known software engineering term has its origins in a story told in The Pragmatic Programmer, a book by Andrew Hunt and David Thomas.
The gist of it is that one should obtain a rubber duck, and use it when your code isn’t working — and you don’t know why.
Explain to the duck what your code is supposed to do, and then “go into detail and explain things line by line.”
Soon, the moment of revelation strikes: You realize, as you speak aloud, that what you meant to do and what you actually did are two very different things.
I often bring up rubber duck debugging in my introductory programming lessons to help students when they can’t understand why their code won’t work.
Despite its roots in programming, the ideas that underpin the rubber duck approach apply to programmers and non-programmers alike.
Why does it work?
Most of us think out loud as we learn with our first books, reading aloud as we go. There’s something illuminating about articulating aloud that helps you “hear” the problem your brain has thus far been unable to detect.
And research by U.S. scholars Logan Fiorella and Richard Meyer has examined how learning can be enhanced through the act of teaching others.
Their experiments found that when students learn the contents of a lesson as though they are going to teach it to others — and then actually teach it to others — they “develop a deeper and more persistent understanding of the material.”
Teaching others forces us to break the material down into conceptual pieces, integrating it with our existing knowledge, and organizing it in logical ways.
Their research also identifies “self-explaining” as an evidence-based learning strategy.
That’s why our little yellow friend is so helpful; in explaining the problem aloud to your rubber duck, you are teaching it as well.
The rubber duck and their blank, cute face
But why a rubber duck?
Well, talking to a human can come with certain limits.
Humans are contextual, with previous thought and experience; they may miss your mistakes because they’ve assumed something about your previous attempts to solve the problem. They may have internal biases that make it hard for them to see where you’ve gone wrong.
A rubber duck, however, has none of this. As silly as it might look, rubber ducking forces you to explain things in precise detail to that blank (cute) face looking back at you.
Of course, it doesn’t have to be a duck. Any old object (or uninterested party, as I seem to keep finding) will do in a pinch. Some researchers even advocate replacing the duck with a large language model such as ChatGPT. The AI chatbot can, they argue, “act as a virtual, hyper-intelligent, ever-present programming partner to a software engineer” wanting to walk through their code line by line to find errors — and suggest fixes, too.
Others have experimented with a modified rubber duck that, when the user presses a button, nods or offers brief, neutral replies to your explanations. The interactivity, the researchers argue, might make people feel more comfortable talking to a duck.
So, next time you’re stuck on a problem at work, suffering writer’s block, or trying to make sense of a convoluted email chain, try turning to a little yellow duck.
See if explaining your problem aloud to them can help you arrive at the answer.
Elliot Varoy, Senior Lecturer, School of Computer Science, University of Sydney
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Follow us on X, Facebook, or Pinterest
