Someone's been reading Aerodynamics for Naval Aviators, haven't they?

The notion that the structure will stall before damage occurs is inherent in aircraft design: the gist is that there is a value at which structural damage won't occur under a given gust load or structural loading imposed by say, full control deflection, or a gust load imposed at a particular speed. That criterial is applied to many elements of design and certification. However, with respect to the end user, pilots, considering Va, or maneuvering speed, pilots were taught for many years that it was the speed below which a full control deflection would result in airflow separation (stall) before a damaging load could be imposed. The problem is that such an explanation is too simplistic, and it's wrong.

One can break an airplane below Va, because when we consider a fullly deflected control surface, for example, we consider the singular load imposed by that isolated event. Pull back on the control stick, raise the elevator, load the elevator surface and hinges, stab attach points and stab, and wing, wing spars, attach points, etc, and in theory...it won't break. But that's just one pull. In that loading, there's bending and flexing, torsion, compression, and tension, and when we apply a second load, say for example, to reverse the control input, we don't simply load in the opposite direction. We impose a series of unintended loads that are not explained by the purpose of Va...more than one control input, especially opposite or reverse inputs, can impose loads that do exceed the structural integrity of the aircraft, causing damage or failure.

The crash of American Airlines 587 in Jamaica Bay (New York) was an eye-opener to some, who believed the myth that the aircraft couldn't be damaged at low speeds. It can. Multiple control inputs, alternating inputs and rapid inputs alter the equation and the complexity of the load, and exceed or act outside the established limitation and the so-called protection of operating under that speed.