1
00:00:05,830 --> 00:00:09,340
Welcome to the third lecture on general philosophy.

2
00:00:09,340 --> 00:00:19,300
And now we're going to be going quite deeply into one of the topics one of the six topics we have already touched on several of them,

3
00:00:19,300 --> 00:00:29,560
including particularly mind and body. But now we're going to spend an entire lecture and indeed a little bit more on induction.

4
00:00:29,560 --> 00:00:36,220
The luminaries we have here are Pierro of Ellis at the left, who'll be introduced shortly.

5
00:00:36,220 --> 00:00:45,610
Descartes. David Hume. Peter Strossen, who for a long time was Univ. and ended up spending many years at Maudlin.

6
00:00:45,610 --> 00:00:56,830
And Simon Blackburn, who was fellow at Pembroke for quite a long time before moving to a certain unmentionable university in the Fenton's.

7
00:00:56,830 --> 00:01:03,100
So Pierro of Ellis, we don't know very much about him.

8
00:01:03,100 --> 00:01:07,150
What we do know comes from this chap, Diogenes Lazarus,

9
00:01:07,150 --> 00:01:17,740
who enjoyed telling Uri Tales of the Great Greek philosophers and in particular, how they died often rather go orally.

10
00:01:17,740 --> 00:01:21,640
In Pirro's case, however, he lived to a ripe old age,

11
00:01:21,640 --> 00:01:29,410
apparently because he had lots of nice friends who followed him around, saving him from the consequences of his scepticism.

12
00:01:29,410 --> 00:01:31,480
So he was a notorious extreme sceptic.

13
00:01:31,480 --> 00:01:42,820
He saw no reason for supposing that precipice is dangerous or indeed hurrying chariots, but his friends kept him out of harm's way and he was OK.

14
00:01:42,820 --> 00:01:48,520
So when you hear about tyranny and scepticism, for example, when reading Hume,

15
00:01:48,520 --> 00:01:56,650
that's referring to this, this somewhat legendary extreme form of scepticism.

16
00:01:56,650 --> 00:02:05,530
Now the works have loads of ancient Greek philosophers whom we typically know about only through Diogenes were lost.

17
00:02:05,530 --> 00:02:13,140
As I mentioned in, I think the first lecture when Constantine converted to Christianity.

18
00:02:13,140 --> 00:02:19,600
Basically, you had a sequence of Roman emperors who just destroyed the old pagan schools, burnt down libraries,

19
00:02:19,600 --> 00:02:28,890
etc. But some of the ancient texts survived in the Byzantine Empire and in the Arabic world.

20
00:02:28,890 --> 00:02:40,920
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.

21
00:02:40,920 --> 00:02:49,140
I want to particularly focus on a few techs and in particular, the techs of sex empiric us,

22
00:02:49,140 --> 00:02:56,830
who wrote a work called Outlines of Piron ism, in which he spelled out a lot of sceptical arguments.

23
00:02:56,830 --> 00:03:05,950
And I mentioned in the first lecture why over this period, from about particularly 1500 to about 16 50, there was a lot of tumult in the West.

24
00:03:05,950 --> 00:03:10,960
A lot of problems were being raised, not least thanks to the Reformation.

25
00:03:10,960 --> 00:03:18,820
So people were debating things, questioning whether the things that they thought they knew were as solid as they were.

26
00:03:18,820 --> 00:03:28,870
And into this came these sceptical manuscripts. Also, with the invention of printing, suddenly these works could be spread far and wide.

27
00:03:28,870 --> 00:03:34,780
So works that had survived literally just one or two copies by chance had survived.

28
00:03:34,780 --> 00:03:38,200
Of these hundreds of texts that we know about through Diogenes,

29
00:03:38,200 --> 00:03:45,200
just a small number survived, but they were printed and reproduced and had a huge effect.

30
00:03:45,200 --> 00:03:50,510
So the outlines of Paris apparent ism I've mentioned by Sixtus imperious.

31
00:03:50,510 --> 00:03:58,820
He actually lived around 200 A.D. that was translated into Latin and printed in fifteen sixty two.

32
00:03:58,820 --> 00:04:03,740
It had a major impact on Montaigne. He, in turn, influenced Descartes.

33
00:04:03,740 --> 00:04:14,950
And but Bale PR Bale. Now, I've mentioned before that not only were there all these issues like the Reformation and so forth,

34
00:04:14,950 --> 00:04:19,480
but also modern science itself invited scepticism.

35
00:04:19,480 --> 00:04:28,600
And of course, still does. If you go back to the ancient Aristotle, the idea was that when you perceived something,

36
00:04:28,600 --> 00:04:32,770
you actually mentally grasped the form of the thing that you were perceiving.

37
00:04:32,770 --> 00:04:37,610
Your mind became, in a sense, like the thing that it was perceiving.

38
00:04:37,610 --> 00:04:46,550
But for the moderns, it was quite different. They saw the world as composed of matter, whose form is quite different from what we see.

39
00:04:46,550 --> 00:04:53,480
So in the case of Descartes, he thought everything, everything material was composed of pure extension of someone like Boyle.

40
00:04:53,480 --> 00:05:03,530
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.

41
00:05:03,530 --> 00:05:10,400
So again, you get an invitation to scepticism. Now we've already seen some snippets from Descartes.

42
00:05:10,400 --> 00:05:15,560
Here are a few more focussed, particularly on scepticism.

43
00:05:15,560 --> 00:05:21,140
So his first meditation very famously presents sceptical arguments.

44
00:05:21,140 --> 00:05:24,500
Now, Descartes is using this for a reason.

45
00:05:24,500 --> 00:05:32,780
Scepticism is extremely convenient to Descartes because it enables him to sweep away the orthodoxy of Aristotelian ism.

46
00:05:32,780 --> 00:05:37,910
He can say, I'm not going to accept anything unless it can withstand the arguments of the sceptics,

47
00:05:37,910 --> 00:05:42,270
and an appeal to Aristotelian authority is not going to do that.

48
00:05:42,270 --> 00:05:51,820
Now, Descartes thinks he actually has arguments that will withstand scepticism, and we're going to briefly look at those in a moment.

49
00:05:51,820 --> 00:05:58,990
So famously, Descartes speculated that he might be dreaming that the world might all be an illusion,

50
00:05:58,990 --> 00:06:03,460
or maybe that some kind of evil demon putting the perceptions into his mind.

51
00:06:03,460 --> 00:06:08,980
So again, maybe the external world simply doesn't exist.

52
00:06:08,980 --> 00:06:15,160
I'm an extreme variety of this kind of scepticism is is called solipsism,

53
00:06:15,160 --> 00:06:22,390
and solipsism is the claim that the only thing that exists in the entire universe is me, my mind.

54
00:06:22,390 --> 00:06:27,740
Everything else is as it were a figment of my imagination.

55
00:06:27,740 --> 00:06:36,980
Now, Descartes himself, of course, is not a solipsistic, he is using scepticism, as I said, as a tool for getting rid of rival theories.

