Software and Computer Engineering

How Can We Know Anything?

Também disponível em português.

In a world with increasing polarization and wide spread of misinformation, it is crucial to question oneself about what is true, what is not true. If we take one further step back, we may also ask the question: how can we know anything?.

The Case For Epistemology

This seemingly obvious question is the central subject of the branch of Philosophy called Epistemology.

Now, I know that Philosophy get the bad wrap of being “the field which deals with abstract, inconsequential, or obvious questions”, but hear me out: If an individual acts according to one's own beliefs and their own understanding of the world, wouldn't it be of their best interest that their beliefs are true, as in, which more accurately reflect the reality of the world?

If you agree this is the case, there's not only room for questioning one's own beliefs, but also -- and, I argue, more importantly -- the process at which one has arrived at their own set of beliefs.

When pondering about how beliefs are formed, it might be helpful to consider how one acquires their own first beliefs.

Our First Beliefs

At a very young age we are, very honestly, hopeless. Children only survive to adult age because their parents or guardians feed them and provide shelter and protection.

Young age is also the time that one first starts learning about and developing their beliefs about how the world works: What should be pursued? What should be avoided? What are the dangers and rewards involved?

As as a child lacks the resources, time -- and the even very understanding of their limitations -- getting to know things quickly becomes not only important, but instrumental to survival.

During childhood, beliefs are mostly bestowed by their parents or guardians through analogy and second-hand experience. Let's dive into those two kinds.

In this article we will explore the merits and limitations of the first kind.

Merits of Analogy

Analogy is such a powerful source of knowledge.

Take, for example, language itself. The first language that an individual ever learns is mainly through analogy. There is no one actively telling them how to talk -- as the very linguistic means to allow for that haven't been developed yet. Nonetheless, a child is able to learn this complex ability through listening and experimentation.

Analogy is clearly observed in the world at large. People will “copy” what they perceive are strategies that have worked for other people, while adapting the details to their own circumstances. This is a very successful source of knowledge, as doing what has worked for others, more often than not, yields similar results for oneself.

Limitations of Analogy

While very powerful, analogy has a fundamental limitation: One can only learn from, well, analogous, situations. If ever presented with unexpected or novel conditions, the learned analogous beliefs and beliefs may be of little help. Let's illustrate this with an example from History

In the 18th century, the then scientists in the Holy Roman Empire -- and later in Paris, France --, were attempting why some things were hot to the touch, while others were cold. The explanation was drawn from an analogy with fluids.

They proposed that all objects contained some undetectable fluid, called “calorific”. On contact, this fluid would flow from objects with much of it (hot objects) towards objects with less of it (cold objects), thus making the hot object colder, and the cold object hotter.

This theory explained many observations at the time, but was ultimately incomplete. The modern Thermodynamic theory is based on mechanics and “jiggling” of atoms, and this newer model explains even more observations in a simpler manner consistent with contemporary Physics.

The point is not to demerit the beliefs of the past: these scholars arrived at conclusions with the tools and knowledge that were available to them at the time. The point is that modern science has succeeded at a higher degree by giving up on the analogy as the starting point.