SCIENCE & CURIOSITY · B2 · 50 min

What If?

A NASA roboticist turned cartoonist takes the most ridiculous questions strangers can invent — and answers them with real physics. The answer is almost always the same: everything dies, but in a fascinating way.

Today: you'll meet the man who answers impossible questions seriously — and you'll master the language for imagining things that will never happen: if the oceans disappeared, the rain would stop; you might survive; it could go either way.
Before we start. Talk these through out loud. Nothing to look up — just say what comes to mind.
Discuss:
  • What would happen if everybody on Earth jumped into the air at exactly the same moment? Commit to an answer before you hear anyone else's.
  • Think of a question you asked as a child that adults refused to take seriously. What was it — and what answer did you get?
  • Is there such a thing as a stupid question? Argue the opposite of whatever you actually believe.

The vocabulary of impossible questions

Ten items. Work through the four activities — say your answer out loud before you reveal it.

A · What do they mean here? Work it out from the context

1. Nobody is ever going to fill the Solar System with soup. The question is entirely hypothetical — and that is exactly the point of asking it.
hypothetical (adj) — imagined, not real; asked in order to explore an idea. A hypothetical situation; purely hypothetical; hypothetically speaking…
2. He does not need the exact figure. He needs to know whether the answer is closer to ten or to ten billion — so he makes a rough estimate on the back of an envelope and follows it.
a rough estimate (n) — an approximate calculation, accurate enough to be useful. To estimate (v); at a rough estimate; a ballpark figure. Note the pronunciation of rough: it rhymes with stuff.
3. When the drop hit the ground, the air itself was pushed outward in a wall of pressure that flattened the forest for miles. Almost every answer in the book ends with a shockwave.
a shockwave (n) — the wave of pressure that travels outwards from an explosion or impact. Also used about news: the announcement sent shockwaves through the industry.

B · Which word means…?

Three words: lethal · negligible · plausible. Say the right one, then reveal.

1. Which word means so small that it makes no real difference?
negligible (adj) — a negligible effect / amount / risk. From the verb to neglect: small enough to ignore. Several of today's answers turn out to be exactly this.
2. Which word means strong enough to kill you?
lethal (adj) — a lethal dose; lethal force; potentially lethal. Slightly more formal and more final than deadly.
3. Which word means believable — it could reasonably be true?
plausible (adj) — a plausible explanation; that sounds plausible. Careful: it says the idea is believable, not that it is true. The opposite is implausible.

C · Choose the meaning — a / b / c

1. A falling object speeds up until the air pushes back as hard as gravity pulls, and then it stops accelerating. That fixed speed is its terminal velocity. Terminal velocity means…
  (a) the fastest a falling object can go through air   (b) the speed at which an object breaks apart   (c) the speed needed to escape the planet
(a) — terminal velocity (n) — the steady speed a falling body reaches when air resistance balances gravity. Terminal here means final, not fatal — though in several of today's answers it is both.
2. Deep water sits between you and the radiation and absorbs almost all of it. The water shields the swimmer. To shield means…
  (a) to warn about a danger   (b) to measure a danger   (c) to protect by blocking something
(c) — to shield (v) / a shield (n) — to protect by putting something in the way. To shield somebody from the truth works the same way.

D · Word building & collocation

1. You know absurd. Build the set: an ___ (noun: the quality of being ridiculous) · ___ expensive (adverb: to an unreasonable degree) · to reduce an argument to ___ (the classic debating move).
An absurdity · absurdly expensive · to reduce an argument to absurdity. That last move — pushing an idea until it becomes ridiculous — is the entire method of this book.
2. Grade these three from mildest to most extreme, and say what each one collocates with: negligible · significant · catastrophic.
negligible → significant → catastrophic. A negligible risk · a significant increase · catastrophic damage / a catastrophic failure. Today's answers live almost entirely at the two ends and almost never in the middle.

Now use them — out loud

Before we read:
  • Using hypothetical, rough estimate and plausible, describe how you would go about answering a question nobody can test.
  • Sort the ten items into words for danger and words for thinking. Which group is bigger — and what does that predict about the book?

The story, in three parts

Read each part. After it, there's a quick check — answer from memory first — then one question to discuss.