56
00:06:36,980 --> 00:06:41,420
He thinks he has an answer, which is going to refute scepticism.

57
00:06:41,420 --> 00:06:45,830
And here I just want to sketch the form of that reply.

58
00:06:45,830 --> 00:06:54,470
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.

59
00:06:54,470 --> 00:06:57,560
But you've got them there to read at leisure.

60
00:06:57,560 --> 00:07:04,760
And I hope that they will give you a guide when you're reading the meditations to some of the highlights.

61
00:07:04,760 --> 00:07:09,920
OK, so Descartes says I'm going to cast away anything that isn't completely certain.

62
00:07:09,920 --> 00:07:16,780
That is how I am going to reply to the sceptic. I'm only going to hold on to things that are certain.

63
00:07:16,780 --> 00:07:22,930
When I think about my own existence, I realise that there there is a source of certainty.

64
00:07:22,930 --> 00:07:29,680
I'm aware of my own thinking. I could not be thinking without existing, so I'm absolutely certain that I exist.

65
00:07:29,680 --> 00:07:34,180
I think therefore I am very famous.

66
00:07:34,180 --> 00:07:45,010
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.

67
00:07:45,010 --> 00:07:51,970
Now, that's a very strange move, if you think about it, because he's identified this one unique special truth.

68
00:07:51,970 --> 00:07:58,380
I think therefore I am. But then he goes on and tries to establish a general rule on the basis of that.

69
00:07:58,380 --> 00:08:08,010
Now it may be that his knowledge of his own existence is somehow special and not just because he clearly and distinctly perceived it to be true,

70
00:08:08,010 --> 00:08:13,620
but because it has other qualities like, for example, I cannot doubt without existing.

71
00:08:13,620 --> 00:08:23,550
The very fact of doubting proves my existence. So that's the kind of thing that suggests that I think therefore I am special.

72
00:08:23,550 --> 00:08:27,270
My own thinking somehow guarantees my existence.

73
00:08:27,270 --> 00:08:35,070
That's not necessarily necessarily going to carry over to other claims that Descartes thinks that he clearly and distinctly perceives.

74
00:08:35,070 --> 00:08:43,320
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.

75
00:08:43,320 --> 00:08:50,820
So I can now draw a general rule that whatever I clearly and distinctly perceive is true.

76
00:08:50,820 --> 00:08:55,320
Then he goes on and argues on the basis of things that he claims to clearly and

77
00:08:55,320 --> 00:09:01,710
distinctly perceive that God exists in particular in the third meditation.

78
00:09:01,710 --> 00:09:06,930
He notices that within his mind, he has an idea of God.

79
00:09:06,930 --> 00:09:12,630
He sees that this idea of God has a sort of perfection to it.

80
00:09:12,630 --> 00:09:18,990
And he claims that that perfection could only come about from a perfect cause.

81
00:09:18,990 --> 00:09:27,520
So the effect that is the idea of God in his mind could not have been put there, except by perfect being.

82
00:09:27,520 --> 00:09:31,600
And an example of one of the principles that he appeals to in this proof,

83
00:09:31,600 --> 00:09:40,450
it is manifest by the natural light that there must be at least as much in the efficient and total calls as in the effect of that cause.

84
00:09:40,450 --> 00:09:47,260
So if I get my idea of God is perfect, that perfection must have come from the cause of that idea.

85
00:09:47,260 --> 00:09:53,210
So there must be a god. And a perfect God couldn't be a deceiver.

86
00:09:53,210 --> 00:10:01,250
So if he's not a deceiver, then it must be that I who have been made by him must be capable of truth.

87
00:10:01,250 --> 00:10:09,950
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.

88
00:10:09,950 --> 00:10:14,240
And we know that God can't be a deceiver because God has got to be perfect.

89
00:10:14,240 --> 00:10:21,770
OK, I've gone through that very quickly, but you may have noticed that there's a little bit of a circularity here.

90
00:10:21,770 --> 00:10:28,890
Descartes seems to be appealing to his clear and distinct perception to prove the existence of God.

91
00:10:28,890 --> 00:10:31,620
But then having prove the existence of God, he says,

92
00:10:31,620 --> 00:10:37,420
now that I know that God exists and is not a deceiver, I can rely on my clear and distinct perception.

93
00:10:37,420 --> 00:10:41,650
So it looks like he's relying on his faculties to prove the existence of God.

94
00:10:41,650 --> 00:10:50,050
Then he's using God's existence to justify his faculties that looks circular.

95
00:10:50,050 --> 00:10:58,010
And you might wonder in the light of this weather, actually, it's possible.

96
00:10:58,010 --> 00:11:03,470
To defeat the sceptic without some such circular argument,

97
00:11:03,470 --> 00:11:15,440
if the sceptic throws doubt on your very thinking faculties, how on earth can you justify them against him?

98
00:11:15,440 --> 00:11:21,170
So I'm going to leave that hanging just for a moment and introduce the next big figure.

99
00:11:21,170 --> 00:11:25,430
Probably these days, the most influential philosopher of the early modern period,

100
00:11:25,430 --> 00:11:34,370
there are probably more people who think of themselves as humans in active in philosophy these days, certainly than Cartesian or Lacunes.

101
00:11:34,370 --> 00:11:40,520
And probably even Campion's part of the reason that David Hume's influence is so great these

102
00:11:40,520 --> 00:11:47,390
days is because he was a he was trying to establish a philosophical view of lots of things,

103
00:11:47,390 --> 00:11:55,910
including ethics and philosophy of religion, as well as epistemology and metaphysics, but was completely independent of God.

104
00:11:55,910 --> 00:12:04,140
And in the 20th and 21st centuries, that has become the standard way of doing philosophy.

105
00:12:04,140 --> 00:12:10,260
So Hume was cold in his day, the great infidel, he's known as a notorious sceptic.

106
00:12:10,260 --> 00:12:19,410
So unlike Descartes, who whose aim is to oppose scepticism, Hume often seems to be on the side of the sceptic.

107
00:12:19,410 --> 00:12:22,290
Now we'll see that it's not actually so clear.

108
00:12:22,290 --> 00:12:31,770
In many ways, he wants to respond to the sceptic, but he's certainly much more sympathetic to sceptical arguments than Descartes or Locke,

109
00:12:31,770 --> 00:12:40,080
or indeed, any other great philosopher of this period.

110
00:12:40,080 --> 00:12:48,210
So in your readings for general philosophy, there are two particular sections of Hume's first enquiry, which are relevant.

111
00:12:48,210 --> 00:12:54,240
This is the enquiry concerning human understanding of 1748. One of them is Enquiry Section four.

112
00:12:54,240 --> 00:13:02,400
That's where he presents his famous argument concerning induction, but the other is Enquiry 12, where he discusses scepticism.

113
00:13:02,400 --> 00:13:06,420
And in this, in the next lecture, I'm going to be appealing to both of those.

