There is not a single, uniform, design philosophy, for airplane cockpits. The cockpit is a "user interface". Some are easy to use. Others are powerful but confusing for beginners. Some implement standards so that knowledge of one can be applied with any. Some are hodgepodges with no consistency within themselves. Some start with an expresable goal but fail to acheave it. Some are utilitarian objects with controls and indicators in a regular pattern or patterns without regard for how they are comprehended or operated. In many cases, one intent at the beginning of design is merged with other ideals, requirements, informal norms and the like. Cheap and simple to build, and easy to maintain for decades, are as real as imperatives as any other.
In a modern Boeing transport, pilot and copilot each have a semicircular 'wheel' that turns left and right, to deflect the ailerons and regulate rate of roll. The wheel is pulled back or pushed forward to deflect the elevators, which control the angle of attack of the airplane's wing, and thus, taking engine thrust into account, the vehicles's speed, if allowed to stabilize.
The wheels and columns carrying them are interconnected, so moving one control moves the other. Unless some mechanical mishap has locked one in place, in which case an intentionally breakable link can be broken, by force, allowing them to move independently.
Note that the same physical object produces two different kinds of effect, one a rate, with zero near the middle, one, a position, which physics causes to control a rate, with zero beyond one extreme, (speed) not intuitively linked to it. Speed is
also, very strongly, influenced by the engines' throttle position.
In a modern Airbus transport, a one-hand handle off on the outside of each seat pivots at it's base, seeming to "point" the airplane this way or that, but physics still separates the effect of roll, a rate, from the effect of pitch, a position, which, at equilibrium control speed, a rete, but not the intuitively obvious one. However, the captain's handle doesn't move the copilot's handle, so there's no visual or tactile link between the position of one control and the other. If each is moved in the opposite sense of the other, the result is zero roll rate and zero pitch angle.
In a General Dynamics (now Lockheed) F-16, there is a one-hand handle to the right of the pilot's seat, which does not move (in any significant way) but senses the direction the pilot is pushing it and moves the same control surfaces to accomplish the same pitch and roll control.
When 'blind' flying instruments for night and bad weather were invented, the authorities in Great Britain specified that all airplanes bought by the British government would have their blind flying instruments in a standard 2 rows by 3 columns layout. When deHavilland designed the Mosquito, the world's fastest production fighter-bomber at that time, they designed the instrument panel by mounting some other airplane's panel in the middle and adding additional switches, dials, etc, on additional panels left, right and below the "basic 6". The 'borrowed' panel is curved at the top to match the top of some fuselage, but not a Mosquito. Every Mosquito built has a panel with that same, mis-matched, curve. Kinda funny, once you know to look for it.
Cultural norms play a large role too. The USAF studied moving tape/fixed pointer indicators in the 1960s, concluded that pilots could operate more accurately with them than "steam gauge" round instruments. But they cost more because of smaller volumes, more moving parts, they don't fit in a simple, round hole in the panel, and they're "different". Modern glass cockpits mix simulated moving tapes with fixed needles and simulated dials with moving needles. The instruments for some of SpaceShip One's cockpit were horizontal graphics on a laptop computer, which may have been time histories with vertical amplitudes...