More on David Hume - Correlation vs. Causality

Another aspect of David Hume's philosophy (see my first post on David Hume here) was a study of psychological association and the laws by which rational concepts are created from sensory experiences. These associations between ideas or events may be classified as "correlation" and "causality." When two events occur close to one another in space and time, the mind tends to "associate" those events, building up an ensemble of data that reflects correlation between those two events. This means when one of the events occurs, we learn to expect the other to occur nearby and close to the same time. We might even train ourselves or others to respond to the expectation of an event by creating another event that is correlated with the first. An example of this is Pavlov's experiment with dogs. He correlated the ringing of a bell with the presentation of food in the dogs' routine. At some point, the dogs would expect food when they heard the ringing of a bell and Pavlov was able to "measure" this effect by the dogs' salivating. But such an effect doesn't need to be manufactured. It may often occur naturally.

Now, the intuitive leap we often make next seems to be a product of an innate desire to establish causality between two events. Nothing in the external world seems to say that one event must necessarily cause another ... and yet, that is what we seem to be psychologically wired to explore. It's hard to say whether the dogs made that intuitive leap in thinking that the bell caused the food to appear, but we as humans certainly tend to do so. When we see events that occur together, we may rightly conclude that those two events are correlated. But we often jump to the conclusion that one event must necessarily have caused the other. And usually we determine which event is the cause of the other by which occurs first in time. This leads to some pretty silly conclusions sometimes.

Here's a ridiculous example. I may observe that the air becomes cooler as the sky becomes colorful near the end of the day. In my mind, I may justifiably correlate these two events. If I'm not observing enough data or paying attention to some of the important aspects of the sky, I may also determine that the colorful sky causes the air to become cooler. What I have missed in my jump from correlation to causality is the presence of an underlying cause driving both correlated events. The sun's rays refracting off the air and water in the atmosphere at an oblique angle cause the colors. The oblique angle condition occurs as the sun is setting, without whose heat, the air becomes cooler. The setting of the sun is the cause of both events. The relationship between the two events is a function of the sun's effects.

A common cause for two correlated events is one possibility. Or perhaps the two events are related causally, but through a sequence of events and with multiple events driving the sequence. There are probably many other conditions that could account for two events being correlated without being directly causally related. It's interesting though that we seem to be eager to establish causality even at the risk of drawing ridiculous conclusions. Something in our nature seems to yearn to understand our surroundings. Maybe we want to feel comfortable or safe in some knowledge of what to expect from our environment. The danger is that we become complacent or lazy in our half knowledge and draw conclusions that the data doesn't support just to feel good about ourselves. I'd rather cling to truth even if it forces me to acknowledge myself an idiot than delude myself with half-truths to make myself feel better.

Be cautious about what you think you know. The beginning of wisdom is skepticism about one's own knowledge. The great philosophers of the ages, beginning with Socrates I believe, recognized this principle and encouraged intellectual humility in their disciples by challenging their preconceptions and prejudices. Maybe they knew something that we, in our modern arrogance, have forgotten.