1
00:00:10,580 --> 00:00:15,060
OK. Last week, we did Hume's regularity theory. I can try this again.

2
00:00:15,060 --> 00:00:21,990
There we are. And I'm just there are a few people who are looking a bit bemused at the end of it.

3
00:00:21,990 --> 00:00:26,670
So I thought that what I'd do is I'd go through the arguments again, just briefly.

4
00:00:26,670 --> 00:00:37,530
So the question that we're asking is, what is causation? And if you remember, I said causation is important because without its prediction,

5
00:00:37,530 --> 00:00:43,530
an explanation and manipulation of our environment would be difficult, if not impossible.

6
00:00:43,530 --> 00:00:50,970
It's our causal beliefs on which we rest of predictions, explanations and manipulations.

7
00:00:50,970 --> 00:00:59,370
So the question that we're trying to ask is, what is causation? And according to Hume, David Hume, the Scottish philosopher.

8
00:00:59,370 --> 00:01:05,280
It's regularity now. That's that's shorthand a bit because it's more than just regularity.

9
00:01:05,280 --> 00:01:15,600
It's also temporal priority and spatial contiguity. But the key for the human theory of causation is its regularity or constant conjunction.

10
00:01:15,600 --> 00:01:20,880
And his reason for that is because that could be our only evidence for causation.

11
00:01:20,880 --> 00:01:27,690
So we see one billiard ball hitting another. We do not don't see causation in the individual case, says Hume.

12
00:01:27,690 --> 00:01:33,690
But we see another billiard ball hitting another and that rolling off and another billiard ball hitting another and it's rolling off.

13
00:01:33,690 --> 00:01:42,720
And the experience of that correlation causes us to think that it's the first billiard ball hitting the other that causes the other to roll off.

14
00:01:42,720 --> 00:01:46,500
So causation is constant conjunction.

15
00:01:46,500 --> 00:01:54,630
And an obvious objection to that is, hang on, surely causation is the necessary connexion between the two events.

16
00:01:54,630 --> 00:01:59,190
That explains the regularity. It's not the regularity itself.

17
00:01:59,190 --> 00:02:03,810
It's the necessary connexion behind it. That explains the regularity.

18
00:02:03,810 --> 00:02:14,850
That's causation. And Hume, based on his theory of ideas and you remember we talked about the difference between impressions and ideas.

19
00:02:14,850 --> 00:02:23,370
That habit of expectation, that observations of regularity and stillness makes us think that there's a necessary connexion.

20
00:02:23,370 --> 00:02:27,540
But actually, you cannot experience necessary connexion.

21
00:02:27,540 --> 00:02:33,030
You can experience a connexion, but not any necessity of that connexion and signals.

22
00:02:33,030 --> 00:02:39,030
We can't experience it. We can't either form a coherent idea of it.

23
00:02:39,030 --> 00:02:44,640
So we're answering the question, what is the causation? According to Hume, it's regularity.

24
00:02:44,640 --> 00:02:50,830
An immediate knee jerk objection is. But causation is the necessary connexion that explains the regularity.

25
00:02:50,830 --> 00:02:57,660
It's not the regularity itself. Hume says no, it is only the regularity.

26
00:02:57,660 --> 00:03:03,620
The only reason that we think it's the necessary connexion is that we think it's necessary connexion.

27
00:03:03,620 --> 00:03:15,180
And we spread that habit on the world. We project that belief onto the world and unthink that causation is itself necessary connexion.

28
00:03:15,180 --> 00:03:21,690
Okay, so that's what we did last week. Now, are there any quick questions of clarification?

29
00:03:21,690 --> 00:03:25,770
I don't want any questions of substance at this point because we haven't got time.

30
00:03:25,770 --> 00:03:33,170
But any quick questions of clarification about the human regularity theory?

31
00:03:33,170 --> 00:03:36,960
Nope. Jolly good. Okay. This week we're going to look at HEWs.

32
00:03:36,960 --> 00:03:42,150
Two definitions, of course. We looked at it last week, but we didn't divide it into the two.

33
00:03:42,150 --> 00:03:52,410
So we'll have a look at each one separately. And then we're going to look at the nature of counterfactuals and counterfactual dependence.

34
00:03:52,410 --> 00:04:00,030
Then we're going to look at counterfactual theories of causation. And just as I called the Regularity Theory of Causation RTC.

35
00:04:00,030 --> 00:04:05,160
I'm calling this one CTT Counterfactual Theory of Causation.

36
00:04:05,160 --> 00:04:08,460
I'm going to look at some problems for the counterfactual theory.

37
00:04:08,460 --> 00:04:17,850
And I'm going to end by looking at the nature, the relationship between the counterfactual theory of causation and the regularity theory of causation.

38
00:04:17,850 --> 00:04:24,660
They were always seen as rivals by the people who developed the counterfactual theory of causation.

39
00:04:24,660 --> 00:04:34,780
But as we'll see, it looks as if the counterfactual theory of causation is actually a version of the regularity theory of causation,

40
00:04:34,780 --> 00:04:39,540
and that rather puts the cat amongst the pigeons for various reasons.

41
00:04:39,540 --> 00:04:43,350
And I'll explain why as we look at that. OK.

42
00:04:43,350 --> 00:04:49,140
Right. So let's have a look at humans. There's Hume's first definition of causation.

43
00:04:49,140 --> 00:04:53,790
We may define a cause to be an object followed by another.

44
00:04:53,790 --> 00:04:59,550
Where are all the objects? Similar to the first are followed by objects similar to the second.

45
00:04:59,550 --> 00:05:03,060
So the object is the first billiard ball hitting the second beli ball.

46
00:05:03,060 --> 00:05:09,640
That's the first object. And the second object is the second billiard ball rolling off.

47
00:05:09,640 --> 00:05:10,100
Okay.

48
00:05:10,100 --> 00:05:19,730
So when we get an object that's similar to a first billiard ball hitting the other, we get an object similar to the second billiard ball rolling off.

49
00:05:19,730 --> 00:05:23,480
That's the regularity theory of causation.

50
00:05:23,480 --> 00:05:35,840
But then Hume continues in the same sentence or in other words, hey, where if the first object had not been the second, never had existed.

51
00:05:35,840 --> 00:05:40,670
Now, that's interesting. I mean, is that equivalent to that?

52
00:05:40,670 --> 00:05:45,380
He does make it sound as if it is when he says or in other words.

53
00:05:45,380 --> 00:05:51,440
But actually, this is surely a statement of necessary connexion, isn't it?

54
00:05:51,440 --> 00:05:55,220
If the first had not been, the seconds never had existed.

55
00:05:55,220 --> 00:06:03,590
In other words, if the first billiard ball hadn't hit the second billiard ball, the second billiard ball would not have rolled off.

56
00:06:03,590 --> 00:06:10,730
That's the counterfactual theory of causation. And it's interesting that Hume seems to take them as equivalent.

57
00:06:10,730 --> 00:06:16,280
The counterfactual theory of causation as equivalent to the regularity theory of causation.

58
00:06:16,280 --> 00:06:23,800
But in about the 1970s, people stopped thinking that.

59
00:06:23,800 --> 00:06:32,270
And before I could describe the counterfactual theory of causation, I've got to look at the way I meant to change that.

60
00:06:32,270 --> 00:06:35,840
I meant to change that because grammar is horrible. It should be.

61
00:06:35,840 --> 00:06:40,040
If Lee Harvey Oswald didn't shoot Kennedy, somebody else did.

62
00:06:40,040 --> 00:06:45,500
So pretend that's what it says. I changed it on my machine and forgot I'd sent this in already.

63
00:06:45,500 --> 00:06:49,790
So if Lee Harvey Oswald didn't shoot Kennedy, somebody else did.

64
00:06:49,790 --> 00:06:58,830
Is that true or not? Put up your hand if you think it's true. If Lee Harvey Oswald didn't shoot Kennedy, somebody else did.

65
00:06:58,830 --> 00:07:03,650
It's got to be true, hasn't it? Okay, that's just a conditional.

66
00:07:03,650 --> 00:07:08,600
That's a conditional sentence which divides into two the antecedent.

67
00:07:08,600 --> 00:07:14,450
If Lee Harvey Oswald didn't shoot Kennedy, then there's an implicit then in there.

68
00:07:14,450 --> 00:07:17,870
Somebody else did. That's the consequence. Somebody else did.

69
00:07:17,870 --> 00:07:25,910
So you've got an antecedent clause. If Lee Harvey Oswald didn't shoot Kennedy and a consequent clause, somebody else did.

70
00:07:25,910 --> 00:07:30,560
And we're quite good at determining the truth value of such statements.

71
00:07:30,560 --> 00:07:38,720
But what about this? If Lee Harvey Oswald hadn't shot Kennedy, someone else would have.

72
00:07:38,720 --> 00:07:42,830
Is that true? Not necessarily.

73
00:07:42,830 --> 00:07:46,970
It might be. But you'd have to tell a story about it, wouldn't you?

74
00:07:46,970 --> 00:07:57,140
So where is that one, at least if you can see what happened? I just you know, I put that one there and forgot to change the antecedent clause.

75
00:07:57,140 --> 00:08:01,850
We can look at the first one and see immediately that it's true. Has to be true.

76
00:08:01,850 --> 00:08:09,080
I mean, Kennedy is dead and he died of a gunshot wound. So if Lee Harvey Oswald didn't do it, somebody else must have done it.

77
00:08:09,080 --> 00:08:14,330
But if Lee Harvey Oswald hadn't shot Kennedy, somebody else would have done.

78
00:08:14,330 --> 00:08:18,680
Now, maybe that is true or maybe it's false.

