3 Scepticism & the Problem of Induction (General Philosophy 2018 - Peter Millican)

welcome to the third lecture on general philosophy and now we're going to be going quite deeply into one of the topics one of the six topics we have already uh touched on several of them including particularly mind and body but now we're going to spend an entire lecture and indeed a little bit more on induction the luminaries we have here are pirro of ellis at the left who will be introduced shortly descartes david hume peter strawson who for a long time was at univ and ended up spending many years at maudlin and simon blackburn who was fellow at pembroke for quite a long time before moving to a certain unmentionable university in the fence so pirro of ellis we don't know very much about him uh what we do know comes from this chap diogenes laertius who enjoyed telling lurid tales of the great greek philosophers and in particular how they died often rather gorilla in pirro's case however he lived to a ripe old age apparently because he had lots of nice friends who followed him around saving him from the consequences of his skepticism so he was a notorious extreme skeptic he saw no reason for supposing that precipices were dangerous or indeed hurrying chariots but his friends kept him out of harm's way and he was okay so um when you hear about pyronian skepticism for example when reading hume that's referring to this this somewhat legendary extreme form of skepticism now the works of loads of ancient greek philosophers um whom we typically know about only through diogenes were lost as i mentioned in i think the first lecture when constantine converted to christianity basically you had a sequence of roman emperors who just destroyed the old pagan schools burnt down libraries etc but some of the ancient texts survived in the byzantine empire and in the arabic world and with the downfall of the byzantine empire a lot of scholars fled to the west and that's standardly credited with kicking off the renaissance i want to particularly focus on a few texts and in particular the texts of sextus empiricus who wrote a work called outlines of pyranism in which he spelled out a lot of skeptical arguments um i i mentioned in the first lecture why over this period from about particularly 1500 to about 1650 there was a lot of tumult in the west a lot of problems were being raised not least um thanks to the reformation so people were debating things questioning whether the things that they thought they knew were as solid as they were and into this came these skeptical manuscripts also with the invention of printing suddenly these works could be spread far and wide so works that had survived literally just one or two copies by chance had survived of these hundreds of texts that we know about through diogenes just a small number survived but they were printed and reproduced and had a huge effect so the outlines of piracy of pyranism i've mentioned um by sextus empiricus he actually lived around 200 a.d um that was translated into latin and printed in 1562 it had a major impact on montaigne he in turn influenced descartes and and bail pierre bail now i've mentioned before that not only were the there all these issues like the reformation and so forth but also modern science itself invited skepticism and of course still does if you go back to the ancients aristotle the idea was that when you perceived something you actually mentally grasped the form of the thing that you were perceiving your mind became in a sense like the thing that it was perceiving but for the moderns it was quite different they saw the world as composed of mata whose form is quite different from what we see so in the case of descartes he thought everything everything material was composed of pure extension for someone like boyle you've got corpuscles in the void but in either case the perceptions that we have of things are very different from the things themselves so again you get an invitation to skepticism now we've already seen some snippets from descartes here are a few more but focus particularly on skepticism so his first meditation very famously presents skeptical arguments now descartes is using this for a reason skepticism is extremely convenient to descartes because it enables him to sweep away the orthodoxy of aristotelianism he can say i'm not going to accept anything unless it can withstand the arguments of the skeptics and an appeal to aristotelian authority is not going to do that now descartes thinks he actually has arguments that will withstand skepticism and we're going to briefly look at those in a moment so famously descartes uh speculated that he might be dreaming that the world might all be an illusion or maybe there's some kind of evil demon putting the perceptions into his mind so again maybe the external world simply doesn't exist um an extreme variety of this kind of skepticism is is called solipsism and solipsism is the claim that the only thing that exists in the entire universe is me my mind everything else is as it were a figment of my imagination now descartes himself of course is not a solipsist he is using skepticism as i said as a tool for getting rid of rival theories he thinks he has an answer which is going to refute skepticism and here i just want to sketch uh the form of that reply and to focus on its essential logical structure so again as with many of these slides i'm going to be skipping over