Philosophy of Model Based Physics

One of the most frustrating parts of my education was how often I was told what I had previously learned was wrong.

"Molecules are the most fundamental particle... now it's atoms... now it's protons, neutrons, and electrons... now it's quarks, leptons, and bosons."

"Light behaves like particles... and also now like waves... light and electricity are both parts of the electromagnetic field."

"Energy is ALWAYS conserved... except as objects fly through the expanding universe."

It seemed strange to me to start teaching science with less accurate outdated concepts and only then cascade into more precise descriptions. Wouldn't it be easier to start with the most fundamental concepts and build up? We don't start building a house by putting up the roof.

The main problem? We haven't the slightest idea what the true foundation of the universe is!

Over the course of history, humans have been forced to learn science through their five senses (sight, sound, taste, touch, smell) and via machines they build (which then conveys information to one of senses).

Much of this process is like trying to understand what makes up a house without any prior tools or knowledge. There are walls, but what lies within them? There are floors but what lies under them? There are many parts of a house that will hurt you if you don't treat them with care.

There are still a lot of questions about the universe we don't understand today... so where do we start when we want to teach? We really can only start with our five senses and observe, model, predict, and test.

When we observe, the knowledge we use to predict an outcome is called a model. When we test our predictions, we learn whether our model is a good one.

Suppose that we observe sticking our arm into a nearby river causes it to feel extremely uncomfortable... a feeling we associate with cold shivery nights. We might create a model, which says that all rivers will cause us to feel this way. Repeated observations of this river (and even nearby rivers) confirm this model, but one day, we find a far away river that results in no such discomfort!

Our model, which we initially thought applied to all rivers was wrong. This is the way of science. To obtain a better model, we must understand why some rivers are different than others... HOWEVER, our initial model remains sufficient for describing all nearby rivers and if we were not to navigate far from home, any inclusion of far away rivers into our model would needlessly complicate our day to day life! Less accurate models are still successful if they provide helpful predictions.

Over the centuries, humanity has developed many models... many of which provide helpful predictions even amidst inaccuracy. Within this blog, I will be teaching many of them. With this discussion, we have hopefully broken an impulse to reject models that have any inaccuracies and instead weigh them based on the situational value they provide.

As we go, each model will be justified by either direct observations or experimental evidence. Attempts will also be made to highlight a model's weaknesses.