79
00:08:18,680 --> 00:08:26,270
The difficulty with that one is as a subjunctive, conditional on a counterfactual conditional.

80
00:08:26,270 --> 00:08:33,740
And traditionally, counterfactuals have been thought of as conditionals where the antecedent is false.

81
00:08:33,740 --> 00:08:40,430
But actually not all counterfactuals as such that the antecedent clause is false in this one.

82
00:08:40,430 --> 00:08:51,080
Indeed. Arguably it is true. So instead we should think of them as subjunctive conditionals.

83
00:08:51,080 --> 00:08:58,060
Okay, so what are the truth conditions of that? And let's have a look at the next one.

84
00:08:58,060 --> 00:09:03,290
OK. If the Germans had won the war, we would be speaking German.

85
00:09:03,290 --> 00:09:09,590
Is that true or not? Put up your hands if you think that one's true.

86
00:09:09,590 --> 00:09:14,480
If the Germans had won the war, we would be speaking German.

87
00:09:14,480 --> 00:09:27,780
Put up your hand if you think that's true. Okay, put up your hand if you think it's false and put up your hand if you haven't got the foggiest idea.

88
00:09:27,780 --> 00:09:34,020
Yes, quite reasonable. Okay. That's the difficulty with counterfactual conditionals.

89
00:09:34,020 --> 00:09:43,710
Actually, we do use them all the time and we talk about them as if they have perfectly straightforward truth conditions.

90
00:09:43,710 --> 00:09:51,590
But actually, they don't. Not straightforward truth conditions in the way ordinary count the conditional statements have.

91
00:09:51,590 --> 00:09:52,980
And with that one,

92
00:09:52,980 --> 00:10:02,340
it looks as if it can't be any state of affairs in this world that makes that statement true because the Germans didn't win the war after all.

93
00:10:02,340 --> 00:10:11,280
So how how can what can make that true or false? There can't be any state of affairs in this world that makes that true or false.

94
00:10:11,280 --> 00:10:20,910
That's that's the thought. And so people thought about giving a counterfactual theory of causation for quite a few years in

95
00:10:20,910 --> 00:10:28,950
between Hume's regularity theory and the counterfactual theories when they were developed in the 70s.

96
00:10:28,950 --> 00:10:38,130
But they didn't because it was explaining causation, something mysterious in terms of counterfactuals, which are even more mysterious.

97
00:10:38,130 --> 00:10:52,950
And so nobody bothered because it didn't seem worthwhile. But then we started to understand the logic or the semantics of counterfactuals.

98
00:10:52,950 --> 00:11:00,720
Okay, so in 1940, Stefan Rudolf Carnap printed his meaning, a necessity.

99
00:11:00,720 --> 00:11:08,260
And in that, he introduced the idea. He used the possible worlds talk that was introduced by Liben.

100
00:11:08,260 --> 00:11:13,860
It's to explain the truth conditions of counterfactuals.

101
00:11:13,860 --> 00:11:23,670
And just to tell you how important truth conditions are, if you're trying to explain the meaning of the word cat to your child, you show it.

102
00:11:23,670 --> 00:11:27,720
Lots of cats. And you don't hear it. Lots of black cats, do you?

103
00:11:27,720 --> 00:11:35,040
Or lots of fat cats. You show it. Lots of tortoiseshell cats, some ginger cats and black cats and lots of different cats.

104
00:11:35,040 --> 00:11:39,570
Until the child is able, you point to a cat and you say, is that a pussy cat?

105
00:11:39,570 --> 00:11:43,980
And the child says, yes. And you point to a dog and you say, is that a pussy cat?

106
00:11:43,980 --> 00:11:51,300
And the charge you hope says no, because it's understood the conditions under which that's a cat are true.

107
00:11:51,300 --> 00:11:58,260
And the conditions under which that's a cat are false. In other words, it's got the meaning of cat.

108
00:11:58,260 --> 00:12:03,960
It's got the meaning of it's a cat. So the conditions of truth and falsity are very important.

109
00:12:03,960 --> 00:12:15,870
And Rudolf Carnap suggested that we can use all possible worlds to understand counterfactual conditionals and just to.

110
00:12:15,870 --> 00:12:21,210
That's him. It's nice to know who you're talking about. So that's Rudolf Carnap.

111
00:12:21,210 --> 00:12:29,880
And after Carnap in 1968. So you can see it took quite a long time for the logic of counterfactuals to be understood.

112
00:12:29,880 --> 00:12:39,000
Robert Stalnaker published a theory of conditionals. And again, with a hand out, you've got all the references that I'm using here.

113
00:12:39,000 --> 00:12:43,890
And that's Stalnaker just again, so you know who you're talking about.

114
00:12:43,890 --> 00:12:52,620
And then after Stalnaker, this came along. And this is probably the most famous book on counterfactuals, 1973.

115
00:12:52,620 --> 00:13:01,230
David Lewis, a book called Simply Counterfactuals. And in this was Lewis who gave a counterfactual theory of causation.

116
00:13:01,230 --> 00:13:06,870
And this is Lewis always looks nice, cheery chap. And he looks like Bill Radcliffe, doesn't he?

117
00:13:06,870 --> 00:13:11,980
Or rather Bill Radcliffe looks like him to those of you who know Bill Radcliffe.

118
00:13:11,980 --> 00:13:20,070
Okay, so it took quite a long time for an account of the semantics of counterfactuals come about.

119
00:13:20,070 --> 00:13:25,330
But this is what it is. If the Germans had won the war, we would be speaking German.

120
00:13:25,330 --> 00:13:31,020
OK, how do you determine the truth value of that? Well, there are two possible worlds.

121
00:13:31,020 --> 00:13:39,270
And what we've got to do is decide which one of them is closer. So is it the world in which the Germans won the war?

122
00:13:39,270 --> 00:13:44,550
And we are speaking German, or is it the world in which the Germans won the war?

123
00:13:44,550 --> 00:13:52,140
And we're not speaking German in which. In other words, which takes us further from reality.

124
00:13:52,140 --> 00:13:58,200
So we know we're taken away from this world. Was we looking at the possible world in which the Germans won the war?

125
00:13:58,200 --> 00:14:03,840
Well, there are two of them. One of them, we're speaking German. And in the other one, we're not speaking German.

126
00:14:03,840 --> 00:14:14,160
Which of those is closer to the actual world? In other words, which of them is closer to actuality than the other?

127
00:14:14,160 --> 00:14:26,570
So that's that's the semantics of possible. Now, if you're thinking what possible worlds?

128
00:14:26,570 --> 00:14:31,040
You're thinking quite reasonable view, there's not a logical explosion here.

129
00:14:31,040 --> 00:14:40,280
And if that bothers you, you're quite right. I mean, it is we've if we're going to accept possible worlds semantics,

130
00:14:40,280 --> 00:14:47,990
if we're going to introduce an ontology of possible worlds in order to explain the truth, conditions of counterfactuals.

131
00:14:47,990 --> 00:14:53,750
And you might think I really am very uncomfortable with this.

132
00:14:53,750 --> 00:14:59,120
It's like introducing possible objects as well as actual objects and things like that.

133
00:14:59,120 --> 00:15:03,200
Now, we haven't got time to talk about that here, so we're not going to talk about it.

134
00:15:03,200 --> 00:15:06,380
But again, I've put it on the notes and on the reading.

135
00:15:06,380 --> 00:15:11,030
There are plenty of things you can follow up on if you want to read a little bit more about that.

136
00:15:11,030 --> 00:15:16,640
But you're quite right. If you're worried about possible worlds, the only thing I can say in defence of it is,

137
00:15:16,640 --> 00:15:26,320
is that it looks as if what physicists seem to think that they need possible worlds and logicians seem to think that they need possible worlds.

138
00:15:26,320 --> 00:15:34,010
And they're not the same thing because physicists, of course, the possible worlds they introduce are empirically possible.

139
00:15:34,010 --> 00:15:40,040
In other words, they they think that they can conduct empirical experiments to show that there are such worlds,

140
00:15:40,040 --> 00:15:50,240
whereas logicians conduct any thought experiments. But it's interesting that both physicists and philosophers think that possible worlds are needed.

141
00:15:50,240 --> 00:15:54,980
I say philosophers. I mean logicians, of course, because it's mathematicians, too.

142
00:15:54,980 --> 00:16:01,620
But that's all we're going to say about possible worlds. We can talk about it later, if you like.

143
00:16:01,620 --> 00:16:09,350
Okay. So the possible worlds theories of the semantics of counterfactual semantics is to do with truth conditions.

144
00:16:09,350 --> 00:16:13,490
Semantics is to do with truth and falsity. And I explained why.

145
00:16:13,490 --> 00:16:15,890
Truth conditions are important.

146
00:16:15,890 --> 00:16:27,050
So the semantics of counterfactuals appeals to possible worlds, and that enables us to think that counterfactuals are no longer mysterious.

147
00:16:27,050 --> 00:16:38,240
And so we can use them to explain causation. As long as there aren't other problems, we've taken away the mystery of counterfactuals.

148
00:16:38,240 --> 00:16:43,010
You might not think so, but. But those in the know think so.

149
00:16:43,010 --> 00:16:54,470
Okay, so now we'll look at counterfactual dependence, because that's the thing on which all counterfactual theories of causation depend.

150
00:16:54,470 --> 00:17:00,740
So here's a definition. So event e counter factually depends on events.

151
00:17:00,740 --> 00:17:04,970
See if and only if that I double F again.

152
00:17:04,970 --> 00:17:14,180
Not a misprint. If and only if. If C were not to occur, E would not occur.

153
00:17:14,180 --> 00:17:23,750
Okay. So if the first billiard ball hadn't second hit the second billiard ball, the second billiard ball would not have rolled off.

