1
00:00:06,340 --> 00:00:13,590
Welcome to the fourth lecture on whom we're going to be talking today about Hume's most famous argument,

2
00:00:13,590 --> 00:00:17,880
indeed probably the most famous argument in the English language philosophy.

3
00:00:17,880 --> 00:00:24,810
His argument concerning induction and following that his theory of belief.

4
00:00:24,810 --> 00:00:31,200
OK, so first of all, a reminder of where we were last time.

5
00:00:31,200 --> 00:00:39,330
We're looking at treaties book one, part three. It's the longest part of the work on knowledge and probability,

6
00:00:39,330 --> 00:00:46,800
but it turns out that the key to probable reasoning is going to be the relation of causation.

7
00:00:46,800 --> 00:00:55,530
So in fact, it's Hume's quest for the analysis of causation that occupies the vast bulk of the part.

8
00:00:55,530 --> 00:01:00,450
Now in section two, a book one, part three.

9
00:01:00,450 --> 00:01:09,270
Hume starts off analysing causation. He identifies contiguity and succession as playing a role.

10
00:01:09,270 --> 00:01:14,910
But he says that the main component of the concept of causation is necessary connexion.

11
00:01:14,910 --> 00:01:22,950
There is a necessary connexion, we think, between a cause and its effect, and that is the key part of the idea.

12
00:01:22,950 --> 00:01:34,440
But it seems quite elusive. So Hume sets off through two neighbouring fields to try to find a clue to where this necessary connexion might come from.

13
00:01:34,440 --> 00:01:37,860
And one of those is to do with the causal maxim.

14
00:01:37,860 --> 00:01:49,490
And then he turns to the discussion of inductive inference, causal inference, and that's how we get to treaties one three six.

15
00:01:49,490 --> 00:01:56,780
OK. It's important to appreciate there are actually three presentations of this argument.

16
00:01:56,780 --> 00:01:58,220
The first one is in the treatise.

17
00:01:58,220 --> 00:02:10,820
That's obviously the one we'll be focussing on mainly, but it features very prominently in the abstract of the treatise, which was published in 1740.

18
00:02:10,820 --> 00:02:17,630
Now we think that the abstract of the treatise was actually written in the autumn of seventeen thirty nine.

19
00:02:17,630 --> 00:02:23,150
And bear in mind that books won. Two of the treaties were published at the end of January 17.

20
00:02:23,150 --> 00:02:32,600
Thirty nine, so Hume was writing the abstract very soon after he published the treatise the first two books of the treaties.

21
00:02:32,600 --> 00:02:40,640
But in the abstract, the argument concerning induction this argument is far more prominent and it is in the treaties.

22
00:02:40,640 --> 00:02:44,060
I mean, the abstract is tiny compared with the first two books of the treaties,

23
00:02:44,060 --> 00:02:50,240
but this argument in the abstract is about 75 percent as long as the original.

24
00:02:50,240 --> 00:03:01,220
So he's only contracted that very slightly, and it's therefore interesting to see that Hume's philosophy is clearly changing during 1739.

25
00:03:01,220 --> 00:03:09,230
He's appreciating that this argument, which really isn't very prominent in the treaties, actually deserves far more prominent.

26
00:03:09,230 --> 00:03:17,270
It's far more significant than simply being a staging post on the way to the analysis of the idea of cause,

27
00:03:17,270 --> 00:03:24,020
which is what it seems to be in the treaties and then in the enquiry 1748.

28
00:03:24,020 --> 00:03:35,270
That very much follows the pattern of the abstract and section four of the enquiry where this argument receives its its fullest expression.

29
00:03:35,270 --> 00:03:45,180
Again, that's really the centrepiece of the enquiry. So the arguments now very well known, very famous, very influential.

30
00:03:45,180 --> 00:03:57,960
But looking at the treaties in its original form, you wouldn't see it as being such a major part.

31
00:03:57,960 --> 00:04:07,080
Another point worth bearing in mind is that the argument in the enquiry in its final form is improved in various ways.

32
00:04:07,080 --> 00:04:11,940
It's improved structurally. For example, it's streamlined.

33
00:04:11,940 --> 00:04:25,380
So when you if you read the two together, you might find it a little awkward to kind of put them side by side and see them as arguing in the same way.

34
00:04:25,380 --> 00:04:32,880
One particular difference is that in the treaties, Hume focuses on causal inference from the impression to the idea.

35
00:04:32,880 --> 00:04:39,450
So we see a followed by B again and again and again, and then we see we get an impression of A.

36
00:04:39,450 --> 00:04:47,760
And we find ourselves expecting B. So we have a vivacious idea of B an expectation that B will follow.

37
00:04:47,760 --> 00:04:54,790
That is an inference from the impression to the idea in a specific case.

38
00:04:54,790 --> 00:04:58,930
But in the abstract, in the enquiry, human broadens it, he says,

39
00:04:58,930 --> 00:05:05,170
I'm concerned with any inference be on the present testimony of the memory and senses.

40
00:05:05,170 --> 00:05:11,430
So anything that I'm not actually perceiving or remembering any matter of fact.

41
00:05:11,430 --> 00:05:19,330
How can I infer such a matter of fact, and he says it has to be true causal inference.

42
00:05:19,330 --> 00:05:25,930
So in the enquiry, he starts out with that epistemological investigation.

43
00:05:25,930 --> 00:05:30,250
How can we discover things which we're not perceiving or remembering?

44
00:05:30,250 --> 00:05:35,110
He says it must be by means of a causal inference. Now let's investigate that.

45
00:05:35,110 --> 00:05:40,700
Whereas in the treaties, he starts out with the inference from the impression to the idea,

46
00:05:40,700 --> 00:05:50,020
a specific causal inference and probable inference in general only comes into the picture later on when he's discussing the uniformity principle.

47
00:05:50,020 --> 00:06:00,290
So the argument in the treaties is actually in in this way more convoluted than the one in the enquiry where he identifies probable reasoning,

48
00:06:00,290 --> 00:06:04,680
the causal reasoning right from the beginning.

49
00:06:04,680 --> 00:06:12,930
OK, so the argument initially the first the first part of the argument is to show that causal inference cannot be a priority.

50
00:06:12,930 --> 00:06:21,330
We've got that whenever we make an inference from one thing to another, according to Hume, that's assuming some causal relation between them.

51
00:06:21,330 --> 00:06:27,530
How can we find out about causal relations will only by experience?

52
00:06:27,530 --> 00:06:34,370
Now in the treaties, humour argues that this is the case because whenever we perceive a cause or

53
00:06:34,370 --> 00:06:39,290
what we take to be a cause we can imagine anything following is its effect.

54
00:06:39,290 --> 00:06:43,880
We can conceive of any different causal relation, take any object.

55
00:06:43,880 --> 00:06:53,660
It could be followed by any other as its effect and the mere conceive ability that something different could happen from what we expect.

56
00:06:53,660 --> 00:06:59,300
That's enough to show that the inference from the course to the effect cannot be a priori.

57
00:06:59,300 --> 00:07:04,440
At least that's what Hume seems to be arguing in the treaties.

58
00:07:04,440 --> 00:07:13,960
So he seems to be assuming that if you have an April try inference, it has to yield total certainty.

59
00:07:13,960 --> 00:07:19,920
Now in the enquiry. He doesn't argue like that.

60
00:07:19,920 --> 00:07:25,860
He argues far more powerfully, he says, in advance of experience.