them quite quickly but you've got them there to read at leisure and i hope that they will give you a guide when you're reading the meditations to some of the highlights okay so descartes says i'm going to cast away anything that isn't completely certain that is how i am going to reply to the skeptic i'm only going to hold on to things that are certain when i think about my own existence i realize that there there is a source of certainty i'm aware of my own thinking i could not be thinking without existing so i'm absolutely certain that i exist i think therefore i am very famous more controversially he says that by examining that particular case of certainty he can see what is required in order for something to be certain now that's a very strange move if you think about it because he's identified this one unique special truth i think therefore i am but then he goes on and tries to establish a general rule on the basis of that now it may be that his knowledge of his own existence is somehow special and not just because he clearly and distinctly perceives it to be true but because it has other qualities like for example i cannot doubt without existing the very fact of doubting proves my existence so that's the kind of thing that suggests that i think therefore i am is special my own thinking somehow guarantees my existence that's not necessarily necessarily going to carry over to other claims that descartes thinks that he clearly and distinctly perceives but at any rate what descartes does is quite clear he says what makes me sure that i exist is that i clearly and distinctly perceive it to be true so i can now draw a general rule that whatever i clearly and distinctly perceive is true then he goes on and argues on the basis of things that he claims to clearly and distinctly perceive that god exists in particular in the third meditation he notices that within his mind he has an idea of god he sees that this idea of god has a sort of perfection to it and he claims that that perfection could only come about from a perfect cause so the effect that is the idea of god in his mind could not have been put there except by a perfect being and an example of one of the principles that he appeals to in this proof it is manifest by the natural light that there must be at least as much in the efficient and total cause as in the effect of that cause so if my my idea of god is perfect that perfection must have come from the cause of that idea so there must be a god and a perfect god couldn't be a deceiver so if he's not a deceiver then it must be that i who've been made by him must be capable of truth so when i clearly and distinctly perceive things they must indeed be true because were that not the case god would have been a deceiver and we know that god can't be a deceiver because god's got to be perfect okay i've gone through that very quickly but you may have noticed that there's a little bit of a circularity here descartes seems to be appealing to his clear and distinct perception to prove the existence of god but then having proved the existence of god he says now that i know that god exists and is not a deceiver i can rely on my clear and distinct perception so it looks like he's relying on his faculties to prove the existence of god then he's using god's existence to justify his faculties that looks circular and you might wonder uh in the light of this whether actually it's possible to defeat the skeptic without some such circular argument if the skeptic throws doubt on your very thinking faculties how on earth can you justify them against him so i'm going to leave that hanging just for a moment and introduce the next big figure probably these days the most influential philosopher of the early modern period there are probably more people who think of themselves as humans in active in philosophy these days certainly than cartesians or lochians and probably even cantians part of the reason that david hume's influence is so great these days is because he was attract he was trying to establish a philosophical view of lots of things including ethics and philosophy of religion as well as epistemology and metaphysics that was completely independent of god and in the 20th and 21st centuries that has become the standard way of doing philosophy so hume was called in his day the great infidel he's known as a notorious skeptic so unlike descartes who whose aim is to oppose skepticism hume often seems to be on the side of the skeptic now we'll see that it's not actually so clear in many ways he wants to respond to the skeptic but he's certainly much more sympathetic to skeptical arguments than descartes or locke or indeed any other great philosopher of this period so in your readings uh for general philosophy there are two particular sections of hume's first inquiry which are relevant this is the inquiry concerning human understanding of 1748. one of them is inquiry section 4 that's where he presents his famous argument concerning induction but the other is inquiry 12 where he discusses skepticism and in this in the next lecture i'm going to be appealing to both of those so in the next lecture we're going to go and see more about skepticism and about how hume and a human can reply to skepticism so very early in section 12 hume attacks descartes there is a species of skepticism antecedent to all study and philosophy which is much inculcated by descartes and others as a sovereign preservative