The Two-Hypothesis Answer
Never answer 'how does this work' — ask for two guesses and a way to tell them apart.
- Difficulty
- Starter
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 89%
Eagleman was asked in Smithsonian magazine about encouraging his son to do hypothesis testing, and he unpacks the mechanics on the podcast. Any time one of his children — he has a son and now a daughter — asks how something works, he does not answer. He asks how they think it works. They come up with an idea. He says okay, cool, can you come up with another idea, and they produce a second. Then comes the move that does the actual work: okay, great, now how do we test that, how would you distinguish those two ideas. He describes this as what it is to come up with something on your own and figure out how to test it. The goal is explicitly not vocational — he says they probably won't become scientists for all he knows — it is that he wants them to be good thinkers, able to figure things out so they're not dependent on dogma or what their neighbour tells them is the answer. The method mirrors what Eagleman does professionally, where a good scientist has lots of hypotheses each of which seems like it will crack something open, and then finds that most are wrong. It also sits directly against his own experience of being handed answers as a young man, when everyone told him to become an actuary or a waste engineer because there was money in it — advice that was boring and had nothing to do with Carl Sagan and real questions.
Origin
Eagleman built the habit raising his son, and now his daughter, out of his own professional practice of generating multiple competing hypotheses and designing experiments to distinguish them.
Core principles
- 01Handing over the answer trains dependence on whoever is nearest, not thinking.
- 02One guess is a belief; two guesses is a comparison, which is the unit of scientific reasoning.
- 03The critical third move is asking how you would distinguish them — that converts opinion into a test.
- 04The goal is not producing scientists, it is producing people who are not dependent on dogma or what their neighbour tells them.
- 05The method works because it costs the asker nothing to guess, so participation is easy.
- 06Repetition over years matters more than the quality of any single exchange.
How to run it
- 1
Withhold the answer
When asked how something works, do not explain, even when you know the answer precisely and could deliver it in one sentence.
Pro tip A short pause before responding is enough to break your own reflex to teach.
Watch out Answering first and then asking for their guess is not the method — they will simply repeat you.
- 2
Ask for their first idea
Turn the question straight back: how do you think it works. Let them produce something, however wrong.
Pro tip Ask with genuine curiosity — the tone determines whether they guess freely or try to give the right answer.
Watch out Any hint of a correct answer in your phrasing turns the question into a quiz.
- 3
Take it without correcting
Accept the first idea as a legitimate candidate. Eagleman's response is okay, cool — no evaluation, no ranking.
Pro tip Repeat their idea back neutrally so they can hear it stated as a hypothesis.
Watch out Correcting the first guess ends the exercise, because there is now a right answer in the room.
- 4
Demand a second idea
Ask them to come up with another one. This is the step that converts a belief into a comparison and prevents attachment to the first thing they thought of.
Pro tip If they struggle, offer that the second idea can be sillier than the first.
Watch out Stopping at one idea trains exactly the one-hypothesis attachment Crick warned Eagleman about.
- 5
Ask how to distinguish them
Pose the central question: how would you test that, how would you distinguish those two ideas. This is where the method stops being conversation and becomes method.
Pro tip Push for an observation that would come out differently under each idea, not just evidence consistent with one.
Watch out Accepting 'I'd look it up' skips the whole point; steer back to what they could observe themselves.
- 6
Run the cheapest test
Where possible actually do the experiment together rather than settling it verbally, so the outcome comes from the world rather than from you.
Pro tip Let the test fail or be inconclusive sometimes; that is a truer picture of research than a clean answer.
Watch out If you adjudicate at the end, you have reinstalled yourself as the authority the whole method was designed to remove.
In the wild
Eagleman describes the exchange he runs with both his children whenever they ask how something works. He turns the question back, accepts their first idea with okay cool, asks for a second, then asks how they would distinguish the two. He is explicit that they probably will not become scientists, and that this is not the point.
→ The stated aim is children who can figure things out and are not dependent on dogma or what their neighbour tells them is the answer.
Growing up, Eagleman was good at science and everyone around him supplied answers about what to do with it — become an actuary because you can do math, become a waste engineer because there's a lot of money in it. He describes all of this as bad advice: proposals that were boring and had nothing to do with Carl Sagan and real questions.
→ He only found neuroscience by taking a class in his final undergraduate semester, and he worries that without such luck, the David Eaglemans of the world end up being waste engineers.
Common mistakes
Correcting the first guess
The method depends on both candidate ideas being live at the same time. The moment you signal that idea one is wrong, the child stops generating and starts trying to retrieve the answer you already have, which is precisely the dependence on authority the exercise exists to prevent.
Stopping after one hypothesis
One idea leaves nothing to distinguish, so there is no test to design and no scientific move to make. Worse, it trains the exact attachment Crick warned about — the dangerous man with only one idea who will fight and die for it.
Turning it into a quiz
If you already have the answer in mind and are fishing for it, the child learns to guess what the adult wants rather than what the world does. The tell is that you evaluate their guesses rather than asking how to test them.
Is it for you?
Best for
Parents, teachers and mentors who want people to become independent thinkers rather than well-informed dependents.
Not ideal for
Safety-critical or time-pressured moments where the correct answer is needed immediately, and genuinely settled factual lookups.
From the transcript
“Now, how do we test that? How would you distinguish those two ideas?”
“So that's what it is to come up with something on your own and figure out how to test it”
“figure things out so they're not dependent on dogma or what their neighbor tells them is the answer, that sort of thing.”
From the episode
#674: Neuroscientist David Eagleman — Exploring Consciousness, Sensory Augmentation, The Lazy Susan Method of Extraordinary Productivity, Dreaming, Improving Hearing with a Wristband, Synesthesia, Stretching Time with Novelty, Lessons from Titans of Science, and Much More