61
00:07:25,860 --> 00:07:32,160
It's not just that we can conceive other things happening, but actually in advance of experience.

62
00:07:32,160 --> 00:07:36,140
Any expectation would be completely arbitrary.

63
00:07:36,140 --> 00:07:44,760
So it's not just that other things are possible, it's that we've no reason whatsoever for expecting one effect rather than another.

64
00:07:44,760 --> 00:07:52,330
And he gives a thought experiment of Adam. And he introduces Adam in the abstract with billiard balls.

65
00:07:52,330 --> 00:07:57,370
And he we have Adam, seeing one billion bill moving towards another.

66
00:07:57,370 --> 00:08:06,610
He's just been newly created by God. He's got perfect human faculties and God asks him what's going to happen when the one bill hits the other?

67
00:08:06,610 --> 00:08:12,500
Well, in advance of experience, Adam wouldn't have a clue. And that's a very powerful thought experiment,

68
00:08:12,500 --> 00:08:21,500
I think you put yourself in the position where you have no experience of bulls at all and you're asked to predict you couldn't do it.

69
00:08:21,500 --> 00:08:26,240
So it's not just that you could conceive of different things happening actually a priori,

70
00:08:26,240 --> 00:08:34,160
you'd have no reason whatsoever for preferring one thing rather than another. So it's a much more powerful argument than the one in the treaties.

71
00:08:34,160 --> 00:08:39,290
The argument in the treatise has led some people to think that humans are deductive,

72
00:08:39,290 --> 00:08:45,860
that he takes for granted that the only legitimate method of inference is deductive.

73
00:08:45,860 --> 00:08:52,700
Because it's enough to be able to conceive something different to undermine any argument, you don't get that impression from the enquiry.

74
00:08:52,700 --> 00:09:02,830
He's just saying, you know, in advance experience, it's completely arbitrary. You'd have no ground, whatever for preferring one thing to another.

75
00:09:02,830 --> 00:09:10,690
So here is the text where he he says that where any object presented to us and where we required to pronounce

76
00:09:10,690 --> 00:09:16,810
concerning the effect which will result from it without consulting past observation after what manner I beseech you,

77
00:09:16,810 --> 00:09:18,940
must the mine proceed in this operation?

78
00:09:18,940 --> 00:09:29,020
It must invent or imagine some event which it ascribes to the object as its effect and its plain that this invention must be entirely arbitrary.

79
00:09:29,020 --> 00:09:38,390
OK, so if we can't infer from calls to effect a priori, how does experience make a difference?

80
00:09:38,390 --> 00:09:45,470
Well, experience makes a difference, Hume claims, because when we see a followed by B repeatedly,

81
00:09:45,470 --> 00:09:54,200
when we see an air, we expect to be his a couple of passages from one three six.

82
00:09:54,200 --> 00:10:04,100
Now notice the second passage here, having found that this is the experience that characteristically makes us infer an effect from a cause, he says.

83
00:10:04,100 --> 00:10:08,090
Oh, we've discovered a new relation between cause and effect.

84
00:10:08,090 --> 00:10:11,210
Right. We had priority and contiguity.

85
00:10:11,210 --> 00:10:19,370
We noticed way back in one three two that they weren't enough, that there was this necessary connexion to be taken into account.

86
00:10:19,370 --> 00:10:25,500
We've now found a new relation, but the relation is constant conjunction.

87
00:10:25,500 --> 00:10:36,420
And it's very clear that this passage here from one three six three is linking back to one three to 11 both the phrasing and the capitalisation.

88
00:10:36,420 --> 00:10:43,350
So it's indicative that humans are seeing constant conjunction as the key to necessary connexion.

89
00:10:43,350 --> 00:10:51,900
And this is where he identifies that one important thing the the text from one three two is sometimes misinterpreted.

90
00:10:51,900 --> 00:10:57,690
People say humans saying there is a necessary connexion to be taken into account.

91
00:10:57,690 --> 00:11:02,400
That's obviously more than constant conjunction. No, he's not saying that.

92
00:11:02,400 --> 00:11:09,630
If you look back at one three two, he's talking about a single case priority and contiguity,

93
00:11:09,630 --> 00:11:13,860
and he's saying there is something else beyond in a causal relation.

94
00:11:13,860 --> 00:11:19,020
Now he's saying, Oh, I found this something else. The something else is constant conjunction.

95
00:11:19,020 --> 00:11:23,760
OK. So it's very important when you read necessarily connexion in one three to don't

96
00:11:23,760 --> 00:11:28,620
take for granted that human there is meaning some what we call now a thick,

97
00:11:28,620 --> 00:11:33,540
necessary connexion. That's something that goes beyond Hume's analysis of causation.

98
00:11:33,540 --> 00:11:42,840
There's no evidence for that at all. And one three six is making clear that he is seeing constant conjunction as providing the key there.

99
00:11:42,840 --> 00:11:53,470
So we'll talk later, obviously, about Hume's analysis of causation, but that that passage from one three two is quite often misunderstood.

100
00:11:53,470 --> 00:12:00,700
OK, so what is the link between necessary connexion and constant conjunction?

101
00:12:00,700 --> 00:12:04,390
Well, perhaps it will appear in the end that the necessary connexion depends on the

102
00:12:04,390 --> 00:12:08,290
inference instead of the inference is depending on the necessary connexion.

103
00:12:08,290 --> 00:12:17,980
That's a clue to what's going to come. Humans going to discover that it's the inference is that we make in response to

104
00:12:17,980 --> 00:12:23,110
observed constant conjunction that give rise to the idea of necessary connexion.

105
00:12:23,110 --> 00:12:30,910
So you might naturally think that when we make a causal inference, we start with the idea of necessary connexion and use that to make the inference.

106
00:12:30,910 --> 00:12:34,900
He's saying it's the other way round. Actually, we find ourselves making an inference.

107
00:12:34,900 --> 00:12:48,810
That's how we get the idea of necessary connexion. Now, as we saw last time, he frames his discussion in terms of faculties,

108
00:12:48,810 --> 00:12:56,550
the next question is whether experience produces the idea by means of the understanding or imagination is the reason.

109
00:12:56,550 --> 00:13:03,360
Or is it associational processes? And remember that understanding and reason are synonyms for Hume.

110
00:13:03,360 --> 00:13:12,930
At least they are. The vast majority of cases they are. And here you can see he is alternating between the terms of elegant variation.

111
00:13:12,930 --> 00:13:17,220
He's going to argue that reason cannot ground the inference,

112
00:13:17,220 --> 00:13:23,250
and he's going to conclude that therefore it's imaginative process is a process he will call custom.

113
00:13:23,250 --> 00:13:37,430
That does say. Well, let's ask ourselves again, we're making causal inference, imagine ourselves in the position of Adam.

114
00:13:37,430 --> 00:13:44,090
A priority. We cannot infer from one thing to another. How does experience make a difference?

115
00:13:44,090 --> 00:13:49,310
Well, it makes a difference because we find ourselves extrapolating from what we've seen.

116
00:13:49,310 --> 00:13:57,420
We've seen a followed by be. We now expect that they will continue to be followed by be.

117
00:13:57,420 --> 00:14:02,130
So we need to appeal to something that's commonly called the uniformity principle.

118
00:14:02,130 --> 00:14:10,080
Hume doesn't call it the uniformity principle, but most Hume scholars do useful term now in the treaties.

119
00:14:10,080 --> 00:14:17,070
The uniformity principle is introduced very explicitly if reason determined us.

