How do metaphysics and epistemology connect?


Markus Gabriel

To person

Dr. phil., born 1980; Professor of Epistemology, Modern and Contemporary Philosophy at the University of Bonn; Deputy Director at the Käte Hamburger Kolleg "Law as Culture"; currently visiting professor at the University of Berkeley / USA; Institute for Philosophy, University of Bonn, Lennéstraße 39, 53113 Bonn. [email protected]

The epistemology deals primarily with the question of what "knowledge" and "knowledge" mean. It cannot be taken for granted from the outset that the two expressions each have only one single meaning, nor that they each have a set of meanings that are systematically related to one another. It could be that we use the terms "knowledge" and "cognition" in all sorts of different ways, without these uses having a common meaning.

I myself advocate epistemological pluralism. By this I mean the assertion that there are different forms of knowledge and cognition. Again, this claim could be interpreted in at least two different ways. On the one hand, one could say that there is practical and theoretical knowledge: practical knowledge would be a form of ability if one knew, for example, how you can roll down a steep slope on roller skates. Theoretical knowledge, on the other hand, would consist in knowing that the slope is steep. In what follows, however, I will advocate another form of epistemological pluralism, namely the assertion that we are by no means always expressing the same term when we claim that someone knows that this or that is the case. So I am claiming that there is a plurality more theoretical Forms of knowledge exist.

This claim seems to be somehow familiar, as it corresponds to an obvious finding, namely that we distinguish between, for example, mathematical, sociological and physical knowledge and associate this difference with different sciences. Different sciences are different in that they make use of different theoretical forms of knowledge.

In addition, I am convinced that the expression "knowledge" also indicates a plurality, that there are different forms of knowledge. I therefore differentiate between "knowledge" and "knowledge" - a difference that is unfortunately often blurred in current German-language epistemology, as it is based on the English-language debate, in which there is no exact linguistic equivalent to the German expression "knowledge". [1]

Although I advocate epistemological pluralism, I believe that there is a narrow core of meanings between "knowledge" and "knowledge", that knowing what this core consists of is by no means sufficient to already know what "knowledge" in this way "means. Anyone who did not know that mathematical knowledge differs from sociological knowledge and could give any list of differences would, in my opinion, know far too little about "knowledge" to be able to claim that they knew the meaning of this expression. The narrow meaning of "knowledge" is, in my opinion, "true justified conviction", which is often referred to in epistemology as the "standard definition", which was first considered by Plato, but also rejected again. [2] Since hardly anyone introduces the alleged standard definition as a fully valid concept of knowledge, I consider it absurd to speak of a "standard definition" and instead call it the "initial definition".

In the following, I will first briefly motivate the initial definition, that is, present reasons that speak for the elements of the initial definition of "knowledge". Then I will explain the difference between "knowledge" and "knowledge". By "knowledge" I understand a "true and therefore truthful reference", that is, a reference to any object that is subject to certain conditions of success. Finally, I will explain that we have to distinguish epistemological knowledge from other forms of knowledge, which already speaks in favor of epistemological pluralism. In current epistemology it is particularly often overlooked that it claims knowledge itself or recognizes something, but that this form of knowledge or knowledge differs significantly from other forms. I explain this difference as a difference in architecture, which means that the elements of the narrow core meaning of "knowledge" and "cognition" are designed differently in the various forms of knowledge and cognition.

Initial definition of "knowledge" as a narrow core of meaning

A "definition" of "knowledge" is usually understood to mean the specification of necessary and sufficient conditions that a state must meet in order to be able to be considered knowledge. The relevant state that comes into question for knowledge should be that of a belief. Whoever knows something is convinced of the matter in question, that is, he maintains his opinion even against resistance. With this we have discovered the first condition, which we can call the "belief condition":
  1. Who knows something is convinced = S is convinced that p.
The correctness of this condition can be seen from the fact that it would be strange to say that S knows that it is raining, but he is not convinced of it. Knowledge goes hand in hand with certainty, which simply means that you hold on to what you think you know, if necessary. You don't just believe it like when I randomly assume that Angela Merkel is in Frankfurt at the moment, without having any reason for this assumption. I can easily be convinced otherwise. On the other hand, if I know that she is in Frankfurt right now because I see a live broadcast on television, I will not be easily dissuaded from this assumption. The second condition for knowledge is the "truth condition". This states that you cannot know anything wrong unless you know that it is wrong. The truth condition can be formulated as follows:
  1. p is true.
If I know it's raining in London right now, then it's true that it's raining in London right now. From the fact that someone really knows something (and does not just claim it) it follows that what he or she knows is true. Finally, the third condition of knowledge is that someone who knows something can defend this knowledge against objections. Anyone who knows something has thought about it and has embedded the knowledge in a context. What is known is somehow justified, some process suitable for establishing the truth belongs to the prehistory of knowledge: one has read it, watched a live broadcast on television, studied a subject or successfully solved a logic puzzle through hard thought. In epistemology, all these different processes are brought together in the rather general concept of justification or justification. One can therefore say that there is a "justification condition":
  1. S ’belief that p is justified or (in a very general sense) justified.
Epistemologists differ particularly on the question of what exactly constitutes a "justification" or a "justification". I myself believe that there is no single answer to this, as there are different forms of knowledge, each of which is subject to different conditions of justification. More on this in the last section.