114
00:13:06,420 --> 00:13:16,740
So in the next lecture, we're going to go and see more about scepticism and about how human and a human can reply to scepticism.

115
00:13:16,740 --> 00:13:22,610
So very early in Section 12, Hume attacks Descartes.

116
00:13:22,610 --> 00:13:27,860
There is a species of scepticism antecedent to all study and philosophy,

117
00:13:27,860 --> 00:13:35,510
which is much inculcated by Descartes and others as a sovereign preservative against error and precipitate judgement.

118
00:13:35,510 --> 00:13:40,970
It recommends and universal doubt not only of all our former opinions and principles,

119
00:13:40,970 --> 00:13:47,450
but also of our very faculties of whose veracity say they we must assure ourselves by a chain of

120
00:13:47,450 --> 00:13:54,650
reasoning deduced from some original principle which cannot possibly be fallacious or deceitful.

121
00:13:54,650 --> 00:14:01,910
He calls the antecedent scepticism because the scepticism is being raised even before we've applied our faculties,

122
00:14:01,910 --> 00:14:09,510
it's casting doubt on our faculties right from the start. Now, Hume responds to that.

123
00:14:09,510 --> 00:14:18,210
But neither is there any such a regional principle, which has a prerogative above others that are self-evident and convincing, or if there were.

124
00:14:18,210 --> 00:14:26,160
Could we have advanced a step beyond it? But by the use of those very faculties of which we are supposed to be already diffident,

125
00:14:26,160 --> 00:14:36,210
the Cartesian doubt there for wherever possible to be attained by any human creature as it plainly is not would be entirely incurable.

126
00:14:36,210 --> 00:14:44,300
And no reasoning could ever bring us to a state of assurance and conviction upon any subject.

127
00:14:44,300 --> 00:14:51,850
Now, personally, I have to say, this seems quite a strong counter to Descartes.

128
00:14:51,850 --> 00:15:02,900
It's saying, look, if you start out distrusting your own faculties and requiring that your own faculties be justified, you've had it.

129
00:15:02,900 --> 00:15:09,010
There's no way you're going to get to do it, except by using your faculties. And since you don't trust your faculties.

130
00:15:09,010 --> 00:15:16,340
No chance. So what's the point of doing that? What is the point of pursuing that kind of sceptical philosophy?

131
00:15:16,340 --> 00:15:20,420
You've simply dug yourself in a pit, you've thrown away the means of getting out of the pit.

132
00:15:20,420 --> 00:15:32,480
You've had it. Futile. Now, let's go on and see an argument of Hume's, which is seriously sceptical,

133
00:15:32,480 --> 00:15:41,800
but not a case of antecedent scepticism of the kind that he accuses Descartes of.

134
00:15:41,800 --> 00:15:48,490
We'll be coming back, incidentally to this issue about antecedent scepticism, particularly in the next lecture.

135
00:15:48,490 --> 00:15:52,120
Now Kim's most famous for his argument concerning induction, indeed,

136
00:15:52,120 --> 00:15:58,690
it's probably the most famous and influential argument in all of English language philosophy.

137
00:15:58,690 --> 00:16:04,990
It appears in three versions in the treats of human nature, the abstract of the treaties and the enquiry.

138
00:16:04,990 --> 00:16:10,360
The version that you read for this course. The enquiry version is the best.

139
00:16:10,360 --> 00:16:16,660
It's the fullest, the most polished. So you don't need to worry too much about the treaties in the abstract.

140
00:16:16,660 --> 00:16:24,360
But if you, the abstract gives a very nice summary a sort of potted version of the argument.

141
00:16:24,360 --> 00:16:33,330
It's considered as very widely being considered a dangerously sceptical argument, Broad called it the scandal of philosophy.

142
00:16:33,330 --> 00:16:38,160
A lot of modern philosophy of science has been developed in response to this argument.

143
00:16:38,160 --> 00:16:44,850
Its interpretation still remains controversial in various ways, but it continues to be extremely influential.

144
00:16:44,850 --> 00:16:50,730
So quite unlike Descartes arguments that we've been talking about here is an argument which is alive and kicking,

145
00:16:50,730 --> 00:17:06,230
which many people actually think is correct, despite the fact that its conclusion can seem wildly paradoxical or sceptical.

146
00:17:06,230 --> 00:17:11,990
So I'm going to be following it broadly as it goes in enquiry for and enquiry section

147
00:17:11,990 --> 00:17:19,220
four starts with the distinction between two different types of proposition. Relations of ideas and matters of fact.

148
00:17:19,220 --> 00:17:30,140
Now relations of ideas are propositions that you can know to be true without consulting experience that a priori we've already come across that term.

149
00:17:30,140 --> 00:17:36,410
You can know them to be true simply by thinking of the meanings of the terms they contain.

150
00:17:36,410 --> 00:17:40,820
So they're necessarily true. There's no way you can conceive them to be false.

151
00:17:40,820 --> 00:17:47,570
Hume gives examples of Pythagoras theorem and three times five equals half of 30.

152
00:17:47,570 --> 00:17:53,900
An example. A very common example these days is all bachelors are unmarried.

153
00:17:53,900 --> 00:18:02,330
An example of what is commonly called an analytic proposition, it's one such that the meanings of the terms determine its truth.

154
00:18:02,330 --> 00:18:04,940
So those are relations of ideas.

155
00:18:04,940 --> 00:18:14,570
A relation of ideas just tells you about the logical connexions between the words that you're using or the ideas that you associate with them.

156
00:18:14,570 --> 00:18:19,730
It isn't telling you anything substantial about the world.

157
00:18:19,730 --> 00:18:27,000
Now against that, we have matters of fact and matters of fact, our statements, but do tell you something about the world.

158
00:18:27,000 --> 00:18:29,510
The things that cannot be known a priori.

159
00:18:29,510 --> 00:18:37,970
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.

160
00:18:37,970 --> 00:18:42,200
So here's an example of a matter of fact The Sun will rise tomorrow.

161
00:18:42,200 --> 00:18:46,100
Here's another one. The Sun will not rise tomorrow.

162
00:18:46,100 --> 00:18:54,020
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.

163
00:18:54,020 --> 00:18:59,510
But it's a question of fact, whether it's true or false, or take this one.

164
00:18:59,510 --> 00:19:03,560
This pen will fall when released in air.

165
00:19:03,560 --> 00:19:11,030
Oh, that turned out to be a true matter of fact. So in modern terms, this is a synthetic proposition,

166
00:19:11,030 --> 00:19:17,210
a proposition whose truth is determined by the facts of experience, not by the meanings of the terms.

167
00:19:17,210 --> 00:19:21,350
OK, so that seems a very intuitive distinction.

168
00:19:21,350 --> 00:19:25,490
It's been very, very influential since Kant.

169
00:19:25,490 --> 00:19:33,520
It's been drawn more in terms of analytics, synthetic rather than relations of ideas and matters of fact, but it's essentially the same distinction.

170
00:19:33,520 --> 00:19:40,510
OK, now it may seem straightforward to understand how I can know the truth of a relation of ideas.

