Why Do Planes Fly?
— The Textbook Explanation Is Wrong
"The top of the wing is curved, so air travels a longer path, moves faster, and the pressure drops." You'll find this explanation in countless books and websites. It's wrong. But the claim that "nobody actually knows why planes fly" is also wrong. The truth is known — it just hasn't been explained properly.
The familiar story goes: "The top of the wing bulges outward. Air passing over it travels farther than the air underneath, but it has to meet up again at the back, so it speeds up. Faster air means lower pressure, so the wing gets pulled up."
Let's line up three simple questions.
1. Why would the air have to meet up again? The air above and below hasn't made any appointment to reunite at the trailing edge.
2. Then why does a paper plane fly? Paper is flat — the top and bottom are the same length.
3. Why can stunt planes fly upside down? Flip the plane over and the bulging surface is now on the bottom.
The standard explanation can't answer any of these. That's because its basic premise is wrong.
When you actually film the airflow, the air passing over the top arrives before the air underneath. There's no rendezvous. And the real difference in speed is even bigger than the standard explanation would predict.
A wing bends the passing air downward. As a reaction to pushing the air down, the wing is pushed up. That's what lift actually is.
Here's the important part: the pressure explanation and this explanation don't contradict each other. They're just two ways of describing the same thing. Let's go through it step by step.
What the wing is doing — throwing air downward
Picture catching water from a hose in your palm and deflecting it downward. The instant you redirect the water, your hand feels an upward kick.
That's exactly what a wing does. It takes the air coming from the front and sends it out the back, angled downward. Air has weight too (about 1.2 kilograms per cubic metre). An airliner ends up bending several tonnes of air downward every single second.
Changing the direction of something takes force. To bend the air downward, the wing pushes the air down. Push something, and it pushes back. That's the upward force — lift.
This view answers all three questions from the start of the article.
- Paper planes: even a flat sheet can bend air downward if it's tilted slightly. That's why it flies.
- Inverted flight: even upside down, raising the nose to create a tilt still bends air downward.
- The curved wing shape: not essential. It's a shape optimised for efficiency, not a requirement for flight.
The reaction to that is what keeps the plane up.
So is the pressure explanation wrong?
This is where people get confused. The pressure explanation is correct. The pressure above the wing really is lower, and below it really is higher. You can measure it directly.
What was wrong was the reasoning for why the flow above speeds up. The parts about "a longer path" and "meeting at the same time" simply weren't true.
So what's the correct picture? A bending flow and a pressure difference are inseparable — they're one and the same thing.
Think of a car going round a curve. Turning requires a force pointing toward the inside of the curve. Air is the same: wherever the flow curves, the pressure is lower on the side toward the centre of the curve. Above the wing, the flow curves downward following the wing's surface, so the pressure drops on the wing's side (the inside of the curve).
In other words:
A way of describing what happens to the air as a whole, outside the wing. Explained through the exchange of momentum.
A way of describing what's pushing on the wing's surface. Explained through the pressure distribution.
A and B are just the same phenomenon viewed from different places. It's not a question of which one is right. In fact, if you add up all the pressure acting on the wing's surface, it matches exactly the size of the reaction force from bending the air.
A few years ago, an article suggesting that "no one can explain why planes fly" went viral, and this phrase spread widely. That's not accurate either.
Lift can be calculated using theory that has been established for over a hundred years, and its predictions match measurements well. The fact that planes fly exactly as designed is itself the proof. What isn't settled isn't "why planes fly" — it's "how to say it correctly in one sentence."
Trying to keep an explanation short always means cutting something out, and misunderstandings grow from whatever gets cut. This article is no exception. That's exactly why we started by clearing away the errors first.
When flight stops — stalling
If "the more you bend the air down, the more lift you get" is true, then tilting the wing more and more should keep helping. And up to a point, it does.
But tilt it too far, and at some point the air separates from the wing's surface. Instead of curving along the wing, it breaks into turbulent eddies behind it. It can no longer be bent, so lift is lost abruptly. This is called a stall.
The key point is that a stall isn't caused by flying slowly — it's caused by tilting too far. Even flying fast, a sudden sharp pull-up of the nose can cause a stall. Recovering from this is one of the very first things taught in flight training.
Something you can check at home
- Hold a sheet of A4 paper just below your lower lip, letting it hang down in front of you
- Blow a strong, straight breath along the top surface of the paper
- The drooping paper lifts up
- Next, hold the paper flat and tilt it slightly, then blow air at it from the front using a fan or similar. Change the tilt and feel with your hand how the paper tries to rise
- Tilt it too far, and at some point it suddenly stops rising. That's a stall
Step 2 is often shown as an example of "faster flow on top means lift," but what's actually happening is that your breath is bending downward, following the shape of the paper. Change the angle at which you hold the paper, and the way it lifts changes noticeably. You can feel for yourself that angle matters more than shape.
