This is something I just have to get off my chest. It’s been bugging me for a while now.
Physics is the endeavour to understand the physical world. Mathematics is a powerful tool employed in this endeavour. It often happens that specific mathematical procedures are developed for specific scenarios found in physics. These developments then often lead to dedicated mathematical methods, even special notations, that we call formalisms.
The idea of a formalism is that it makes life easier for us to investigate physical phenomena belonging to a specific field. An example is quantum mechanics. The basic formalism has been developed almost a hundred years ago. Since then, many people have investigated various sophisticated aspects of this formalism and placed it on a firm foundation. Books are dedicated to it and university courses are designed to teach students all the intricate details.
One can think of it almost like a kitchen appliance with a place to put in some ingredients, a handle to crank, and a slot at the bottom where the finished product will emerge once the process is completed. Beautiful!
So does this mean that we don’t need to understand what we are doing anymore? We simply need to put the initial conditions into the appropriate slot, the appropriate Hamiltonian into its special slot and crank away. The output should then be guaranteed to be the answer that we are looking for.
Well, it is like the old saying: garbage in, garbage out. If you don’t know what you are doing, you may be putting the wrong things in. The result would be a mess from which one cannot learn anything.
Actually, the situation is even more serious than this. For all the effort that has gone into developing the formalism (and I’m not only talking about quantum mechanics), it remains a human construct of what is happening in the real physical world. It inevitably still contains certain prejudices, left over as a legacy of the perspectives of the people that initially came up with it.
Take the example of quantum mechanics again. It is largely based on an operator called the Hamiltonian. As such, it displays a particular prejudice. It is manifestly non-relativistic. Moreover, it assumes that we know the initial state at a given time, for all space. We then use the Hamiltonian approach to evolve the state in time to see what one would get at some later point in time. But what if we know the initial state for all time, but not for all space and we want to know what the state looks like at other regions in space? An example of such a situation is found in the propagation of a quantum state through a random medium.
Those that are dead sold on the standard formal quantum mechanics procedure would try to convince you that the Hamiltonian formalism would still give you the right answer. Perhaps one can use some fancy manipulations of the input state in special cases to get situations where the Hamiltonian approach would work for this problem. However, even in such cases, the process becomes awkward and far from efficient. The result would also be difficult to interpret. But why would you want to do it this way, in the first place? Is it so important that we always use the established formalism?
Perhaps you think we have no choice, but that is not true. We understand enough of the fundamental physics to come up with an efficient mathematical model for the problem, even though the result would not be recognizable as the standard formalism. Did we become so lazy in our thoughts that we don’t want to employ our understanding of the fundamental physics anymore? Or did we lose our understanding of the basics to the point that we cannot do calculations unless we use the established formalism?
What would you rather sacrifice: the precise physical understanding or the established mathematical formalism? If you choose to sacrifice the former rather than the latter, then you are not a physicist, then you are a formalist! In physics, the physical understanding should always be paramount! The formalism is merely a tool with which we strive to increase our understanding. If the formalism is not appropriate for the problem, or does not present us with the most efficient way to do the computation, then by all means cast it aside without a second thought.
Focus on the physics, not on the formalism! There I’ve said it.