.HOME
[Guest Blog - Causality and Systems]
August 1st, 2007On Causality and the Concept of Systems
Lovelace
Lovelace here again, and this time I’m going to be discussing the importance of the concept of systems. This revelation was brought to me via a dispute against Hume’s argument against causality, so I’ll discuss that in brief; then I will head into why the concept of the system is incredibly, absurdly important to intelligence.
Hume and Causality
The following is an excerpt from my paper on the matter.
Hume begins his examination of causation with a simple example: that of one billiard ball striking another. He notes that, via sense-data, we can determine that one billiard ball moved, struck another, and that the second moved; however, he points out that there is no “sense” of connected-ness. “Consequently, there is not in any single particular instance of cause and effect, anything which can suggest the idea of … necessary connection.” (Ariew and Watkins, 515) By “necessary connection,” Hume means to say that there is no proof of causality: that is, there is nothing inherent to this particular set of occurrences to indicate that the former action caused the latter.
Hume then examines this peculiarity in detail. He notes that all elements of an extended - that is, physical - object are self-contained: they are complete within themselves, and give no hint as to anything that could arise from them. He notes that the universe and the objects held within move and follow each other in succession, without ever revealing to us - from within themselves - the manner in which they are caused to do so. “Heat is a constant attendant of flame,” (515) but we have not the faculties to determine how this is the case. His conclusion from these examples, then, is that the idea of causality cannot be derived from objects or their actions because there can be found no origin for this concept within the objects themselves.
At this point, I feel it necessary to question Hume’s reasoning. Certainly, we cannot readily determine what events an object may bring about via its action or inaction in regards to other objects, but then why should we be able to do so? In the case of the billiard balls, we have no way to determine the effects of the first upon the second merely by examining either. It is the system we must examine - that which contains both billiard balls, and any other information that would be relevant to the exercise: namely, physics. Taken in the context of the system, we can surely deduce that, given the first ball striking the second, the second ball must move, as per the laws of physics. Not only that, but the understanding of physics must certainly give rise to knowledge of an inherent property of the billiard balls: that of potential and kinetic energy. Given the system, we are given an understanding that the objects in question possess potential energy, which is capable of transferring itself between two objects when struck against one another. If kinetic energy causes an object to move, and can be transferred - with some loss - from one object to another via physical contact, then surely we have found the cause of the second ball’s motion.
The Importance of the System
It was in writing this paper that the concept of the system hit home for me. The realization that we must acknowledge multiple elements - and the manners in which they interact - as a system in order to comprehend any number of the events that occur in relation to these elements is profoundly important. It allows us, as intelligent beings, to take seemingly disparate things and correlate and connect them. We see an event, and we understand that it occurs because of interactions between elements. If we ignored or were incapable of understanding systems, no event would ever seem correlated with any physical object. We would see cause and effect, but be unable to comprehend the idea of cause or the idea of effect. It would all seem to be a meaningless jumble of occurances, with no way to determine what will or will not happen.
A system, essentially, is anything in which objects may or may not exist and in which events may or may not occur. The more important aspect, however, is that the system has rules, which govern these objects and events. These rules can be simple - for instance, there is gravity - or they can be complex, such as our own universe. A system can also have within it a number of systems, which can also contain systems, ad infinitum. An atom is a system, contained within the cells, which are systems contained within the body, which is a system contained within the system of the earth, which is a system contained within the solar system, etc.
However, a system does not necessarily have to be a physical thing. Tetris is a good example of a system that does not physically exist. In Tetris, there are objects (the blocks, the walls) and events (clearing a row, end of game, gradually speed increases over time). There are also rules - the methods by which these objects and events are governed within the game (Tetris is a poor example for rules, as the very nature of the objects involved almost negate the need for rules).
So, the world is full of systems, of both the mental and physical sort. The system of Tetris is no different from the system of the human body, in the basic concept; while they differ in individual elements, the base concept is still the same.
…and that is the key concept. That, at their core, systems are inherently similar due to the very fact that they are systems. Rules, concepts, and ideas that are understood, gained, and learned within one system can be transferred to another.
Let us say that you play Tetris, and you understand very well how the blocks go together, and how to form cohesive rows without leaving any gaps. Then, let us say you move to the system that is a FedEx truck, which operates in a very similar manner. Certain aspects of your understanding of Tetris are very much so applicable to your ability to work within the FedEx system. While it is not a perfect translation, there is still enough there that you will perform better at FedEx than you would if you had never played Tetris before.
The point is that interacting with any given system allows us to learn and adapt, adding new concepts to our mental repertoires that makes us more adaptable and versatile. Depending our experiences with systems, we will be able to take and combine that knowledge in new ways to approach a problem. The more breadth of systems a person has had experience with, the more adaptable he will be in solutions.
At its base, this concept is deeply rooted in the idea that knowledge is based upon comparisons. As we age and encounter new objects and ideas, we learn them best by comparing and contrasting them to previous experiences. Our knowledge gained within systems follows a similar idea, allowing us to compare and contrast a new system with all previous systems.
Hofstadter has written a good deal on the idea of analogies, and how important they are to intelligence. I feel that Hofstadter is quite correct - the ability to construct analogies between disparate elements (or systems, but I’ll get to that in a moment) is a rather important skill. It allows us to determine when things are similar, even though they are not exact matches. It allows us to recognize that two people are humans, even though they do not necessarily share the same particular features. They are similar enough - analogous - and that is good enough to allow us to make that judgement.
The importance of systems, then, is to recognize the disparate elements in the analogy as a cohesive whole. When we examine a human, there are a multitude of features we must take into account: eyes, ears, mouth, nose, body shape, number of appendages, number of digits, and so on. While not all humans fit the norm, the vast majority do - this allows us to take a glance at a figure and say, “this is a human,” without seeing any particulars. However, it is the acknowledgement that all of those features fit together to form a system.
The concept of the system is so utterly important! It allows us to recognize that individual elements comprise a whole; it allows us to understand how elements interact with one another. In the case of Hume, he could not understand how there was any causal connection between the two billiard balls. It was because he lacked cognizance of the idea of the system that he failed to recognize causality: in the context of a system, with the elements being the billiard balls and physics being the ruleset, it becomes clear as day that causality exists.
It becomes clear that the ability to draw analogies and recognize systems are key to intelligence, and are therefore key in the construction of an artificial intelligence. I am uncertain if there is more to intelligence than these two key concepts, but even so, it brings us that much closer to a mechanical intelligence.
 |
Loading ...