against error and precipitate judgment it recommends and universal doubt not only of all our former opinions and principles but also of our very faculties of whose veracity say they we must assure ourselves by a chain of reasoning deduced from some original principle which cannot possibly be fallacious or deceitful okay he calls it antecedent skepticism because the skepticism is being raised even before we've applied our faculties it's casting doubt on our faculties right from the start now hume responds to that but neither is there any such original principle which has a prerogative above others that are self-evident and convincing or if there were could we have advance a step beyond it but by the use of those very faculties of which we are supposed to be already diffident the cartesian doubt therefore were it ever possible to be attained by any human creature as it plainly is not would be entirely incurable and no reasoning could ever bring us to a state of assurance and conviction upon any subject now personally i have to say this seems quite a strong counter to descartes it's saying look if you start out distrusting your own faculties and requiring that your own faculties be justified you've had it there's no way you're going to be able to do it except by using your faculties and since you don't trust your faculties no chance so what's the point of doing that what is the point of pursuing that kind of skeptical philosophy you've simply dug yourself in in a pit you've thrown away the means of getting out of the pit you've had it futile now let's go on and see an argument of humes which is seriously skeptical but not a case of antecedent skepticism of the kind that he accuses descartes we'll be coming back incidentally to this issue about antecedent skepticism particularly in the next lecture now hume's most famous for his argument concerning induction indeed it's probably the most famous and influential argument in all of english language philosophy it appears in three versions in the treats of human nature the abstract of the treatise and the inquiry the version that you read for this course the inquiry version is the best it's the fullest the most polished so you don't need to worry too much about the treatise and the abstract but if you the abstract gives a very nice summary a sort of potted version of the argument it's considered has very widely been considered a dangerously skeptical argument broad um called it the scandal of philosophy a lot of modern philosophy of science has been developed in response to this argument its interpretation still remains controversial in various ways but it continues to be extremely influential so quite unlike descartes arguments that we've been talking about here is an argument which is alive and kicking which many people actually think is correct despite the fact that its conclusion can seem uh wildly paradoxical or skeptical so i'm going to be following it broadly as it goes in inquiry 4. and inquiry section 4 starts with the distinction between two different types of proposition relations of ideas and matters of fact now relations of ideas are propositions that you can know to be true without consulting experience they're a priori we've already come across that term you can know them to be true simply by thinking of the meanings of the terms they contain so they're necessarily true there's no way you can conceive them to be false hume gives examples of pythagoras theorem and three times five equals half of 30. an example a very common example these days is all bachelors are unmarried an example of what is commonly called an analytic proposition it's one such that the meanings of the terms determine its truth so those are relations of ideas a relation of ideas just tells you about the logical connections between the words that you're using or the ideas that you associate with them it isn't telling you anything substantial about the world now against that we have matters of fact and matters of fact are statements that do tell you something about the world they're things that cannot be known a priori you can't know them to be true just by looking at the meanings of the terms and their truth and falsity are therefore equally conceivable so here's an example of a matter of fact the sun will rise tomorrow here's another one the sun will not rise tomorrow notice that when hume talks about a matter of fact he means if you like a question of fact he doesn't mean it is actually true but it's a question of fact whether it's true or false or take this one this pen will fall when released in air ah that turned out to be a true matter of fact so a modern term for this is a synthetic proposition a proposition whose truth is determined by the facts of experience not by the meanings of the terms okay so that seems a very intuitive uh distinction it's been very very influential since can't it's been drawn more in terms of analytic synthetic rather than relations of ideas and matters of fact but it's essentially the same distinction okay now it may seem straightforward to understand how i can know the truth of a relation of ideas i simply look at my own ideas i think through the consequences of the meanings of my terms and i work out that it must be true but how can i know a matter of fact well you may think in some cases that's very straightforward i can know there's a glass in front of me because i see it i can know that there's a glass