3.3 The Problem of Induction
HTML-код
- Опубликовано: 12 сен 2024
- A series of lectures delivered by Peter Millican to first-year philosophy students at the University of Oxford. The lectures comprise the 8-week General Philosophy course and were delivered in late 2009.
Slides for all his lectures can be found here:
podcasts.ox.ac....
Learn more about Oxford:
www.ox.ac.uk
It sounded like he was speaking at a hospital. If I was the lecturer, I would stop the lecture for a moment and say: "I ask those who heed to cough, or are having trouble with their throat or are sick to go outside and drink some water."
Hi @Natasha Lim, our podcasts are free for anyone to download via iTunes or podcasts.ox.ac.uk, including in some cases, slides and notes as well.
All of human knowledge and understanding is so flimsy. It's amazing.
This man wears a magnificent wool helmet.
Peter Millican is a great lecturer.
Here, reasoning means drawing logical conclusions from premises established by us. Only those things can follow from demonstration or pure reasoning which we ourselves define. We call an object of a particular kind a pen. We name the act of unclasping "releasing". The existence of said things are completely within our control. However, we do not define the laws of nature. As the laws of nature are not creations of the mind, we cannot predict how they work unless we've experienced them.
He is good and made very simple.
Basically, Hume's saying that we only think this from experience, he goes on to explain that if you hadn't experienced anything, you would have no idea of what would happen if you released a pen.
what the enemies of induction, such as hume, simply don't realise (or confess to), is that they necessarily require the use of induction to present their objection to its validity.
behind all objections to induction lies the assumption that refutation is justified for inductive inferences as soon as a single dis-confirming instance is perceived or can be conceived of. i.e. there is a hidden premise buried in the assumed justification to refute a claim simply because a dis-confirming instance is possible. it is induction that teaches us that as soon as a dis-confirming instance arises it contradicts the law identity and for _that_ reason it should be disregarded as truth preserving. this might be better described a meta-induction...where we are not only required to analyse the statements in an argument but also the information the statements contain, and how they are arrived at.
induction does not always require large samples of instances either. many samples might only contain two possibilities as allowed by the law of identity, such as binary relation; ‘on or off’, ‘true or false’. therefore when hume moves from premises to conclusion he must rely on experience that such a move is justified...how is this move achieved but through experience?
more importantly, by invalidating induction he simultaneously destroys deduction. the way the relation of ideas is described here assumes a platonic understanding of ideas, i.e. they are unconnected to the physical world, they can and do exist without the sequences of events that brought them about in a physical world. often this is done by the device of 'conceiving of' unobservable possibilities such as using the thought experiment which assumes a mind fully furnished with ideas but with no historical physical career.
take a single example of unmarried bachelors; we have a statement that reads..."all bachelors are unmarried" and "Mr A is unmarried", which allows us to 'deduce' that because bachelordom is synonymous with unmarriedness "Mr A is therefore a bachelor". this so-called _'relation of ideas'_ is simply a non-contradictory restatement of the law of identity as it pertains the binary relationship, married vs unmarried. therefore we don't need an infinite series of instances to determine that married and unmarried are the only two choices to choose from, and that because they are distinct from one another given their contradictory nature in relation to one another, it is an inductive step that we take when we assume bachelor will always mean unmarried...i.e. that bachelor is another name for unmarried. but the 'idea' not given any consideration in this example is the word 'all'. if ‘all’ bachelors are unmarried we admit the absolute numerical significance of the assertion and do so because experience and thus induction that informs us that the concept ‘all’ has an identity that excludes the possibility of ‘some’ instances not conforming.
the whole concept of movement from valid premises to a valid conclusions requires induction to set up the standard. put differently, why are we justified in believing that valid premises must necessarily lead to a valid conclusion? and HOW did we come to realise this? it is the law of identity that facilitates the philosophers conviction in this regard. but how do we come to recognise the law of identity? we can only do this by experience...specifically through sheer repetition of instances....thus induction.
induction is therefore the process of building up _'ideas'_ by identifying non-contradictory nature of instances, generalising similarities while omitting measurement, then predicting how they should behave given their nature with inference to past experience, and then testing the conclusions to reality. if the tests prove positive the ideas become a rung in a ladder moving up the information chain. deduction conversely assumes the 'ideas' are valid and can therefore move down the information ladder to further undiscovered non-contradictory consequences of the law of identity.
Very fascinating subject.
I really wish that guy in the background would stop coughing!!!!
I didn't realise he was coughing until you mentioned it. Thanks. Hahaha
I agree. reasoning. I've always held scientific induction to be a form of a priori statistical reasoning. The majority of arguments against induction criticize it's fallibility, but it has been shown that reasoning may, indeed, be fallible - for example, the existence of non-euclidean geometry, and the variability of time/space , the existence of superimposed quantum states, all go against the basic sensual notions.