154
00:17:23,750 --> 00:17:28,520
So if, if C hadn't occurred E wouldn't have occurred either.

155
00:17:28,520 --> 00:17:32,570
That's counterfactual dependence. Any questions about that.

156
00:17:32,570 --> 00:17:39,890
Because that's pretty crucial. Is that a question about to arrive, Sean?

157
00:17:39,890 --> 00:17:48,920
Or were you just an intake of breath only? Good thing.

158
00:17:48,920 --> 00:18:02,170
Perfectly reasonable. No, you need oxygen. In what way?

159
00:18:02,170 --> 00:18:15,900
We're trying to establish that he has always said, but if we in advance assume that.

160
00:18:15,900 --> 00:18:24,760
Si, si. You see, I can sort of see why you think that.

161
00:18:24,760 --> 00:18:30,790
But I think the mistake has already been made and it's been made by you.

162
00:18:30,790 --> 00:18:36,250
I think you're assuming that necessary connexion comes and causation.

163
00:18:36,250 --> 00:18:43,510
What we've got here is is an account of which is we'll see in a minute that this isn't an account of causation.

164
00:18:43,510 --> 00:18:48,790
It's a counterfactual depending on which an account of causation depends.

165
00:18:48,790 --> 00:19:02,080
But the idea is that C is what makes a difference. And the difference it makes is if it failed to occur, the other wouldn't have occurred.

166
00:19:02,080 --> 00:19:10,000
So there's no we're not assuming causation there, we're explaining it is the thought.

167
00:19:10,000 --> 00:19:16,330
Not happy yet. OK. Let's continue. If you're still not happy at the end, we can talk about it again.

168
00:19:16,330 --> 00:19:20,610
OK. So this is a definition of counterfactual dependence.

169
00:19:20,610 --> 00:19:26,910
E counterfactual. Depends on see if and only if if C were not to occur.

170
00:19:26,910 --> 00:19:34,900
E would not occur either. If the first billiard ball hadn't hit the second, the second wouldn't have rolled off.

171
00:19:34,900 --> 00:19:46,480
Okay. So all a counterfactual theories of causation depend on counterfactual dependence at some level.

172
00:19:46,480 --> 00:19:55,330
But there's a problem with thinking that counterfactual dependence is a causation, a theory of causation.

173
00:19:55,330 --> 00:20:03,220
And that's because although counterfactual dependence is sufficient for causation, it is not necessary for causation.

174
00:20:03,220 --> 00:20:11,200
And if counterfactual dependence were the same thing as causation, obviously it would have to be sufficient and necessary.

175
00:20:11,200 --> 00:20:20,520
But it's sufficient but not necessary. So two events can be causally related without one counterfactual depending on the other.

176
00:20:20,520 --> 00:20:26,260
And let me give you an example of that. Here's Lucy.

177
00:20:26,260 --> 00:20:29,950
Lucy, you see, she's got a name written underneath her. So, you know, it's Lucy.

178
00:20:29,950 --> 00:20:35,860
This is Lucy who's throwing a rock, as you can tell. And in throwing the rock.

179
00:20:35,860 --> 00:20:40,540
She shatters the window. Okay. There's a shattered window there.

180
00:20:40,540 --> 00:20:48,820
But Lucy's rock breaks the window. But the window shattering isn't counterfactuals dependent on Lucy's rock throwing.

181
00:20:48,820 --> 00:20:58,010
And the reason for that is because if Lucy hadn't broken the window, Brian would have done Bruns a much better shot than Lucy.

182
00:20:58,010 --> 00:21:03,070
He's much more practised at throwing rocks. Even those rock is a lot smaller.

183
00:21:03,070 --> 00:21:09,080
And he was waiting there beside Lucy and he threw just after she did.

184
00:21:09,080 --> 00:21:12,010
But her rock shattered the window in his didn't.

185
00:21:12,010 --> 00:21:19,390
But it's not true that if Lucy hadn't thrown her rock, the window would not have shattered because yes,

186
00:21:19,390 --> 00:21:25,080
it would, because Brian's rock would have shattered the window. Are you with me?

187
00:21:25,080 --> 00:21:32,080
So the counterfactual dependence is just not there.

188
00:21:32,080 --> 00:21:38,140
So there's causation and no counterfactual dependence. They cannot be the same thing.

189
00:21:38,140 --> 00:21:44,430
Now, do you all see that? I'm very happy to explain it again. If you'd like me to.

190
00:21:44,430 --> 00:21:50,670
Yes. That's all right.

191
00:21:50,670 --> 00:22:10,870
Si, si. I said different. It is said if this window is within that event then finally then.

192
00:22:10,870 --> 00:22:15,070
Because I see where you're going.

193
00:22:15,070 --> 00:22:23,350
And I think your intuitions are very sound. Let me try and explain this again.

194
00:22:23,350 --> 00:22:27,490
And I think you'll see where I'm going. But you're not wrong in saying what you've said.

195
00:22:27,490 --> 00:22:34,270
They are two different events. There's Lucie's rock throwing and bronze rock throwing, and they are two different.

196
00:22:34,270 --> 00:22:38,770
But what we want to say is that it's Lucie's rock throwing cause the window to shatter.

197
00:22:38,770 --> 00:22:40,480
That's true, isn't it?

198
00:22:40,480 --> 00:22:50,050
But what we're trying to postulate is that that's because the shattering of the window is counter factually dependent on Lucy's throwing the rock.

199
00:22:50,050 --> 00:22:59,110
But it's not because it's not true. That had Lucy not thrown the rock, the window would not have shattered.

200
00:22:59,110 --> 00:23:08,550
But what we're saying is that it's Lucy's role within that dance.

201
00:23:08,550 --> 00:23:12,420
But then that depends. Okay.

202
00:23:12,420 --> 00:23:16,600
I tell you what. Let me carry on for a few slides.

203
00:23:16,600 --> 00:23:20,860
And I think that we might explain your intuition.

204
00:23:20,860 --> 00:23:25,900
And if not, well, we'll look at it again. But I don't think that your intuition is wrong.

205
00:23:25,900 --> 00:23:32,570
I think you need to get the logic absolutely straight and that you happen to the moment, which isn't at all surprising.

206
00:23:32,570 --> 00:23:43,090
Therefore, I'm assuming that it was necessary for Ryan to throw a rock once, Lucy, because it was necessary.

207
00:23:43,090 --> 00:23:47,420
He didn't have to throw the rock there, therefore. Lucy and the road.

208
00:23:47,420 --> 00:23:51,650
No. There was no I've no idea what you're talking about.

209
00:23:51,650 --> 00:23:57,170
She shouldn't. This says the windows shatter is not dependent on comfort.

210
00:23:57,170 --> 00:24:04,470
And this is what if Lucy did not throw out the window broken?

211
00:24:04,470 --> 00:24:11,790
Well, the window wouldn't be broken then, but it would have broken only because Brun threw it.

212
00:24:11,790 --> 00:24:22,170
Yeah. Well, I think that's exactly the same. Is that the same problem? I was thinking after listening to two gentlemen here, I was thinking it works.

213
00:24:22,170 --> 00:24:33,480
If you knew the window breaking, rather let me you.

214
00:24:33,480 --> 00:24:39,240
I mean, I think I like the way that you're both thinking. Let me see if I can explain it.

215
00:24:39,240 --> 00:24:50,600
Where's the board rubber? Because a bit of tissue doesn't do it, frankly. Um.

216
00:24:50,600 --> 00:25:06,140
A causes B is equivalent to a B counter factually depends on A.

217
00:25:06,140 --> 00:25:09,500
Okay, that's what we're postulating. That's the hypothesis we're putting.

218
00:25:09,500 --> 00:25:14,030
So A cause is B is equivalent to B counterfactuals.

219
00:25:14,030 --> 00:25:29,360
Depends on A or had A not have occurred, B would not have occurred.

220
00:25:29,360 --> 00:25:34,640
Okay. Now an objection to that is go.

221
00:25:34,640 --> 00:25:40,580
I'll get out of the way as soon as I finish writing it because I realise I'm blocking it.

222
00:25:40,580 --> 00:25:45,150
Lucie's rock throwing are high.

223
00:25:45,150 --> 00:26:03,120
Fante's rock throwing caused the window to shatter but the window shattering.

224
00:26:03,120 --> 00:26:11,910
Did not counter factually depend on Lucie's through.

225
00:26:11,910 --> 00:26:16,140
That's the way the argument is going. So this is the claim.

226
00:26:16,140 --> 00:26:25,080
This is the hypothesis that we're considering is A causes B the same as be counter factually.

227
00:26:25,080 --> 00:26:29,380
Depends on A or had A not occurred, B would not have occurred.

228
00:26:29,380 --> 00:26:34,440
OK. So that's the hypothesis. And the counter example is.

229
00:26:34,440 --> 00:26:39,840
Well, no, that can't be right because Lucy's rock throwing cause the window to break.

230
00:26:39,840 --> 00:26:50,730
We accept that. OK. That that's given by the counterexample. But the window shattering isn't counterfactuals dependent on Lucy's rock throwing?

231
00:26:50,730 --> 00:26:59,340
Because had the window not shattered as a result of Lucy's rock throwing, it would have shattered as a result of Brian's rock throwing.

232
00:26:59,340 --> 00:27:05,280
OK. Let me let me continue, because I think we'll we might solve your problem in just a second.

233
00:27:05,280 --> 00:27:09,420
Let's try it. But first of all. OK. So this is the problem.

234
00:27:09,420 --> 00:27:14,460
Again, counterfactual dependence is sufficient for causation, but not necessary.