120
00:14:17,070 --> 00:14:21,030
In other words, if Reason was responsible for a causal inference,

121
00:14:21,030 --> 00:14:28,680
it would proceed upon that principle that instances of which we have had no experience must resemble those of which we have had experience,

122
00:14:28,680 --> 00:14:35,330
and that the course of nature continues always uniformly the same.

123
00:14:35,330 --> 00:14:41,400
Now, I think the phrasing of that is slightly unfortunate and misleading.

124
00:14:41,400 --> 00:14:52,560
Because it gives the impression that Hume here is saying something conditional, if reason determined does, then it would proceed upon that principle.

125
00:14:52,560 --> 00:14:59,110
Now this is the only case in which the uniformity principle is put in a conditional way.

126
00:14:59,110 --> 00:15:09,790
Later in the treatise, he says probability in a probable argument is founded on the presumption of a resemblance in the enquiry.

127
00:15:09,790 --> 00:15:15,640
All our experimental conclusions proceed upon the supposition that the future will be conformable to the past.

128
00:15:15,640 --> 00:15:23,330
That's not conditional it saying we do assume the uniformity principle.

129
00:15:23,330 --> 00:15:29,720
And we find that in other passages, too, so so that that initial presentation in the treaties is actually unique,

130
00:15:29,720 --> 00:15:39,300
it's the only case where he says if reason determined us, then it would be using the principle.

131
00:15:39,300 --> 00:15:48,240
Notice also that the original presentation in the treatise looks rather implausible, he says the course of nature continues always uniformly the same.

132
00:15:48,240 --> 00:15:51,690
It doesn't continue always uniformly. The same does it?

133
00:15:51,690 --> 00:16:04,460
What we assume is that the same basic laws are applying the forms of the uniformity principle in the enquiry of Vega and hence actually better.

134
00:16:04,460 --> 00:16:14,480
We take the past as a rule for the future, we suppose that similar causes will have similar effects.

135
00:16:14,480 --> 00:16:24,830
Now, importantly, Hugh does not think that when we make an inductive inference, we explicitly ponder the uniformity.

136
00:16:24,830 --> 00:16:29,340
OK. Indeed. Later in the treaties, he will say quite explicit.

137
00:16:29,340 --> 00:16:37,950
That's not the case. And most inductive inferences are unreflective, we see, and we just expect to be.

138
00:16:37,950 --> 00:16:48,340
We don't think about the uniformity principle. So there is a question about what role the uniformity principle is playing for him.

139
00:16:48,340 --> 00:16:57,490
And I think I'm not alone in this that the most plausible way of taking this is to say that when we make an inductive inference,

140
00:16:57,490 --> 00:17:00,820
we manifest our commitment to the uniformity principle.

141
00:17:00,820 --> 00:17:07,420
The very fact that we're making the inference shows that we are taking the past as a rule for the future.

142
00:17:07,420 --> 00:17:14,230
So it's not that the uniformity principle is acting as some premise in our reasoning,

143
00:17:14,230 --> 00:17:22,310
but we are taking it for granted when we make an inductive inference. OK, so we still get the same question.

144
00:17:22,310 --> 00:17:27,290
Does the uniformity principle, does our supposition of the uniformity principle?

145
00:17:27,290 --> 00:17:40,990
Does that have a rational basis? Now in the treaties, immediately after he's introduced the uniformity principle,

146
00:17:40,990 --> 00:17:49,840
he focuses on the question of whether it can be founded on some kind of ratty assassination,

147
00:17:49,840 --> 00:17:54,880
some kind of stepwise argument in order therefore to clear up this matter.

148
00:17:54,880 --> 00:18:00,520
Let us consider all the arguments upon which such a proposition may be supposed to be founded.

149
00:18:00,520 --> 00:18:04,000
And as these must be derived either from knowledge or probability,

150
00:18:04,000 --> 00:18:11,620
let us cast our eye on each of these degrees of evidence and see whether they afford any just conclusion of this nature.

151
00:18:11,620 --> 00:18:19,450
And he's saying knowledge and probability. The very next sentence makes clear he means demonstration and probable argument.

152
00:18:19,450 --> 00:18:29,050
So we've got two different kinds of arguments that could be used to support the uniformity principle, demonstrative and probable.

153
00:18:29,050 --> 00:18:38,260
So let's just have a few slides on this distinction, the distinction between the demonstrative and probable reasoning.

154
00:18:38,260 --> 00:18:46,090
It's more or less the same as the modern distinction between deductive and inductive argument.

155
00:18:46,090 --> 00:18:53,350
This doesn't mean formally deductive, by the way. Human, like most philosophers of his period, was completely disdainful of formal logic.

156
00:18:53,350 --> 00:19:00,670
So when we say an argument is deductive, fully valid when talking about Hume, we do not mean formally valid.

157
00:19:00,670 --> 00:19:07,930
What we mean is it's an argument in which if the premises are true, the conclusion cannot possibly be false.

158
00:19:07,930 --> 00:19:13,720
OK, so the conclusion follows from the premises with absolute and complete certainty.

159
00:19:13,720 --> 00:19:20,740
But that needn't be in virtue of its logical form. OK.

160
00:19:20,740 --> 00:19:29,050
Improbable reasoning, by contrast, the links that constitute the argument are merely probable, not demonstrative.

161
00:19:29,050 --> 00:19:38,030
They don't follow with complete certainty. Now, Hume, again here he I've already noted his fondness for elegant variation,

162
00:19:38,030 --> 00:19:41,930
like talking about reason and the understanding or imagination and the fancy,

163
00:19:41,930 --> 00:19:46,250
and I'm afraid with regard to probable reasoning, he does something of the same.

164
00:19:46,250 --> 00:19:53,600
So in the treaty as he normally does refer to it as probable reasoning in the same way that Locke does in the enquiry, he doesn't.

165
00:19:53,600 --> 00:20:03,200
And the reason is that Hume thinks the talking of the word probable when we're talking about completely certain inferences like,

166
00:20:03,200 --> 00:20:09,230
you know, I hold up a stone and I let go of it. And I'm absolutely certain that it's going to fall, OK?

167
00:20:09,230 --> 00:20:14,390
It's not demonstrative, but but inductively certain that it's going to fall.

168
00:20:14,390 --> 00:20:18,320
And Hume thinks it's in this to talk about that as merely probable.

169
00:20:18,320 --> 00:20:22,310
He says it's a proof I've got uniform experience.

170
00:20:22,310 --> 00:20:27,140
So he doesn't actually want to use the word probable to talk about probable arguments in the way that locked it.

171
00:20:27,140 --> 00:20:35,540
He wants a different name for the for the broader category of inductive arguments, some of which he thinks yields complete psychological certainty.

172
00:20:35,540 --> 00:20:37,660
And some of which don't.

173
00:20:37,660 --> 00:20:45,370
So he calls them reasonings concerning matter of fact or reasonings concerning matter of fact, an existence or moral reasoning.

174
00:20:45,370 --> 00:20:50,910
So if you see the term moral reasoning, he's talking about this same category.

175
00:20:50,910 --> 00:21:01,760
So when I'm discussing Hugh, Mike quite often used the word factual reasoning, and that's just shorthand for reasoning concerning matter of fact.

176
00:21:01,760 --> 00:21:06,470
OK, so we've got two different forms of reasoning, demonstrative and probable,

177
00:21:06,470 --> 00:21:13,370
and I've said they're more or less the same as deductive and inductive in in our modern usage.