Summary
Planes fly because the wing bends air downward, and is pushed back as a reaction. The pressure difference above and below happens at the same time, but it's just another way of looking at the same phenomenon. The only thing that was wrong was the premise that "air above and below meets at the back at the same time."
The correct explanation doesn't start with "the shape of the wing,"
but with "how much the air was bent downward."
As a byproduct of bending air downward, vortices form at the wingtips. Migrating birds use these vortices to save energy by flying in V formation. You can read more about this in this article.
For readers who want more — terminology, equations, and links to the school curriculumLabels show whether a topic ranges from middle-school science to active research
- MSCovered in middle-school science
- HSCovered in "Basic Physics" in high school
- HS+Covered in high-school "Physics," or treated as advanced/sidebar material in textbooks
- UnivNot taught in high school — content from university-level specialised courses (fluid dynamics, aeronautical engineering)
- ResearchNot even taught as settled fact at university — topics researchers are actively investigating
MSTerminology: words around the wing
- Lift: the force that holds a plane up. A plane also experiences thrust (pushing it forward), drag (pulling it back), and gravity (pulling it down).
- Angle of attack: how much the wing is tilted relative to the oncoming airflow. The single most important factor determining the amount of lift.
- Downwash: the air flowing diagonally downward behind the wing. This is the result of "bending the air downward" described in the main text.
- Stall: when the angle of attack is too large, the flow separates from the surface and lift drops sharply. Determined by angle, not speed.
- Airfoil: the cross-sectional shape of a wing. Not essential for flight, but engineered to reduce drag and resist stalling.
HSWorking it out with equations: how fast must you go to lift off?
Explaining in words why planes fly can lead to endless argument. But "how fast, in km/h, must you go to lift off?" can be answered by setting up a single equation and solving it.
L = ½ ρ v² S C
| L Lift (the upward force) | Unit: N (newtons) |
| ρ Air density | About 1.2 kg/m³ at ground level |
| v Speed | Unit: m/s |
| S Wing area | Unit: m² |
| C Wing effectiveness | A number set by shape and tilt. Around 1 |
The most important term is v². Speed matters squared. Everything else is a simple multiplication, so doubling the wing area only doubles the lift, but doubling the speed quadruples the lift.
| Aircraft mass | Take it as 1000 kg |
| Force needed to lift off (weight) | 1000 × 9.8 = 9800 N |
| Wing area | S = 16 m² |
| Wing effectiveness | Take C = 1.0 |
| Solve the equation for v² | v² = L ÷ (½ ρ S C) |
| Calculate the denominator first | 0.5 × 1.2 × 16 = 9.6 |
| Divide | 9800 ÷ 9.6 ≈ 1021 |
| Find the square root | 32 × 32 = 1024, so about 32 m/s |
| Convert to km/h | 32 × 3.6 ≈ 115 km/h |
Without reaching 115 km/h, it won't lift off. That's why a runway is needed. This calculation shows that the order of causation is "it flies because it reached that speed," not "it flies because the wing has a special shape."
During landing, speed drops. But because v² matters, lift falls off sharply as speed drops.
| Halve the speed | 2 × 2 = 4, so lift drops to a quarter |
| To keep enough lift | S or C needs to increase almost fourfold |
That's why panels extend out from the back of the wing during landing. Flaps increase both S (area) and C (effectiveness) at once, to make up for the drop in v². The same thing happens on takeoff, to gain lift over a short distance.
Once you look at flaps as "which letter in the equation is being boosted to balance the books," that moving panel starts to mean something different.
Lift can also be explained as the reaction to pushing air downward. Force = change in momentum per second. If a wing bends a mass m of air downward at speed w every second, the upward force on the wing is m w.
| Mass of a large airliner | Take it as 200,000 kg (200 tonnes) |
| Lift required | 200,000 × 9.8 = 1,960,000 N |
| If bending air down at 5 m/s | 1,960,000 ÷ 5 = 392,000 kg |
| Air bent downward per second | about 390 tonnes |
The plane is pushing down a mass of air far heavier than the aircraft itself, every single second. This shows the plane isn't so much "floating" as it is "continuously pushing." It's also why the airspace right behind a plane is dangerous — that downwash and its vortices linger.
* This is a simplified estimate meant to give a sense of scale. The real airflow around a wing is far more complex than this.
HS+How far does Bernoulli's principle actually hold?
Bernoulli's principle, p + ½ρv² = constant, holds along a single streamline, when viscosity and energy losses can be ignored. This condition is broadly satisfied in the flow around a wing.