235
00:27:14,460 --> 00:27:22,260
That's what we're saying. Two events can be causally related. So Lucy's rock throwing and the window shattering can be causally related.

236
00:27:22,260 --> 00:27:29,580
Without the window shattering cause, lead counts factually depending on Lucy's rock throwing.

237
00:27:29,580 --> 00:27:36,270
I think your in your mind, you're accusing me of a lot of logical subterfuge here, aren't you?

238
00:27:36,270 --> 00:27:41,310
I can see that you think there's a logical sleight of hand going on.

239
00:27:41,310 --> 00:27:43,290
Okay, let's carry on.

240
00:27:43,290 --> 00:27:53,910
Lewis's first counterfactual theory of causation solves this problem by defining causation in terms of chains of counterfactual dependence.

241
00:27:53,910 --> 00:27:56,820
Now, somebody last week I can't remember who it was.

242
00:27:56,820 --> 00:28:05,280
Was it you actually who who came to me afterwards and said, surely what's important about causation is chains of causation.

243
00:28:05,280 --> 00:28:09,740
It was you. Primary was first.

244
00:28:09,740 --> 00:28:17,400
Right? Well, and I said to that this week, I would be giving exactly that.

245
00:28:17,400 --> 00:28:21,810
So Lewis solves the problem of counterfactual dependence not being necessary

246
00:28:21,810 --> 00:28:26,790
for causation by defining causation in terms of the ancestral it's called.

247
00:28:26,790 --> 00:28:30,870
That's the proper term, the ancestral of counterfactual dependence.

248
00:28:30,870 --> 00:28:36,940
In other words, chains of counterfactual dependence. So C is the cause of E.

249
00:28:36,940 --> 00:28:41,550
I'm very annoyingly I used lowercase letters before and I'm now using uppercase letters.

250
00:28:41,550 --> 00:28:45,510
That's very bad. But anyway, we'll ignore it.

251
00:28:45,510 --> 00:28:56,350
C is of course, of a if and only if there's a causal chain of counterfactual dependence leading from C to E.

252
00:28:56,350 --> 00:29:01,360
Or actually should be leading back from E to C. But anyway, in a month.

253
00:29:01,360 --> 00:29:11,610
So there is a chain from the shattered window to Lucy's rock throwing and there isn't such a chain from the shattered window to Brian's rock throwing,

254
00:29:11,610 --> 00:29:24,000
is the thought. So although you can say the window shattered, the window would not have shattered if Lucy's rock had not been there at T minus one.

255
00:29:24,000 --> 00:29:27,600
And if it hadn't been there at T minus one and so on.

256
00:29:27,600 --> 00:29:32,520
But there isn't such a chain back to Brian's rock throwing.

257
00:29:32,520 --> 00:29:40,560
So the first ever counterfactual theory of causation, which was the result of this possible world.

258
00:29:40,560 --> 00:29:53,190
Semantics of counterfactuals was C causes e if and only if there's a chain of counterfactual dependencies leading back from E to see that C.

259
00:29:53,190 --> 00:30:01,900
So if. See, had not occurred, you would not have occurred if the sea had not occurred, D would not have occurred.

260
00:30:01,900 --> 00:30:09,070
If he had not occur, E would not have occurred and so on. So it shows it suggests counterfactual, sorry.

261
00:30:09,070 --> 00:30:23,890
Causation is transitive, doesn't it? So if C causes E and E causes D, then C causes T as well, that will become problematic later on.

262
00:30:23,890 --> 00:30:31,720
Okay. Before I look at the objections, does anyone want to ask a question of clarification at that point?

263
00:30:31,720 --> 00:30:43,020
Is there the unofficial publication introduced by the fact that the window shattering does not have any time associated with it,

264
00:30:43,020 --> 00:30:50,070
that there would be much simpler to look at if it specified the window, shut it down?

265
00:30:50,070 --> 00:30:56,310
At which point you would need to specify that the ropes heated at the same time?

266
00:30:56,310 --> 00:31:02,200
Well, I think this is going to be the same problem that's going to be, I hope, solved in just a minute.

267
00:31:02,200 --> 00:31:06,950
I mean, it isn't the problem. So what isn't the problem then?

268
00:31:06,950 --> 00:31:11,790
The problem disappears. If you specify the of.

269
00:31:11,790 --> 00:31:16,560
Oh, I see. Okay. So if you say the window shattering happened at T. Okay.

270
00:31:16,560 --> 00:31:27,450
Will you give me another few slides. And I hope that will will solve that one because I'm going to talk about exactly that problem here.

271
00:31:27,450 --> 00:31:32,010
Okay. Any other questions of clarification before I move on to look at the problems?

272
00:31:32,010 --> 00:31:39,750
Bob, speak up. Isn't it the case that every time somebody claims counterfactual dependency,

273
00:31:39,750 --> 00:31:47,240
it's always possible to say that if Lucy shattered the window, somebody else might?

274
00:31:47,240 --> 00:31:52,020
Nobody is there. Somebody else might. Is there any cause at all?

275
00:31:52,020 --> 00:31:58,890
Something else? Well, I mean, you're all you all I can say to that is, is that that just shows us that that's counterfactual.

276
00:31:58,890 --> 00:32:06,000
Dependence isn't necessary for causation. Is isn't necessary or sufficient.

277
00:32:06,000 --> 00:32:11,010
Necessary, isn't it. Yeah. The bureau might even be irrelevant.

278
00:32:11,010 --> 00:32:17,040
Yes. Because we all we've got at the moment is a hypothesis and we're testing the hypothesis.

279
00:32:17,040 --> 00:32:22,620
And we've seen one reason to reject it. Maybe there are other reasons to reject it.

280
00:32:22,620 --> 00:32:26,670
Okay, so two objections were first brought to the CTC.

281
00:32:26,670 --> 00:32:30,150
So we've got the hypothesis. We've got an objection.

282
00:32:30,150 --> 00:32:37,440
We've got a possible solution, which is that we add chains of causation, but we're now looking to see if the change of causation will do it.

283
00:32:37,440 --> 00:32:43,800
And there are two reasons to think it won't. And the first problem is the problem of trumping.

284
00:32:43,800 --> 00:32:49,650
Now, are there any ex military men or indeed women in this audience?

285
00:32:49,650 --> 00:32:57,750
I'm rather inclined to ask you to leave. If I've got the wrong strong number of stripes, etc.

286
00:32:57,750 --> 00:33:02,490
It's a relevant right.

287
00:33:02,490 --> 00:33:07,230
Good. Thank you. I realise it's irritating if you're in the know. But um.

288
00:33:07,230 --> 00:33:12,150
Yes, okay. Chumping. This is why trumping is important.

289
00:33:12,150 --> 00:33:15,570
You can see why I'm a bit nervous about this. Okay.

290
00:33:15,570 --> 00:33:20,940
Let's say we've got the troops and we've got a sergeant and we've got a major there.

291
00:33:20,940 --> 00:33:26,370
Now it's true that these this may not be a sergeants or a major, but nevertheless, we're going to continue.

292
00:33:26,370 --> 00:33:28,710
So there's a sergeant and major.

293
00:33:28,710 --> 00:33:36,920
And if, like me, you know nothing about these things, you ought to know that a major trumps a sergeant every time if you're a military man.

294
00:33:36,920 --> 00:33:45,540
And so if both the major and the sergeants yell advance at the same moment, the troops will advance.

295
00:33:45,540 --> 00:33:51,270
And whose order? Are they obeying the majors?

296
00:33:51,270 --> 00:33:56,070
Exactly. Because the sergeants order is neither here nor there.

297
00:33:56,070 --> 00:34:02,720
If the if the major sites advance, the troops advance. And it's his order that they're being.

298
00:34:02,720 --> 00:34:09,750
And the problem we've got here is that there's no chain of counterfactual dependency that leads back from the

299
00:34:09,750 --> 00:34:19,290
major's order that isn't matched by a chain of counterfactual dependencies that leads back to the sergeants order.

300
00:34:19,290 --> 00:34:28,890
And there we really it does show that you're not wrong when you say aren't we just assuming something in here?

301
00:34:28,890 --> 00:34:38,550
Because all we've done is given to cause. And naturally, that second one is going to show it even more because earlier we looked at early pre-emption.

302
00:34:38,550 --> 00:34:44,580
So Lucy's rock throwing pre-empted, Brian's rock throwing, they happened at different times.

303
00:34:44,580 --> 00:34:53,070
But actually, if we look at late pre-emption, here's Lucy and her rock and her Broch rock breaks the window.

304
00:34:53,070 --> 00:34:58,640
But a split second later, Brian's rock sails through the window, too.

305
00:34:58,640 --> 00:35:08,190
Okay. And because it's a split second later, this is where you're bit about the time is absolutely right, because it's a split second later.

306
00:35:08,190 --> 00:35:14,610
There's no chain of counterfactual dependencies that leads back from Lucy's throw that

307
00:35:14,610 --> 00:35:20,490
isn't matched by a chain of counterfactual dependencies that leads back to Brian's throw.

308
00:35:20,490 --> 00:35:26,430
Now we know that it's Lucy's throw that breaks the bloody window.

309
00:35:26,430 --> 00:35:31,280
And we're trying to say that causation is counterfactual dependent.

310
00:35:31,280 --> 00:35:35,620
But it clearly isn't. Is it because counterfactual dependence?

311
00:35:35,620 --> 00:35:46,010
So let's sort of keep the argument flow going here so that we in fact.

312
00:35:46,010 --> 00:35:58,100
Let me move it a little bit. Where's the pen?

313
00:35:58,100 --> 00:36:06,710
So we've got the hypothesis. Causation is counterfactual dependence.