178
00:21:13,370 --> 00:21:22,850
Now that's actually a little bit controversial because some scholars have thought that when Hume talks about demonstrative argument,

179
00:21:22,850 --> 00:21:29,430
he must be confining it to deductive arguments that a priori.

180
00:21:29,430 --> 00:21:33,960
And there are two main reasons why people have thought this.

181
00:21:33,960 --> 00:21:37,920
I think, by the way, it's certainly wrong and I'm going to explain why.

182
00:21:37,920 --> 00:21:48,000
But the the two reasons why people have thought this is that Hume says that no matter of fact, can be demonstrated.

183
00:21:48,000 --> 00:21:57,430
And some people have taken this to mean that according to him, no matter of fact, can be the conclusion of a demonstrative argument.

184
00:21:57,430 --> 00:21:59,650
Now, I think those are just different things.

185
00:21:59,650 --> 00:22:11,770
OK, so suppose I've got an argument which is disruptively valid and it's got three premises P1, P2, P3 and Conclusion C. And it's a valid argument.

186
00:22:11,770 --> 00:22:19,720
OK. Can I be said to have demonstrated C? No, because P1, P2, P3 might be controversial.

187
00:22:19,720 --> 00:22:30,340
You might not believe those in which case you're not going to agree that I've demonstrated c all I've done is demonstrated C from P1, P2, P3.

188
00:22:30,340 --> 00:22:35,350
So you never says that a matter of fact can't be the conclusion of a demonstrative argument, he says.

189
00:22:35,350 --> 00:22:41,930
No matter of fact can be demonstrated. So. So much for that first reason,

190
00:22:41,930 --> 00:22:49,220
the second reason is this that Hume does say explicitly in both the treaties and the enquiry that

191
00:22:49,220 --> 00:22:58,430
only in mathematics does demonstrative argument apply or does it apply to any significant extent.

192
00:22:58,430 --> 00:23:08,290
So in the enquiry, it seems to me that the only objects of the abstract sciences or of demonstration are quantity and no.

193
00:23:08,290 --> 00:23:16,070
The team gives his reason for this. The reason why demonstration, he thinks, is more or less restricted to mathematics.

194
00:23:16,070 --> 00:23:25,790
Is because only in mathematics do you get ideas that are sufficiently precise with intricate relations that you can follow a long chain of reasoning.

195
00:23:25,790 --> 00:23:32,690
So outside mathematics, Hume thinks demonstrative arguments tend to be trivial.

196
00:23:32,690 --> 00:23:35,930
Like, you can conclude that where there is no property,

197
00:23:35,930 --> 00:23:42,590
there is no injustice by defining justice as a violation of property, he gives this as an example in the enquiry.

198
00:23:42,590 --> 00:23:47,960
But that's just trivial. So it's only in mathematics.

199
00:23:47,960 --> 00:23:55,520
He thinks that demonstration that he's particularly complex demonstration really gets a foothold.

200
00:23:55,520 --> 00:24:01,940
The the obvious crucial case that we need to look at in order to sort this out is applied mathematics,

201
00:24:01,940 --> 00:24:11,120
because in applied mathematics, you get complex mathematical reasoning, but it starts from a posterior right premises.

202
00:24:11,120 --> 00:24:21,730
So a philosopher who accepts the existence of applied mathematical demonstrations is not assuming that demonstration is a priori.

203
00:24:21,730 --> 00:24:29,290
Well, Hume's most explicit discussion of applied mathematics, he calls it mixed mathematics is Enquiry Section four.

204
00:24:29,290 --> 00:24:35,470
It's a law of motion discovered by experience that the moment or force of any body in motion.

205
00:24:35,470 --> 00:24:43,120
Think of that as momentum is in the compound ratio or proportion of its solid contents tense and its velocity

206
00:24:43,120 --> 00:24:49,870
momentum is mass times velocity and consequently that a small force may remove the greatest obstacle.

207
00:24:49,870 --> 00:24:57,810
If by any contrivance, we can increase the velocity of that force so as to make it an overmatched for its antagonist.

208
00:24:57,810 --> 00:25:04,640
So here's an illustration. We've got a large body moving to the left at four metres per second.

209
00:25:04,640 --> 00:25:13,250
We've got a very small body. If we move it to the right fast enough, it will displace the large body.

210
00:25:13,250 --> 00:25:22,940
It'll make it move in a different direction, it seems to me. Hume is clearly expressing something like this.

211
00:25:22,940 --> 00:25:29,780
OK, so mathematics can assist us in working out the implications of the law.

212
00:25:29,780 --> 00:25:33,800
But the discovery of the law itself is owing merely to experience and all the

213
00:25:33,800 --> 00:25:38,030
abstract reasonings in the world could never give us any knowledge of it.

214
00:25:38,030 --> 00:25:45,050
So humans being absolutely explicit here, the law of conservation of momentum is discovered by experience,

215
00:25:45,050 --> 00:25:59,190
and then mathematics can help us to apply it. So it's very clear Hume does not restrict demonstrative reasoning to a priori reasoning.

216
00:25:59,190 --> 00:26:02,310
OK, so as I mentioned in the treaties,

217
00:26:02,310 --> 00:26:14,290
having brought in the uniformity principle and said that all our inductive inferences depend on this depend on this extrapolation from past to future.

218
00:26:14,290 --> 00:26:23,040
He's now asking. Can the uniformity principle be supported by either demonstrative argument or probable argument?

219
00:26:23,040 --> 00:26:32,100
As I say, by deduction or inductive argument, as we would now say now in the enquiry, the argument is more complete.

220
00:26:32,100 --> 00:26:39,070
He doesn't just restrict his discussion of the uniformity principle to demonstration and probably argument.

221
00:26:39,070 --> 00:26:46,240
He also takes the trouble to rule out that sensation could provide a basis for the uniformity principle.

222
00:26:46,240 --> 00:26:51,280
And he also also rules out intuition what we would call self-evident.

223
00:26:51,280 --> 00:26:58,000
There is no known connexion between the sensible qualities. That's the sensory qualities and the secret powers.

224
00:26:58,000 --> 00:26:59,020
And consequently,

225
00:26:59,020 --> 00:27:06,940
the mind is not led to form such a conclusion concerning their constant and regular conjunction by anything which it knows of their nature.

226
00:27:06,940 --> 00:27:16,450
So think back to the Adam thought experiment. Adam sees the sensory qualities of the ball and the shape the colour.

227
00:27:16,450 --> 00:27:27,370
He cannot infer from that what's going to happen. But nor can he infer from that that what will happen will be consistent over time.

228
00:27:27,370 --> 00:27:33,820
So therefore, sensation alone does not give us a basis for expecting consistent behaviour.

229
00:27:33,820 --> 00:27:39,300
It doesn't give us a basis for extrapolating from past to future.

230
00:27:39,300 --> 00:27:42,870
But also, the connexion between past and future is not intuitive,

231
00:27:42,870 --> 00:27:50,000
it's not self-evident that what happens in the future will resemble what happened in the past.

232
00:27:50,000 --> 00:27:55,200
So the enquiry argument, as I say, is more complete.

233
00:27:55,200 --> 00:28:00,690
And rather nicely, the letter from a gentleman to his friend in Edinburgh, which you published in 1745,

234
00:28:00,690 --> 00:28:09,900
so we've got the enquiry appearing 17 48 in 1745, Hume applies for a chair of Edinburgh.

