2.2 Types of uncertainty

Uncertainty in its broad definition from the previous section covers a spectrum of situations. There can be clear logical differences between uncertain beliefs based upon the type of uncertainty that is described. We focus on two types of uncertainty:

1. Aleatory uncertainty: uncertainty due to variation.
2. Epistemic uncertainty: uncertainty due to a lack of knowledge.

Aleatory uncertainty describes situations with natural variability, sometimes referred to as chance. With this type of uncertainty, we cannot be certain about something because there is a component of randomness to it. For example, before a roulette wheel is spun, you are asked what number the ball will land on. Assuming the spin is not fixed to favour any particular number, then the result will be random in the sense that there is no strategy that is guaranteed to win more often than selecting a number at random. Due to the randomness involved in the spin, your guess will be uncertain. This type of uncertainty cannot be reduced because it is inherent to the process that causes it.

Epistemic uncertainty concerns situations about which we have a lack of knowledge. The reason for this type of uncertainty is that we do not know all of the necessary information. Suppose I ask you to turn your back so that you do not see the result of the roulette spin. I spin the wheel and see the result, and then I ask you to guess the number. Your guess is now uncertain not because of randomness, but because you do not have the information that I have. There is no uncertainty for me but there is for you, and your uncertainty could be eliminated by learning what I know. Epistemic uncertainty can be reduced by obtaining more information.

The idea of uncertainty being personal is key to thinking about events which have happened in the past. Past events have either occurred or not, and so they are factually certain. However, an individual’s beliefs about those past events can be epistemically uncertain because they have insufficient information. Individuals will base their beliefs on the information that they have available to them and, since this information and how it is valued could vary from person to person, it is possible to have varying degrees of epistemic uncertainty between individuals about the same factually certain event. Imagine that another player is introduced into the roulette example above. I ask you both to turn your back whilst I spin the wheel and I see that the result is a 10. I then tell you nothing about the result but I tell the new player that the result is an even number. You still have the same uncertainty as before but the new player has less uncertainty than you because they have more knowledge about the result; you both have different degrees of epistemic uncertainty about the same factual event.

Epistemic uncertainty is reflected in the courts through the phrase ‘beyond reasonable doubt,’ the burden of proof for the prosecution in UK criminal trials. It is not ‘without any doubt’; there is room for some uncertainty about the facts. Events have factually occurred but perhaps only the aggrieved and/or the defendant are certain about the truth. On the other hand, the fact finder begins a trial with a high degree of uncertainty about events with the presumption of innocence on behalf of the defendant. Evidence about the facts is presented to the fact finder in an attempt to reduce their epistemic uncertainty. Upon hearing all of the evidence, the fact finder makes a decision as to whether the prosecution sufficiently reduced their uncertainty to ‘beyond reasonable doubt’ in favour of guilt. When the fact finder is a jury, then each juror must have their personal epistemic uncertainty reduced to this level during the initial stage of deliberation.

Uncertainty is present in all parts of the legal system as it is in every other aspect of society. This includes uncertainty in scientific evidence.