314
00:36:06,710 --> 00:36:13,370
And we've got an objection. No, there's counterfactual dependence.

315
00:36:13,370 --> 00:36:20,750
Sorry. There's causation and no counterfactual dependence.

316
00:36:20,750 --> 00:36:37,720
OK. That was early pre-emption. We've got an attempted solution to the objection, which is causation is a chain of counterfactual dependencies.

317
00:36:37,720 --> 00:36:51,480
And we've got objections. No, there's causation and sorry.

318
00:36:51,480 --> 00:37:01,890
Hold on. Yes, that's right, causation. Hang on.

319
00:37:01,890 --> 00:37:10,440
There's counterfactual dependence and no causation.

320
00:37:10,440 --> 00:37:12,480
That's right. So there's counterfeit.

321
00:37:12,480 --> 00:37:21,480
The window seems to counterfactuals depend upon Bryans throw, but brons throw doesn't cause the window to shatter.

322
00:37:21,480 --> 00:37:25,770
So how can causation be counterfactual dependent's? It can't.

323
00:37:25,770 --> 00:37:30,330
Do you see where we are in the argument as it goes through?

324
00:37:30,330 --> 00:37:33,120
So we're postulating a hypothesis.

325
00:37:33,120 --> 00:37:41,430
And each time we're putting up objection's, which question the hypothesis, we're putting up a new theory in the hope that this will do the trick.

326
00:37:41,430 --> 00:37:53,290
And in each case, so far it hasn't. So Lewis deals with this problem in a way that's going to so suit you and possibly you.

327
00:37:53,290 --> 00:38:04,650
And he introduces the notion of an alteration. So here's the definition of an alteration and event E that's identical to another event.

328
00:38:04,650 --> 00:38:13,170
ETR Except it occurs at a slightly different time, slightly different place, in a slightly different manner, etc.

329
00:38:13,170 --> 00:38:22,260
In other words, it's it's an event very like another event, but not the same one, not the identical event.

330
00:38:22,260 --> 00:38:30,000
And I mean, since the notion of an alteration. He talks of one event influencing another.

331
00:38:30,000 --> 00:38:37,680
So if an alteration of Lucy's rock throwing occurred, an alteration of windows shattering would occur.

332
00:38:37,680 --> 00:38:42,690
So if her rock had been thrown just two seconds later, the window would have shattered.

333
00:38:42,690 --> 00:38:51,690
Two seconds later. But an alteration of Brauns rock throwing would have no effect on the window shattering.

334
00:38:51,690 --> 00:39:00,690
Not on that window shattering. There might have been a different window shattering, which was the one that's caused by Bryans throw.

335
00:39:00,690 --> 00:39:06,630
Now, at this point, you might be getting a little bit bloody philosophers, you know.

336
00:39:06,630 --> 00:39:14,700
What are they talking about here? Okay, so this is the second counterfactual theory of causation.

337
00:39:14,700 --> 00:39:23,610
C causes E if and only if there's a chain of stepwise influence leading from E to C.

338
00:39:23,610 --> 00:39:30,960
So we've got a chain of stepwise influence that leads from Lucy's rock throwing to the window shattering

339
00:39:30,960 --> 00:39:37,310
because an alteration of Lucy's rock throwing would have led to an alteration of windows shattering.

340
00:39:37,310 --> 00:39:47,130
But there's no chain of stepwise influence leading from Brian's rock throwing to the shattering of the window.

341
00:39:47,130 --> 00:39:53,220
But as you might expect, objections were soon brought brought to this CTC.

342
00:39:53,220 --> 00:39:57,750
So the solution here. So, again, we've got a hypothesis.

343
00:39:57,750 --> 00:40:01,710
We've got objections to that. Early pre-emption is an objection.

344
00:40:01,710 --> 00:40:05,940
So the chains of counterfactual dependence were the solution.

345
00:40:05,940 --> 00:40:11,460
We've got two objections to that. Can't be because of trumping and late pre-emption.

346
00:40:11,460 --> 00:40:17,460
We've got a solution to that, which is stepwise influence.

347
00:40:17,460 --> 00:40:21,690
So we've moved away from counterfactual dependence itself.

348
00:40:21,690 --> 00:40:29,700
And we've moved away to counter factual influence. Of course, all this took some years.

349
00:40:29,700 --> 00:40:34,950
I mean, isn't it nice that you can all understand it and take it on board in five minutes?

350
00:40:34,950 --> 00:40:43,110
But it took quite a lot of time for this to actually happen. But as you can imagine, there are objections to this.

351
00:40:43,110 --> 00:40:51,220
So let's have a look at what these objections are. So let's have a look at choice of cause.

352
00:40:51,220 --> 00:41:01,620
OK, here's courtesy of Steve Photography. A camper lighting a fire.

353
00:41:01,620 --> 00:41:08,340
Now, you can imagine what happened. The lighting of the fire led to the conflagration that destroyed the forest.

354
00:41:08,340 --> 00:41:14,880
That's what happened. So the campers lighting the fire caused the forest to burn down.

355
00:41:14,880 --> 00:41:20,640
Okay, straightforward. And there's a woops.

356
00:41:20,640 --> 00:41:29,640
So I'm giving it away there. The trouble with that is that although there's a chain of stepwise influence that leads back from the fire,

357
00:41:29,640 --> 00:41:33,760
the forest burning down to the campus, lighting the fire.

358
00:41:33,760 --> 00:41:43,050
There's also a chain of stepwise influence that leads from the fire burning down to the birth of the camper, isn't there?

359
00:41:43,050 --> 00:41:44,000
It's true, isn't it?

360
00:41:44,000 --> 00:41:56,100
That had the camper not been born, the forest would not have burnt down, or at least that would have been an alteration of the forests burning down.

361
00:41:56,100 --> 00:41:59,460
It would have been a different type of burning down, wouldn't it?

362
00:41:59,460 --> 00:42:06,240
And so there's a chain of stepwise influence from the forest fire to the campers birth.

363
00:42:06,240 --> 00:42:12,810
Doesn't that mean that the campers birth was the cause of the forest burning down?

364
00:42:12,810 --> 00:42:18,010
Huh? No, course it wasn't.

365
00:42:18,010 --> 00:42:22,870
I mean, none of us would say that, would we? Or would we? Maybe Chris would.

366
00:42:22,870 --> 00:42:27,610
He's got his own slightly devious son.

367
00:42:27,610 --> 00:42:33,610
I mean, you can take your father and say big man.

368
00:42:33,610 --> 00:42:43,120
Well, that's that's just taking the objection even further. But it is exactly to force the what loses the force is the idea.

369
00:42:43,120 --> 00:42:53,980
It's a hypothesis that we're testing. So if counterfactual dependence is is basically the relation that that is causation.

370
00:42:53,980 --> 00:42:59,680
Well, then, Hans, the big bang not occurred. The forest would not have burnt down.

371
00:42:59,680 --> 00:43:09,350
So the big bang is the cause of the forest burning down. Well, if there's something wrong with that, surely.

372
00:43:09,350 --> 00:43:13,520
So we change nation. What's up?

373
00:43:13,520 --> 00:43:17,900
How is that different from a chain of causation? Which isn't that.

374
00:43:17,900 --> 00:43:25,470
Oh, I see. Well, we wouldn't normally wouldn't say that the camp has birth had caused the forest fire.

375
00:43:25,470 --> 00:43:31,700
No, I would say that without it, etc, etc., there wouldn't be a campfire to cause the forest fire.

376
00:43:31,700 --> 00:43:37,500
Therefore, we would change causation. No, there's a chain of counterfactual dependence.

377
00:43:37,500 --> 00:43:46,580
Yes. And this is what makes it look as if it's not a chain of counterfactual dependence that is causation.

378
00:43:46,580 --> 00:44:03,440
OK, let's. Sorry. Go on. Just things I think that if you take this long history that's also deterministic.

379
00:44:03,440 --> 00:44:13,100
And no, it needn't be deterministic. I think I said last week that that we often want to say that there's something probabilistic in causation.

380
00:44:13,100 --> 00:44:18,890
And you're absolutely right. We do. We're saying we can just add the word probably here.

381
00:44:18,890 --> 00:44:26,150
Had the well had the big bang not occurred, the forest fire probably wouldn't have occurred.

382
00:44:26,150 --> 00:44:30,470
Well, all we have to is, is the word probability. We've got it.

383
00:44:30,470 --> 00:44:35,540
But had the campers birth not occurred, the forest fire probably wouldn't have occurred.

384
00:44:35,540 --> 00:44:43,210
Allows for probabilistic causation. Let's bring this back to Brian.

385
00:44:43,210 --> 00:44:53,630
Brian. And then Brian Simons inspired that encourage about throwing you.

386
00:44:53,630 --> 00:44:58,360
Right. Does it go back to her?

387
00:44:58,360 --> 00:45:02,380
No. It would then go back to no. It would then be Brian's rock that shattered the window.

388
00:45:02,380 --> 00:45:07,850
And see, that doesn't change. You would say it was a chain that leads back to Brian's rock.

389
00:45:07,850 --> 00:45:14,060
So it's Brian's rock that caused it. He wouldn't have thrown it if he wouldn't get through it.

390
00:45:14,060 --> 00:45:18,740
So you could say that it was Lucy's getting ready to throw the rock that caused.

391
00:45:18,740 --> 00:45:26,660
Yes, you could say that. And actually, that takes me nicely onto the second objection here, which is the lack of transitivity.

392
00:45:26,660 --> 00:45:34,190
Objection. Okay. We were talking about bombs. These aren't bombs on my back here, but that's a bomb.

