What Knowledge Is and How to Check It

Knowledge is an ideal form in which five components are given.

  Component Example
1 The real world within boundaries A city
2 A model of reality A map
3 A list of tasks A list of pairs: from A to B
4 A real observer A taxi driver
5 Equipment A taxi car

If there are not five components, but only four or fewer, then this is not knowledge and should not be checked as knowledge.

If all five components are present, then knowledge is present.

However, not all knowledge is true. Knowledge can also be doubtful or false.

To understand which of the three classes our knowledge belongs to, we apply the P-procedure.

The P-Procedure

We run a cycle over all tasks, where the pair A and B means the task of driving from point A to point B:

  1. Solve the task (A, B), using only the map.
  2. Map points A and B from the terrain onto the map, obtaining points A1 and B1.
  3. Build a route on the map from A1 through X1 to B1.
  4. Map it back onto the terrain, obtaining points A, X, B.
  5. Drive through these points to point B.
  6. Ask: did we arrive?

We count the percentage of successful trips.

We choose two thresholds: high and low. For example, 95% and 50%.

If success is above 95%, the knowledge is true.

If success is below 50%, the knowledge is false.

Otherwise, the knowledge is doubtful.

Berkeley’s Paradox

The two properties mentioned above - adequacy to the world and consistency - are both used in the P-procedure.

Adequacy is used when we map the terrain onto the map and back.

Consistency is used when we search for a route on the map.

In this way, we use the first three parts of knowledge.

But this is not enough.

For the whole procedure to work, we also need the fourth and fifth parts: an observer and an instrument of observation.

Did anyone say that adequacy and consistency are not enough?

Yes. Bishop Berkeley did.

If a person is convinced that the world around him is an illusion, it is impossible to persuade him.

The problem is solved if we make the person an observer and give him equipment.

Let him perform the experiments himself.

By supplementing theoretical knowledge with a real observer, we obtain a reliable criterion for evaluating the quality of knowledge.