235
00:28:09,900 --> 00:28:17,940
He wants to be a professor of philosophy at Edinburgh New medical philosophy, so he would be talking about the soul and so, so forth.

236
00:28:17,940 --> 00:28:23,220
Not surprisingly, the clerics of Edinburgh weren't best pleased with this.

237
00:28:23,220 --> 00:28:26,160
They didn't want somebody who had written the treatise of human nature,

238
00:28:26,160 --> 00:28:34,060
which seemed rather atheistic to be becoming a professor of new mystical philosophy at Edinburgh.

239
00:28:34,060 --> 00:28:41,620
And so they stirred up a hubbub against him, and Hume wrote this letter from a gentleman to his friend in Edinburgh in order to

240
00:28:41,620 --> 00:28:48,600
defend himself in order to say the treaties isn't as bad as it's been painted.

241
00:28:48,600 --> 00:28:54,970
Now, the letter from a gentleman was written while Hume was actually acting as a tutor for a madman,

242
00:28:54,970 --> 00:28:58,950
the man the Marquess of Annandale, out in the country.

243
00:28:58,950 --> 00:29:05,940
He did not have the treaties with him, so he wrote this letter to the people in Edinburgh.

244
00:29:05,940 --> 00:29:17,700
The relevant people? Actually, it was Henry Hume who I mentioned in the first lecture who who got this published on his behalf.

245
00:29:17,700 --> 00:29:25,230
He wrote this letter defending himself about the philosophy of the treaties, but without actually being able to quote the treaties itself.

246
00:29:25,230 --> 00:29:30,240
So the letter is actually a very interesting document for getting insights into Hume's thought,

247
00:29:30,240 --> 00:29:34,470
and this sentence is, I think, particularly illuminating. It is.

248
00:29:34,470 --> 00:29:42,900
So you're saying in order to explain my position on the causal maxim because that had been picked up by his critics,

249
00:29:42,900 --> 00:29:48,030
I mentioned the causal maxim plays a big role in arguments for the existence of God in the cosmological argument.

250
00:29:48,030 --> 00:29:54,420
So that was a matter of considerable interest says. In order to explain my position on the causal maxim,

251
00:29:54,420 --> 00:30:04,570
I need to tell you that it's common for philosophers to distinguish the kinds of evidence into intuitive, demonstrative, sensible and moral.

252
00:30:04,570 --> 00:30:13,570
And all I was doing in treaties one, three three was denying that the causal maxim can be based on either intuitive or demonstrative evidence.

253
00:30:13,570 --> 00:30:20,860
I wasn't denying it. On the contrary, I said that the causal maxim is based on moral evidence on probable argument.

254
00:30:20,860 --> 00:30:28,680
No, you didn't. Hume, you forgot to say that in the treaties, but evidently he intended to say.

255
00:30:28,680 --> 00:30:37,480
OK, now what's particularly interesting here is that we've got four forms of evidence they're intuitive, demonstrative, sensible, moral and human.

256
00:30:37,480 --> 00:30:44,390
And the enquiry is explicitly ruling out each of these as a basis for the uniformity principle.

257
00:30:44,390 --> 00:30:51,700
And again, we've got that letter being written no doubt at the time that he was working on the enquiry.

258
00:30:51,700 --> 00:30:55,190
OK, so. In the treaties and in the enquiry,

259
00:30:55,190 --> 00:31:02,810
demonstrative argument for the uniformity principle is ruled out in the same way we can at least conceive a change in the course of nature,

260
00:31:02,810 --> 00:31:11,180
which proves that such a change is not absolutely impossible and that yields a refutation of any pretended demonstration against it.

261
00:31:11,180 --> 00:31:22,470
If you could demonstrably demonstratively prove the uniformity principle, you know, e.g. by arguing from past uniformity to future uniformity.

262
00:31:22,470 --> 00:31:27,810
And then that would mean that it was inconceivable that the course of nature would change.

263
00:31:27,810 --> 00:31:33,840
But it's not inconceivable. Therefore, you can't have such an argument and probable argument.

264
00:31:33,840 --> 00:31:42,040
You can't have a probable argument for the uniformity principle because probable argument itself depends on the uniformity principle.

265
00:31:42,040 --> 00:31:49,420
Now, this is where the intelligence in the treaties argument comes out, because as I said in the treaties,

266
00:31:49,420 --> 00:31:53,410
the argument starts from the inference, from the impression to the idea.

267
00:31:53,410 --> 00:31:57,850
He hasn't actually mentioned probable argument in general yet.

268
00:31:57,850 --> 00:32:03,310
He only comes to mention it when he discusses the ways of supporting the uniformity principle.

269
00:32:03,310 --> 00:32:11,260
And then when he comes to that, he has to go back to square one and say, Can the uniformity principle be sought supported by a probable argument?

270
00:32:11,260 --> 00:32:19,090
Well, no, because probable argument depends on causation, and causation depends on experience,

271
00:32:19,090 --> 00:32:26,780
and therefore it depends on the uniformity principle and therefore we've got a circularity. In the enquiry, right at the beginning of the argument,

272
00:32:26,780 --> 00:32:36,470
Hume has said I am concerned about all inferences from things directly perceived or remembered to other matters of fact.

273
00:32:36,470 --> 00:32:42,500
So all inferences concerning matter of fact, I found it on causation and therefore on uniformity.

274
00:32:42,500 --> 00:32:50,980
So he's already proved that general point before he comes to ask the question about the uniformity principle.

275
00:32:50,980 --> 00:32:58,390
So I've got here a couple of slides that summarise the structure of the argument to some extent,

276
00:32:58,390 --> 00:33:02,560
you know, I'm shoehorning the argument into a particular structure,

277
00:33:02,560 --> 00:33:11,800
but I think it's quite illuminating that the structure of the argument can be revealed by thinking in terms of the founded on relation.

278
00:33:11,800 --> 00:33:19,780
So I'm not claiming that Hume expresses it in this way. I'm not claiming that in every case, Hume uses the term founded on.

279
00:33:19,780 --> 00:33:24,960
But if you think through the logic of the arguments in this way, I think it can help.

280
00:33:24,960 --> 00:33:32,910
So what we've got there at the top is that he argues that causal inference is founded on experience,

281
00:33:32,910 --> 00:33:37,140
that reasoning from experience is founded on the uniformity principle.

282
00:33:37,140 --> 00:33:43,100
And he concludes, therefore, that causal inference is founded on the uniformity principle.

283
00:33:43,100 --> 00:33:48,790
OK. But the uniformity principle can't be founded on demonstration.

284
00:33:48,790 --> 00:34:00,930
Nor can it be founded on probability. So that's the weather you can see that you can't if you look at the the yellow box in the middle there,

285
00:34:00,930 --> 00:34:05,340
that's saying it's not the case that the uniformity principle is founded on probability,

286
00:34:05,340 --> 00:34:10,680
but you can see that in order to reach that lemma, he's having to go through at the bottom.

287
00:34:10,680 --> 00:34:17,200
The same argument that he started with, we've got a repetition within the argument.

288
00:34:17,200 --> 00:34:23,980
And then because uniformity principle is not founded on demonstration and is not founded on probability, that's the two yellow boxes.

289
00:34:23,980 --> 00:34:29,380
He concludes that the uniformity principle is not founded on reason and because causal inference

290
00:34:29,380 --> 00:34:33,310
is founded on the uniformity principle in the uniformity principle is not founded on reason.

