If you’ve been following along, you know I’ve been pursuing the topic of critical thinking – the analysis of how we think and how we think about thinking in a rigorous fashion – and dragging you along with me. We had a recent review for Aristotle’s Organon, we posted about teaching critical thinking, and we examined normative and descriptive judgements and our tendencies to apply those techniques without realizing it. I’ve come to think that there are three major prongs to this concept of critical thinking and its effective implementation. First, there is understanding how we think and why we think in that way. That is the role of investigations like in the paper behind our post on normative and descriptive judgements, and of examinations of logical fallacies, even imperfect ones. Second is the kind of nuts-and-bolts approach to logic that Aristotle provides in the Organon: syllogism, dialectic, enthymeme, paradeigma, et cetera. These are the tools with which to improve our thinking, to put systems in place that help mitigate our flawed natural thought-processes.

The third prong takes a step back and looks at how we know things. Socrates told us that the beginning of wisdom is in knowing we know nothing, but that’s rather a shortcut from thinking deeply about knowing and not knowing. The study of how we know is known as epistemology, and its extensive philosophical tradition offers three main schools of thought that each have innumerable permutations presented by every philosopher who has ever philosophized: skepticism, rationalism, and empiricism. Each has had its heyday of dominance, each has waxed and waned, but together they summarize how we know.

In the modern, colloquial sense of the word, skepticism is thought of like reasoned doubt. It is in this sense that I deploy it to describe science as a “discipline of skepticism,” in which we begin from a state of ignorance and must find evidence to reduce that ignorance. Epistemologically, however, the school of skepticism posits that it is impossible to know anything. This is the truest to Socrates’ quote, and was arguably the dominant school of thought for the ancient Greeks who give us the foundation of so much in these fields, which is why they were so focused on asking questions and less so on answering them. Skepticism in the formal sense claims that since our sense are imperfect means of approaching the world, and are minds are similarly fallible, we can never truly know anything. To a skeptic, it is not that something must be proven thoroughly before it is known – it is that nothing can ever be known.

We could end it there. Skepticism’s claims resonate with me, this idea that every means we have for probing reality are fundamentally flawed and filtered throughout our horribly unreliable brains. It’s also a path of futility, though, and when I start thinking too skeptically, I remind myself of how I philosophize around the idea of a simulated universe. There’s a theory that our whole existence, our whole universe, is just a hyper-accurate computer simulation on some alien’s laptop. My response to this claim is that it doesn’t matter. If it is truly impossible to perceive any flaw, and indication that would separate what we think of as being “reality” from this perfect simulation, then is there a difference? What would then make that simulated reality any less real? That’s how I avoid falling into what I term the skepticism trap – by saying that if we can know things as well as possible with the means at our disposal, that is enough to claim we know, even if it is never able to approach some fundamental truth.

That brings us to the rationalist school of knowing, which posits that we can know things only through the might of human reason. This was the dominant school of thought during the Middle Ages in Europe, and is a whole concept of knowing built around the same idea as the syllogism. Granted, the syllogism is a huge idea – just read Prior Analytics. In rationalism, we approach knowledge by starting from a basic premise which all can accept as self-evident, and derive a conclusion from there through application of a reasoning framework, like syllogism or dialectic. The famous example, of course, is starting with the premises “Socrates is a man,” and “man is an animal,” and therefore concluding that “Socrates is an animal.” There are many permutations of the concept, but that is the core of rationalism, this idea that we can know by pure reason.

There are, as you can probably imagine, some serious flaws in this approaching to knowing, and if you can’t imagine, I suggest that you read some of the theological and philosophical works that came out of Medieval Europe, which use rationalist techniques to arrive at some patently bizarre conclusions. Some of Martin Luther’s writings demonstrate the flaws quite nicely, for instance. It’s a bit like building a house of cards on a foundation of sand; your foundation isn’t necessarily stable, and neither is your construction material, but when the conditions are right you can build something quite spectacular. You must somehow know a premise from which to begin your reasoning, which rationalism does not quite address, and the process of building from there involves additional assumptions that are not fully supported.

Not that empiricism is a perfect answer, either. Empiricism is the dominant theory of knowing today, and it will sound familiar to anyone familiar with the basic tenets of the scientific method; it is all about gaining knowledge through the senses, through taking measurements of reality and drawing conclusions based on those measurements, whether those measurements are taken with our personal senses or with space telescopes or particle accelerators. In other words, empiricism is knowing through data, and our modern world loves data. It is obsessed with it. We generate enormous amounts of information, we measure everything, and we record it all in giant databases to perform analyses upon in order to bring us closer to an understanding of reality, in order to know.

Except, of course, that data is flawed, and so are the analyses. It is still collected by flawed human beings and analyzed with our squishy, imperfect brains. There are inevitably biases unaccounted for, variables uncontrolled for, confounding factors unnoticed, patterns imagined where they don’t exist, and that’s not even touching the issues of statistics which are studied in Bernoulli’s Fallacy. In a world of computers empiricism is treated like a panacea, or maybe a religion, but it is as flawed as its fellows, rationalism and skepticism. The flaws are just harder for us to perceive because our bias is towards empiricism.

At this point, you can probably guess where this discussion is leading. If no single epistemological theory is adequate to answer the question of how we know, then we must combine them. Combine the constant questioning and doubt of skepticism with the strict, regimented logic of rationalism and the data-driven grounding in reality of empiricism and you can mitigate many of the flaws in each. It is no surprise, then, that some of the greatest triumphs of our understanding of the universe combine these schools of knowing: Einstein’s original papers on relativity, for instance, are more an exercise in rationalism than empiricism (as much fundamental physics is), the conclusions in which have since been supported and vetted through empirical approaches…and even then, we must allow for doubt, the possibility that there is a better theory out there that will explain, for instance, dark energy without invoking mysterious substances we cannot detect.

Skepticism may be right. We may never be able to fully know anything in a fundamental way…but that’s okay. Because, as I am so fond of saying, it’s not about finding the right answers; it’s about starting with wrong answers and pursuing less wrong answers. Just because we can’t know Truth with a capital “T” doesn’t mean we can’t know anything at all. We can have knowledge, even if it is flawed, by combining these approaches to knowing. How do I know? I don’t. It’s not the Answer. But it might just be a less wrong answer.