behind me because i remember seeing it okay so let's not worry about that we're not going to do what descartes was doing we're not going to raise skeptical questions about perception or about memory what we're going to ask is how can we know any matter of fact which we are not either currently perceiving or remembering so for example how can i know that that pen will fall when i let go of it okay let's take a paradigm example of this kind of inference and hume chooses billiard balls so suppose we see a yellow billiard ball moving towards a red one and colliding with it we expect the red one to move don't we why well hume says that we suppose a causal connection a causal connection between the impact of the yellow ball and the movement of the red one okay fine but how do we learn what causes what how do we learn about causal relations and here hume brings in a thought experiment involving adam adam is the very first man he's just been created by god he's a perfect specimen of a human being all his faculties work just as they should because god's just made him okay um so adam's sitting there and god up comes up to him and says look adam here's a billiard ball oh what's that never seen one of those before no you haven't but here's one anyway here's another one uh now i'm gonna roll says god one of the billiard balls towards the other um what do you think will happen adam when that billiard ball hits that one and hume's thought experiment is actually adam would not have a clue adam's never seen this before he doesn't have the faintest idea what will happen when one hits the other because he can imagine lots of different things that might happen one might just stop it might go through right through it might one might jump in the air or go backwards or all sorts of things how could adam without experience possibly know what was going to happen when one billiard ball hit the other he couldn't and that does indeed seem a plausible result of that thought experiment okay now suppose god did this a number of times you know so adam by now he's seen a hundred billion balls and every time the second one has moved and now god says to adam uh what will happen when this ball when the yellow ball hits the red one and adam says oh the red one will move okay what what makes the difference why why does adam's experience of unbelievable hitting another how does it help him in making a prediction about others well it must be because he's able to extrapolate he's able to extrapolate from the experience he has had to the experience he's going to have so hume claims very plausibly all inference to matters of fact beyond what we perceive or remember is based on causation and all our knowledge of causation comes from experience but we can only learn from experience on the assumption that observed phenomena the billiard balls we've seen in the past provide a guide some sort of at least vaguely reliable guide to unobserved phenomena the billiard balls we're going to see in the future so we have to be able to extrapolate from observed to unobserved on the basis that the unobserved will resemble the observed okay do we have a rational basis for making that assumption now that assumption is commonly called the uniformity principle it's not hume's term but hume scholars normally use it to refer to this principle uh here are a couple of ways in which hume expresses it in the inquiry all our experimental conclusions he just means conclusions from experience all our experimental conclusions proceed upon the supposition that the future will be conformable to the past okay how then can we know that the future will be conformable to the past what possible means of justification do we have for the principle of uniformity okay spoiler alert we don't have any hume is going to claim by the way i think he's right that we don't have any source of independent justification for the uniformity principle okay so i now need to take you on a little detour to look at the various sources of evidence that hume thinks might conceivably be available for the uniformity principle or somebody might think were available for the uniformity principle so here's a a little snippet from a document that hume wrote in 1745 that's three years before he published the inquiry it's a shame he doesn't say anything quite this clear in the inquiry um it is common for philosophers to distinguish the kinds of evidence into intuitive demonstrative sensible and moral okay so when hume talks about intuition he means things that are self-evident so if something's an intuitive truth it's self-evident when he talks about sensible evidence he means what we would call sensory evidence evidence from the senses and when he talks about demonstrative and moral reasoning or sometimes he uses the term probable rather than moral reasoning he means two different types of inference uh which were distinguished by john locke so again i'm a brief detour here's an example from locke of demonstrative reasoning uh it's a proof that the angles within a triangle add up to the angles on a straight line and the claim is uh that we can see intuitively we've got two parallel lines we can see intuitively it's self-evident that angle a is equal to at angle e it's self evident that angle b is the same as angle d and it self evidently follows from the first two propositions that a plus b plus c is equal to e plus d plus c so the sort of idea of a demonstrative argument for