Notes on time and causality

A collection of philosophical essays on the subject of causality, determinism and time in general

On causality as the defining principle of all knowledge and on the subjective nature of all knowledge

If I throw a die or think of a random number and then tell you what that number is, I doubt that you would call this piece of information “knowledge” (unless your goal is to argue with me). This is because this piece of data, when taken by itself, is highly unlikely to be relevant to anything that will happen in the future. I’d define knowledge as all “information that is relevant in the future i.e. can be used to predict it”. So the concept of knowledge, depends on the concept of time - in order to know what knowledge is, we have to know what time is, indeed, in order to have a concept of the future at all, we must be able to perceive time. So let’s examine how do we do that. There are many ways to go about it, I, as a programmer, often tend to think about things in terms of input and output:

0. We can view the *input* which our brain receives, and from which the perception of time and continuity is derived, as a collection of frames, pictures of different states of reality which are then, somehow, united in the output. 

In other words, the mystery of time can be reduced to the following riddle - we have two pictures and we have to find the elements of the first one in the second one, like a converse version of the “Find the ten differences” game. Seems easy, but it’s not.

1. In order for the perception of time to be realised (and knowledge to be created), the list of frames must be interpreted as signifying some sort of change of object from one state to another, like change of position (motion), change of shape, colour etc.

- If the frames are all alike (like if you are standing facing a wall) you would not be able to perceive change (and time). 

- But if the frames are all different i.e. don't have anything to do with each other, you also wouldn't be able to perceive change. 

To perceive change, then, we must be able to interpret the frames in such a way that there is an aspect of them which is different for each frame but at the same time stays the same for all of them i.e. we have to postulate the identity of objects and events (we don’t need to think about objects separately from events, as they technically are nothing but longer events).

2. The basic form of identity of events and objects (objects being just collections of events) is based on the concept of *causality* - when observing event `A` in a given frame followed by another event `B` at another frame, we presume that `A ⇒ B` (`B` is caused by `A`).**

What this means is that identity is just a manifestation of causality, that they are basically the same thing - when I see a given object standing on my desk and then I see a similar object in the next moment I assume that it is probably the same object i.e. the object being there at time x causes it to be there at time x + 1.

There are other ways for defining the identity of objects (we can say, for example, for example, that an object is the same only if it is composed of the same atoms), but this is the main way that identity is perceived by people in practice, hence a ship that has all its parts changed at a course of a given period is still the same ship it was at the beginning of the said period, although different in terms of the material that it is composed of.

If we think more about the ship, however, we would discover that identity, when defined in this way, is not set in stone.

3. Causality is in the eye of the beholder i.e. `A ⇒ B` is not a fact about the world, but a mental image. This is so because `B` is party defined by its internal characteristics but it is also partly defined by just being "the thing that comes after `A`" (if `B` just happens without any sign of `A` before it, to what extent would it still be `B`?), in the same way as `A` is partly defined as the thing that comes after some other event.

This is the most essential of all my propositions, so let’s elaborate: imagine that we know that A ⇒ B and we observe A and then observe following from it another event that we would call B' that resembles B in terms of some of its internal characteristics, but is also different in some other characteristics (note that this is not a mere thought experiment, but a general description of the perception process itself, as all events are different from one another).

In this case, we have two choices:

The first kind of thinking is called empirical, the second one - dogmatic. Only when thinking empirically, do we obtain information about the world. Only when thinking dogmatically are we able to use the information that we gained by making predictions. Needless to say, empirical and dogmatic thinking are complemental to each other and go hand in hand. They are like input and output, Like question and answer…

With this, we established that dogmas like A ⇒ B are not truths, but just rules for structuring information. We may naively think of them as true because they mostly “work” i.e. allow us to achieve a given goal or false (getting us into trouble), but the fact is that they are not - instances that follow a given rule may only follow it by accident, or because we perceive them as following it. Instances that don’t follow the rule are simply not instances of that rule - no rule is right or wrong.

By the same token, we may naively think of the causality maxim (of A ⇒ B) as true (true as in “valid law of nature”, let’s say), because when we perceive A, then (in the most cases) we also perceive B and it is easier to explain that by postulating causality than to just say it happens by accident (Occam’s razor). This may lead us to believe that causality is some kind of law that exist in the world, or rather a meta-law, which implies the existence of all kinds of other laws. In this case, we would be overlooking the following:

4. `B` is not a specific state of affairs, it is just a mental image, a pattern *we* begin to search for given our previous knowledge of `A ⇒  B`.

We search for B and often do find it even without there being perfect candidates - If we already think that A ⇒ B we will see B everywhere we see A. In this case, we say that someone sees B even when it is not really there, but the fact of the matter is we cannot possibly see anything that is there in the way that we see B in this example.

Causality is neither a rule, nor a meta-rule, but a belief that every thinking being should hold to some extent in order to be a thinking being.

The last statements probably sound too counter-intuitive, so to be taken as true without some objections, so I will attempt to address some of them using the somewhat forgotten form of philosophical dialogue. Let’s imagine that the physicist Isaac Newton, (who pioneered the modern scientific method) had a chat with the philosopher David Hume (who challenged the principles on which this method is based).

Hume and Newton

Hume: Causality is not a quality of the world, but merely a belief. It’s a very general belief and one that every thinking being should hold to some extent or another in order to be a thinking being at all, but still, it is just a belief.

Newton: That is nonsense. You can clearly see that the world adheres to certain laws which are unrelated to whether you are observing it. This means that causality is a characteristic of the world in itself i.e. a law.

Hume: If causality were a law, there ought to be a way to test it, as we do with all other scientific laws, right?

Newton: Yes of course, and as a matter of fact, we do that pretty often. All scientific theories are based on the causality maxim. A scientific theory is nothing, but the assumption that some statement that has the form A ⇒ B is true. And then we have experiments that try to test this statement by making A happen many times and seeing if B follows. In other words, every science experiment that tests whether a given theory works also tests whether causality itself works.

Hume: This is true, but many if not most science experiments fail to some extent or another. Doesn’t that, allow us to conclude that causality also fails?

Newton: The only reason that experiments fail is that we simply don’t know enough to conduct them properly. Blaming causality for our failure is just ridiculous. If our theory is exactly right, it will produce the expected outcome every single time.

Hume: That sounds too theoretical for me. Let’s get an example. Can you construct a thought experiment, even a very elementary one that always gets the expected outcome? If you do that, you win.

Newton: Very well then. I suggest the following experiment which uses everyday logic and objects: a pistol is turned to a window. The pistol’s trigger is pressed, therefore the window is broken.

Hume: But what if there is no bullet in the pistol? Or if the pistol is broken in some way?

Newton: OK, let’s say that a bullet should necessarily be fired and that the aim isn’t off. Then the glass gets broken.

Hume: What if there is a hole in the glass?

Newton: No there can’t be. The bullet hits the damn glass, OK?

Hume: OK, but how fast is the bullet moving?

Newton: Let’s just say that the bullet hits the glass with sufficient speed.

Hume: Sufficient for what?

Newton: Sufficient to break it.

Hume: Look what you did here. You just defined the situation in such a way that the effect is embedded in the setup.

Newton: So you are saying that the effect can be embedded in the setup? Well, that sounds quite objective, does not really look like a belief to me.

Hume: Look, there may be possible cases where you will be able to guess whether the glass gets broken, but this does not make the general principle true. Because there is no general principle, in a first place, only a mental image and situations which remind you of the mental image.

References