393
00:45:34,190 --> 00:45:38,060
And the story behind this, if I can get it right,

394
00:45:38,060 --> 00:45:50,840
is enemy who hasn't got another name is just called enemy like a bomb and puts it outside Lucy's door in the courtroom.

395
00:45:50,840 --> 00:45:54,260
Whether it is Lucy, maybe it's somebody else, but it doesn't really matter.

396
00:45:54,260 --> 00:46:01,340
Lucy's door, Brian comes along, sees the bomb about to go off and pinches out the fuse.

397
00:46:01,340 --> 00:46:05,300
And so Lucy survives. OK, no problem.

398
00:46:05,300 --> 00:46:14,140
So far, so there's a chain of stepwise influence from Billie's pinching out the fuse to Lucy's of Susie's survival.

399
00:46:14,140 --> 00:46:18,410
Thought it was someone else. Susan's survival. Except that OK.

400
00:46:18,410 --> 00:46:23,600
Had Billy not pulled out, the few pinched out the fuse. Lucy would not have survived.

401
00:46:23,600 --> 00:46:31,490
But it's also true that by the transitivity of stepwise influence had enemy not placed the bomb.

402
00:46:31,490 --> 00:46:36,710
Billy would not have pinched out the flute fuse. And Lucy Susie would not have survived.

403
00:46:36,710 --> 00:46:42,530
So the placing of the bomb that caused the survival of Lucy.

404
00:46:42,530 --> 00:46:48,500
Oh, no. That is really bad. We don't want that at all.

405
00:46:48,500 --> 00:46:53,240
There's something badly wrong with this account of causation.

406
00:46:53,240 --> 00:47:08,690
Do you see where I'm going? So the objections here are that there's too many causes.

407
00:47:08,690 --> 00:47:27,320
In other words, the camp has birth as well as the camper lighting the fire and it makes enemies placing the bomb a cause cause of Susie's survival.

408
00:47:27,320 --> 00:47:32,240
And again, we might. We think. No, no, no, no, no.

409
00:47:32,240 --> 00:47:36,620
That will not do. This can't be the right theory of causation.

410
00:47:36,620 --> 00:47:41,330
Would anyone like to to think of the next step?

411
00:47:41,330 --> 00:47:45,080
How are we going to get over this? There is a way of getting over this.

412
00:47:45,080 --> 00:47:55,890
Can anyone think that it might be. You might think it's blindingly obvious that this is a lousy theory of causation.

413
00:47:55,890 --> 00:47:58,530
But believe me, it took quite a long time.

414
00:47:58,530 --> 00:48:10,380
Something that's obvious when you explain it like this in its outlines is much less obvious as you're actually working through it.

415
00:48:10,380 --> 00:48:17,300
No one can see the solution to this one. Go and have a go.

416
00:48:17,300 --> 00:48:24,720
I assume that handwaving was. I have a hypothesis. It was my brain.

417
00:48:24,720 --> 00:48:33,680
It's funny, my brain works through my feet. I walk in if you have to stop.

418
00:48:33,680 --> 00:48:43,410
Stop saying the word. Well, funnily enough, you're going in the right direction.

419
00:48:43,410 --> 00:48:47,550
It's not stopping the chains so much. Should I go on?

420
00:48:47,550 --> 00:48:53,070
Can anyone be negative in this subway stop?

421
00:48:53,070 --> 00:49:01,110
I'm inclined to say that's always true. Branching or stopping.

422
00:49:01,110 --> 00:49:07,350
Otherwise, you're going to market. You don't want to go. The transitivity does that for us.

423
00:49:07,350 --> 00:49:15,150
I mean, lots of this is cause this problem has caused lots of people to say that causation isn't transitive,

424
00:49:15,150 --> 00:49:20,470
but Lewis has a completely different way of solving it. Okay, let's have a look at his way of solving it.

425
00:49:20,470 --> 00:49:28,010
And Lewis says he wants no truck with these principles of invidious selection.

426
00:49:28,010 --> 00:49:37,530
I just like that sentence. So I thought I'd quoted and he thinks that we've got to distinguish causation and explanation.

427
00:49:37,530 --> 00:49:44,610
Now, that's a really important distinction. And if this were a flip chart, I'd turn the page and write something down.

428
00:49:44,610 --> 00:49:48,420
But as it's not a flip chart, this is very irritating, isn't it?

429
00:49:48,420 --> 00:49:53,430
Do you think there's one behind her? No.

430
00:49:53,430 --> 00:50:00,000
There's a table. Right.

431
00:50:00,000 --> 00:50:13,820
Okay. Causation. An explanation. I said right at the beginning of these lectures that when we explain things, we tend to appeal to causal chain.

432
00:50:13,820 --> 00:50:17,640
So why did so-and-so get cancer?

433
00:50:17,640 --> 00:50:21,840
Lung cancer? Answer because he smoked. Well, hang on a second.

434
00:50:21,840 --> 00:50:25,200
So-and-so got lung cancer and she doesn't smoke. OK.

435
00:50:25,200 --> 00:50:27,450
There must be more to it than smoking.

436
00:50:27,450 --> 00:50:37,740
But you see that when we explain something, we appeal to some event in the causal history of what it whatever it is that we're trying to explain.

437
00:50:37,740 --> 00:50:45,140
And what Lewis says is that the causation and it's not just Lewis in Stanley and I'm the philosopher.

438
00:50:45,140 --> 00:50:55,230
It's important for or hugely important for making the distinction between causation and explanation is actually a philosopher called Donald Davidson.

439
00:50:55,230 --> 00:51:07,170
And it is a very important explanation because causation is, we think, a relation that exists in the world quite independently of us.

440
00:51:07,170 --> 00:51:12,960
If we weren't here. Causation would still exist. So the world would go on.

441
00:51:12,960 --> 00:51:17,500
It's just that we wouldn't be here to look at it if we weren't here.

442
00:51:17,500 --> 00:51:22,950
And let's take away animals as well for anyone else who? Anyone who thinks that animals go in for explanation.

443
00:51:22,950 --> 00:51:26,610
There wouldn't be any explanation, would there?

444
00:51:26,610 --> 00:51:37,350
They'd be causation, but nobody would be making use of all of their observations of causation in explanation of anything.

445
00:51:37,350 --> 00:51:41,520
So explanation is very much an epistemological relation.

446
00:51:41,520 --> 00:51:48,660
It's a relation that makes things intelligible to us. Causation is a metaphysical relation.

447
00:51:48,660 --> 00:51:53,180
It's a relation between events quite independently of us.

448
00:51:53,180 --> 00:51:57,030
OK. So there might be causation, but no explanation.

449
00:51:57,030 --> 00:52:09,990
But it looks as if there couldn't be explanation and no causation, because explanation depends upon identifying a causal chain or in some event,

450
00:52:09,990 --> 00:52:14,490
in the causal history of the events that you're trying to explain.

451
00:52:14,490 --> 00:52:20,670
Do you see the difference between causation and explanation? Absolutely vital difference.

452
00:52:20,670 --> 00:52:28,470
And Lewis would say we look at the campus birth.

453
00:52:28,470 --> 00:52:33,990
And of course, it is an event in the causal history of the forest fire.

454
00:52:33,990 --> 00:52:41,130
When we look at the and again, if I had a flip chart so I would draw lots of little circles in a tree like shape

455
00:52:41,130 --> 00:52:45,960
with the one at the top being a forest far and somewhere down here at the bottom,

456
00:52:45,960 --> 00:52:50,970
there would be the birth of the camper and somewhere even further down would be the big bang.

457
00:52:50,970 --> 00:52:57,360
And this is the causal history. All these events combined to produce the forest fire.

458
00:52:57,360 --> 00:53:07,530
Now, all those are events in the causal history, but we'd be off to pick out the campers birth in explanation.

459
00:53:07,530 --> 00:53:19,210
Of the forest fire. We use our principles of invidious selection as hers to choose the campus, lighting the fire light, lighting his campfire.

460
00:53:19,210 --> 00:53:24,610
That's what explains to us the the forest fire.

461
00:53:24,610 --> 00:53:28,750
But the forest burning down, not the birth of the camper.

462
00:53:28,750 --> 00:53:34,840
So whereas the campers lighting the fire makes the forest far intelligible to us.

463
00:53:34,840 --> 00:53:39,100
The campers birth doesn't. If you say how did this forest fire happens?

464
00:53:39,100 --> 00:53:44,180
And I say, because Fred was born on 21st of June, 1957.

465
00:53:44,180 --> 00:53:53,590
You were in, say, a. That's no explanation. But if I say that Fred lit a fire half an hour before the forest fire started.

466
00:53:53,590 --> 00:54:01,060
That makes the forest far intelligible to you. Do you see the difference between explanation and causation?

467
00:54:01,060 --> 00:54:08,980
So Lewis thinks that both those objections to his second counterfactual theory of causation.

468
00:54:08,980 --> 00:54:13,800
So the choice, of course, one. So, again, look at the bomb example.

469
00:54:13,800 --> 00:54:23,320
And it's quite true. The enemy is placing the bomb was an event in the causal history of Susie's survival.

470
00:54:23,320 --> 00:54:28,420
Because had it not happened, Billy couldn't have pinched out the bomb.

471
00:54:28,420 --> 00:54:34,930
And if Billy's punching out, the bomb caused Susie's survival, which would all agree to.

472
00:54:34,930 --> 00:54:42,850
Then so did an enemy's placing the bomb. But in explanation of Susie's survival, we're not going to cite the placing of the bomb,

473
00:54:42,850 --> 00:54:48,910
Alwi, because that wouldn't make Lucy's survival intelligible to us at all.