lock which hume takes over is an argument where every link in the inferential chain goes with absolute and complete certainty so we've got what we would call a deductive argument an argument where if the premises are true you simply could not possibly doubt the truth of the conclusions as following because there's no way the premise could be premises could be true without the conclusion being true now probable reasoning or what hume calls moral reasoning notice by the way when he talks about moral reasoning it's nothing to do with ethics all right this is an old use of the term it's still retained when people talk about moral certainty they don't mean ethical they mean certainty for the purposes of human life or something like that so in probable reasoning you have an argument where the links of the chain are not infallible so in demonstrative reasoning you go from the premises through the steps of the argument to the conclusion with absolute certainty all the way the sort of thing you do in maths the sort of thing you do in logic in probable reasoning the links of the chain are not so certain and this is locke talks about this but gives very little in the way of example an example he gives is this one tell a country gentlewoman that the wind is southwest and the weather lowering and like to rain and she will easily understand tis not safe for her to go abroad thin clad in such a day after a fever she clearly sees the probable connection of all these this southwest wind and clouds rain wetting taking cold relapse and danger of death well a bit dramatic but you you get the idea these are not demonstrative connections the woman is drawing probable connections uh from one thing to another okay and i've mentioned a point there at the bottom which i'll i'll come back to a little later so lock's distinction between demonstrative and probable reasoning pretty much corresponds with what we call deductive and inductive reasoning deductive reasoning is where the premises guarantee the truth of the conclusion uh an inductive argument is one that draws conclusions about typically about the unobserved on the basis of the observed by extrapolation okay so now let's come back remember we were asking what justification could one possibly have for the uniformity principle and we've got four possible sources of evidence according to hume and hume says this is quite standard as it is in his day is it self-evident that the future will resemble the past no it's not self-evident can it be proved demonstratively by deductive reasoning can you prove that the future will resemble the past that because billiard balls in the past have acted in such in such a way that you can logically prove that the next one will act that way no you can't both of these are clearly impossible because we can easily imagine things turning out differently and if we can consistently imagine things turning out differently that's enough to show that we haven't got a logical a demonstrative proof what about sensory knowledge could we know through the senses that future billiard balls will operate in the same way as past billiard balls well no we can't think of the adam thought experiment adam by looking at the billiard ball didn't have a clue what was going to happen until he actually saw it and by the same reasoning he couldn't have had a clue that the billiard balls would act consistently just looking at the superficial qualities of the billiard ball didn't tell him anything about how it's going to behave or whether it will behave consistently so that's no good what about what hume calls moral reasoning or probable reasoning and i'm calling factual inference for short because he often calls it uh reasoning concerning matter of fact well the problem there is that if we use that to justify uniformity we're just going in a circle we have said that all arguments concerning existence again he means all moral probable reasoning all factual inferences are founded on the relation of cause and effect that our knowledge of that relation is derived entirely from experience and that all experimental conclusions proceed upon the supposition that the future will be conformable to the past to endeavor therefore the proof of this last supposition by probable arguments or arguments regarding existence which are just the same thing must be evidently going in a circle and taking that for granted which is the very point in question okay so we cannot justify extrapolating from past to future either on the basis that it's self-evident it isn't or with a deductive argument because there isn't one or on the basis of sensation because that doesn't tell us that the future is going to resemble the past nor on the basis of inductive argument because inductive argument takes for granted that the future will resemble the past so we've got absolutely no justification apparently for the assumption that the future will resemble the past and since all of our inferences beyond what we immediately observe or remember depends upon this assumption or at least the assumption that the unobserved will resemble the observed we act it looks like we haven't got any justification for any belief about anything any matter of fact that goes beyond our immediate perceptions and memory that's a serious worry and notice that hume here he's not doing things like doubting the external world he's not doubting our perception he's not doubting our memory he's simply saying