291
00:34:33,310 --> 00:34:40,090
It follows that causal inference isn't founded on reason. By the way, that last stage of the argument is an interesting one.

292
00:34:40,090 --> 00:34:48,790
Some interpretations of Hume's argument, I think, cannot make sense of that final stage, but I'm not going to focus on that here.

293
00:34:48,790 --> 00:34:54,250
Is the argument in the enquiry? It's much more elegant and it's more complete.

294
00:34:54,250 --> 00:35:00,570
You can see right at the top, he starts with probable inference. He's founded on causation.

295
00:35:00,570 --> 00:35:05,820
Any reasoning from causation is founded on experience, therefore probable inference is founded on experience.

296
00:35:05,820 --> 00:35:12,450
Any reasoning from experience is founded on the uniformity principle. Therefore, probable inference has to be found on the uniformity principle.

297
00:35:12,450 --> 00:35:17,670
So he's got that right at the start, and you can see at the bottom there, we've got that.

298
00:35:17,670 --> 00:35:22,770
The uniformity principle isn't founded on intuition, and it's not founded on sensation either.

299
00:35:22,770 --> 00:35:25,560
And it's not founded on demonstration,

300
00:35:25,560 --> 00:35:34,230
and it can't be founded on probability because probability probable argument is founded on it on the uniformity principle.

301
00:35:34,230 --> 00:35:39,480
So when Hume concludes that the uniformity principle isn't founded on reason in the enquiry,

302
00:35:39,480 --> 00:35:47,260
he's drawing that conclusion from the fact that none of the four forms of evidence can support it.

303
00:35:47,260 --> 00:35:58,310
OK. The sceptical conclusion notice, I'm putting the word, I'm putting the word sceptical in quotes here.

304
00:35:58,310 --> 00:36:03,590
It's debateable how sceptical Hume actually is about induction.

305
00:36:03,590 --> 00:36:07,820
He is. He continues to be a supporter of induction.

306
00:36:07,820 --> 00:36:18,320
He's certainly not saying that we should not use induction for drawing inferences about the world he thinks is the only way we can.

307
00:36:18,320 --> 00:36:22,880
But on the other hand, there is something sceptical about the conclusion of this argument.

308
00:36:22,880 --> 00:36:31,760
He is denying that induction has a rational basis of a kind that somebody might have thought it would have.

309
00:36:31,760 --> 00:36:35,510
Even after experience has informed us of constant conjunction,

310
00:36:35,510 --> 00:36:40,790
it is impossible for us to satisfy ourselves by our reason why we should extend that experience beyond

311
00:36:40,790 --> 00:36:46,550
those particular instances which have fallen under our observation and later in the treaties one,

312
00:36:46,550 --> 00:36:53,060
three, 12 20, as I mentioned there. He makes the same point even more emphatically.

313
00:36:53,060 --> 00:36:58,250
We have no reason for drawing any inference beyond things we've experienced.

314
00:36:58,250 --> 00:37:09,690
And there's a couple of quotes from the enquiry there. I think the last quote here probably gives the most precise version of Hume's conclusion.

315
00:37:09,690 --> 00:37:13,250
And by the way, Don Garrett agrees with me on that.

316
00:37:13,250 --> 00:37:17,810
We interpret the argument differently in various ways.

317
00:37:17,810 --> 00:37:30,950
But in recent work that he's published, he has very much focussed on on this issue that there is a step in all inductive inferences,

318
00:37:30,950 --> 00:37:36,860
which is not supported by any argument or press the understanding. So.

319
00:37:36,860 --> 00:37:49,880
This is tricky because when we interpret Hume's conclusion, Hume's Hume's conclusion that inductive inference is not founded on reason in some sense.

320
00:37:49,880 --> 00:37:58,730
How you interpret that is going to depend on how you interpret human reasons, so we come back to this question about how we should understand.

321
00:37:58,730 --> 00:38:06,750
Faculty talk and as I mentioned in the last lecture, there's a lot of variation here.

322
00:38:06,750 --> 00:38:11,190
So we've got the extrapolation from observed to unobserved is a step taken by the mind,

323
00:38:11,190 --> 00:38:17,880
which is not supported by any argument or process of the understanding. What exactly is that ruling out?

324
00:38:17,880 --> 00:38:27,480
Well, according to Beecham and Rosenberg and Beecham and Mathis, somebody else, he wrote, with all humans ruling out is a priori demonstration.

325
00:38:27,480 --> 00:38:35,760
Pretty sure that's wrong. According to David Owen, all that's being ruled out is lock lock in stepwise inference,

326
00:38:35,760 --> 00:38:40,950
according to Lerve non associative reasoning, according to Don Garret.

327
00:38:40,950 --> 00:38:47,460
For the human reason just means ratty assassination argument.

328
00:38:47,460 --> 00:38:50,850
It's not normative, it's purely descriptive.

329
00:38:50,850 --> 00:39:04,200
And you can see now why Don Garrett is very keen on this law's quotation because many of our inductive inferences do involve complex rattus a nation,

330
00:39:04,200 --> 00:39:11,130
I suppose. For example, we draw inductive conclusions about the motion of the planets.

331
00:39:11,130 --> 00:39:15,020
That's going to involve a lot of Rattus in Asian.

332
00:39:15,020 --> 00:39:21,860
But there is a step taken by the mind, namely the infants and past the future, the supposition that laws will continue the same.

333
00:39:21,860 --> 00:39:25,760
That step is not founded on Ratti as a nation.

334
00:39:25,760 --> 00:39:40,780
So by focussing on the individual step within the argument, Don Garrett is able to square Hume's conclusion with his interpretation of reason.

335
00:39:40,780 --> 00:39:51,320
Now, a related question is whether houM views his discussion of induction as being epistemological or psychological.

336
00:39:51,320 --> 00:40:00,710
And Don Garrett puts the emphasis on psychology, he says Hume is talking about the processes that go on in our mind when we make inductive inferences.

337
00:40:00,710 --> 00:40:11,840
And according to Don, we don't make those inferences because we have been persuaded by some rational nation to do it.

338
00:40:11,840 --> 00:40:16,850
My interpretation of humour and reason is that reason is the cognitive faculty,

339
00:40:16,850 --> 00:40:24,800
so I take Hume's conclusion to actually have epistemological bite when we draw an inductive inference.

340
00:40:24,800 --> 00:40:30,830
We're not doing it because we have cognisant that the future will resemble the past.

341
00:40:30,830 --> 00:40:35,590
We're reading that into the world rather than reading it off the world.

342
00:40:35,590 --> 00:40:43,030
Now, actually, I mean, there's quite a subtle difference between those two. I'm not going to discuss that debate.

343
00:40:43,030 --> 00:40:53,190
There's plenty you can read in the literature on that, and I will be giving you a handout with readings next time.

344
00:40:53,190 --> 00:40:58,710
But if we ask, is Hume's discussion epistemological or psychological?

345
00:40:58,710 --> 00:41:01,420
It seems to me the answer is both.

346
00:41:01,420 --> 00:41:11,020
Hume is drawing a conclusion about the psychology of our inductive inference, but he's doing it by revealing a certain impossibility.

347
00:41:11,020 --> 00:41:17,800
He's not just saying, look at the way we do inductive inference, we do it immediately and instinctively.

348
00:41:17,800 --> 00:41:22,060
He does say that later in the treaties. But that's not what he's saying in one, three six.