Part 1 — The man who takes silly questions seriously

Randall Munroe trained as a physicist and went to work for NASA, building robots at the Langley Research Center in Virginia. Then, in 2006, he did something that sounds like the setup to a joke: he left the space agency to draw stick figures full time. The webcomic he had been posting under the meaningless four-letter name xkcd had found an audience of scientists, programmers and mathematicians who recognised their own humour in it, and it turned out to pay better than robots.

What he had not planned on was the post. Readers began sending him questions — not sensible ones, but the kind a bright twelve-year-old asks at the dinner table and is told to stop asking. What would happen if you hit a baseball travelling at nearly the speed of light? Could you build a jetpack out of machine guns? Most people would treat these as jokes. Munroe treated them as homework. In 2012 he started a companion blog, What If?, where each week he took one absurd question and answered it with genuine physics, genuine arithmetic and stick-figure drawings. The book that followed in 2014 became an international bestseller, and a second volume appeared in 2022.

The rule that makes the whole thing work is simple and slightly heroic: he never says that's a stupid question. He assumes the question deserves a real answer, and then he goes and finds one — reading research papers, emailing specialists, and occasionally admitting that the honest reply is a shrug. The comedy is not in mocking the question. It is in what happens when a ridiculous premise is followed, patiently and politely, all the way to its conclusion.

Quick check — from memory, then reveal:

1. What was Munroe doing before he drew comics, and why did he stop?

2. What is the one rule that makes the project work?

1. Building robots for NASA at Langley. He left in 2006 because his webcomic xkcd had found a large audience and could support him.

2. He never dismisses a question as stupid. Every question gets a real answer, researched properly — and the humour comes from following an absurd premise honestly to the end.

Discuss:
  • "He never says that's a stupid question." Where in ordinary life would that rule change things most — and where would it be exhausting?

Part 2 — How you answer a question nobody can test

The method has a name and a hero. During the first nuclear test in 1945, the physicist Enrico Fermi is said to have torn up a sheet of paper and let the scraps fall as the shockwave arrived. He watched how far sideways the blast pushed them, did a calculation in his head, and produced an estimate of the explosion's power that was close to the answer the instruments took weeks to confirm. That trick — getting from almost no information to a roughly right number — is now taught as Fermi estimation, and it is the engine underneath every answer in this book.

It works in three moves. First, turn the vague question into numbers: not what if it rained a lot, but how many cubic kilometres of water, falling from what height. Second, find the number that dominates — in almost every problem, one quantity is so much larger than the others that the rest can be ignored, and a rough estimate is therefore enough. Third, follow the physics wherever it goes, even when it stops being sensible. You are not trying to be exact. You are trying to find out whether the answer is a puddle or a crater.

The pleasure of the book is in the detours this produces. A question about rain becomes a question about how fast air can get out of the way. A question about lasers becomes a question about how bright moonlight actually is. Munroe keeps going long after the original asker has stopped caring, and somewhere in that extra distance the reader accidentally learns some real science. The joke and the lesson are the same object.

Quick check — from memory, then reveal:

1. What did Fermi do with the paper, and why is it remembered?

2. What are the three moves of the method?

1. At the 1945 test he dropped paper scraps and measured how far the shockwave blew them sideways, estimating the blast's power in his head — close to the figure the instruments later gave. It shows you can get a useful answer from almost no data.

2. Turn the question into numbers; find the one quantity that dominates and ignore the rest; then follow the physics wherever it leads. Exactness is not the goal — the right order of magnitude is.

Discuss:
  • Try it now, out loud: roughly how many words do you speak in a day? Reason your way there in three steps — no phone.

Part 3 — Why everything ends badly

Read enough of these answers and a pattern appears. An extraordinary number of them finish with a fireball. This is not because the author enjoys destruction, although he clearly does. It is because of what energy does when you scale something up. Double the speed of an object and you do not double its energy — you quadruple it. Push that to a fraction of the speed of light and an ordinary baseball carries the energy of a bomb. Human intuition, built for stones and sticks, has no feeling for this at all.