what justification do we have for going beyond those and his answer seems to be we have none we have to depend on this principle of extrapolation for which we can give literally no reason whatever what's particularly scandalous is i'm i cannot give you an answer to this which is going to show that hume is wrong he may well be right okay so the upshot is that our empirical reasoning all of our empirical reasoning wherever we reason from observed to unobserved it's based on a brute assumption of uniformity for which we can't give any ground it's not based on any insight into the nature of things hume draws the conclusion that human reason basically differs only in degree from animal reasoning it's just like you know the dog who when he hears the door go expects to be taken for a walk or give them food or whatever it's just animal habit well basically we're the same we see one billiard ball going towards another we expect the second one to move by habit lock as i mentioned in an earlier slide attributed probable reasoning as well as demonstrative reason reasoning to the perception of connections uh hume says that's nonsense there we don't perceive any connection between one billiard ball moving and the other one moving it's just habit and no causal interactions are really intelligible they're all based on this familiarity what hume calls custom you can see why um hume became very popular in the 20th century i talked in the last lecture about quantum mechanics general relativity and so on people became aware that actually the world is rather weird and our intuitions do not provide a good guide to it in previous generations people may have thought that they had rational insight into the way the world works in the 20th century we were disabused of that complacent assumption we discovered actually that the only way we can find out how the world works is by experiment and observation but what hume has shown is that even relying on experiments and observation we we have to take for granted this basic assumption that the unobserved will resemble the observed okay the general philosophy reading list includes several attempts to justify induction to answer hume um are any of these successful well let's take a little look at them i've i've listed some there we'll be talking a little bit about um the vancleave and reichenbach moves next time but for now i want to focus on the first two so the analytic justification of induction is basically saying that induction is rational by definition right if you ask how can we justify extrapolation from observed to unobserved well that's just the rational thing to do if you don't extrapolate from observed to unobserved you're being irrational so the gambler who says yes i've lost this game 100 times in a row i'm bound to win next time that's just irrational okay um by definition so we we don't need any argument to justify it that's enough probabilistic attempts to justify induction appeal to mathematical probability and there's quite a variety of those we'll be looking at one of those the inductive justification of induction is basically saying induction has worked in the past it will work in the future now hume obviously says that's just circular all right well some philosophers have tried to find ingenious ways to get around that circularity the pragmatic justification of induction says we are pragmatically justified in relying on induction we can't prove that it'll work or even that it's likely to work but we've got good practical reason for relying on it maybe we can give some theoretical argument to back that up by showing that induction will work in giving us predictions if any method will that's the line that reichenbach takes so let's have a quick look at a couple of these so the analytic justification as i say says that induction is just rational by definition what we mean by a rational method of inference is one that infers the same about the future as has been observed in the past so suppose somebody relies on astrology or reading tea leaves well if you investigate these things it finds out you find that they're hopelessly unreliable so we do not consider those rational methods of inference because they've proved not to be reliable that shows that we are judging methods of inference inductively we are treating those that yield truth consistently over time that attribute the same properties to things over time as rational so induction is arguably constitute constitutive of what it is to be rational you can only be rational by definition if you think if you reason inductively okay well maybe that's plausible but does it actually touch hume i don't think it does because what hume will say is this we do of course take for granted that the future will resemble the past we all do it all the time we couldn't survive if we didn't it's absolutely built into us to take for granted that the future will resemble the past and so obviously we think of that as constitutive of human rationality and we people who think differently we're going to treat us mad so what that's just a question of how we judge it it doesn't mean we've got any independent justification for it that just shows that we are indeed taking it for granted what hume is asking or challenging is the thought that we've got some kind of insight that justifies extrapolation he's not questioning that we use the word reason in a particular way because the way in which we use that word