349
00:41:22,060 --> 00:41:29,680
And it's not what he's saying an enquiry for what he's saying there is we could not possibly do it any other way.

350
00:41:29,680 --> 00:41:37,860
We could not possibly do it using reason. And that is, it seems to me, an epistemological conclusion.

351
00:41:37,860 --> 00:41:45,150
So humans drawing an epistemological conclusion, but using it then to draw a psychological conclusion,

352
00:41:45,150 --> 00:41:50,640
we could not possibly do it that way and therefore we don't do it that way.

353
00:41:50,640 --> 00:42:00,840
OK. So briefly, now I want to run through Hume's alternative theory.

354
00:42:00,840 --> 00:42:06,070
How do we actually do inductive inference?

355
00:42:06,070 --> 00:42:13,950
Well, of course, he says it's due to something like the Association of Ideas.

356
00:42:13,950 --> 00:42:26,410
So we've seen that when our minds wander. They standardly move according to three associative relations,

357
00:42:26,410 --> 00:42:32,710
if I'm thinking about something on my mind may be drawn to something that resembles it or something that's contiguous to it,

358
00:42:32,710 --> 00:42:43,890
or something that is related causally to it. And now he's going to explain that inductive inference is due to a similar kind of associational process.

359
00:42:43,890 --> 00:42:52,500
It's not quite the same. It's not mere association of ideas. But it's rather similar when I see a followed by B again and again and again,

360
00:42:52,500 --> 00:42:58,920
it sets up an associational link between my ideas of A and B so that if I think of a, I may think of B.

361
00:42:58,920 --> 00:43:05,740
Yes, that would be normal association of ideas. But when I actually see an impression of a.

362
00:43:05,740 --> 00:43:14,500
The fulsome vivacity of that impression is conveyed to the associated idea, and as a result, I come to actually believe that be will happen.

363
00:43:14,500 --> 00:43:24,440
That's not just association of ideas, it's a process that Hume calls custom.

364
00:43:24,440 --> 00:43:28,980
And. Hume is not negative about custom.

365
00:43:28,980 --> 00:43:32,970
You might think he would say, like Locke would have said, oh,

366
00:43:32,970 --> 00:43:39,060
if induction isn't due to reason, if it's due to custom, then it's somehow not respectable.

367
00:43:39,060 --> 00:43:43,500
He does not say that custom is the great guide of human life.

368
00:43:43,500 --> 00:43:47,490
It is that principle alone which renders our experience useful to us.

369
00:43:47,490 --> 00:43:55,710
It's only thanks to custom that we are able to draw inferences about the unobserved custom leads us to

370
00:43:55,710 --> 00:44:02,070
just assume that the unobserved will resemble the observed and therefore enables us to make inferences.

371
00:44:02,070 --> 00:44:12,680
He custom came up before in Hume's discussion of general ideas in one one seven, and here he does draw that link.

372
00:44:12,680 --> 00:44:19,640
OK, now recall that before Hume came to induction, he was setting an agenda.

373
00:44:19,640 --> 00:44:27,350
He wanted to explain the original impression of the senses or memory from which causal inference standardly starts.

374
00:44:27,350 --> 00:44:31,730
He wanted to explain the transition, the actual inference, which is what he's just done.

375
00:44:31,730 --> 00:44:38,180
And then he wants to explain the nature of the idea that such an inference delivers.

376
00:44:38,180 --> 00:44:48,480
That is the belief that B will occur. And so he's now going to talk about the nature of belief.

377
00:44:48,480 --> 00:44:58,620
So. The definition of an opinion or belief is that is an idea related to or associated with a present impression.

378
00:44:58,620 --> 00:45:05,690
So I see a I've got an impression of a that enlivens the idea of B.

379
00:45:05,690 --> 00:45:15,310
That is a belief that the will occur. It's an enlivened idea of B.

380
00:45:15,310 --> 00:45:22,400
And. I think it's worth noting that the humour here is addressing an issue that would

381
00:45:22,400 --> 00:45:27,380
later be identified by Fraga when Fraga was developing his logical system,

382
00:45:27,380 --> 00:45:32,910
which of course was the ancestor of the logical systems that we studied today.

383
00:45:32,910 --> 00:45:42,040
And. He pointed out that there's a difference between contemplating a proposition and actually asserting it.

384
00:45:42,040 --> 00:45:51,730
So if you have a conditional like implies, Q, P and Q are both propositions, you're not actually asserting P and you're not asserting Q.

385
00:45:51,730 --> 00:45:57,490
You might not even be asserting P implies Q, you might just be contemplating that conditional.

386
00:45:57,490 --> 00:46:02,200
So there's a difference between contemplating a proposition and actually asserting it.

387
00:46:02,200 --> 00:46:09,830
And what Hume is pointing out here is there's a difference between contemplating a proposition and actually believing it.

388
00:46:09,830 --> 00:46:14,060
So what is the difference between those two?

389
00:46:14,060 --> 00:46:22,430
Well, it can't be down to having some idea of existence, which somehow we attach to the proposition, you know,

390
00:46:22,430 --> 00:46:31,310
maybe I'm I'm I'm thinking about some proposition and then I attach the idea of existence to it to bring it about that.

391
00:46:31,310 --> 00:46:39,700
That's a belief that's not going to be the answer. And if we contemplate a proposition and then come to believe it.

392
00:46:39,700 --> 00:46:44,460
It needs to be the same proposition, so we can't change the content of it.

393
00:46:44,460 --> 00:46:53,850
So the difference, Hugh says, must lie in the manner in which we conceive it, the force and vivacity with which we contemplate it.

394
00:46:53,850 --> 00:47:01,600
So the difference between contemplation and belief is going to come down to fulsome and vivacity.

395
00:47:01,600 --> 00:47:06,190
So an opinion, therefore or belief, may be most accurately defined.

396
00:47:06,190 --> 00:47:12,280
A lively idea related to or associated with a present impression.

397
00:47:12,280 --> 00:47:24,320
So we get a definition of belief in terms of the theory of customary inference.

398
00:47:24,320 --> 00:47:35,380
So. We've seen that causal reasoning has to start from an impression of the senses or memory.

399
00:47:35,380 --> 00:47:40,600
I put that in quotes because strictly the ideas we have in memory are ideas, not impressions.

400
00:47:40,600 --> 00:47:46,070
But as we saw before, Hume thinks they have additional vivacity because they're in the memory.

401
00:47:46,070 --> 00:47:53,030
And that's enough to enable them to serve as impressions for the purpose of causal inference.

402
00:47:53,030 --> 00:48:03,250
And we've seen that belief is distinct from mere contemplation of an idea in terms of false and veracity.

403
00:48:03,250 --> 00:48:11,950
So it seems that when something becomes believed, what's happening is that we have what I've called a hydraulic theory of belief.

404
00:48:11,950 --> 00:48:23,290
We've got the false vivacity of the impression is conveyed by the associative link to the idea, and that's what gives it this greater vivacity.

405
00:48:23,290 --> 00:48:30,490
I've already alluded to this, but here we have Hume coming to this theory when any impression becomes present to us.

406
00:48:30,490 --> 00:48:39,940
It not only transports the mind to such ideas as it related to it, but likewise communicates to them a share of its force and vivacity.

407
00:48:39,940 --> 00:48:44,590
And Hume gives various illustrations to show that indeed,

408
00:48:44,590 --> 00:48:52,930
Association of Ideas does have this feature that it communicates force and vivacity from one idea to another.

