We don’t always think about what really makes our beliefs justified or rational but we tend to be pretty good at identifying which beliefs are justified or rational anyway. We can often take a uncontroversial example of a justified belief then assess why another belief is or isn’t justified by how similar (or analogous) it is to the uncontroversial one. For example, the belief in the law of gravity is a good example of a highly justified belief. It is rational to hold such a belief (in part) because of how well it helps us predict the future. However, the belief that energy can be created or destroyed seems to be unjustified insofar as it has never helped us predict the future.
Although we can often identify how justified a belief is through common sense and analogies, many philosophers believe it can be a good idea to have a theory of knowledge that can be used to help us identify justified beliefs. (Philosophical study of knowledge and justification is called “epistemology.”) When we want to know if a belief is justified (probably true), we can assess various characteristics of the belief. If the characteristics of one belief are sufficiently better than the characteristics of other beliefs, then it is the best belief. It is the most justified and the most likely to be true given our current level of knowledge. Scientific hypotheses, such as the law of gravity, are considered to be “scientific theories” and are among the most certain of our scientific beliefs. (We could even say that we know that the law of gravity is real.) Hypotheses found in philosophy tend not to be so certain, but the most justified philosophical hypotheses, such as the hypotheses that we can reason about morality, are hypotheses that we should agree with. It is much more rational to agree with such sufficiently justified philosophical theories than to disagree with them.
I have already discussed various aspects of justification in “Four Sorts of Justification” and “Objections to Moral Realism Part 2: Intuition is Unreliable,” but I will now discuss the positive characteristics that I believe are found in the most justified hypotheses. The more positive characteristics a theory has, the more likely it is to be true, and the more rational it will be to agree with it. These positive characteristics have been called “cognitive values,” “epistemic merits,” and “theoretical virtues.” I will call them “theoretical virtues” because they are virtues that can be had by theories, hypotheses, and beliefs; rather than virtues that people can have.
I propose the list of theoretical virtues to include the following:
- Logical consistency
If something is self-evident, then we can justifiably believe it’s true just by fully understanding the concepts involved. Self-evidence is meant to help provide us with justified beliefs without requiring direct observation. For example, “1+1=2” seems to be self-evident. Self-evidence does not necessarily provide us with absolute truth or certainty, but it does give us beliefs with at least a low level of confidence.
It is controversial whether or not a belief can be self-evident. Coherence theorists argue that logical consistency and comprehensiveness are enough to justify our beliefs. The important point to make with supposedly self-evident beliefs is that they are very difficult to prove false. We are very certain that these beliefs are true despite a lack of evidence. If “1+1=2” isn’t self-evident, then we need to know how we can be so sure that it’s true.
Perhaps the most important theoretical virtue is logical consistency. No contradictions should be found within our beliefs or at least one of our beliefs are false. I can’t think that killing people is always good, but think that killing me is wrong. If our beliefs are logically consistent, then they might all be true, and we can say that they are consistent. When our beliefs are logically inconsistent, then they are incoherent.
One of the most valued forms of theoretical virtue is observation, which is often taken to be “evidence.” If I perceive that a cat is on the mat, then, all else equal, my belief is justified.
We also know what it’s like to have our psychological experiences by having them. For example, I know what sweet means by actually tasting sweet food. This is also taken to be a kind of observation, but examining what it’s like to have psychological experiences can also be called introspection or phenomenology.
Observation alone is meaningless. Some background assumptions are necessary to make sense of our observations. When I see a cat on a mat I have to already assume that my visual experience is caused by an organism best described as a “cat,” for example. Otherwise my visual experience is just a bunch of blotches of colors and meaningless shapes.
One of the most important theoretical virtues in science is predictability. If a hypothesis can make unlikely and specific predictions of future events, then (a) the hypothesis was not falsified and (b) the hypothesis is more likely to be true. Highly justified scientific theories, such as Einstein’s theory of physics, make very precise and unexpected predictions that have been successful again and again. (For example, Mercury’s orbit around the sun could only be accurately predicted by Einstein’s theory.)
Predictability is one way for a hypothesis to be falsifiable. If we can prove a hypothesis is false, then the hypothesis might fail to make certain predictions.
Even moral theories have elements of predictability. A moral theory can predict that an action will be right or wrong, but we can find that the moral theory fails to identify various actions as wrong. For example, some people believe that utilitarianism can be used to wrongly justify the use of disrespectful forms of oppression, such as the poor treatment of factory workers.
A comprehensive hypothesis can be used to explain or predict more phenomena. The more comprehensive a theory is, the better. For example, I might hypothesize that “punching someone in the face who cuts in line at the Bank” is wrong, but it would be more comprehensive (and therefore preferable) to hypothesize that “hurting people is wrong unless it’s necessary for the greater good.” The more comprehensive moral hypothesis can explain why more actions are wrong than the less comprehensive one. We could identify that not only is “punching” wrong in a situation; but murder, rape, and torture are also wrong (everything else equal).
“Simplicity,” also known as “generality” or “Occam’s Razor” is the view that we should prefer hypotheses and explanations that are simpler and more general. For example, we could hypothesize that “Rape, torture, murder, and physical abuse are wrong unless it’s necessary for the greater good” but that hypothesis is too complected and a simpler (and more preferable) hypothesis is possible, such as “It’s wrong to harm people unless it’s necessary for the greater good.”
At least one reason that simplicity is preferable is because it tends to be more comprehensive. A more general hypothesis will be able to predict more. For example, it might be possible to harm people in ways other than “rape, torture, murder, or physical abuse.” The more simple hypothesis could identify isolation, ostracism, and insults to be forms of “harm” and therefore wrong in most situations. We could continue to expand the more complected hypothesis to correctly identify such actions as wrong, but it might always be missing some other harmful actions.
We are often rationally justified to believe in a hypothesis even though there is a chance it is false. Even our best scientific theories don’t ideally instantiate all the theoretical virtues, but such theories are still highly justified. Many people believe scientific theories, such as evolution, have failed to make various predictions or explanations that somehow disprove the theory; but we can’t demand perfection from any theory or belief. Instead, we tend to compare and contrast theories that have been proposed. Theories that instantiate theoretical virtues the best are found to be the most justified and most likely to be true.
Many people think that whatever they believe is probably true, but our beliefs must also conform to theoretical virtues in order to be rationally justified. It is irrational to have a belief if the belief if the belief clearly fails to instantiate the six theoretical virtues as well as competing beliefs.
Philosophical hypotheses tend to be much less justified than our best scientific theories, but such theories can still be clearly superior to competing hypotheses given their theoretical virtues; and it is therefore rationally necessary to believe in certain philosophical theories. If you want your beliefs to be true, then philosophy is often the most appropriate way to have the most justified beliefs possible.
Sometimes we don’t know for sure which scientific or philosophical hypothesis is the most justified. It can sometimes take hundreds of years of debate for philosophers to determine which hypothesis is probably true. Even so, the competing hypotheses used by non-philosophers are often proven by philosophers to be unjustified. We should prefer the hypotheses that philosophers have found to be possibly true rather than to believe in hypotheses that have been proven to be unjustified.