171
00:19:40,510 --> 00:19:49,360
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.

172
00:19:49,360 --> 00:19:57,280
But how can I know a matter of fact? Well, you may think in some cases, that's very straightforward.

173
00:19:57,280 --> 00:20:04,900
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.

174
00:20:04,900 --> 00:20:08,440
OK, so let's not worry about that.

175
00:20:08,440 --> 00:20:17,230
We're not going to do what Descartes was doing. We're not going to raise sceptical questions about perception or about memory.

176
00:20:17,230 --> 00:20:27,250
What are we going to ask is how can we know any matter of fact which we are not either currently perceiving or remembering?

177
00:20:27,250 --> 00:20:36,580
So, for example, how can I know? That that pin will fall when I let go of it.

178
00:20:36,580 --> 00:20:44,560
OK, let's take a paradigm example of this kind of inference, and Hume chooses billiard balls.

179
00:20:44,560 --> 00:20:49,780
So suppose we see a yellow billiard ball moving towards a red one and colliding with it?

180
00:20:49,780 --> 00:20:54,370
We expect the red one to move, don't we?

181
00:20:54,370 --> 00:21:06,350
Why? Well, Hume says that we suppose a causal connexion, a causal connexion between the impact of the yellow ball and the movement of the red one.

182
00:21:06,350 --> 00:21:12,650
OK, fine. But how do we learn what causes what, how do we learn about causal relations?

183
00:21:12,650 --> 00:21:18,400
And here Hume brings in a thought experiment involving Adam.

184
00:21:18,400 --> 00:21:23,560
Adam is the very first man, he's just been created by God.

185
00:21:23,560 --> 00:21:33,520
He's a perfect specimen of a human being. All his faculties work just as they should because God's just made him.

186
00:21:33,520 --> 00:21:39,820
OK. So, Adam, sitting there and God up comes up to him and says, Look, Adam, here's a billiard ball.

187
00:21:39,820 --> 00:21:44,320
Oh, what's that? Never seen one of those before. No, you haven't. But here's one anyway.

188
00:21:44,320 --> 00:21:48,280
Here's another one. Now I'm going to roll, says God.

189
00:21:48,280 --> 00:21:58,850
One of the billiard balls towards the other. What do you think will happen, Adam, when that billiard ball hits that one?

190
00:21:58,850 --> 00:22:03,140
And Humes thought experiment is actually Adam would not have a clue.

191
00:22:03,140 --> 00:22:05,420
Adam's never seen this before.

192
00:22:05,420 --> 00:22:13,940
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.

193
00:22:13,940 --> 00:22:22,070
One might just stop. It might go through right through it. One might jump in the air or go backwards or all sorts of things.

194
00:22:22,070 --> 00:22:30,800
How could Adam, without experience, possibly know what was going to happen when one billiard ball hit the other?

195
00:22:30,800 --> 00:22:38,330
He couldn't. And that does indeed seem a plausible result of that thought experiment.

196
00:22:38,330 --> 00:22:45,530
OK, now, suppose God did this a number of times. So, Adam, that by now he's seen 100 billiard balls and every time the second one is moved.

197
00:22:45,530 --> 00:22:54,920
And now God says to Adam What will happen when this ball, when the yellow ball hits the red one and Adam says, Oh, the red one will move.

198
00:22:54,920 --> 00:23:02,060
OK, well, what makes the difference? Why, why does Adams experience of one believable hitting another?

199
00:23:02,060 --> 00:23:06,460
How does it help him in making a prediction about others?

200
00:23:06,460 --> 00:23:19,570
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.

201
00:23:19,570 --> 00:23:25,750
So Hume claims very plausibly all inference to matters of fact beyond what we perceive or

202
00:23:25,750 --> 00:23:33,570
remember is based on causation and all our knowledge of causation comes from experience.

203
00:23:33,570 --> 00:23:39,120
But we can only learn from experience on the assumption that observed phenomena,

204
00:23:39,120 --> 00:23:47,400
the billiard balls we've seen in the past provide a guide, some sort of at least vaguely reliable guide to unobserved phenomena.

205
00:23:47,400 --> 00:23:49,470
The billiard balls we're going to see in the future.

206
00:23:49,470 --> 00:23:58,170
So we have to be able to extrapolate from observed to unobserved on the basis that the unobserved will resemble the observed.

207
00:23:58,170 --> 00:24:03,410
OK. Do we have a rational basis for making that assumption?

208
00:24:03,410 --> 00:24:07,760
Now, that assumption is commonly called the uniformity principle, it's not Hume's term.

209
00:24:07,760 --> 00:24:15,380
But Hume scholars normally use it to refer to this principle. Here are a couple of ways in which she expresses it in the enquiry.

210
00:24:15,380 --> 00:24:20,360
All our experimental conclusions, he just means conclusions from experience.

211
00:24:20,360 --> 00:24:27,930
All our experimental conclusions proceed upon the supposition that the future will be conformable to the past.

212
00:24:27,930 --> 00:24:35,320
OK. How then, can we know that the future will be conformable to the past?

213
00:24:35,320 --> 00:24:45,230
What possible means of justification do we have for the principle of uniformity?

214
00:24:45,230 --> 00:24:50,030
OK. Spoiler alert. We don't have any.

215
00:24:50,030 --> 00:25:02,030
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.

216
00:25:02,030 --> 00:25:07,820
OK, so I now need to take you on a little detour to look at the various sources of evidence

217
00:25:07,820 --> 00:25:13,550
that Hume thinks might conceivably be available for the uniformity principle.

218
00:25:13,550 --> 00:25:18,300
Or somebody might think were available for the uniformity principle.

219
00:25:18,300 --> 00:25:26,040
So here's a little snippet from a document that Hume wrote in 1745 that's three years before he published the enquiry.

220
00:25:26,040 --> 00:25:30,510
It's a shame he doesn't say anything quite this clear in the enquiry.

221
00:25:30,510 --> 00:25:40,640
It is common for philosophers to distinguish the kinds of evidence into intuitive, demonstrative, sensible and moral.

222
00:25:40,640 --> 00:25:46,160
OK. So when he talks about intuition, he means things that are self-evident.

223
00:25:46,160 --> 00:25:56,200
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,

224
00:25:56,200 --> 00:26:04,210
evidence from the senses and when he talks about demonstrative and moral reasoning,

225
00:26:04,210 --> 00:26:08,290
or sometimes he uses the term probable rather than moral reasoning.

226
00:26:08,290 --> 00:26:14,020
He means two different types of inference, which were distinguished by John Locke.

227
00:26:14,020 --> 00:26:22,250
So again, I'm brief detour. Here's an example from lock of demonstrative reasoning.

228
00:26:22,250 --> 00:26:28,820
It's a proof that the angles within a triangle add up to the angles on a straight line,

229
00:26:28,820 --> 00:26:33,980
and the claim is that we can see intuitively we've got two parallel lines.