The opposite failure is just as common, and it is the reason some of the answers are so disappointing. Ask what happens if every person on Earth does something at once and the instinct is that it must be enormous. It almost never is. The planet outweighs all of humanity by a factor so vast that our combined effort disappears into the rounding error. Eight billion people are, from the Earth's point of view, negligible — and the honest answer to several famous questions is simply nothing much.

And then there is the third category, the most useful one: the questions where fear points the wrong way. People are terrified of a swimming pool full of nuclear waste and relaxed about a warm afternoon on a motorway; the arithmetic reverses both. What the book is really training is a suspicion of your own gut. Intuition is excellent at human scale and hopeless above and below it — and the cure, every single time, is not to feel harder but to work it out.

Quick check — from memory, then reveal:

1. Why do so many answers end in an explosion?

2. Why is "everybody on Earth does X at once" usually an anticlimax?

1. Energy rises far faster than speed — doubling the speed quadruples the energy — so anything moving extremely fast carries the energy of a weapon. Our intuition, built at human scale, cannot feel that.

2. Because the Earth is unimaginably more massive than all of humanity put together, so our combined effort is negligible. The honest answer is usually "nothing much".

Discuss:
  • "Intuition is excellent at human scale and hopeless above and below it." Name one thing you are afraid of that is probably safe, and one you are relaxed about that probably isn't.

Grammar — imagining the impossible

Every question in this book runs on one structure and three modal verbs. Get them right and you can speculate about anything; get them wrong and the whole idea collapses into confusion about whether you mean it.

The second conditional — for situations that are imaginary, impossible, or extremely unlikely:
If + past simple, would + infinitive
• "If the oceans disappeared, the rain would stop almost everywhere."
• "What would you do if a second Moon appeared tomorrow?"
The past form does not mean past time. It marks distance from reality — which is exactly what these questions need.

Two refinements worth having: careful English prefers were for every person — "if I were you", "if the Sun were closer". And "If you were to…" pushes the idea even further away: "If you were to swim down to the bottom, the story would change."
would · might · could — how certain are you?
• would = you are confident. The physics is not in doubt. "The impact would flatten the forest."
• might / could = it is possible, not certain. "The tunnel might collapse." · "You could survive, in theory."
• Negatives behave differently: wouldn't = certainly not; might not = possibly not. "You wouldn't feel it" is a promise; "you might not feel it" is a hope.

Adverbs that fine-tune it — they go straight after would: would almost certainly die · would probably survive · would presumably notice · would, in theory, work.

1. Put the verbs in the right form

Complete each sentence with the correct form of the verb in brackets.

1. If the Earth (stop) turning tomorrow, the air (keep) moving at the speed it has now.

2. If you (swim) near the surface of that pool, you (receive) less radiation than you do on a beach.

3. What (happen) if an entire storm (fall) as one enormous drop?

1. stopped · would keep    2. swam · would receive    3. would happen · fell

Notice number 3: the question word takes the would half, and the if half still goes into the past. The order of the two halves changes nothing.

2. would or might? Read the note in brackets first

The bracket tells you how certain the speaker is. Only one of the two fits.

1. A drop that size flatten the trees for miles. (the physics is not in doubt)

2. If everybody stayed apart for two weeks, the common cold disappear — though a few long infections could bring it back. (possible, but genuinely uncertain)

3. Eight billion people jumping at once move the planet in any way you could measure. (certainly not — use the negative)

4. Somewhere above the height of the Himalayas you lose consciousness. (confident — everyone does, eventually)

1. would    2. might (could also works here)    3. wouldn't    4. would

Number 2 is the one that matters: the sentence itself admits the doubt, so a confident would contradicts it.

3. Complete the answer

Fill each gap, then read the whole thing aloud as if you were explaining it to somebody who asked.

Suppose you (decide) to drill a tunnel straight through the Earth and jump in. If you left the air in the tunnel, you (not get) very far: you would reach terminal velocity, slow down near the centre, and (end) up floating in the middle of the planet, which (be) an unusual place to spend an afternoon. If, on the other hand, somebody (pump) all the air out first, the fall (take) roughly forty-two minutes, and you (arrive) at the far side with almost no speed at all. At that moment you (need) to grab something. If you (miss), you (fall) back the way you came — and then forwards again, forever.

decided · wouldn't get · end · would be · pumped · would take · would arrive · would need · missed · would fall

Watch gap 3: it shares the would from earlier in the list, so the bare infinitive is enough — "you would reach…, slow down…, and end up…". Repeating would is not wrong, but no English speaker does it.