actually builds in our assumption that induction is a reasonable way to go that we've got good evidence for it humor's argued that actually that complacent assumption is false we don't have any good reason for supposing that the future will resemble the past so the analytic justification you know may be plausible as an account of how we use words it doesn't touch hume what about probabilistic justification now hume in his argument takes something for granted which could be questioned he takes for granted that all probable arguments all arguments that aren't demonstrative in force have to rely on experience because remember he said you can't have a demonstrative argument that the future will resemble the past for the following reason we can imagine the future not resembling the past we can coherently conceive of the future being different that's enough to show that you can't demonstrate that the future will be the same and then with regard to probable argument the reason you can't justify the future resembling the past by inductive or probable reasoning hume says is that all such reasoning takes for granted that the future will resemble the past so you'd be reasoning in a circle but maybe hume's overlooked something maybe it is possible to have probabilistic reasoning which is a priori where your reasoning from basic probabilistic principles that we see intuitively to be true rather than on the base of experience and if such an argument were possible then hume's argument has not ruled it out okay so there is a gap in hume's argument so probabilistic justifications aim to get through that gap now a lot of attempts have been made i've listed a few of them there roy harrod came up with an ingenious argument which was later developed by simon blackburn and we're going to be looking uh a little bit at that that this is on your reading list okay so here's the ingenious argument suppose i'm crossing a desert i don't know how big the desert is um and i'm making predictions about how long the desert will go on so i'm setting out on a straight path across the desert and i'm making predictions about how long my journey will continue through the desert and suppose i make predictions like this the desert ahead will extend for at least 10 percent as long as the distance i've already traveled through it okay now imagine i i make such predictions again and again and again so i'm walking along i say the desert will last for 10 more the desert will last for 10 more the desert will last for 10 more so i'm making lots and lots of these predictions and of course as i go further and further through the desert my predictions are predicting yet more desert so if i go for 10 miles i'm predicting another mile if i go for 20 miles i'm predicting another two miles if i go for 30 miles i'm predicting another three miles and so on okay now at some point the desert ends and so the predictions i make towards the end of that journey will have turned out false it will not in fact have been true that the desert had another 10 to go beyond what it had already gone and the dividing line the point at which the predictions will stop being true will be 10 11 of the way through the journey okay so if i've gone let's let's suppose that the desert is 110 miles wide then after a hundred miles i will be making the prediction the desert will go on for another 10 miles and that prediction will be right but the very next one i make will be wrong agreed okay so imagine that i'm making these um these predictions sufficiently frequently that we can ignore rounding error ten elevenths of my predictions will come out right okay if i make well let's call them one tenth extrapolations i.e the journey is going to go go on for another tenth as long as it's gone so far ten out of eleven of those will turn out true okay so maybe this justifies making such predictions if i make such a prediction i've got a 10 11 chance of being true so i can justify extrapolation at least to that limited extent and hume is wrong well that's not quite right it's not quite right because imagine that i've imagined that i've traveled miles through the desert and i'm making a prediction the desert will go on for another ten miles i've no idea how long it's going to go on all right but i make that kind of prediction when i say to myself well look what i know is that no matter how long it goes on 10 11 of these predictions will be right so i've got a 10 11 chance that this particular prediction will be right hang on a minute that's not right is it because a load of predictions that i made in the past have already been fulfilled right all the predictions for example that i made up to 90 miles they've already come and gone right when i passed the 99th mile that confirmed the prediction that i made at 90 miles that it would go on for another nine so loads of the predictions have already been fulfilled so now when i ask what is the chance that this prediction will be fulfilled i can't go by this 10 11 chance because that a large proportion or a significant proportion of those ten elevenths that are gonna come out true have already come out true so it's a bit like i mean imagine if i collect the tickets from i suppose there's a weekly lottery and i collect the tickets from the previous nine lotteries okay and then on the tenth week i buy a lottery ticket and i put it into a bag together with those uh the nine previous winning tickets and i say there's