Are models necessary to learn about reality?

Reality is complex. Social or Natural. Complex system comprises of two phenomenon: randomness/chaos and regularity/order. Just like our usual life, isn’t it? We end up to one end of either chaos or order, we mess our lives. Reality is not messed up. It hangs in between.

There are infinitely many things to see and learn from. When you see your friend’s face, you are skipping trillions of air molecules, their waves pattern, atoms within them, their interplay, mosaic of messy lights in different wage lengths, billions of microbes hanging in between you and your friend (even in your own eyes). Even when we are able to see things, we do no see them completely or accurately or directly. Unfortunately, we cannot claim with 100% certainty, given the scientific methods we have, that there is no Chinese tea pot somewhere in the asteroid belt.

Of course we cannot “see” (I mean observe) everything. To know everything, we need to be able to see everything. Without seeing, our knowledge about reality will be at the best incomplete.

The biggest epistemological tragedy is that even when we can observe things completely and directly learning can be tricky, messy and misleading. Mere telescopes never guarantee reliability of inference we draw from the data on the movements of Mars. Observation is just the beginning of  epistemological journey. The enlightenment gave us not just the tools of observation, but the severe methodological pathway from observations towards truth.

How do we know that elephants can fly when we see an elephant flying? How many elephants need to fly before our eyes to ensure the conclusion that elephants can really fly? How many eye witness is necessary (or even sufficient) to make the claim? Can we interpret the data/observations we have on its own?  Can we analyze things in its entirety without any control and deletion of reality.

Take an example of a map. Can we study our expected journey between two points on the face of earth without ignoring infinitely many granular things that the reality has to show us? How can we focus on our journey when there are wild lives on the sides of the highway, people talking in different languages, different colors it offers to our eyes, shops, sidewalks, air temperature and movements and what not on and around the highway.

We need to focus and get rid of irrelevant (with respect to what we want to learn about) information from the reality. Of course, we cannot get rid of them in reality. So we get rid of them inside our brain in a simulated world. This simulated world where most of the reality is ignored is a model. Just like a shining and slim dummies in showrooms.

Dummies in a showroom are models of reality

Though the dummies are not exact replica of our bodies, they help us make decision with regard to the fitness of clothes hanging on their lifeless bodies. We can imagine how well the particular color fits our skin. The dummies are not exact but they are adequate enough to make decisions.

But the same model might not be adequate enough if we want to understand the anatomy of human body. We might need much detailed model to adequately explain the real anatomy we observe in human bodies. So, models need to be adapted and changed as our horizon of knowledge broadens. This epistemological fact has been frequently used to attack science saying, “science keeps changing truth so is unreliable” or “science has been proven wrong many times” every time scientists develop a new model. Truth is not changing. Truth is one. But the model we use is changing and improving. Some models like Newton’s law of motion are still being used although much better alternatives are available because the model is still adequate to learn about planetary motion and trajectory of a baseball. It does not mean Newton was proved wrong.

Much better models of human body is available, but showrooms still use the same faceless, limbless, or headless dummies without any counterparts to internal organs.

Same thing can be said about google maps. The google map is not the exact copy of the surface of the planet, but they are adequate enough to infer about the journey we want to make and distance we want to cover in our day to day life. If instead, we have the exact copy of the surface of the earth (which is the earth itself), we will not be able to observe anything and the copy becomes useless.

Flat earth model has been proven to be inadequate to explain what we observe as the earth

Models are not only useful but also necessary humanity have invented to learn about the reality. When models can adequately explain the reality and help us make decisions, models are useful and becomes part of our knowledge bag. When models fail to reflect the reality. We reject them. Flat earth model is totally inadequate to explain what we observe as the earth. But the spherical model, though still not exact, is adequate enough to explain what we observe. We can accomplish many learning goals with this model. So we make it part of our understanding of planet earth.

But again, new models of earth are being developed, has been developed. But the spherical model is enough to compute the surface area and mass accurately enough. If not, we use better detailed models.

Economists use models all the time. All claims, forecasts and prescriptions are based on models. Arguments for free market economy, low tax rates or high tax rates, deregulation or more regulation are all based on models. The effectiveness of prescriptions cannot be judged unless we judge the adequacy of the models.

Models are nothing more than assumptions about the reality we observe. If the assumptions are valid (adequate enough), the inference we will draw are reliable. Less assumptions means more complex models. More assumptions means simpler models. No assumptions means we have the exact replica of the reality which is useless as a learning tool. The models have to be simple but not simpler.

Ironically, we cannot learn about the reality if we have all the information of the same reality in a single frame. Holistic approach in its broadest sense is not an approach at all. Only partial understanding can lead us towards full understanding.