_003 David Hume & Induction

 Many people think that science is the true path to knowledge, and that belief and assumptions are false and unjustified. But David Hume, an 18^th Century Scottish philosopher claimed that even science makes unfounded assumptions in order to access truth.

Hume believes that we can only gain new knowledge through synthetic statements (statements which are not true by definition), which he calls matters of fact, which require observation of the world to validate their truth (he is writing before Kant’s “a priori synthetic statements”).

Yet we claim to know things we haven’t empirically observed ourselves, like that the Earth is round. Furthermore, we say that we know that certain things will happen, like that the Sun will rise tomorrow morning, even though we can’t possibly have observed an event that hasn’t happened yet. So we use induction:

Induction is a type of logic which goes as follows: we assume that if something is true for all we have observed, then it is true for all cases. We observe things regularly enough until we can safely say that it will also apply to cases that we haven’t observed yet.

1) Event X has happened to object Y each time we have observed Y

2) Event X will always happen to object Y

One problem with this is “black swans”, we used to think all swans were white, knowledge gained through observation, which we assumed applied to all swans, but it turned out that Australia has black swans. No matter how many cases we observe, we can’t say it applies to all cases.

This also assumes that the future resembles the past, which is just a belief. But surely nature is governed by laws such as gravity and motion which means the future will resemble the past?

This falls victim to the same problem – since we are assuming that the laws of nature themselves will continue to function in the future in the same way as the past. We assume that apples will fall from trees because of gravity, but how can we assume that gravity will still perform in the future as it has done? This is an unjustified belief.

One argument is actually to use induction in order to justify this claim:

1) We have observed that future events resemble past events

2) Future events will always resemble past events

But this is an example of a circular argument; we cannot justify induction with induction itself. It’s the equivalent of claiming that everything in the bible is true, because it says so in the bible.

Another solution is to make a weak claim: rather than saying that X will always happen to Y, we say that X is very likely to happen to Y, which becomes more likely as we observe more regularly. This generates other problems, but we’ll save those for another day.

To conclude, science works well and consistently, but we can never fully trust that its claims apply to every case that we haven’t observed.

Word Count: 498

Hume, David, [EHU] 1975, Enquiries concerning Human Understanding and concerning the Principles of Morals, reprinted from the posthumous edition of 1777 and edited with introduction, comparative table of contents, and analytical index by L. A. Selby Bigge, MA. Third edition with text revised and notes by P. H. Nidditch. Oxford, Clarendon Press.

http://plato.stanford.edu/entries/induction-problem/