a nine ten chance of me winning this lottery because i've got it's one of ten tickets and nine of them are winners no can't do that the nine winners have already been and gone i'm interested in whether this particular ticket will win so i can't i can't load the dice as it were like that well harrod suggests that there's a way of correcting for this um basically you average over all the different positions that you could be on your way through the through the desert um and he suggests you you get a slightly more modest prediction of 100 over 121. let's not worry about that i mean it's a it's a reasonable point but blackburn points out that it doesn't actually solve the problem because even if i average for all the different positions that i could be through my journey the fact remains that at 100 once i've got to 100 miles there is still the fact that i know that all these previous predictions the ones i made up to 90 miles and indeed a bit beyond 90 miles all those have already been gone so averaging over every possible predict place i could be within the that within the overall desert you know am i in the first half of it or the first quarter or the last 10th or the last 11th i just don't know all i know is that a lot of predictions that i've already made have been and gone i don't know what proportion of those already successful predictions i don't know what proportion they form of the whole unless i know how big the desert is so what blackburn does is he says well here's a way of saving harrod's argument rather than focus on the believer focus on the skeptic and say to the skeptic look you are denying that we can justifiably rely on these one-tenth extrapolations but look harrod's argument still holds that in general if people make inferences of this kind extrapolations of this kind 10 out of 11 of them will be successful so you as a skeptic are casting doubt on something which 10 out of 11 will be successful how do you justify that well it's a neat effort but i don't think it works let's first of all a couple of general points first of all i want to suggest that the the focus that harrod introduces and blackburn retains on the proportion of extrapolations that will come out true is simply giving a false impression of the logic of the situation it's like with those lottery tickets that you put into a bag the extrapolations are sequentially ordered they're not a random group it's not like you've got a load of predictions and you randomly pick one out of it and 10 out of 11 of these are true the extrapolations come strictly in order they are made in order and they are fulfilled in order um and the the extrapolation that i make now can only be true if all of the ones before were true so thinking in terms of choosing one from a group is just a misleading way of representing the situation another important point against blackburn is that appealing to a general practice of making this kind of inference is hopeless for a different reason and that is that the 10 out of 11 proportion applies to every individual uniformity right every journey you make if you make these 10 11 extrapolations sufficiently frequently 10 out of 11 of them will be true sorry i think i need to repeat that steve any journey at all if you make one tenth extrapolations sufficiently frequently then 10 out of 11 of them will come out true in that individual journey so it's not like you've got 10 11 true in some large sample some large group from which you are picking an individual example rather for every individual journey you do the proportion is 10 out of 11. so when i ask is my next extrapolation likely to come out true statistically the only sequence that matters is the current one the current journey what what may be true about other journeys just is not relevant despite blackburn and the only extrapolation that matters is the one i'm about to give appealing to the statistics of extrapolations in general won't help as i say they're sequentially ordered it's not like choosing one from a set unless i know how far i am through the desert or the period of uniformity if we're applying this to induction i can't know what proportion of extrapolations have already been fulfilled so i can't apply the statistics anyway so unless i can justify extrapolation from observed to unobserved it seems i'm no better off than when i started out so it's an ingenious attempt to justify induction it's trying to show that one-tenth extrapolations modest extrapolations from experience are justified statistically if that were true we would be able to justify induction at least to some extent we could say the world has been uniform for at least you know a hundred or a thousand or maybe 13 billion years therefore it's likely to continue to be uniform for a tenth as long again but sadly it doesn't work next time we'll be looking more at skepticism and briefly considering a couple of other attempts to respond to hume skepticism thank you

Channel: Philosophy Overdose

Views: 1,613

Rating: 4.9444447 out of 5

Keywords: Philosophy, David Hume, Hume, Epistemology, Problem of Induction, Scepticism, Causation, Metaphysics, Skepticism, Empiricism, Constant Conjunction, Induction, Analytic Philosophy, History of Philosophy, Necessary Connection, Metaphysics of Causation, Causality, Nature of Causation, Knowledge, Inductive, Descartes, Theory of Knowledge

Id: CbAaFoYHyzI

Length: 53min 3sec (3183 seconds)

Published: Tue Nov 17 2020