chaz720.net - journal

November 30, 2001

We, as people, like to know if something is going to happen, and why particular events occur in our lives. When we are young, the consequences of an action tend to be shrouded from us, as we have never seen such an action take place. In order to draw a conclusion about what will happen following an event, we must rely on our past experiences, and as adolescence, we have few. As we grow older, and our past becomes full of various happenings, we start getting more comfortable at being able to predict these cause and effect relationships, and we make these predictions more often then we sometimes realize. Predictions of this kind, those based on repeated observations of similar occurrences, are referred to as inductive reasoning. But just how good are these predictions we make? What kinds of assumptions are taken in order to draw these everyday conclusions, and are they safe? As you can see, a number of problems emerge in even the simplest of inductive conclusions, and although they may seem very reasonable, Hume believes it is not a very rational way to think.

All inductive arguments follow a pattern similar to the following example given by Hume. First you are given a piece of bread to eat and told it will nourish you. You can see that it has a crust, and that it has a soft center. It gives off a distinct smell that you can distinguish from others, and it�s rough size and shape can all be taken in through the senses. When you do eat it, the texture and taste are recorded into memory along with the other properties, and after all is said and done, it gives you a sense of nourishment, just as promised. Given the bread a second, third, and forth time, you tend to notice a bit of a trend. Same look, same smell, same feel, and same taste, provides you with the same nourishment as the first piece of bread. Now, let�s say you are feeling hungry at some point in the future, and at the very moment it just so happens you stumble upon a croissant. First you can smell it and it smells similar to bread. You pick it up and it looks and feels something along the lines of a bread-like substance. So you figure eating it will provide you with the nourishment you were seeking. Right there you have drawn a conclusion that has absolutely nothing to do with associating the object�s physical properties to it�s ability to provide nourishment, but rather you projected the wholesome properties of bread onto the croissant simply because they appeared similar. While it may seem very casual to you, you just ingested a foreign object based on the assumption that it had similar qualities of bread, even though you have never encountered an object of it�s kind before. And this is the essence of inductive reasoning.

Hume sees a rather large hole in any conclusion reached via inductive reasoning, and although he admits that it can often be reliable and useful, he does not believe it is philosophically sound. Russell presents a good example to explain why inductive reasoning is irrational. Take a chicken whose farmer has fed it every day of its life. This may go on for years until the chicken grows to the proper mature age. And finally one day the farmer, instead of feeding it, wrings its neck, cooks it, and eats it for dinner. Through inductive reasoning the chicken was under the impression that the farmer was going to continually feed it for all eternity, and instead the chicken ended up feeding the farmer. In this case, and I don�t think the chicken would argue otherwise, he only really had to be wrong once. A rational conclusion would not have left the rather inconvenient alternative of death out on the table, however as it would appear, there was no way for the chicken to draw a rational conclusion from nothing but a bunch of happy memories of being fed. A similar course of debunking can be taken when looking at the belief that the sun will surely rise tomorrow. Yes, the sun has risen every day of our lives. Yes, as best we can gather the sun has risen every day prior to our existence. However, as reassuring as it would be to expect that the sun is going to rise tomorrow, we can�t form a rational conclusion that it will. Remember, just like the chicken, we only have to be wrong once. Hume argues that all inductive reasoning follows just as the roasted chicken and the post-apocalyptic earth, in that as many times as something has happened in the past, it is foolish to believe that means it will always happen in the future.

Now that we�ve seen a couple examples where inductive reasoning fails, let�s take a look at why it doesn�t always work out. There is a concept called �the uniformity of nature�. Briefly, it is the belief that every single thing that happens in nature, happened in accordance with some law to which there is no exceptions. The key here is to find the appropriate laws that will stand through just about anything amidst all the ones that only appear to be true. The sun rises every morning, not because it�s a law in and of itself, but because the earth spins in a predictable pattern in accordance with Newton�s laws of motion. So if something smashed into the earth causing the sun never to rise again, such an occurrence would not be in err of the laws of motion, even though it would set straight any beliefs that the sun always rises in the morning. The more specific law folds while the more general law stands. Hume would go further to say that even such a seemingly concrete thing, like the laws physics, cannot be absolutely counted upon, because they are still only valid if current trends continue. And even though we cannot conceive of a situation where they would fail, every time we make use of them we are sticking our necks out.

While it is essentially impossible to refute Hume�s claim in an absolute sense, there are a number of ways in which we can augment the concept of inductive reasoning to make his argument rather trivial. The first of which is Russell�s theory of increasing probability. This theory states, in a nutshell, that the more times you witness the occurrence of two things together that are believed to be related, the more likely it is that the relation of the two things is certain, with no bounds. In other words every time we see the sun rise in the morning, it gets that much more likely that the sun will always rise the following morning as well. As you can see, while this argument isn�t nearly as concrete as �the sun will always rise in the morning� it does escape Hume�s attack in that even if the sun doesn�t rise tomorrow, that�s not to say that it wasn�t a good place to have put your money. Another argument raised against Hume�s inductive reasoning attack lies in the consequences of a belief based on inductive reasoning being wrong. As far as the situation of the sun goes, it is our belief that the sun is going to rise tomorrow. A conclusion such as this is very reasonable, and this is primarily because of the grim consequences if the conclusion were to turn out wrong. As long as the sun keeps rising in the morning the conclusion works beautifully. If at any point the sun ceases to rise in the morning, I strongly doubt that very many people are going to wish that they had seen it coming, as prior knowledge of such an event is pretty darn useless.

However, rather predictably through inductive reasoning, these refutes of Hume�s criticism have their own problems. First we look at Russell�s theory of probability. This states that, rather than guaranteeing a relationship after so many observations, subsequent observations merely make the relationship more and more probable. This distinction makes inductive reasoning all but useless for answering scientific or philosophical questions because of the uncertainty that it lends itself to. No one really wants to know what the meaning of life probably is, nor do they want to know what is most likely the theory of relativity. The second argument has a rather obvious problem of only working in situations where a mistake is catastrophic. But with respect to casual conclusions reached with induction, it offers little recourse if the conclusion turns out to be wrong. If we were to flip a coin ten times, all ten times it was heads, and we concluded that all coin tosses result in heads, only to get tails on the next toss, our world would not be turned upside down, but this argument doesn�t hold an recourse for us. While it appears that these arguments refute Hume�s criticisms, what they actually do is redefine inductive reasoning straight out of Hume�s scope of concern. They both seem to try to define inductive reasoning as �good enough� however imperfect it may be. But whether or not it�s good enough for you is your choice. Hume chooses no, and that is perfectly valid.

54523252 | 1.0 | 8ms

| |