Take the structure straight into the next tab — every prediction there should be a full second-conditional sentence, out loud, before you reveal anything.

The 20 questions

Twenty real questions, twenty real answers. The rule is strict and it matters: say your prediction out loud as a full sentence before you press reveal. Use the structure — "If you did that, it would… / it might…" — and commit to it. Guessing privately and then reading the answer teaches nothing; being wrong out loud teaches a great deal.

Keep score. After each set of five, count how many you were roughly right about. Four rounds, twenty questions — the final score is the interesting part, and so is which kind you keep getting wrong.
Round one — questions 1 to 5
1What happens if you swing at a baseball pitched at 90% of the speed of light?
The ball hits the air molecules in front of it too fast for them to move out of the way, so they fuse. You get a thermonuclear fireball that destroys the pitcher, the batter and most of the surrounding neighbourhood. Under the rules of baseball, the batter is technically hit by the pitch — and is awarded first base.
2If everyone on Earth stood in one place and jumped at exactly the same instant, what would happen?
Nothing. All of humanity shifts the Earth by less than the width of an atom, and it pushes straight back when we land. The catastrophe is what comes afterwards: eight billion people standing in one place with no food, no transport and no way home.
3What if a whole rainstorm fell as one enormous drop?
A drop kilometres across compresses the air beneath it and arrives at tremendous speed. The shockwave flattens forest for miles; the water then races outwards as a wall. The town underneath is gone, but people twenty miles away survive to watch a mushroom cloud of steam.
4Could you survive a swim in a pool where spent nuclear fuel is stored?
Near the surface, yes — comfortably. Water shields you so well that you would receive less radiation than you do walking around outside. Swim down to within a metre or two of the fuel and the dose becomes lethal. Realistically, security would shoot you long before the radiation did.
5What if the Earth stopped turning, but the atmosphere kept moving?
A wind of roughly a thousand miles an hour scours the surface near the equator, stripping away everything and starting firestorms. Only the regions near the poles, where the ground was barely moving anyway, are survivable. Days also become six months long.
Discuss before round two:
  • Score yourself out of five. Which did you get wrong — and was your error too dramatic or not dramatic enough?
Round two — questions 6 to 10
6If everyone pointed a laser pointer at the Moon at once, would it change colour?
No — the effect is completely negligible beside the sunlight already falling on it. Munroe then keeps scaling the lasers up, past anything that exists, until the Moon's surface is boiling. The interesting part is how absurdly powerful they have to get before anything happens at all.
7How long could humanity last against a robot uprising?
Indefinitely. Robots cannot reliably manage stairs, doorknobs, rough ground or rain, and most run out of battery within an hour. Walk upstairs, shut the door, wait. The honest answer is an anticlimax — which is itself worth noticing.
8What if you built a periodic table out of solid blocks of every element?
The first two rows are fine. Then the sodium and potassium catch fire, the halogens poison the air, and by the time you reach the radioactive elements the table is producing enough heat and radiation to melt itself and everyone standing near it. A catastrophic way to teach chemistry.
9How fast can you hit a speed bump and survive?
Faster than you would guess. At normal speeds it is a jolt; well above them the bump stops being a bump and becomes a ramp. The bump itself is not what kills you — losing control of the car and landing is. The danger sits one step further along the chain than people expect.
10If everybody on Earth avoided everybody else for two weeks, would the common cold die out?
It is genuinely plausible. Ordinary cold infections are short, so a fortnight of complete isolation would break the chain. But a small number of people with weakened immune systems carry the virus for months — so it would probably creep back out afterwards.
Discuss before round three:
  • Four of the ten so far have been anticlimaxes. Why is "nothing happens" such a hard answer to guess — and who benefits from us assuming the opposite?
Round three — questions 11 to 15
11Could you build a jetpack out of downward-firing machine guns?
Astonishingly, yes. A rifle firing downwards produces several times more recoil than is needed to lift its own weight, so a cluster of them could lift a person — for a few seconds. Use the enormous cannon carried by a ground-attack aircraft instead and you could hover for about twenty seconds, until the ammunition runs out.
12If you began rising steadily at one foot per second, what would kill you first?
Cold and thin air. You clear the trees within a minute and are freezing within an hour or two; somewhere around the height of the Himalayas the oxygen becomes too thin and you lose consciousness — a few hours in. You die a long way below the stratosphere, and the view is excellent.
13What if the Sun simply switched off?
The first surprise is a list of benefits: no solar storms, no satellite damage, and fewer immediate deaths than you would expect. Then the temperature falls below freezing everywhere within about a week and keeps falling. Everyone dies — but slowly, and with unusually good astronomy.
14What would happen if you drilled a tunnel through the Earth and jumped in?
With the air pumped out (and the heat politely ignored), you fall for about forty-two minutes and arrive at the far side with almost no speed — grab the edge or you fall back forever. Leave the air in and you reach terminal velocity, stall near the centre, and end up floating in the middle of the planet.
15What if you drained every ocean on the planet?
Through a modest drain it takes hundreds of thousands of years. Vast new land appears and the continents join up — but the climate collapses: with no ocean there is almost no rain, and the new interiors become desert. The gift and the disaster are the same event.
Discuss before round four:
  • Three of these five are slow deaths rather than explosions. Which is the more frightening kind of answer, and why do films almost always choose the other one?
Round four — questions 16 to 20
16What if every person really did have exactly one soulmate?
A nightmare. With ordinary chance encounters across a lifetime, only a tiny fraction of people would ever meet theirs; almost everyone dies alone, and the world reorganises itself into a desperate global search operation. The romantic idea does not survive contact with the arithmetic.
17How much physical space does the internet take up?
Less than you would think. Stored on the tiny memory cards that go in a phone, the world's data would fit inside something the size of a shipping container — which leads to a memorable conclusion: for sheer volume of data, a lorry full of memory cards still beats any cable.
18What if you tried to print out the whole of Wikipedia?
Tens of thousands of thick volumes and a room of printers running for months. The real problem is not paper but time: thousands of edits land every hour, so the encyclopaedia is out of date long before the first shelf is dry.
19Could you build a bridge to the Moon out of Lego bricks?
No. You would need far more bricks than have ever been manufactured, and long before you left the atmosphere the ones at the bottom would be crushed flat by the weight above them. Plastic has a limit, and it is a few kilometres, not a few hundred thousand.
20What if you filled the Solar System with soup, out to the orbit of Jupiter?
That much soup weighs millions of times more than the Sun. It collapses instantly under its own gravity into an enormous black hole. No survivors, no soup, and — as the book notes with some regret — nobody left to taste it.

