[AstralStorm]

Correct. Still incomplete. Angle of attack causes a vortex at the trailing edge which has nothing to do with raw air speed and everything to do with fluid dynamics (which involves speed but is much more complex)

Short version is that you created a hole (lower pressure area) in air which it now tries to fill. Air and gasses have finite limited velocity known as speed of sound, which is why you get these pressure differentials while the wing is moving. With a flat wing, they're rather small and low pressure vortex is located behind the wing. In an angled wing, some of it is located below the wing and the air trying to fill the low pressure area exerts a lift force on the wing. (It's unlike a balloon. Bernoulli has very limited impact, unlike essentially wind.)

[jameshart]

Isn't the intent of a smooth aerofoil design to prevent the formation of vortices on the trailing edge? They're inevitable at the wingtip, but in controlled flight most wings are trying to produce laminar flow, right?

In my understanding, if you increase angle of attack sufficiently to generate vortices on the upper surface, then you aren't efficiently transferring downward momentum to the air your wing is shedding, and you lose lift, which causes aerodynamic stall. Am I missing something?

[tomsto]

I’m not sure I fully agree. Do you not get this trailing edge vortex with an asymmetrical aerofoil at 0 angle of attack? (Just less strongly because less pressure difference between suction and pressure sides)

[AstralStorm]

You do. And you get small vortices on wing surfaces too.