Ports of Call  

 

Part II

Hume claims: "...even after we have experience of the operations of cause and effect, our conclusions from that experience are not founded on reasoning, or any process of the understanding." (Pg. 328) This is the crucial claim in the formunation of the problem of induction. Let us consider the nature of inductive arguments.

Hume writes:

If a body of the like quality and consistency with that bread, which we formerly ate, be presented to us, we make no scruple of repeating the experiment and forsee, with certainty, like nourishment and support. Now this is a process of thought of which I would willingly know the foundation. (Pg. 329)

The process of thought which Hume is describing is an inductive inference. Put in the form of an argument it goes something like this:

1. Piece of bread1 nourished me.
2. Piece of bread2 nourished me.
3. Piece of bread3 nourished me.
.
.
.
N. Piece of breadn nourished me.
---------------------------------------------------------
N+1 Piece of breadn+1 will nourish me.
All bread will nourish.

Pg. 329 Past And Future

The bulk of cases of induction have to do with the leap from the evidence you have to a conclusion about the future. Very likely even cases involving the present can be reduced to this case. (So, I will ignore them.) Hume is trying to figure out what justifies us in taking this inductive leap. What makes premises which assert things about the past relevant to claims about the future? It appears that what we need to make that leap justifiable is to know that nature is uniform or that the future will resemble the past in relevant respects. Thus, suppose one were to claim that all the evidence about bread in the past provides no evidence which is relevant to conclusions about whether bread will nourish in the future. How could such a challenge be answered? One might answer this challenge by adding another premise to the argument. The argument would then look like this:

1. Piece of bread1 nourished me.
2. Piece of bread2 nourished me.
3. Piece of bread3 nourished me.
.
.
.
N. Piece of breadn nourished me.
The future will resemble the past in relevant respects.
---------------------------------------------------------
N+1 Piece of breadn+1 will nourish me.
All bread will nourish.

Now we can say that the premises provide evidence which is relevant to the conclusion. But how do we know that the premise that the future will resemble the past in relevant respects is true? According to Hume all truths are either relations among ideas or matters of fact. This division into these two kinds is often called Hume's fork. The fork is a crucial piece in the problem of induction. For such causal truths must be either relations among ideas or matters of fact. Hume explores both possibilities. If a proposition is a relations among ideas and true, then its negation should be a contradiction. But the claim that the future will not resemble the past in relevant respects does not seem to be a contradiction. So it is not a relation among ideas. Therefore it must be a matter of fact. But if it is a matter of fact we need experience to show that it is true. So we start collecting experiences.

1. Future (1) resembled past (1) in relevant respects.
2. Future (2) resembled past (2) in relevant respects.
3. Future (3) resembled past (3) in relevant respects.
.
.
.
n. Future(n) resembled past(n) in relevant respects.
--------------------------------------------------------
n + 1 Future(n + 1) will resemble past(n + 1)
All futures resemble the past in relevant respects.

The problem is that once again one can argue that the truth of the premises is irrelevant to the conclusion because the premises are about the past and the conclusion is about the future. So in order to justify this inductive argument we need to know that the future will resemble the past in relevant respects! So this reasoning is hopelessly circular. Since the claim that nature is uniform is not a relation among ideas and we cannot justify the claim if it is a matter of fact, we are left with the unpalatable conclusion that we have no justification for the claim that the future resembles the past in relevant respects, and thus we have no justification for any inductive leap.

How deep does the problem of induction go? That answer is that the problem goes very deep indeed. First, inductive inference is far more than common-place in every day life. We simply could not go on living without assuming that bread will nourish, that the air we breath will continue to sustain us and the ground we walk on to support us. The problem of induction also effects all branches of learning which deal with matters of fact. Thus, it seems to undermine the justification of all scientific laws (insofar as these are not mere historical curiosities and are supposed to apply to the future).. Students are sometimes inclined to invoke probability theory to justify scientific laws. Unfortunately the problem of induction applies to probability theory as much as it does to anything else. Claims about relative frequencey of events are always inductive leaps from observed past instances to future ones. What justifies this leap? Surely only the claim that the future will resemble the past in relevant respects.

Hume's Solution

Hume basically claim that our belief that the future will resemble the past is largely a matter of habit or custom. We are forced by nature to believe these things, even though reason provides not justification for them. Given the examples cited above, this has a considerable amount of plausibility to it. Some conclusions are forced on us. (Imagine if you were constantly worried that all the oxygen in the room you are in might migrate to a corner of that room leaving you gasping for air, or that the floor might suddenly cease to support you, or that your lunch might actually kill rather than nourish you. Perhaps scientific laws are less necessary for survival, but insofar as the way of coming to such conclusions has been forced on us by nature, it may well be less than surprising if the same process holds true in cases where our survival is not so much of an issue.

 

BACK   4 of 5   NEXT