230
00:26:33,980 --> 00:26:39,890
We can see intuitively, it's self-evident that Angle A is equal to an angle e.

231
00:26:39,890 --> 00:26:46,430
It's self-evident that angle B is the same as angle D,

232
00:26:46,430 --> 00:26:54,790
and it's self-evidently follows from the first two propositions that a plus B policy is equal to E plus d plus c.

233
00:26:54,790 --> 00:27:02,620
So the sort of idea of a demonstrative argument for Lock, which Hume takes over,

234
00:27:02,620 --> 00:27:09,010
is an argument where every link in the inferential chain goes with absolute and complete certainty.

235
00:27:09,010 --> 00:27:15,670
So we've got what we would call a deductive argument and argument where if the premises are true,

236
00:27:15,670 --> 00:27:20,260
you simply could not possibly doubt the truth of the conclusions as following,

237
00:27:20,260 --> 00:27:27,110
because there's no way the premise could be premises could be true without the conclusion being true.

238
00:27:27,110 --> 00:27:34,160
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.

239
00:27:34,160 --> 00:27:37,700
All right. This is an old use of the term. It's still retained.

240
00:27:37,700 --> 00:27:47,030
When people talk about moral certainty, they don't need ethical. They mean certainty for the purposes of human life or something like that.

241
00:27:47,030 --> 00:27:52,640
So improbable reasoning. You have an argument where the links of the chain are not infallible.

242
00:27:52,640 --> 00:28:00,770
So in demonstrative reasoning, you go from the premises through the steps of the argument to the conclusion with absolute certainty all the way,

243
00:28:00,770 --> 00:28:09,690
the sort of thing you do in maths, the sort of thing you do in logic. Improbable reasoning the links of the chain are not so certain.

244
00:28:09,690 --> 00:28:19,230
And this is lock talks about this, but gives very little in the way of example, an example he gives is this one telecom tree gentlewoman.

245
00:28:19,230 --> 00:28:24,540
The wind is southwest and the weather flowering, unlike to rain,

246
00:28:24,540 --> 00:28:31,320
and she will easily understand it is not safe for her to go abroad thin clad in such a day after a fever.

247
00:28:31,320 --> 00:28:41,680
She clearly sees the probable connexion of all these these south west wind and clouds, rain wetting, taking cold relapse and danger of death.

248
00:28:41,680 --> 00:28:45,810
Well. A bit dramatic, but you get the idea.

249
00:28:45,810 --> 00:28:54,210
These are not demonstrative Connexions. The woman is drawing probable connexions from one thing to another.

250
00:28:54,210 --> 00:29:00,970
OK. And I mentioned a point there at the bottom, which I'll come back to a little later.

251
00:29:00,970 --> 00:29:10,960
So Locke's distinction between demonstrative and probable reasoning pretty much corresponds with what we call deductive and inductive reasoning,

252
00:29:10,960 --> 00:29:17,110
deductive reasoning is where the premises guarantee the truth of the conclusion.

253
00:29:17,110 --> 00:29:28,300
An inductive argument is one that draws conclusions about typically about the unobserved on the basis of the observed by extrapolation.

254
00:29:28,300 --> 00:29:36,550
OK, so now let's come back. Remember, we were asking what justification could one possibly have for the uniformity principle?

255
00:29:36,550 --> 00:29:45,880
And we've got four possible sources of evidence, according to Human, and Hume says this is quite standard as it is in his day.

256
00:29:45,880 --> 00:29:50,020
Is it self-evident that the future will resemble the past?

257
00:29:50,020 --> 00:29:57,220
No, it's not self-evident. Can it be proved demonstratively by deductive reasoning?

258
00:29:57,220 --> 00:30:00,940
Can you prove that the future will resemble the past?

259
00:30:00,940 --> 00:30:08,620
That because billiard balls in the past have acted in such and such a way that you can logically prove that the next one will act that way?

260
00:30:08,620 --> 00:30:15,670
No, you can't. Both of these are clearly impossible because we can easily imagine things turning out differently.

261
00:30:15,670 --> 00:30:26,060
And if we can consistently imagine things turning out differently, that's enough to show that we haven't got a logical or demonstrative proof.

262
00:30:26,060 --> 00:30:34,640
What about sensory knowledge, could we know through the senses that future billiard balls will operate in the same way as past billiard balls?

263
00:30:34,640 --> 00:30:38,000
Well, no, we can't think of the atom thought experiment.

264
00:30:38,000 --> 00:30:44,450
Adam, by looking at the billiard balls, didn't have a clue what was going to happen until he actually saw it.

265
00:30:44,450 --> 00:30:52,210
And by the same reasoning, he couldn't have had a clue that the billiard balls would act consistently.

266
00:30:52,210 --> 00:30:56,230
Just looking at the superficial qualities of the billiard ball didn't tell him

267
00:30:56,230 --> 00:31:00,610
anything about how it's going to behave or whether it will behave consistently.

268
00:31:00,610 --> 00:31:06,840
So that's no good. What about what Hume calls moral reasoning or probable reasoning,

269
00:31:06,840 --> 00:31:14,130
and I'm calling factual inference for short because he often calls it reasoning concerning matter of fact.

270
00:31:14,130 --> 00:31:21,810
Well, the problem there is that if we use that to justify uniformity, we're just going in a circle.

271
00:31:21,810 --> 00:31:26,160
We have said that all arguments concerning existence again,

272
00:31:26,160 --> 00:31:31,620
he means all moral probable reasoning or factual inferences are founded on the

273
00:31:31,620 --> 00:31:35,310
relation of cause and effect that our knowledge of that relation is derived

274
00:31:35,310 --> 00:31:40,050
entirely from experience and that all experimental conclusions proceed upon

275
00:31:40,050 --> 00:31:44,640
the supposition that the future will be conformable to the past to endeavour.

276
00:31:44,640 --> 00:31:53,940
Therefore, the proof of this last supposition by probable arguments or arguments regarding existence, which suggests just the same thing,

277
00:31:53,940 --> 00:32:00,810
must be evidently going in a circle and taking that for granted, which is the very point in question.

278
00:32:00,810 --> 00:32:05,910
Okay, so we cannot justify extrapolating from past to future.

279
00:32:05,910 --> 00:32:14,700
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,

280
00:32:14,700 --> 00:32:20,040
because that doesn't tell us that the future is going to resemble the past, nor on the basis of inductive argument,

281
00:32:20,040 --> 00:32:24,120
because inductive argument takes for granted that the future will resemble the past.

282
00:32:24,120 --> 00:32:33,520
So we've got absolutely no justification, apparently, for the assumption that the future will resemble the past.

283
00:32:33,520 --> 00:32:41,140
And since all of our influences is beyond what we immediately observe or remember depends upon this assumption,

284
00:32:41,140 --> 00:32:46,420
or at least the assumption that the unobserved will resemble the observed we act.

285
00:32:46,420 --> 00:32:50,830
It looks like we haven't got any justification for any belief about anything,

286
00:32:50,830 --> 00:32:56,910
any matter of fact that goes beyond our immediate perceptions and memory.