Now sort them

All twenty answers fall into four families. Working from memory, put each question into one:

A · it ends in a fireball    B · nothing much happens    C · you die, but slowly    D · it's really just arithmetic
Sort all twenty, out loud, giving a reason for each. Then reveal one possible sorting and argue with it — several of them genuinely belong in two families.
A — fireball: 1 baseball · 3 rain drop · 5 stopped Earth · 8 periodic table · 20 soup.
B — nothing much: 2 everyone jumps · 6 lasers · 7 robots · 9 speed bumps · 4 the fuel pool.
C — slow death: 12 rising · 13 the Sun · 15 the oceans · 14 the tunnel · 16 soulmates.
D — arithmetic: 17 the internet · 18 Wikipedia · 19 Lego · 11 the jetpack · 10 the cold.
Defensible arguments exist for moving 4, 11, 14 and 16 — make one.
Talk it through:
  • Final score out of twenty. Which family did you consistently misjudge — and what does that say about your intuition?
  • Rank the twenty from "most disappointing answer" to "best answer". Defend your top and bottom choice.
  • Which single question would you most want a longer answer to, and what exactly do you still want to know?

Going deeper

A note on the record: these are estimates, not predictions. The point of the exercise is the reasoning, not the decimal places — and the author is the first to say so, often in a footnote, usually while calculating something worse.

What the questions actually teach. First, the scale problem. Human beings evolved to judge quantities in a narrow band — a thrown stone, a day's walk, a crowd in a square — and outside that band our instincts are not merely imprecise, they are backwards. A baseball feels harmless and becomes a bomb; eight billion people feel unstoppable and move nothing. Neither error is corrected by thinking harder about how it feels. Both are corrected by arithmetic, which is the least glamorous superpower available and the most reliable.

