David Hume and the Problem of Induction

We often use inductive reasoning—especially in science. But David Hume, an 18th century Scottish Enlightenment philosopher, identified a puzzle about such reasoning. It rests on an unjustified assumption. So induction might be a good form of reasoning, but we don't have good reason for believing it. This is the problem of induction.

Media Sources:

[ThinkingAboutStuff]

Can you think why we should believe that "the future will behave like the past" without using induction?

[timonh.]

You couldnt make a better way of getting to the point so quick ! Thanks!

[mutantthegreat]

I encourage everyone to look up Greg Bahnsen's 1980s debates where he puts forward a strong argument for the law of induction.

[duder]

The problem of induction can actually be made stronger when we take memories into consideration. We can say that our memory is fallible, so how do we know we are remembering the past correctly? If we say we remember our memory working we are just using circular reasoning. Therefore the statement “the future will behave like the past, ” (the uniformity of nature) is even more unjustified since we can’t know if we can accurately recall the past. So induction is even worse than Hume originally thought because he thought induction could give us probabilities but with this it is impossible to create probabilities.

[Mart-Bro]

Great video. Good script and excellent editing. Nice one

[jus]

i usually never comment but this was a really good/concise video thank you.

[СлавиНиколов-гн]

great video! you should do one on Hume's critique of causality!

[kangar]

This looks more like a video about circular reasoning than induction.

[tonybanks]

It seems to me like Hume is really pointing to a semantic issue. We KNOW the way we interact with physical events makes sense, but we don't express this fact in a sound way. He reveals that deductive logic is the wrong way of expressing it.

Would Hume agree with this alternative way of expressing the truth of our epxerience? :

1) fitting a model to data is a sound operation
2) Among competing models of our reality we find the hypothesis of regularity of laws of nature (RLN)
3) RLN fits the data better than competing models
4) It thus makes sense to presupose RLN
5) If RLN and event A occured lots of time in the past, then it will probably occur in the future

[alekriley]

Underrated channel
Keep up the good work 👍

[TheCategor]

2:40 "We have no good reason for thinking that's true" - Of course you have!! You have entire history to show exactly that - future behaves either exactly like the past, or rhymes extremely closely with the past.

[jichaelmorgan]

This applies even better to soft sciences like economics.

[Skaxarrat]

Quick, to the point and we don't need to read a philosophy book. Thanks!

[danila]

This video saved me 1 point on my Philosophy test

[michaelluckmann]

I went to boil some water. Left the room with it on the burner for a while. When I came back the water had frozen.

[lewis]

"The future will behave like the past" has probability behind it, so it can be supported without circular reasoning.
There have been an infinite number of instances where copper conducted electricity in the future from a a reference point of a specific point in time the same way as the past.
There have been zero instances where it didn't.

the false assumption to copper conducting electricity is that it is a function of time.

[christopherhamilton]

It’s not the ‘future’ that’s behaving like the past though. It’s a physical phenomenon that is occurring.

[havenbastion]

Replication confers certainty and Truth is that which we can be most certain of. Next problem?

[Pravin.Shidore]

It's nothing to do with reasoning. I'm just curious about history of electromagnetic & how electricity & magnetism become united. 🙅🙅🤦🤦🤦

[DrEnginerd]

The analogy doesn’t hold though. The analogy needed to be “Amy in the past paid you back once, she says she will pay you back again”.