474
00:54:48,910 --> 00:54:55,630
Whereas Billy's pulling out, the fuse does make her survival intelligible.

475
00:54:55,630 --> 00:55:05,740
David? Well, the terms were explanation for what will stop intelligibility.

476
00:55:05,740 --> 00:55:12,880
What the term is intended, as well as purely practical.

477
00:55:12,880 --> 00:55:20,380
Yeah, well, there are certain things make other things intelligible to us and certain things don't.

478
00:55:20,380 --> 00:55:26,080
And so it might be that my the the firing of mine.

479
00:55:26,080 --> 00:55:32,170
Okay, so I. What's your name. Mike. Mike. I shoot Mike.

480
00:55:32,170 --> 00:55:36,970
Okay. What caused the death of Mike. Answer. Marianne's pulling the trigger.

481
00:55:36,970 --> 00:55:44,260
But it might be that my pulling the trigger is identical to neurone n firing in my brain.

482
00:55:44,260 --> 00:55:49,810
But if you say well what caused Mike's death and I say Marion's neurone then firing.

483
00:55:49,810 --> 00:55:56,650
It's not an explanation, is it not Nesser. You have a lot of other knowledge.

484
00:55:56,650 --> 00:56:04,270
Seems to be very much a matter of the judgement. Yes, but intelligibility is intelligibility is to do with us.

485
00:56:04,270 --> 00:56:08,640
And what we can understand, what might not be what might be intelligible.

486
00:56:08,640 --> 00:56:13,120
A smart. Ah well.

487
00:56:13,120 --> 00:56:20,170
Neurone ends. Firing Marianne's heads to a neurophysiologists might be a perfectly good explanation of Michael's death.

488
00:56:20,170 --> 00:56:26,380
Given that Michael was standing where he was standing and so on and so forth, whereas it's not at all an explanation to ask.

489
00:56:26,380 --> 00:56:31,560
So yes, indeed, explanation is is relative to who is doing.

490
00:56:31,560 --> 00:56:39,820
But let me use a simple example. You and your five year old granddaughter and your one hundred and five year old mother.

491
00:56:39,820 --> 00:56:47,240
Forgive me if I am insulting you with the ages I've chosen. Oh, jolly good.

492
00:56:47,240 --> 00:56:55,450
You're all watching television together. But I think you'll agree that each of you is getting very different things from what you're watching.

493
00:56:55,450 --> 00:57:03,220
And that's because something's intelligible to you. Only in the context of your picture of the world.

494
00:57:03,220 --> 00:57:07,960
So something you can understand is not something your five year old granddaughter is going to be able to

495
00:57:07,960 --> 00:57:14,710
understand and possibly even your hundred and five year old mother has things she can teach you yet.

496
00:57:14,710 --> 00:57:22,570
Maybe she understands more than what? Because she's got a even bigger experience.

497
00:57:22,570 --> 00:57:39,310
Maybe if she started to get a little violent, she uses slow slowness, which we'd separate causation and explanation.

498
00:57:39,310 --> 00:57:46,180
He looks right. Be aware of this all three.

499
00:57:46,180 --> 00:57:52,990
It is almost infinite yet. And it does go right back to the Big Bang.

500
00:57:52,990 --> 00:57:59,910
And before that, if there was anything before, there is no causal structure.

501
00:57:59,910 --> 00:58:10,390
Remember? There are only certain bits of it that are intelligible to us as the cause.

502
00:58:10,390 --> 00:58:14,450
But isn't this just about the convention? Well, no, it isn't convention.

503
00:58:14,450 --> 00:58:26,450
What it's a matter of is wait for it regularity. I mean, we can see a regularity between the lighting of a fire and the fires getting out of hand.

504
00:58:26,450 --> 00:58:28,190
That's the sort of thing we understand.

505
00:58:28,190 --> 00:58:37,500
We've seen that correlation before, but there's no correlation that we've observed between the birth of people and the forest fire.

506
00:58:37,500 --> 00:58:46,930
No, there could be something else. Like what? Well, the fire being caused by the camera, maybe because he was particularly angry.

507
00:58:46,930 --> 00:58:50,520
Oh. And turn the gas up too high. Yes. Yes.

508
00:58:50,520 --> 00:58:56,720
I mean. Well, do you remember I said last week that the striking causes the fire, the match to light,

509
00:58:56,720 --> 00:59:03,560
but actually it wouldn't if there wasn't also oxygen, etc., but that would be a married convention if we were lucky.

510
00:59:03,560 --> 00:59:07,910
Well, that's not a matter of convention. It's definitely not a matter of convention.

511
00:59:07,910 --> 00:59:11,870
It's a matter of empirical observation, isn't it?

512
00:59:11,870 --> 00:59:18,410
We know that if you were to strike a match in a vacuum with the camera, if he really was angry,

513
00:59:18,410 --> 00:59:23,960
if if the law case was deciding whether he was negligent or not cause the fire.

514
00:59:23,960 --> 00:59:26,120
Oh, I see what your argument is about.

515
00:59:26,120 --> 00:59:33,080
Well, when we start looking at the law and start apportioning causation according to the law, we do have to start looking at convention.

516
00:59:33,080 --> 00:59:42,500
That's true. But it but the thing that's important is causation is completely independent of us, whereas explanation is never independent of us.

517
00:59:42,500 --> 00:59:47,990
And the law takes it just one bit further. His explanation has to do with intelligibility.

518
00:59:47,990 --> 00:59:57,530
Law has to do with convention as well. The legend, whatever the legislation is, and case studies and not case studies case.

519
00:59:57,530 --> 01:00:02,570
What's the word? Precedence and things like that as well. Did I say another hand?

520
01:00:02,570 --> 01:00:08,480
Yes. Can you speak up then?

521
01:00:08,480 --> 01:00:25,340
In court is a vast, almost infinite, or probably infinite number of things found in the course or the course of history.

522
01:00:25,340 --> 01:00:32,210
And you say pulsations independent. Of course, no causation is independent of us.

523
01:00:32,210 --> 01:00:42,260
Causation is independent. But the thing we pick out as the cause when we're explaining something is dependent on us.

524
01:00:42,260 --> 01:00:47,690
Causation is independent of the causes. No, the cause is also in.

525
01:00:47,690 --> 01:00:57,260
Is it also independent of our or. There's no such thing as the cause except when you're talking about explanation.

526
01:00:57,260 --> 01:01:05,720
Because there are many, many. And again, I wish I had a flip chart because this is drawable.

527
01:01:05,720 --> 01:01:11,330
I can't even turn it round. This is very frustrating. And I don't want to rub that off because I.

528
01:01:11,330 --> 01:01:26,030
Well, actually, let me rub it off because I think I'm doing this at the end anyway, so.

529
01:01:26,030 --> 01:01:37,910
So here's the forest fire. We're calling that E now that's got a string of different events in the causal chain.

530
01:01:37,910 --> 01:01:45,550
Each of them has a string of events in the causal chain leading to it.

531
01:01:45,550 --> 01:01:51,350
And there's no reason why there are three on each, except that I'm afraid of running out of room.

532
01:01:51,350 --> 01:01:56,180
Each of those is going to be caused by a string of events.

533
01:01:56,180 --> 01:02:07,660
Now, if we take this event here, if that had not occurred, that would not have occurred.

534
01:02:07,660 --> 01:02:18,520
But that doesn't mean we can cite that as the cause of that, because the cause tends to they can't be the cause,

535
01:02:18,520 --> 01:02:25,060
because it's only the cause in the context of every other cause that's come.

536
01:02:25,060 --> 01:02:30,740
But we could say it's the cause if it explains it to us.

537
01:02:30,740 --> 01:02:39,280
It probably doesn't if it's that far back. We'll pick something here as the cause.

538
01:02:39,280 --> 01:02:49,430
Are you with me? Yes, it seems inconsistent to have the same.

539
01:02:49,430 --> 01:02:55,060
Well, okay. I mean, just when we say something is the cause,

540
01:02:55,060 --> 01:03:01,660
that's usually part of an explanation because there's no such thing in the world as a verb cause it's there.

541
01:03:01,660 --> 01:03:06,520
There are always multiple causes. I mean, there is is a definite article.

542
01:03:06,520 --> 01:03:15,340
It picks out one and only one event. And there's no event that has one and only one cause.

543
01:03:15,340 --> 01:03:20,680
So the cause only has meaning within the context of an explanation.

544
01:03:20,680 --> 01:03:25,870
And the minute you have an explanation, you have principles of invidious selection.

545
01:03:25,870 --> 01:03:31,790
In other words, it depends on what's intelligible to us. Okay.

546
01:03:31,790 --> 01:03:36,780
There's nothing inconsistent about that. I think it's fairly straightforward.

547
01:03:36,780 --> 01:03:41,920
Did I. Woman at the back. Did I see your hand up?

548
01:03:41,920 --> 01:03:45,690
No. Okay, Bob, I'm not happy with this thing.

549
01:03:45,690 --> 01:03:57,560
There's a causation explanation. I'm happy that you can say that the birth of the camera or the big bang is part of the cause of the cause.

550
01:03:57,560 --> 01:03:59,360
It's not a very important one.

551
01:03:59,360 --> 01:04:11,880
It's quite I'm not happy with the idea that the enemy planting the bomb is part of the cause in any sense at all of the of Luce's survival.

552
01:04:11,880 --> 01:04:17,470
That seems to be just. Yes. No, I. I have some sympathy with that.

553
01:04:17,470 --> 01:04:23,860
But this is the argument that we'd make for that child.

554
01:04:23,860 --> 01:04:34,120
Well, that's the argument. Hard enemy not planted the bomb.

555
01:04:34,120 --> 01:04:43,930
Billy would not have pinched out.