Second, the dignity of the silly question. Every one of these began as the kind of question a child asks and an adult deflects, and every one of them, taken seriously, opens into real science: fusion, fluid dynamics, radiation shielding, orbital mechanics, epidemiology. The absurd premise is not a distraction from the learning — it is the delivery mechanism. Curiosity that has been laughed at once tends not to come back, which makes "that's a stupid question" one of the more expensive sentences in education.

Third, the limits, stated honestly. A rough estimate can tell you whether an answer is a puddle or a crater; it cannot tell you the depth of the puddle, and it collapses entirely where the physics is unknown or where human behaviour is involved. The soulmate question is the clearest example — the mathematics is straightforward and the conclusion is still absurd, because people are not billiard balls. The method is a torch, not a map. Knowing the difference is the whole skill.
Talk it through:
  • "Curiosity that has been laughed at once tends not to come back." Test it against your own life — what did you stop asking about, and when?
  • Where else does a rough estimate beat a strong feeling — money, risk, health, time? Give a concrete case and do the sum out loud.
  • The soulmate answer is mathematically clean and humanly wrong. What does that reveal about the limits of the whole method?
  • Is there any question that genuinely should not be answered seriously? Make the strongest case you can, then argue against yourself.
  • Another conversation in this collection argues that our appetites and our evenings are deliberately engineered; this one argues that our instincts were simply never calibrated for the job. Which explains more of an ordinary day?

Speak

All of it out loud, against the clock — and let the structure carry it: if that happened, it would… it might… it almost certainly wouldn't.

Useful language: Suppose you… · If you were to… · It would almost certainly… · It might well… · The effect would be negligible · That sounds plausible, but… · At a rough estimate… · Follow it through: if X, then Y
Time remaining: 4:00
Task A · Tell it back (4 minutes)

Tell the whole story without notes: the physicist who left NASA to draw stick figures → the questions readers wouldn't stop sending → Fermi and the falling paper → turn it into numbers, find the number that matters, follow the physics → why so many answers end in fire → why so many end in nothing at all → what that says about intuition.

Use at least five key words (hypothetical, rough estimate, shockwave, negligible, lethal, plausible, terminal velocity, shield, absurd, catastrophic) and at least three second conditionals.

Task B · Your own question (4 minutes)

Invent a "what if" question nobody has asked. Then answer it using the method — out loud, showing your reasoning: what has to be turned into a number, which number dominates, and where the physics leads.

Then defend it: is your answer a fireball, an anticlimax, a slow death, or just arithmetic?

Task C · The argument (4 minutes)

Take a side and argue it properly: is answering absurd questions a good use of a scientist's time? Round one, for. Round two, against. Concede one real point on each side.

Prefer a different angle? Take one of these:

  • Explain one of the twenty answers to a curious ten-year-old — no jargon, sixty seconds.
  • Choose the answer you found most disappointing and argue that it is actually the most interesting one.
  • Make the case for a school subject built entirely from ridiculous questions. What would be lost?
  • Estimate something unmeasurable out loud — the number of doors in your city, the weight of everyone's keys — and show every step.
Without looking anything up. Say the answers aloud:
  • What are three different ways to talk about the future in English — and when would you choose each one?
  • What is the difference between I used to live near the sea and I would swim every morning? Which one works with states, and which with repeated actions?
  • Give two ways to soften a strong opinion so that it sounds less absolute.
  • From today: three of the twenty answers, from memory, each in one full second-conditional sentence.

To finish — the bigger questions

  1. You scored yourself out of twenty. Was your intuition too dramatic or not dramatic enough — and where else in life does that same bias show up?
  2. "Intuition is excellent at human scale and hopeless above and below it." Name one public argument that would improve if everybody involved did the arithmetic first.
  3. The method answers "what would happen" beautifully and "what should we do" not at all. Where is the boundary between the two, and who guards it?
  4. Before you go — from memory: one sentence with if… would, one with might, one with wouldn't — all about something impossible — and explain why the past form appears in a sentence about the future.