Humility versus confidence
Epistemic attitudes and building justifiable confidence in what we know
Long time readers of Humble Knowledge are very well acquainted with my core argument. I have repeatedly argued that it is important to be humble about what we know. However, this has prompted questions about the impact on our ability to make decisions. Presumably, if we have to always be humble, and never confident, about what we know, then decisions are difficult and always provisional.
While this objection captures something important, it misses the more complex interplay between our broader philosophical attitudes and the more personal convictions necessary for decision making. When we understand these, epistemic humility turns into a net positive for good decision making.
Philosophy vs psychology
My focus on humility around knowledge is always in the context of epistemic attitudes - fundamental philosophical positions about what we can know. Previous posts have explored four of these epistemic attitudes: certainty, confidence, humility and skepticism. What is slightly awkward is that these concepts, which I have drawn from a few different sources, use some psychological terms to describe philosophical positions. Confidence and humility, especially, are more often used to describe a person and their character or state of mind than a fundamental philosophical position. This may have created some confusion.
An epistemic attitude describes a fundamental belief towards the achievement of knowledge by humans writ large. So, for example, I have previously defined epistemic confidence as the conviction that humans have achieved genuine and reliable knowledge (at least about certain topics). By contrast, we often also talk about our confidence in (or humility about) something that we know, which describes a psychological state held by a person (or a group). I might be very confident that I know my friends' birthdays, or the history of Mongolian Empires, or the theory of relativity - or I may be personally unsure and humble about what I know about these things.
Confidence as an epistemic attitude and as a personal mental state will often, but need not, coincide. I may be convinced that humans have achieved genuine and reliable knowledge about something (and have the epistemic attitude of confidence) but I personally am unsure about what is true. For example, I might be confident that the theory of relativity is true and that others know how it works, but personally I am not confident about what it says.
On the other hand, I can have the philosophical attitude that genuine knowledge is difficult to acquire for everyone, even with the best of care and attention (i.e. epistemic humility) but be personally confident that I've done all the work necessary and do genuinely know something with no room for doubt. This is most obvious in cases that are close to my every day experience, like knowing my friends' birthdays. I might have asked everyone for their birthday and written them down carefully - so am justifiably confident.
While epistemic attitudes and personal psychological attitudes are different things and need not coincide, there is often a correlation. If I believe the fundamental philosophical position that humans have achieved genuine and reliable knowledge and no-one can reasonably doubt it (i.e. epistemic certainty), then it is likely that I will be more confident and certain about what I personally know. On the other hand, if I believe that knowledge is hard, I will tend to be less likely to make bold and absolute claims about facts.
Justifiable confidence
If we are thinking about decision making, then there is another important factor in the interplay between knowledge and human psychological states. As humans, our psychological confidence in something is often unrelated to any justification for that confidence. I can be justifiably confident in my knowledge of the history of Mongolian Empires (having done lots of research and looked at primary evidence) or I can be unjustifiably confident (I've only watched a couple of short YouTube videos).
To make decisions, we need a sufficient level of confidence in what we know to make a call, but we also want our confidence to be justified. There are many stories about hubris coming unstuck for good reasons. It is therefore worth asking which epistemic attitudes better dispose us to having justifiable confidence.
Intuitively, the types of things that give us justification for what we know are fairly clear.1 Ideally, we trust that someone knows what they are talking about if they have done things like taken significant time to learn the topic, tested things (or experienced them) themselves, understand the range of important views on the topic and can back up their knowledge with evidence. As a result, we rightly emphasise research practices like cross-checking against multiple sources, questioning motivations of the people making claims and going back to primary evidence where possible.
What is striking is that all these processes of justification assume that knowledge is hard to find and so we need to work to get it right. Put simply, if we want people to ensure they are justified in what they know, we ask them to adopt an attitude of epistemic humility towards their knowledge.
We would prefer that they don’t start from the conviction that we have already achieved reliable knowledge (epistemic confidence) let alone that no-one can reasonably doubt the knowledge we have (epistemic certainty). Each of these predispose us towards taking things on face value and not truly verifying the information we are given. Instead, good practice to build justified knowledge, and therefore justified confidence, assumes a practical epistemic humility.
Intriguingly, this means that epistemic humility both predisposes us to be less psychologically confident about any of our knowledge but also to do the work necessary to have justified confidence. Epistemic confidence and certainty, however, push us in the other direction - we will be more confident but with less justification.
So, in a decision-making context, adopting the general philosophical attitude of humility inclines us to do the type of work that allows us to be better justified in what we know. It also doesn't prevent us from being personally confident (either for good or bad reasons) as my beliefs about the state of human knowledge writ large are rarely relevant to a decision in front of me now. What matters is whether I'm confident enough to make a decision, and ideally we want that to be a justified confidence that is calibrated with reality.
So, despite initial appearances, epistemic humility is beneficial for decision making - at least in the long run. It encourages us to be clearer about what we know with justification and what we don’t, which helps connect our decisions more closely to reality.
I’ll note that there are deep philosophical arguments about the nature of justification in knowledge. But our common intuitions hold true under almost all of these and are sufficient here.
Love it. This draws a number of things together rather neatly.
Reading it led me to envisage two connected operating systems - a knowledge system and a decision/action system. The starting point of the knowledge system is a question mark, which only sometimes resolves into a + or a - (and then often only contextually, such as the operation of the earth's gravitational pull).
The starting point of the decision/action system is also a question mark. But, unlike the knowledge system which is designed to remain a ? unless our knowledge of reality dictates otherwise, this system requires a + or - answer. That is, the question must be resolved.
One way of bridging the gap between the two systems is to be more easily convinced about how much is knowable and true (epistemic certainty). This allows the knowledge and decisions systems to align, to provide decision-making certainty. Such certainty is, according to the world of epistemic humility, will often be falsely based as it is predicated on the wrong approach to gaining knowledge.
The practical problem with this wrongness is that those who fall into its trap are unable to tolerate any other possibility at the knowledge level, and often at the decision level.
Bridging the knowledge-decision gap in the world of epistemic humility effectively requires some form of accepted heuristic. This means that knowledge the decision system bases action on may not be true, but what is believed holds well enough for the decision to be valid. This, in my view, is what is happening in your birthday example. Having knowledge of time has proved very difficult. In response, humans have created an artificial model of time which is commonly understood. Provided humans use this model in a consistent way, the deeper knowledge problem is resolved even if it is not solved.
The great advantage of this system is that it prevents categorical statements of truth forming when they are not justified. This, in turn, would hopefully create greater tolerance and understanding in the uncertain processes of human decision making that would lead to world peace and infinite leisure time for all.