409
00:48:52,930 --> 00:49:00,850
And later in the treatise, we see him applying a similar idea to probability if I've seen is sometimes

410
00:49:00,850 --> 00:49:05,320
followed by bees and sometimes followed by CS with a particular probability,

411
00:49:05,320 --> 00:49:07,360
a particular frequency in the past.

412
00:49:07,360 --> 00:49:16,720
Then when I see an A, the force and vivacity of that impression is conveyed to the different outcomes in proportion to their past frequency.

413
00:49:16,720 --> 00:49:23,950
So I will come to have a proportional belief in B and C.

414
00:49:23,950 --> 00:49:31,660
Now, all of this isn't entirely satisfactory. It looks philosophically rather crude.

415
00:49:31,660 --> 00:49:39,550
I mean, what is this force and vivacity? How does it really fit into the theory of ideas?

416
00:49:39,550 --> 00:49:48,850
Is it the case that you really can distinguish? And I contemplated idea from a belief in terms of a single feature, fulsome vivacity,

417
00:49:48,850 --> 00:49:53,770
which is the same thing that distinguishes ideas of memory from those of imagination,

418
00:49:53,770 --> 00:49:59,920
which is the same thing as those is what distinguishes impressions from ideas.

419
00:49:59,920 --> 00:50:13,810
This seems somewhat implausible. Moreover, more crudely, something I brought up in the second lecture when we were talking about the theory of ideas,

420
00:50:13,810 --> 00:50:21,260
a fictional story can be much more forceful and vivacious than a dull historical account.

421
00:50:21,260 --> 00:50:32,420
So in exactly the same way as watching paint dry can be much duller than imagining being attacked by spiders.

422
00:50:32,420 --> 00:50:34,700
It's simply not true,

423
00:50:34,700 --> 00:50:47,510
the things that we take to be actual perceptions or beliefs or reports about real things are always going to be more vivid than imagined things.

424
00:50:47,510 --> 00:50:51,860
Now, in the appendix to the treatise, which was published the very end of October,

425
00:50:51,860 --> 00:51:00,260
beginning of November 17 40, this was published along with Book three of the Treatise of Morals.

426
00:51:00,260 --> 00:51:09,440
We get some symptoms of unease. Hume is saying an idea assented to feels different.

427
00:51:09,440 --> 00:51:14,690
And when I talked about that feeling, I use terms like false and vivacity.

428
00:51:14,690 --> 00:51:19,370
But you know? You know what I mean when I talk about belief?

429
00:51:19,370 --> 00:51:23,630
It feels different, doesn't it? You know the feeling. You know what it's like to have a belief.

430
00:51:23,630 --> 00:51:31,910
That's what I'm talking about, that feeling. And well, that's all very well.

431
00:51:31,910 --> 00:51:36,960
But it's not so clear now that the hydraulic theory is going to work.

432
00:51:36,960 --> 00:51:46,290
If I am asked to contemplate the difference between a belief and an idea between a memory and a mere imagining,

433
00:51:46,290 --> 00:51:52,410
between an impression, an idea, yeah, I know what those things feel like in a sense.

434
00:51:52,410 --> 00:51:57,210
Is it actually one in the same thing marking the difference there?

435
00:51:57,210 --> 00:52:03,110
No, it isn't. So the hydraulic theory is probably going to have to go.

436
00:52:03,110 --> 00:52:09,320
In the enquiry, Hume says quite explicitly, you cannot define belief.

437
00:52:09,320 --> 00:52:21,410
So that's a contrast with the treaties. He repeats some of the words that he used in the appendix when he was showing his unease.

438
00:52:21,410 --> 00:52:28,940
And when he comes to discuss probability in section six of the enquiry, instead of giving a hydraulic theory,

439
00:52:28,940 --> 00:52:40,000
he says that such belief arises from an inexplicable contrivance of nature.

440
00:52:40,000 --> 00:52:49,980
But frankly, the hydraulic theory is not really the main point of Hume's analysis of induction.

441
00:52:49,980 --> 00:52:58,980
What's revolutionary about Hume's analysis of induction is his identification that when we make an inference,

442
00:52:58,980 --> 00:53:04,380
an inductive inference, whenever we infer from observed to unobserved,

443
00:53:04,380 --> 00:53:11,190
we are taking something for granted which we simply cannot give any basis for other

444
00:53:11,190 --> 00:53:19,140
than the fact that we are naturally inclined to to reason that way to think that way.

445
00:53:19,140 --> 00:53:30,150
So what we call reasoning about matter of fact is actually, in some ways, more like feeling than it is like rational inference,

446
00:53:30,150 --> 00:53:35,850
we feel ourselves committed to the view that the same will happen in the future as the past.

447
00:53:35,850 --> 00:53:45,630
We cannot give a reason for that fundamental assumption, which is absolutely basic to all of our thinking about the world.

448
00:53:45,630 --> 00:53:52,000
That's all probable reasoning is nothing but a species of sensation, OK, that's a bit of hyperbole.

449
00:53:52,000 --> 00:53:58,080
But when I'm convinced of any principle, it is only an idea which strikes more strongly upon me.

450
00:53:58,080 --> 00:54:00,600
When I give the preference to one set of arguments above another,

451
00:54:00,600 --> 00:54:05,790
I do nothing but decide from my feeling concerning the superiority of their influence.

452
00:54:05,790 --> 00:54:14,250
As I say, that's somewhat hyperbolic. But the crucial point is that human nature is playing a fundamental role when we make inductive inference.

453
00:54:14,250 --> 00:54:21,660
It is nothing like the kind of conception of rational thinking which had dominated previously.

454
00:54:21,660 --> 00:54:29,700
And here we've got a quotation from Locke's essay where Locke is saying is saying about probable inference.

455
00:54:29,700 --> 00:54:37,170
Inference consists in nothing but the perception of the connexion. There is between the ideas in each step of the deduction,

456
00:54:37,170 --> 00:54:46,530
whereby the mind comes to see either of a certain agreement or disagreement of any two ideas, as in demonstration in which it arrives at knowledge.

457
00:54:46,530 --> 00:54:56,220
Or this is the crucial bit the probable connexion on which it gives or withholds its assent as in opinion.

458
00:54:56,220 --> 00:55:04,380
So in a probable argument, there are probable connexions between the different steps of the inference for as reason

459
00:55:04,380 --> 00:55:09,750
perceives the necessary and indubitable connexion of all the ideas of or proofs by proofs,

460
00:55:09,750 --> 00:55:16,260
he means intermediate ideas one to another in each step of any demonstration that produces knowledge.

461
00:55:16,260 --> 00:55:21,870
So it likewise perceives the probable connexion of all the ideas or proofs one

462
00:55:21,870 --> 00:55:26,670
to another in every step of a discourse to which it will think assent due.

463
00:55:26,670 --> 00:55:32,430
So Locke thinks that when we do a probable argument, we are perceiving probable connexions.

464
00:55:32,430 --> 00:55:36,360
Hume says that's nonsense when we do a probable argument.

465
00:55:36,360 --> 00:55:43,440
There is at least one step where we infer from observer to unobserved, where we are not perceiving anything at all.

466
00:55:43,440 --> 00:55:45,990
There's no cognition going in there.

467
00:55:45,990 --> 00:55:55,760
It's an operation of the imagination we are reading into the world an expectation of uniformity for which there is no rational basis.

468
00:55:55,760 --> 00:56:01,173
Thank you.