287
00:32:56,910 --> 00:33:05,300
That's a serious worry. And notice that Hume here, he's not doing things like doubting the external world.

288
00:33:05,300 --> 00:33:13,100
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?

289
00:33:13,100 --> 00:33:24,260
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.

290
00:33:24,260 --> 00:33:31,850
What's particularly scandalous is I cannot give you an answer to this, which is going to show that Hume is wrong.

291
00:33:31,850 --> 00:33:36,680
He may well be right.

292
00:33:36,680 --> 00:33:44,600
OK, so the upshot is that our empirical reasoning, all of our empirical reasoning, wherever we reason from observed to unobserved.

293
00:33:44,600 --> 00:33:50,150
It's based on a brute assumption of uniformity for which we can't give any ground.

294
00:33:50,150 --> 00:33:53,810
It's not based on any insight into the nature of things.

295
00:33:53,810 --> 00:33:59,030
Hume draws the conclusion that human reason basically differs only in degree from animal reasoning.

296
00:33:59,030 --> 00:34:06,410
It's just like, you know, the dog. When he hears the door go, he expects to be taken for a walk or given food or whatever.

297
00:34:06,410 --> 00:34:10,250
It's just animal habit. Well, basically, we're the same.

298
00:34:10,250 --> 00:34:16,660
We see one billiard ball going towards another. We expect the second one to move by habit.

299
00:34:16,660 --> 00:34:20,920
Look, as I mentioned in the earlier slide, attributed probable reasoning,

300
00:34:20,920 --> 00:34:25,510
as well as demonstrative reasoning reasoning to the perception of Connexions.

301
00:34:25,510 --> 00:34:32,440
Hume says that's nonsense that we don't perceive any connexion between one billiard ball moving and the other one moving.

302
00:34:32,440 --> 00:34:39,640
It's just habit. And no causal interactions are really intelligible.

303
00:34:39,640 --> 00:34:46,250
They're all based on this familiarity, what Kim calls custom.

304
00:34:46,250 --> 00:34:52,670
You can see why Hume became very popular in the 20th century.

305
00:34:52,670 --> 00:34:59,180
I talked in the last lecture about quantum mechanics, general relativity and so on.

306
00:34:59,180 --> 00:35:06,950
People became aware that actually the world is rather weird and our intuitions do not provide a good guide to it.

307
00:35:06,950 --> 00:35:12,650
In previous generations, people may have thought that they had rational insight into the way the world works.

308
00:35:12,650 --> 00:35:17,030
In the 20th century, we were disabused of that complacent assumption.

309
00:35:17,030 --> 00:35:23,750
We discovered actually that the only way we can find out how the world works is by experiment and observation.

310
00:35:23,750 --> 00:35:27,920
But what humans shown is that even relying on experiments and observation,

311
00:35:27,920 --> 00:35:36,400
we have to take for granted this basic assumption that the unobserved will resemble the observed.

312
00:35:36,400 --> 00:35:46,960
OK, the general philosophy reading list includes several attempts to justify induction to answer him or any of these successful.

313
00:35:46,960 --> 00:35:50,860
Well, let's take a little look at them. I've listed some there.

314
00:35:50,860 --> 00:35:57,070
We'll be talking a little bit about the Van Cleave and Reichenbach moves next time.

315
00:35:57,070 --> 00:36:01,300
But for now, I want to focus on the first two.

316
00:36:01,300 --> 00:36:08,800
So the analytic justification of induction is basically saying that induction is rational by definition, right?

317
00:36:08,800 --> 00:36:13,150
If you ask how can we justify extrapolation from observe to unobserved?

318
00:36:13,150 --> 00:36:20,380
Well, that's just the rational thing to do if you don't extrapolate from observe to unobserved, you're being irrational.

319
00:36:20,380 --> 00:36:26,830
So the gambler who says, Yes, I've lost this game 100 times in a row, I'm bound to win next time.

320
00:36:26,830 --> 00:36:31,600
That's just irrational. OK? By definition.

321
00:36:31,600 --> 00:36:37,030
So we don't need any argument to justify it. That's enough.

322
00:36:37,030 --> 00:36:42,220
Probabilistic attempts to justify induction appeal to mathematical probability.

323
00:36:42,220 --> 00:36:46,080
And there's quite a variety of those, we'll be looking at one of those.

324
00:36:46,080 --> 00:36:50,610
The inductive justification of induction is basically saying induction has worked in the past.

325
00:36:50,610 --> 00:36:54,930
It will work in the future. Now Hume obviously says that's just circular.

326
00:36:54,930 --> 00:37:01,740
All right. Well, some philosophers have tried to find ingenious ways to get around that circularity.

327
00:37:01,740 --> 00:37:08,290
And the pragmatic justification of induction says we are pragmatically justified in relying on induction.

328
00:37:08,290 --> 00:37:17,660
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.

329
00:37:17,660 --> 00:37:27,560
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.

330
00:37:27,560 --> 00:37:33,930
That's the line that Reichenbach takes. So let's have a quick look at a couple of these.

331
00:37:33,930 --> 00:37:46,380
So the analytic justification, as I say, says that induction is just rational by definition, what we mean by a rational method of inference.

332
00:37:46,380 --> 00:37:53,270
Is one that infers the same about the future as has been observed in the past.

333
00:37:53,270 --> 00:37:58,910
So suppose somebody relies on astrology or reading tea leaves.

334
00:37:58,910 --> 00:38:06,040
Well, if you investigate these things, it finds out you find that they're hopelessly unreliable.

335
00:38:06,040 --> 00:38:13,690
So we do not consider those rational methods of inference because they've proved not to be reliable.

336
00:38:13,690 --> 00:38:19,060
That shows that we are judging methods of inference inductively.

337
00:38:19,060 --> 00:38:24,760
We are treating those that yield truth consistently over time.

338
00:38:24,760 --> 00:38:32,210
That's attribute the same properties to things over time as rational.

339
00:38:32,210 --> 00:38:40,070
So induction is arguably constitute constitutive of what it is to be rational.

340
00:38:40,070 --> 00:38:48,920
You can only be rational by definition if you think if you reason inductively.

341
00:38:48,920 --> 00:38:55,260
OK, well, maybe that's plausible. But does it actually touch you?

342
00:38:55,260 --> 00:39:03,520
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.

343
00:39:03,520 --> 00:39:12,400
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.

344
00:39:12,400 --> 00:39:17,680
And so obviously, we think of that as constitutive of human rationality.

345
00:39:17,680 --> 00:39:23,090
And we people who think differently, we're going to treat as mad.

346
00:39:23,090 --> 00:39:29,870
So, well, that's just a question of how we judge it. It doesn't mean we've got any independent justification for it.

347
00:39:29,870 --> 00:39:34,830
That just shows that we are indeed taking it for granted.

348
00:39:34,830 --> 00:39:43,170
What Hume is asking or challenging is the thought that we've got some kind of insight that justifies extrapolation.

