What is Rationality?

The Rationalist Movement, embodied by communities such as LessWrong and SlateStarCodex uses the term Rationality in a very particular way.  As such, it is useful to be able to express this meaning to people, when trying to describe what the movement is, and what it is trying to accomplish.  On a personal level, I have struggled greatly to adequately explain to other people not already familiar with the community and its values, the core of what Rationality is, and why they should care.  There are several places trying to explain this but all are quite long-winded:

What Do We Mean By “Rationality”? (1671 words)

Twelve Virtues of Rationality (2205 words)

Biases: An Introduction (1787 words)

Whilst all of these are perfectly reasonable, if either technical, mystical or both, they are difficult to boil down to something short and pithy.  Anything that I have previously read that is a short and pithy description of Rationality can either easily be woefully misinterpreted, resulting in “so Rationalists want us to all be emotionless robots then? Count me out!”, or is really a description of a small subset of the community, making it sound far more specific and exclusive than it actually is.  Another two articles that come remarkably close, doing a very good job of explaining what the Rationalist Movement is and is not are these on Putanumonit and SlateStarCodex:

The Secret Society for Suppressing Stupidity (3815 words)

Why I Am Not Rene Descartes (Archive) (3480 words)

These articles focus heavily on defending Rationality against its detractors, which is again very reasonable, however unfortunately these still don’t easily boil down into a short, hard to misinterpret definition.  Therefore, I will make an attempt to describe what I consider Rationality and the Rationalist Movement to be, and not to be, in a way that can be boiled down without losing all subtlety.

I. What it is not

Firstly, I shall deal with what Rationality is not.  Unfortunately, there are several schools of philosophy that Rationality is often conflated with, so I will go through each of them, and briefly detail the differences.  If you are not in the mood for a foray into the delights of 17th and 18th century western philosophy, I recommend skipping straight to section II, where I will start discussing what Rationality is instead.

The main stumbling block is that it is rather unhelpfully named – Rationality is not the practice of the 17th and 18th century philosophy of Rationalism.  In fact, if you go to the Wikipedia article on Rationalism, at the very top of the page is a disambiguation note “Not to be confused with Rationality”. Although Rationalism has several things in common with the current Rationalist Movement such as an appreciation for logical argument, it was actually developed in opposition to Empiricism, positing that reality has a fundamentally logical structure, and therefore that purely logical thinking can deduce truths about reality without reference to experiment or our senses.  Rationality is also not the same as, nor does it advocate Rational Planning. In the mid 20th century, high modernism was very much in vogue, and the term “rational” was used to silence opposition to huge and often disastrous reforms. This Rational Planning can be seen as the logical conclusion of 18th century Rationalism, and its outcomes can be seen as the exact reason Empiricism was right to oppose its disdain for evidence and experiment.

Nor can we say that Rationality is purely the practice of Empiricism either however.  With its focus on evidence, falsification and the Scientific Method, Empiricism also has a great deal in common with the current Rationalist Movement.  Unfortunately however, historically it takes a rather dim view of any attempts at theoretical inquiry. This is not to say there were not people that managed to subscribe to both Rationalism and Empiricism, but as with many philosophical disagreements, the tyranny of small differences usually wins out.  If we were to describe Rationality as the practice of the philosophy of Empiricism, it would carry a great deal of historical baggage, suggesting an opposition to theorising, rather than a symbiosis with it.

What about Skepticism?  Again, lots in common with the Rationalist Movement – the suspension of judgement due to the inadequacy of evidence is an important principle that could be seen as a cornerstone of modern Rationality, but the philosophy of Skepticism once again goes too far to be able to be seen as representative.  The goal of modern Rationality is partly to allow people to make better decisions, and to more reliably achieve their goals, but Skepticism borders on an almost nihilistic view of knowledge, obsessing over uncertainty at the expense of practical considerations.

So far we have established several things that Rationality is not.  Rationalism, Empiricism and Skepticism all come close in different areas, but all overshoot the mark, making claims that are too bold.  The philosophy of Pragmatism established in the 1870s by Charles Sanders Peirce manages to bridge the gap between deductivist Rationalism, inductivist Empiricism and nihilist Skepticism nicely, placing focus on the practical uses and effects of ideas.  The benefit of experiment is clear when looked at from a practical perspective, but theorising is itself worthwhile if it results in a practical benefit. Uncertainty is a key concern in practical situations, but Pragmatism’s aim is to mitigate or manage this uncertainty.