So the relationship "faster flow means lower pressure" is itself correct, and the pressure above a wing really is lower. The problem is getting the direction of cause and effect backwards. It's not that "the path is longer, so it speeds up" — rather, the shape of the entire flow around the wing is determined all at once, and both the speed distribution and the pressure distribution fall out of that together.
Another key point is the pressure gradient across a curving flow. When a flow curves with radius of curvature R, a pressure difference of ∂p/∂n ≈ ρv²/R is needed, pointing toward the centre of the curve. Above the wing, the flow curves toward the wing, so the pressure drops the closer you get to the wing's surface. "The flow is curving" and "the pressure is low" are two faces of the same fact.
UnivCirculation, and viscosity's surprising role
The framework for quantifying lift is the Kutta–Joukowski theorem. Using the circulation Γ around the wing, the lift per unit span can be written as L' = ρ V Γ.
The catch is that theory alone can't fix the value of the circulation. For the same wing shape, mathematically you could construct a solution for almost any circulation value. What pins it down uniquely is the Kutta condition, which requires that the flow leaves the trailing edge of the wing smoothly.
And the reason this condition holds in reality comes down to the viscosity of air. In an idealised fluid with no viscosity at all, flow would be allowed to curl around the trailing edge, and both lift and drag would come out as zero (d'Alembert's paradox). Paradoxically, then, even in fast flows where "ignoring viscosity" is normally a fair approximation, the very existence of lift depends on viscosity.
Viscosity only really matters in an extremely thin layer at the wing's surface (the boundary layer), and what happens there governs the behaviour of the whole wing. A stall happens when this boundary layer separates from the surface.
ResearchWhat's still hard, even today
- Turbulence calculations still rely on approximations. The equations governing fluid flow (the Navier–Stokes equations) can be written down, but solving for the full turbulent flow around a real aircraft directly is computationally impossible. In practice, the effects of turbulence are replaced with models, and the choice of model changes the results. Predicting the region just before a stall, or where flow separation occurs, still depends heavily on wind-tunnel and flight testing.
- Whether these equations even always have a solution is unresolved. The existence and smoothness of solutions to the Navier–Stokes equations remains an unsolved problem in mathematics, with a prize still on offer. The basic equations behind a phenomenon that tens of thousands of planes rely on every day are still not fully understood mathematically.
- How insects fly couldn't be explained by aircraft theory. For insects like bumblebees, calculations based on steady-wing theory fall short of the lift they actually need. In the 1990s, it was discovered that a vortex forming at the leading edge during flapping greatly boosts lift, which changed the picture. The famous claim that "bumblebees shouldn't be able to fly" was really a case of the theory falling short, not the insect defying physics. This finding is now being applied to the design of small flying vehicles.
- Even "what's the best way to teach this" isn't settled. Education researchers have repeatedly pointed out that the flawed explanation has lingered for a long time in textbooks and museum displays. Proposals and debate continue over how to explain lift in a way that's both accurate and easy to understand.
Links to the school curriculum (by level)
| Level | Subject/unit | Where in this article |
|---|---|---|
| MS | Science: balance of forces / action-reaction / the weight of air | Bending air downward, and its reaction |
| HS | Basic Physics: laws of motion / Physics: momentum | Force = change in momentum, estimating lift |
| HS+ | Physics: fluids (advanced material in most textbooks) | Conditions for Bernoulli's principle, curvature and pressure gradient |
| Univ | Fluid dynamics / aeronautical engineering | Circulation, the Kutta condition, boundary layers, d'Alembert's paradox |
| Research | Computational fluid dynamics / biological fluid dynamics (unresolved) | Turbulence models, stall prediction, the equations' mathematical open problems, insect flight |
- Anderson, J. D., Fundamentals of Aerodynamics (a standard aerodynamics textbook, including the refutation of the equal-transit-time theory).
- Babinsky, H., How do wings work?, Physics Education 38(6), 497–503, 2003 (a survey of the incorrect explanation and an account based on the curving of the flow).
- McLean, D., Understanding Aerodynamics: Arguing from the Real Physics, 2012 (a detailed discussion of the equivalence between pressure-based and momentum-based explanations).
- Ellington, C. P. et al., Leading-edge vortices in insect flight, Nature 384, 626–630, 1996 (insect flight).
- NASA Glenn Research Center, Incorrect Lift Theory (a page explaining the common incorrect explanation).
* The figures used in the estimates are simplified approximations meant to give a sense of scale, and differ from actual engineering design calculations.
* This article is a general-audience science explainer. The figures given are meant to aid understanding of how things work, and are not standards for actual aircraft design or operation. For matters relating to aviation, please consult the relevant aviation authorities and operators.