I think the difference is that relations of ideas are based on the identity principle. They are prepositions that do not add anything to their subject. For example, "a triangle has 3 sides" is a relation of ideas because it is inherent in the idea of triangle that it has three sides. This is not valid for "the sun rises each morning", because the idea of the sun does not contain the idea of rising, that is something we extrapolate from experience, because see the sun rise each morning.
way to ruin the vids for everyone else cougher #1
+Jobie Wan thought i was the only 1 that noticed
Thought experiment: if Hume's perception is to see a yellow billiard ball strike a red one then (A) is Hume, in fact, actually playing snooker or pool and not billiards and (B) should his friends tell him?
The assumption made with typical indifference, is the personal beliefs such as witnessing of an event, word meanings and expectations of what it is that others wish to hear or the conveyor of the information wants others to believe. The biologists tell us that most of the brain is there to deal with the human social environment.
Wether it's a fact or not, the word "rational" implies that an observer is applying a "sense of proportions", which by default, also includes the idea of limited knowledge, and therefore the kind of mathematical incompleteness that should cause doubt, scientifically. That's what experience and experiments are for, to limit the proportion of likely errors.
The place for Induction, as it appears to be for the era in which Hume is doing an analysis, was then, and still should be, in Mathematics where it's essential to hypothesize, creat a synthetic proposition, and analyze the outcomes before testing in the real world, and that's uniquely human. Self-satisfying beliefs are the animal equivalent of expectations of a particular situation according to the "fight or flight" type reaction. Is this Pavlovian induction?
According to experience-first, accompanying education, everyone has to learn to trust their own intelligence and understanding. The scientific reason is at greater personal cost socially but returns the investment many times over if everyone else does the same. (?)
i wonder what Hume would say about Bayesian inference....
Maybe it could be better explained if I said that the sun would not stop being the sun if it did not rise, while a triangle would stop being a triangle if it did not have 3 sides
Where can I get the power point slides for this lecture? I can't find them in the link.
Hi C B, they are there, but admittedly it's a bit tricky! At the time of writing, look down the list of items (or search for 'slides') to find the entry for slides for particular lectures. Click on * that * link (nearly there) and there's a download link under 'Download Media' on the right of that page.
cracking lecture
Aren't these in the description ?!
lOVE IT!
you can on itunes U
It's ironic: many intellectuals criticize people with "faith," yet rely on faith for nearly every bit of knowledge they have ever acquired. Induction is a type of faith.
True. But the problem starts when the faithfuls try to impose their faith upon others, seek extra privileges for their faith and try to suppress inquiry because of their faith. Just saying.
I think a lot of people have a chip on their shoulder, atheists, Christians, Muslims, agnostics, all of them. They can't stand people that don't agree with them. There really is no such thing as "tolerance."
Perhaps, but at least the "faith" in uniformity is justifiable and necessary. Hume did make a distinction between what follows from that assumption vs what follows from other superstitious forms of faith.
very clear
"One can not go farther downhill where it concerns the realm of thought" (GWF Hegel on Hume's philosophy)
You guys do know how this inductive reasoning discussion right? Aristotle was wrong, Francis Bacon has it right 'its a matter of weaker argument versus stronger', and Isaac Newton is right science needs both inductive and deductive. Also, Renee Descartes trying to prove God with inductive reasoning was wrong, Thomas Aquinas whom believes in God said Renee Descartes was wrong, granted, I agree.
So, yeah the speaker is right, old news, God can not be proven as a matter of fact or analytically, which God has always been a possibility, however, we have reason to believe there is one, plus, is God worth believing, has the concept of God helped Humanity or has it helped the unfolding of civilization? What is consciousness? Consider Dark Matter, consider the size of Universe.
Yes, in a way it is a valid thing for Atheist to say we believers are making a positive claim, so we have to prove it, well, I say if God is intentionally not revealing his existence to us then it's futile to ask 'that'. It's always been more of a matter of belief, of choosing what to believe, because it's incumbent for an intelligent mind to ponder 'such' question.
So, that's the problem with inductive reasoning, however, it's not a problem for me to believe in a monotheistic God, instead to me God is: a solution, redemption, salvation, justice, peace of mind!! Glory to God, the Alpha and Omega.
Either way, I know my place, ultimately wretched.
Also, I'll add this, candidly, there are thing's or decisions making that I do not wish to not put myself in God's place, ie. pain and suffering, like why can't we regrow an arm, but perhaps it has to be that way, there's a philosophy for having to choose the lesser evil, not being able to have it one way without the other, which I stray or stay away from, because I've grown weak, and like I said we are ultimately wretched.
Not a 'load' at all.. Try to release your pen inside a spaceship with zero gravity and hopefully you'll get it. ;)
The fact that all bachelors are unmarried is untestable, as are all facts, therefore I dont see how "All bachelors are unmarried" can be an idea, because ideas are usually testable.
Can't you go around asking unmarried men if they were bachelors ? Could that be considered a test ?
But isn't the relationship between the word bachelor and the idea "unmarried" inherently based in language? So the statement that all bachelors are unmarried probably just echoes the idea that this assemblage of symbols refers to this particular idea by common agreement.
Go Induction go!!!!
That requires money because education is not free sorry