Effectronica posted:

Let us begin by setting out the basic situation. Let us assume that evolution by means of natural selection happens and is predominant. Let us assume that our minds are created by the operations of our brain. Let us also assume, for the time being, that what we can perceive through verifiable, repeatable empiricism is all that is real. So in other words, let's assume that biology is true, and that our general theory of scientific knowledge is true, and that only things that can be scientifically analyzed are real.

So let's drop some of the formality and get to the basic issue. How do our beliefs interact with our behaviors? How does the mind interact with the body? There are four basic options:

• Beliefs have no influence on behavior.
  
• Beliefs cause behavior but not by reason of their semantic content (or in plainer language, their meaning).
  
• Beliefs cause behavior but are maladaptive, evolution-wise.
  
• Beliefs cause behavior and are evolutionarily adaptive, but not inherently true or false.
  

So now let's consider whether we have reason to believe that our beliefs about the universe are generally true.

In the first case, beliefs are invisible to evolution, which can only select for behaviors, so there is a low probability that our beliefs are true.

In the second case, beliefs are selected for or against, but the truthfulness of beliefs is in their semantic content, which is not selected and so there is a low probability that our beliefs are true.

In the third case, beliefs are selected against and so not only is there a low probability of their truth, but we also should probably not exist.

In the fourth case, evolution can act on beliefs directly and in relation to their semantic content. But there are many possible false beliefs that may still inspire evolutionarily adaptive behaviors. For example, let's say that a prehistoric hominid, Paul, sees a tiger. He can run because he believes tigers are dangerous, which is a true, adaptive belief. But he can also run because he wants to be eaten but feels that this tiger is an unlikely prospect to do so, which is a false but adaptive belief. Or he can run because he believes tigers are dangerous though not because they will eat him, which is an adaptive belief and still a false one for our purposes, which are about our perception of the universe, though it's partially true. So while the probability that our beliefs are true is higher than in any of the other cases, it's still low.

But I am saying "low probability" without really doing any math. So let's make up a simple model, where we have a model human who has 100 beliefs, any of which may be true or false, and which are randomly formed as true or false, independently of one another. What is the probability that she forms her first 50 beliefs as true ones? Remember, these are independent events! So the formula for the probability is (0.5)^50, which amounts to 8.88e-16, or 8.88e-14% chance. Remember, this is for half the beliefs, not all of them. Even forming 10 true beliefs in a row has a 0.098% chance of occurring for her. While on average half of her beliefs will be true, they will be randomly scattered throughout this set of beliefs, generally in small groups.

It is possible to construct more complex methods of evaluating truthfulness of beliefs and so on, but the basic problem right here is that most of our beliefs seem to be true ones. Snakes can be poisonous or otherwise dangerous, we generally are pretty good about determining whether there's a beer in our fridge or not, and so on. Our senses tell us that our beliefs are truer than they must be by probability. So: "biology is true, and that our general theory of scientific knowledge is true, and that only things that can be scientifically analyzed are real". Pick two. Either our senses can't be trusted, or creationism is true but there is no god directing it, or, and this is frankly the most reasonable- the supernatural exists. Something, which is beyond our perceptions and knowledge, is a phenomenon which ensures that our beliefs are more likely to be true than they should be given natural selection.