Almost everyone who has been told how a wing works has been told something wrong.

The explanation goes like this: the top of the wing is curved and the bottom is flat, so air going over the top has further to travel, so it must go faster to meet up with the air underneath, and faster air has lower pressure, so the wing is sucked upwards. It appears in textbooks, museum displays and a great many flight-training courses.

NASA has a page dedicated to explaining why it is wrong. The page is called “Incorrect Lift Theory”.

The four forces

Start with the simple part, which is genuinely simple. Four forces act on an aircraft in flight.

Peso pulls toward the centre of the Earth and acts through the centre of gravity. Lift acts perpendicular to the direction of flight, through the centre of pressure. Drag acts directly against the direction of flight, also through the centre of pressure. Spinta acts in whatever direction the engines are pointed.

If they balance, the aircraft continues at constant velocity. If they do not, it accelerates in the direction of the largest force. That is Newton’s second law and there is nothing mysterious about it.

Note one thing that trips people up: lift is defined relative to the flight path, not to the ground. In a steep turn, lift is not pointing upward. That is precisely why aircraft descend in turns unless the pilot does something about it.

Why the equal transit story fails

NASA gives three separate reasons, and any one of them is fatal.

Symmetric wings make lift. A flat plate makes lift. A paper aeroplane has no curvature at all. If the upper surface having a longer path were the mechanism, none of these could work. Some modern low-drag airfoils actually have a longer bottom surface. And aircraft fly inverted, which under the equal-transit story should press them into the ground.

The premise is made up. There is no physical reason why two air molecules that separate at the leading edge must arrive at the trailing edge together. NASA calls it a non-physical assumption. Cambridge University filmed the experiment: pulses of smoke released upstream of a wing show that the air going over the top arrives at the trailing edge first, well ahead of the air underneath.

The numbers do not work. Even if you accept the premise, it gives the wrong answer.

It is worth being precise about what is wrong here, because the correction is often overstated in the other direction. Bernoulli’s principle is not wrong. NASA says so explicitly. Faster-moving air really does have lower pressure. The error is in how the equal-transit story works out the speed.

The Cambridge smoke-pulse demonstration. Watch where the two pulses are when the upper one reaches the trailing edge.

What actually happens

A wing makes lift by turning the air. That is the whole of it.

The wing deflects the oncoming flow downward, and by Newton’s third law the air pushes the wing upward by an equal and opposite amount. The downward-moving air behind a wing is called downwash, and you can see it: it is what makes helicopters kick up dust and what makes the wake behind a large aircraft dangerous.

Crucially, both surfaces do the turning. The underside deflects air downward in the obvious way. But the upper surface also turns the flow downward, by curving it around and back down toward the trailing edge, and on a normal wing at a normal angle the upper surface does more of the work than the lower one. Leaving out the upper surface produces the other classic wrong explanation, the one where the wing works like a skipping stone.

Pressure and velocity still matter. The air near the upper surface does move faster and its pressure is lower, and integrating that pressure distribution around the wing gives you the lift. Bernoulli is a perfectly good way to calculate the answer. It just is not the cause. The velocity differences exist because the flow is being turned.

Angle of attack does the steering

If lift comes from turning the flow, then the amount of turning is what controls the lift. That is the angle of attack: the angle between the wing’s chord line and the oncoming air.

Increase it and you turn more air, so you get more lift. The relationship is very nearly a straight line up to around ten degrees, which is why thin airfoil theory can say that the lift coefficient is roughly 2π times the angle of attack in radians and be usefully accurate.

Then the flow stops following the upper surface, separates, and the lift collapses. That is the stall, and predicting exactly when it happens is not something anyone can do reliably with equations alone. NASA’s own guidance is that it is very difficult mathematically and that engineers rely on wind tunnels. That is why the photograph above exists.

A concise summary of the modern explanation, without the equal-transit myth.

So why does the wrong version survive?

Because it is memorable, it involves a shape you can draw, and it produces a number that is in roughly the right direction. It gives people the feeling of understanding without requiring them to think about momentum.

Babinsky has described starting lectures by giving the wrong explanation and asking who has heard it, and watching almost every hand in the room go up. That is not a failure of the audience. It is what happens when a tidy story outcompetes an accurate one for a century.

The correct version is not much harder: a wing pushes air down, and the air pushes the wing up. Everything else, the pressure distribution, the curvature, the angle of attack, is detail about how efficiently it manages to do that.

Sources: NASA Glenn Research Center, Beginner’s Guide to Aeronautics, “Incorrect Lift Theory”, “Lift from Flow Turning”, “The Four Forces” and “Inclination Effects on Lift”, by Tom Benson; University of Cambridge, “How wings really work”, 25 January 2012.