So you only know something if you can defend your convictions against any number of objections that are justified on your part or provide reasons for them in response to critical inquiries. Our beliefs have a network structure, they form a system in a relatively undemanding sense. If I am convinced that it is raining at the moment, it is because I was looking out the window, that the sky is cloudy, that at this time of year it rains a lot in the place where I am right now. Since you cannot know anything that you are not at least convinced of, our knowledge also forms a network structure. We only ever know something in a larger context, which is why we never only learn something individual, but always learn connections in order to expand individual insights into larger units of knowledge. For example, if you learn something about contemporary China at school, you learn it in a context that systematically links all of the individual insights we have about China.

Knowledge consists in being able to put insights into context. Let's take a simple example: A passer-by asks us if we know when the next regional train will leave Bonn to Cologne. If we know this, we know it in a context, for example because we took such a regional train with a certain regularity, because we know where you can see a relevant timetable, that a regional train runs from Bonn to Cologne at all. We usually also know where the train station is and a lot more, like what color a regional train is in that area, what models are in use (at least roughly) and so on.

If we know something, there is a connection between a network of beliefs and the fact that we can use this network to cite background beliefs when our knowledge is called into question or is in demand. This expresses the initial definition of knowledge. In doing so, one should not ignore the truth condition. Because you have to differentiate between a knowledge claim and actual knowledge. Anyone who claims to know something that is wrong does not know. Claims to knowledge can go wrong, but not knowledge itself. That is why they also say: "I thought I knew, but I was wrong", and not just stick to a claim to knowledge. A claim to knowledge is therefore fallible (fallible), knowledge is not. Knowledge is the name for the case of success, for a successful claim to knowledge.

To summarize, the initial definition of "knowledge" is that S knows that p if the following three conditions are met:
  1. S is convinced that p.
  2. p is true.
  3. S ’belief that p is legitimate.
In an influential article from 1963, the American epistemologist Edmund L. Gettier showed that the initial definition of knowledge cannot be universally valid. [3] If you accept them in this abstract formulation, so-called Gettier cases arise, as has been said since then. Here's a simple Gettier case. Let us assume that S is convinced that he had a salad for lunch last Monday. His reasons for this belief are that he has a bill in his pocket for a salad that bears the appropriate date. He also says he remembered eating a salad. Perhaps this is confirmed by a generally very reliable and sincere friend who wants to see him. His conviction is therefore entirely justified. Let us now further assume that S really had a salad last Monday. As a result, all three conditions of the initial definition are met.

However, the calculation is still wrong. Instead, S had a salad in the restaurant in question on Tuesday and the waiter accidentally issued a bill with the wrong date. His friend didn't even see him on Monday, but also made a mistake in his memory about the date. Also, S doesn't really remember eating a salad on Monday, his memories are all memories of Tuesday's salad. Accordingly, all three conditions are met, but there is still no suitable connection between the list of reasons that make S ’authorization and the fact that he actually ate a salad on Monday. The fact p is therefore only true by chance in relation to S’s belief system. Nevertheless, all conditions are met. Consequently, the initial definition is not sufficient, since counter-examples can easily be found.

After almost 50 years, no one has so far succeeded in concretizing the initial definition, generally recognized, in such a way that counter-examples can no longer be cited. There may be many reasons for this. I myself believe that the decisive reason is that the initial definition correctly reproduces a narrow core meaning of "knowledge", but that real knowledge has additional structures that make it resistant to relevant Gettier cases. This thesis should not be confused with the fact that claims to knowledge can be immunized against Gettier cases or complicated objections. But real knowledge cannot be shaken by objections, because otherwise it would not be knowledge, but only a claim to knowledge. This may not help much, since we cannot easily distinguish between knowledge claims and real knowledge. Because to do this, we in turn have to make a claim to knowledge that can fail. However, the fact that we are prone to error when it comes to the question of what we really know does not in itself mean that we may not know anything. We just don't know all of a sudden what we know, because nobody has an overview and control of all forms of knowledge. Every knowledge system is and remains finite, an infinite knowledge that overlooks and assesses all knowledge is in principle impossible. [4]