II. Narrowing it down

We now have a neat Venn diagram with the three philosophies of Rationalism, Empiricism and Skepticism as the circles and Pragmatism at the intersection of all three in the centre. Within Pragmatism, we see many familiar Rationalist concepts.  The idea of knowledge as “that upon which a man is prepared to act”. Awareness of the “philosophical fallacy” (archive) (the danger of taking categories for granted, without acknowledging that the divisions are entirely invented). Embracing the Scientific Method as a means to encroach upon the unknown.  Endeavouring to make praxis out of epistemology (implementing the philosophy’s methods and conclusions in reality, rather than leaving it as an entirely academic exercise). Pragmatism also naturally lends itself to a consequentialist view of ethics, due to the importance it places on real world effects.  This leads us to a reasonably good definition of Rationality, which encompasses the practical application of the Scientific Method, experiment and evidence, as well as logical thought, and an awareness of any limitations and uncertainty, to achieving our goals. It still feels like we are missing something however.  The philosophy of Pragmatism does get very close to the ideology of Rationality, but it slightly undershoots instead – there are other key things that are quite fundamental to the modern Rationalist Movement, without which the ideology feels rather vague and incomplete.

Perhaps because it is so often conflated with 18th century Rationalism, and therefore blamed for the excesses of high modernism, there is a deep vein of intellectual honesty running through the modern Rationalist Movement.  The idea that the truth should win out, and that we should aim for what is true to be what we think, not for what we think to be what is true, results in a culture that among other things places value on disagreements being collaborative.  To quote Richard Feynman:

“It’s a kind of scientific integrity, a principle of scientific thought that corresponds to a kind of utter honesty—a kind of leaning over backwards.  For example, if you’re doing an experiment, you should report everything that you think might make it invalid—not only what you think is right about it: other causes that could possibly explain your results; and things you thought of that you’ve eliminated by some other experiment, and how they worked—to make sure the other fellow can tell they have been eliminated.”

Because this kind of modesty is not usually something that comes naturally to people, it is this quality that makes it relatively easy to differentiate between people that are part of the Rationalist Movement, and people with an extreme political opinion that they justify as being “rational”.  Anyone claiming to “destroy their opponent” in a discussion using “facts, logic and rationality” can reasonably be assumed to be one of the latter, not the former.

There is also a particular focus within the Rationalist Movement on cognitive biases.  Whilst this is not a fundamental principle, in the way that evidence and logical thought are, it is a cornerstone of the culture.  Even if people choose to ignore their biases, which Rationalists may decide to do from time to time for various reasons, there is a view that it is useful to know what they are, and that people have them (even though knowing about them might make them worse).  The ability to avoid these biases when appropriate can be helpful in pursuing logical thought, and in achieving goals more effectively. This is not to say that the Rationalist Movement is entirely devoid of emotion however – it is a false dichotomy to view emotions as irrelevant to Rationality.  A better dichotomy would be to view thought processes as analytic or empathetic – both are valuable and have real world consequences. While Rationalists are often stereotyped as being far more analytic than empathetic, it is more that there is a balance to be found between both. Often in everyday life, emotion can overwhelm other considerations, so people trying to find a balance between analysis and empathy can be considered heartless, which is unfortunate.

Furthermore, no description of Rationality would be complete without mentioning Bayes.  Again, not a fundamental principle, but Bayesian probability theory is a very important idea for the Rationalist Movement.  As a culture with a great focus on the importance of the Scientific Method, there is a great frustration with abuses of statistics often committed in both the scientific community and the world at large.  Bayesian probability theory is seen by many in the Rationalist Movement as a better, simpler, more reliable and less open to abuse method of performing statistical analysis.

III. Bayesian Pragmatism

Based on this reasoning, in future my go-to definition of Rationality will be:

“The philosophy of Pragmatism (the practical application of the Scientific Method, experiment and evidence, as well as logical thought, and an awareness of any limitations and uncertainty, to achieving our goals), with a particular focus on Bayesian statistics and cognitive biases.” (41 words)

Rationality still an unhelpful name though.  Maybe “Bayesian Pragmatism” will catch on instead?  …no, I didn’t think so either – we’re stuck with it.