349
00:39:43,170 --> 00:39:48,170
He's not questioning that we use the word reason in a particular way.

350
00:39:48,170 --> 00:39:56,300
Because the way in which we use that word actually builds in our assumption that induction is a reasonable way to go,

351
00:39:56,300 --> 00:40:02,360
that we've got good evidence for it. Humans argued that actually that complacent assumption is false.

352
00:40:02,360 --> 00:40:09,000
We don't have any good reason for supposing that the future will resemble the past.

353
00:40:09,000 --> 00:40:17,810
So the analytic justification, you know, may be plausible as an account of how we use words, it doesn't touch him.

354
00:40:17,810 --> 00:40:28,770
What about probabilistic justification? Now, Hume in his argument.

355
00:40:28,770 --> 00:40:36,780
Take something for granted, which could be questioned he takes for granted that all probable arguments,

356
00:40:36,780 --> 00:40:42,770
all arguments that aren't demonstrative in force have to rely on experience.

357
00:40:42,770 --> 00:40:51,890
Right. Because remember, he said. You can't have a demonstrative argument that the future will resemble the past for the following reason.

358
00:40:51,890 --> 00:40:59,420
We can imagine the future not resembling the past. We can coherently conceive of the future being different.

359
00:40:59,420 --> 00:41:05,850
That's enough to show that you can't demonstrate that the future will be the same.

360
00:41:05,850 --> 00:41:14,400
And then with regard to probable argument, the reason you can't justify the future resembling the past by inductive or probable reasoning,

361
00:41:14,400 --> 00:41:20,040
Hume says, is that all such reasoning takes for granted that the future will resemble the past.

362
00:41:20,040 --> 00:41:26,500
So you'd be reasoning in a circle. But maybe humans overlooked something.

363
00:41:26,500 --> 00:41:36,820
Maybe it is possible to have probabilistic reasoning, which is a priority where your reasoning from basic probabilistic principles

364
00:41:36,820 --> 00:41:43,390
that we see intuitively to be true rather than on the basis of experience.

365
00:41:43,390 --> 00:41:48,930
And if such an argument were possible, then Hume's argument has not ruled it out.

366
00:41:48,930 --> 00:41:57,120
OK, so there is a gap in Hume's argument. So probabilistic justifications aim to get through that gap.

367
00:41:57,120 --> 00:42:02,160
Now, a lot of attempts have been made, I've listed a few of them there.

368
00:42:02,160 --> 00:42:09,290
Roy Harrod. Came up with an ingenious argument, which was later developed by Simon Blackburn,

369
00:42:09,290 --> 00:42:15,440
and we're going to be looking a little bit at that, that this is on your reading list.

370
00:42:15,440 --> 00:42:23,500
OK, so here's the ingenious argument. Suppose I'm crossing a desert, I don't know how big the desert is.

371
00:42:23,500 --> 00:42:29,150
And I'm making predictions about how long the desert will go on.

372
00:42:29,150 --> 00:42:40,320
So I'm sitting out on a straight path across the desert, and I'm making predictions about how long my journey will continue through the desert.

373
00:42:40,320 --> 00:42:42,690
And suppose I make predictions like this,

374
00:42:42,690 --> 00:42:51,100
the desert ahead will extend for at least 10 percent as long as the distance I've already travelled through it.

375
00:42:51,100 --> 00:42:56,170
OK, now imagine I I make such predictions again and again and again, so I'm walking along.

376
00:42:56,170 --> 00:43:02,170
I say the desert will last for 10 percent more. The desert will last for 10 percent more, the desert will last for 10 percent more.

377
00:43:02,170 --> 00:43:04,150
So I'm making lots and lots of these predictions.

378
00:43:04,150 --> 00:43:10,930
And of course, as I go further and further through the desert desert, my predictions are predicting yet more desert.

379
00:43:10,930 --> 00:43:16,690
So if I go for 10 miles, I'm predicting another mile. If I go for 20 miles, I'm predicting another two miles.

380
00:43:16,690 --> 00:43:22,090
If I go for 30 miles, I'm predicting another three miles. And so on.

381
00:43:22,090 --> 00:43:26,940
OK, now at some point. The desert hands.

382
00:43:26,940 --> 00:43:32,430
And so the predictions I make towards the end of that journey will have turned out false.

383
00:43:32,430 --> 00:43:40,520
It will not, in fact, have been true that the desert had another 10 per cent to go beyond what it had already gone.

384
00:43:40,520 --> 00:43:48,350
And the dividing line, the point at which the predictions will stop being true will be 10 11.

385
00:43:48,350 --> 00:43:58,040
So the way through the journey? OK. So if I've gone, let's let's suppose that the desert is 110 miles wide.

386
00:43:58,040 --> 00:44:05,050
Then after 100 miles, I will be making the prediction. The desert will go on for another 10 miles.

387
00:44:05,050 --> 00:44:10,330
And that prediction will be right, but the very next one I make will be wrong.

388
00:44:10,330 --> 00:44:18,450
Agreed. OK, so imagine that I'm making these these predictions sufficiently frequently that

389
00:44:18,450 --> 00:44:23,220
we can ignore rounding error 10 11 months of my predictions will come out,

390
00:44:23,220 --> 00:44:32,460
right? OK, 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.

391
00:44:32,460 --> 00:44:37,950
So far, 10 out of 11 of those will turn out true.

392
00:44:37,950 --> 00:44:41,970
OK, so maybe this justifies making such predictions.

393
00:44:41,970 --> 00:44:47,040
If I make such a prediction, I've got a 10 11 chance of being true.

394
00:44:47,040 --> 00:44:55,900
So I can justify extrapolation, at least to that limited extent and humans wrong.

395
00:44:55,900 --> 00:45:07,300
Well, that's not quite right. It's not quite right, because imagine that I've imagined that I've travelled 100 miles through the desert.

396
00:45:07,300 --> 00:45:12,280
And I'm making a prediction the desert will go on for another 10 miles.

397
00:45:12,280 --> 00:45:17,210
I've no idea how long it's going to go on, all right, but I make that kind of prediction.

398
00:45:17,210 --> 00:45:23,630
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.

399
00:45:23,630 --> 00:45:28,590
So I've got a 10 11 chance that this particular prediction will be right.

400
00:45:28,590 --> 00:45:38,130
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?

401
00:45:38,130 --> 00:45:42,690
All the predictions, for example, that I made up to 90 miles.

402
00:45:42,690 --> 00:45:46,020
They've already come and gone. Right?

403
00:45:46,020 --> 00:45:53,640
When I passed the 99th mile that confirmed the prediction that I made it 90 miles, that it would go on for another nine.

404
00:45:53,640 --> 00:45:58,680
So loads of the predictions have already been fulfilled.

405
00:45:58,680 --> 00:46:03,120
So now when I ask, what is the chance that this prediction will be fulfilled,

406
00:46:03,120 --> 00:46:09,480
I can't go by this 10 11 chance because that a large proportion or significant

407
00:46:09,480 --> 00:46:16,040
proportion of those 10 11s that are going to come out true have already come out true.

