On Causality
Hume inferred that associations of subsequent events yielded to the idea of causation, hence making the source of the practice rooted at the custom. From this inference, even a defense of conservativism is built after Hume, which is a surprising way of thought.
Two events A and B are connected under the umbrella of causation if they are observed frequently. Here maybe the idea of probability and the theory of probability is beneficial, but I will deliberately stay away from that, at least until I have no other efficient direction.
Custom when chosen as the answer to the root of causality, is an intelligent move but I think there is something missing in that formulation. Kant, ingeniously divided the scope into two domains where in the first lies the appearances (phenomena) and in the second resides the things in themselves (noumena). Causality then is applied to the first and there remained a room for the free will in the second. A really systematic way of thinking which persuaded me greatly (and gently).
Abstracting the principle of causality from all other subjects, my aim to write on this topic was at the beginning restricted to the concept itself but soon I realized that the area is not very independent from the discussion on free will i.e. the trends such as compatibilism, incompatibilism, determinism and indeterminism. So, apart from the roots of causality, I will (hopefully) also state something on the issue of compatibilism/incompatibilism.
We have two choices; either we will detect the root of causality on the senses and perception i.e., we will select the Hume’s strategy, therefore determine the custom as the source of the principle or we will derive it from our pure understanding i.e., making it a priori. Deriving from the first source makes it statistical, getting from the second source determines it as categorical. First one is an established or rather at least chosen by the followers of Hume, second one is somehow risky and needs explication. How can we derive the idea of causality from the inner sources of our understanding and how can it be a priori?
Accepting the Kant’s distinction between the appearances the things in themselves could be a good starting point (I am reading CPR currently but have not knowledge of the exact point of view of Kant on the subject; so I am taking his distinction between phenomena and noumena as a tool on the matter). If causality applies to the appearances but not to the things in themselves, then causality can be a mode of our understanding or rather it can be a filter applied to the perception of events. Beliefs deduced from pragmatic thinking can be at the root of causality, too, on the other hand. That is, if perceiving the tuple (A, B) successively has some benefit on my life, then I associate a weight to it and use this in my thinking. Then such a weight assignment becomes a custom and I develop an intuition regarding this.
Thus, although not an a priori explanation of causality, pragmatism can be the source of this problematic concept. But, by this way, we get a practical rather than a pure grounding since the idea of benefit is highly related to the outer mode of our sensation, making the explanation nearer to the Hume rather than a pure law rooted at our understanding. Okay, let’s take a deep breath and summarize our reasoning: first we divided the sources as inner and outer, labelled the outer as the Hume way (custom). Then we moved to the inner possibilities and reminded ourselves the Kantian distinction between appearances and things in themselves. Having accepted the rule (of causation) is applied to the world of appearances, we thought that it can be a mode of our understanding or rather a filter applied to the perception of events. Taking the second one as true, we have speculated that causality is something related to pragmatic thinking. If my observations of subsequent (A, B) tuple yields a benefit to me, then I assign a weight to the pair and hence develop an intuition regarding it. If assuming A => B gives a benefit to me, then I assign a weight to the connection so that I can use it afterwards.
Back to the idea of compatibilism. I was once a compatibilist, thinking that having an explanation of my will and my decision makes me comfortable, compared to the chaotic nature of indeterminist system of explanation. This idea is still hold true today by me, especially when it is synthesized with the Kantian notion of the distinction between appearances and the things in themselves. Causality belongs to the world of phenomena and the will belonging to the world of things in themselves, then the incompatibility is automatically resolved.
If causality is developed by assigning weights to the necessity of subsequent events, then the intuition of it is emerged by making it a connection if the necessity quantity converges to 1. This intuition itself is never solely dependent on specific events so that we have a general notion of causality developed by collecting multitude of event pairs. Although derived by the analysis of appearances through the filters of understanding, the process is not categorical but still statistical, or rather mathematical. Thus the modality of causality lies somewhere between statistical and categorical, i.e., between outer and inner analysis. Intuition and convergence to 1 makes it nearer to the inner analysis but its statistical nature, that is the collection of the (A, B) pairs and a model derived from these pairs makes it closer to the outer analysis, hence to the custom solution of Hume. What is apart from Hume’s solution is that the model presented is mathematical rather than pure empirical. And it may even be the case that, the model’s core is innate and the model’s implementation is analogical to a software written in Python and Prolog. High-level part of the model belonging to Prolog and low-level part of it belonging to Python (sklearn).
To sum up: I see the causality as an intuition regarding to the appearances. It is yielded by the innate core of a survival-pragmatical programming. The collection of events are analyzed through the eyes of the species and weight assignments are done according to the observations’ quantity of benefit. It is innate; thus it is indirectly empirical.