Concept Clarification 2: 'Why?'

In this concept clarification post, I wanted to tackle something part pragmatic, part philosophical. Because of this, don't take this as more than my thoughts on this rather detailed topic.

That something is the question 'Why does X happen?'.


It isn't uncommon to see media headlines akin to 'Scientist show X causes Y!', particularly when X or Y are really flashy things: death, cancer, global warming, autism, etc. Jorge Cham, creator of Ph.D. Comics, covered the evolution of a scientific finding to a mass media headline far more eloquently than I ever could here.

But the basic idea of people wanting to know 'why' something happens is constant. It's certainly appealing to ask, too: just talk to a child for a bit. The trouble is, 'why' is an ill-defined question. In science and engineering, we don't really ask 'why'. If anything 'how' is a much closer approximation to the question generally being asked. The reason is simple: for physical systems, there is really no bottom to the question of 'why', you can continue to ask it infinitely (or very nearly). This is much the same as how a preschooler can continue to ask 'why' well after a parent has lost their patience for answering.

Let's take a 'simple' example: 'Why does this bar bend when I push on it?'

• A mechanical engineer may answer something like: 'Because that's the bar's response to an applied load.'
• But a materials scientist may say something like: 'Because of the relaxation of the crystal structure to accommodate the change in energy that results from the work exerted on the system.'
• Even further down, a solid-state physicist may say: 'Because the changing positions of adjacent nuclei influences aspects of the electronic structure and this is reflected in the move towards more energetically favorable positions.'
• And at some point, a particle physicist would probably say something about quarks and gluons that no one would understand, but it would inevitably end in them wanting to smash things together at really, really high speeds to show what they mean.

The basic problem comes down to this: they are all right, at least in the reference frame in which they are dealing. A design engineer (and even mechanicians) are generally considering things at a much larger scale, in which many smaller scale phenomena can be abstracted or neglected. Meanwhile, a particle physicist isn't generally concerned with large-scale material behavior (on the order of many moles of atoms, a measure on the order of \(10^{23}\) atoms), so they are free to concern themselves with some details neglected by others.

But they are also wrong in that no one of those answers is the full answer. There's a connectivity between them that is missing. This connectivity is most readily seen in derivations of the mathematical relationships that describe physical phenomena. At some point, every field decides certain aspects can be neglected because they simply aren't important, and this comes out in terms which can disappear or entirely different equations. The neglect of quantum mechanical phenomena by traditional engineering disciplines, for example, because classical mechanics works well enough for their purposes. 

Now, there are times when the things traditionally neglected can't be neglected: such as fracture mechanics and radiation damage in materials for microelectronics (among other things): two of my areas of work during my Ph.D. But the answer to 'why' still must be qualified by who is doing the asking and how the answer is being obtained. This also means that the answer may not satisfy all who may ask the question.

The above is why I prefer to ask 'how': because it doesn't pretend to be the full answer to the question that 'why' asks, even if I phrase the answer to 'how' as 'why'.