408
00:46:16,040 --> 00:46:22,360
So it's a bit like I mean, imagine if I clicked the tickets from one supposes a weekly lottery.

409
00:46:22,360 --> 00:46:28,680
All right. And I clicked the tickets from the previous nine lotteries.

410
00:46:28,680 --> 00:46:38,250
OK. And then on the tenth week, I buy a lottery ticket and I put it into a bag together with those the nine previous winning tickets.

411
00:46:38,250 --> 00:46:45,270
And I say there's a nine 10 chance of me winning this lottery because I've got it's one of 10 tickets and nine of them are winners.

412
00:46:45,270 --> 00:46:49,830
No, can't do that. The nine winners have already been and gone.

413
00:46:49,830 --> 00:46:59,030
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.

414
00:46:59,030 --> 00:47:03,530
Well, Harrods suggests that there's a way of correcting for this.

415
00:47:03,530 --> 00:47:11,600
Basically, you average over all the different positions that you could be on your way through the through the desert,

416
00:47:11,600 --> 00:47:17,180
and he suggests you get a slightly more modest prediction of 100 over 121.

417
00:47:17,180 --> 00:47:25,120
Let's not worry about that. I mean, it's a it's a reasonable point.

418
00:47:25,120 --> 00:47:30,100
But Blackburn points out that it doesn't actually solve the problem,

419
00:47:30,100 --> 00:47:38,580
because even if I average for all the different positions that I could be through my journey.

420
00:47:38,580 --> 00:47:48,270
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,

421
00:47:48,270 --> 00:47:51,930
the ones I made up to 90 miles and indeed beyond 90 miles.

422
00:47:51,930 --> 00:48:02,490
All those have already been gone. So averaging over every possible predict place, I could be within the that within the overall desert,

423
00:48:02,490 --> 00:48:08,970
you know, am I in the first half of it or the first quarter or the last 10th or the last 11?

424
00:48:08,970 --> 00:48:16,330
I just don't know. All I know is that a lot of predictions that I've already made have been and gone.

425
00:48:16,330 --> 00:48:21,310
I don't know what proportion of those already successful predictions.

426
00:48:21,310 --> 00:48:29,490
I don't know what proportion they form of the whole unless I know how big the desert is.

427
00:48:29,490 --> 00:48:34,350
So what Blackburn does is he says, well, here's a way of saving Harvard's argument,

428
00:48:34,350 --> 00:48:46,550
rather than focus on the believer, focus on the sceptic and say to the sceptic, Look, you are denying.

429
00:48:46,550 --> 00:48:53,050
That we can justifiably rely on these one extrapolations.

430
00:48:53,050 --> 00:49:01,600
But look, Howard's argument still holds that in general, if people make inferences of this kind extrapolations of this kind.

431
00:49:01,600 --> 00:49:12,630
10 out of 11 of them will be successful. So you, as a sceptic, are casting doubt on something which 10 out of 11 will be successful.

432
00:49:12,630 --> 00:49:21,150
How do you justify that? Well, it's a neat effort, but I don't think it works.

433
00:49:21,150 --> 00:49:27,130
Let's, first of all, a couple of general points.

434
00:49:27,130 --> 00:49:34,450
First of all, I want to suggest that the the focus that Harold introduces and Blackburn retains on the proportion of

435
00:49:34,450 --> 00:49:40,150
extrapolations that will come out true is simply giving a false impression of the logic of the situation.

436
00:49:40,150 --> 00:49:47,480
It's like with those lottery tickets that you put into a bank. The extrapolations are sequentially ordered.

437
00:49:47,480 --> 00:49:55,970
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.

438
00:49:55,970 --> 00:50:03,040
The extrapolations come strictly in order they are made in order and they are fulfilled in order.

439
00:50:03,040 --> 00:50:08,920
And the extrapolation that I make now can only be true with all of the ones before were true.

440
00:50:08,920 --> 00:50:17,520
So thinking in terms of choosing one from a group is just a misleading way of representing the situation.

441
00:50:17,520 --> 00:50:28,140
Another important point against Blackburn is that appealing to a general practise of making this kind of inference is hopeless for a different reason.

442
00:50:28,140 --> 00:50:37,080
And that is that the 10 out of 11 proportion applies to every individual uniformity, every journey you make.

443
00:50:37,080 --> 00:50:44,520
If you make these 10 11 extrapolations sufficiently frequently, 10 out of 11 of them will be true.

444
00:50:44,520 --> 00:50:49,230
All right. Sorry. I think I need to repeat that, Steve.

445
00:50:49,230 --> 00:50:51,150
Any journey at all.

446
00:50:51,150 --> 00:51:02,830
If you make one tenth extrapolations sufficiently frequently, then 10 out of 11 of them will come out true in that individual journey.

447
00:51:02,830 --> 00:51:09,090
So it's not like you've got 10, 11, it's true in some large sample,

448
00:51:09,090 --> 00:51:16,840
some large group from which you are picking an individual example rather than every individual journey you do.

449
00:51:16,840 --> 00:51:26,300
The proportion is 10 out of 11. So when I ask, is my next extrapolation likely to come out true, statistically,

450
00:51:26,300 --> 00:51:31,310
the only sequence that matters is the current one, the current journey.

451
00:51:31,310 --> 00:51:37,550
What what may be true about other journeys just is not relevant despite Blackburn.

452
00:51:37,550 --> 00:51:47,090
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.

453
00:51:47,090 --> 00:51:53,480
As I say, they're sequentially ordered. It's not like choosing one from a set.

454
00:51:53,480 --> 00:52:01,100
Unless I know how far I am through the desert or the period of uniformity, if we're applying this to induction.

455
00:52:01,100 --> 00:52:09,030
I can't know what proportion of extrapolations have already been fulfilled, so I can't apply the statistics anyway.

456
00:52:09,030 --> 00:52:16,320
So unless I can justify extrapolation from observed to unobserved, it seems I'm no better off than when I started out.

457
00:52:16,320 --> 00:52:20,830
So it's an ingenious attempt to justify induction.

458
00:52:20,830 --> 00:52:29,620
It's trying to show that one tenth extrapolations, modest extrapolations from experience are justified statistically.

459
00:52:29,620 --> 00:52:33,910
If that were true, we would be able to justify induction, at least to some extent.

460
00:52:33,910 --> 00:52:40,840
We could say the world has been uniform for at least, you know, 100 or a thousand or maybe 13 billion years.

461
00:52:40,840 --> 00:52:48,140
Therefore, it's likely to continue to be uniform for a tenth as long again.

462
00:52:48,140 --> 00:52:51,500
But sadly, it doesn't work.

463
00:52:51,500 --> 00:52:59,810
Next time, we'll be looking more at scepticism and briefly considering a couple of other attempts to respond to Hume's scepticism.

464
00:52:59,810 --> 00:53:03,402
Thank you.