556
01:04:43,930 --> 01:05:01,600
Sorry, I should have put up here. Susy's survive, Billie's pinching out the fuse caused Susie's survival.

557
01:05:01,600 --> 01:05:05,950
I mean, that would accept, wouldn't you just took that. Okay.

558
01:05:05,950 --> 01:05:11,170
Would you also accept how the enemy not planted the bomb? Billy would not have pinched out the fuse.

559
01:05:11,170 --> 01:05:18,370
Yes. Well, then by the transitivity of causation, it's got to be the enemies planting the bomb.

560
01:05:18,370 --> 01:05:22,540
That is the cause of Susie's survival, as we know. That's not the case.

561
01:05:22,540 --> 01:05:25,360
Well, hang on a sec. Well, no, we don't. No, it's not the case.

562
01:05:25,360 --> 01:05:37,360
I mean, it's just what we know is that we would never cites the planting of the bomb as an explanation of Susie's survival or of the cause.

563
01:05:37,360 --> 01:05:42,380
But I would say yes.

564
01:05:42,380 --> 01:05:47,140
This is bomb planting season. No, no, no. Right. Okay.

565
01:05:47,140 --> 01:05:53,250
If C causes E and E causes.

566
01:05:53,250 --> 01:05:57,700
D does. C cause.

567
01:05:57,700 --> 01:06:06,550
D? No. Of course not. Of course, means factually counter, factually defending the conclusion.

568
01:06:06,550 --> 01:06:17,260
I would say, you see, it is so easy to use counterfactuals on it.

569
01:06:17,260 --> 01:06:22,730
It does not follow. But what do you use counterfactual dependent on.

570
01:06:22,730 --> 01:06:26,680
See. This is what I would say. Okay.

571
01:06:26,680 --> 01:06:33,250
But I'm not asking that moment. I'm asking you whether you think causation is transitive.

572
01:06:33,250 --> 01:06:40,150
I mean, is that true? If it was if by by causes woman sufficient causes.

573
01:06:40,150 --> 01:06:46,270
Probably, yes. What we do think of causation is sufficient. So a cause is sufficient for its effect.

574
01:06:46,270 --> 01:06:50,460
We do things like that. And if that's true, then that's.

575
01:06:50,460 --> 01:06:58,750
And if that's true, then how is that not true? Because that's isn't that isn't counterfactual preparedness necessarily cause.

576
01:06:58,750 --> 01:07:02,290
Well, we're not. Don't forget, we're no longer talking about counterfactual dependence.

577
01:07:02,290 --> 01:07:07,600
We're talking about chains of stepwise influence. I'm not.

578
01:07:07,600 --> 01:07:11,470
We might have to engage in this by email or something like that.

579
01:07:11,470 --> 01:07:20,290
But you're absolutely right that this has caused the lot of people to think that that must be false.

580
01:07:20,290 --> 01:07:30,700
That causation is not transitive. After all, the trouble is, it seems a little hard to deny that.

581
01:07:30,700 --> 01:07:34,900
And if you think that's hard to deny, then you think the causation is transitive.

582
01:07:34,900 --> 01:07:45,430
And if you think causation is transitive, then it's difficult to deny that you just it is Sufi's survival death because.

583
01:07:45,430 --> 01:07:49,990
Well, you can, but that is what we mean. I don't think that helps much.

584
01:07:49,990 --> 01:07:57,260
Does it just not what it seems to me.

585
01:07:57,260 --> 01:08:08,710
Certainly we apply. We will all be over most of the all of this.

586
01:08:08,710 --> 01:08:15,250
No, you. You're absolutely wrong about that because all the network comes in because it had the big bang not happened.

587
01:08:15,250 --> 01:08:22,960
The forest fire would not have happened had the camper not been born, then the forest fire would not have happened.

588
01:08:22,960 --> 01:08:27,250
So we've still got in every event. That's part of the causal chain.

589
01:08:27,250 --> 01:08:39,040
What I'm saying is as soon as you think about this specific kind of like.

590
01:08:39,040 --> 01:08:42,760
Well, that's to do with explanation, not with causation,

591
01:08:42,760 --> 01:08:51,910
because the minute we've identified the cause and we've made it intelligible to ourselves, we tend not to worry about the rest.

592
01:08:51,910 --> 01:09:00,490
Of course, we will have to if if we want to start manipulating things, if we're scientists, we want to produce an effect.

593
01:09:00,490 --> 01:09:10,630
We'll have to make sure that the whole cause is no, then not just the cause, not just the part of the cause.

594
01:09:10,630 --> 01:09:16,090
To say the cause is nonsense. There is no such thing as the cause.

595
01:09:16,090 --> 01:09:23,920
The cause is always shorthand for an explanation. I mean, that's what I was saying a minute ago.

596
01:09:23,920 --> 01:09:32,980
The cause is complete nonsense. This was never an event that is caused by only one other event.

597
01:09:32,980 --> 01:09:56,900
Can you speak up? But you see, why do you need to say no need?

598
01:09:56,900 --> 01:10:02,470
Yeah. I mean, lots of people would have sympathy with with that claim.

599
01:10:02,470 --> 01:10:07,330
One way of solving this would be to say, I can't remember exactly how it goes.

600
01:10:07,330 --> 01:10:13,810
But you can't count as the cause. Something that would not have happened had.

601
01:10:13,810 --> 01:10:18,210
Sorry, I can't remember. But but you're absolutely right that there's a sort of.

602
01:10:18,210 --> 01:10:21,700
Yeah, absolutely. And there's a name for it.

603
01:10:21,700 --> 01:10:23,000
And I can't remember what that name is.

604
01:10:23,000 --> 01:10:31,180
I'll look it up next week if you want to e-mail me and remind me and I'll put you onto a paper that will explain that.

605
01:10:31,180 --> 01:10:36,910
But Louis' distinction between causation and explanation is his response.

606
01:10:36,910 --> 01:10:39,670
You would never cite the bomb being planted,

607
01:10:39,670 --> 01:10:51,610
enemies planting the bomb in explanation of Susie's survival because there's no correlation between the planting bombs and the survival of the target.

608
01:10:51,610 --> 01:10:57,240
So that's just not on. OK. Let's just. I'm worried about the time here.

609
01:10:57,240 --> 01:11:06,910
So another objection. So we've dealt with or Lewis thinks that we've dealt with the first objections to his second TTC,

610
01:11:06,910 --> 01:11:13,210
but we won't say, well, how can the causal history of an event be constituted of interrelated events?

611
01:11:13,210 --> 01:11:18,940
So how can causation have the structure that I drew on the board a minute ago with lots of

612
01:11:18,940 --> 01:11:24,360
events being caused by lots of other events and all of them being part of a causal history.

613
01:11:24,360 --> 01:11:28,420
And when we allow omissions and absences to be causes.

614
01:11:28,420 --> 01:11:37,300
So it was the Queen's failure to water my tomato plants that led to my tomatoes, plants dying,

615
01:11:37,300 --> 01:11:42,880
or at the very least, it was my failing to water my tomato plants that caused the tomato plants dying.

616
01:11:42,880 --> 01:11:51,490
And similarly, it was Bob's failure to put on the window locks that caused the burglary, etc.

617
01:11:51,490 --> 01:12:01,060
So we allow missions to be caused as well, unless you're going to allow negative events and some philosophers do.

618
01:12:01,060 --> 01:12:07,780
In other words, Billy's failing to put on the window locks is an event in itself.

619
01:12:07,780 --> 01:12:11,740
He's not doing it. And actually, that's difficult, isn't it?

620
01:12:11,740 --> 01:12:17,950
Because how many negative events are happening right now? There's you not having breakfast for a start.

621
01:12:17,950 --> 01:12:23,210
There's you not being in the pub. There's you not et cetera, et cetera, et cetera.

622
01:12:23,210 --> 01:12:27,970
So it's so if you allow negative events again, you're exploding your ontology.

623
01:12:27,970 --> 01:12:31,510
Maybe that's all right because we've already allowed possible world.

624
01:12:31,510 --> 01:12:38,830
So there's there is a possible world. You'll be glad to hear where you are in the pub.

625
01:12:38,830 --> 01:12:48,550
So we do seem to allow omissions and absences to be causes.

626
01:12:48,550 --> 01:12:55,960
But again, if we distinguish causation and explanation, we can explain this.

627
01:12:55,960 --> 01:13:07,390
Am I going to say no? Okay. So if we want to explain the burglary, we cite the failure of whoever it was to put the window locks on.

628
01:13:07,390 --> 01:13:17,620
If we want to explain your sobriety, we say it was your failure to be in the pub at the particular moment, et cetera.

629
01:13:17,620 --> 01:13:21,100
But again, it's explanation that matters here.

630
01:13:21,100 --> 01:13:31,150
So we the only reason that we can cite absences as causes is not because there are any such things as negative events,

631
01:13:31,150 --> 01:13:35,930
but because non events, non happenings.

632
01:13:35,930 --> 01:13:45,310
Make intelligible to us certain events. And so, again, we can cite causation and explanation.

633
01:13:45,310 --> 01:13:50,760
OK. Let's have a look at the I'm going to we can talk about this in a minute,

634
01:13:50,760 --> 01:13:58,680
but I'd just like to do the final five minutes of this session and then we can go onto all the questions, get them all at once.

635
01:13:58,680 --> 01:14:08,070
Okay. So the counterfactual theory of causation was always taken as a rival to the regularity theory of causation.

636
01:14:08,070 --> 01:14:17,070
When Lewis first postulated the counterfactuals area of causation, it was explicitly as a rival to the regularity theory.

637
01:14:17,070 --> 01:14:21,570
So he looked at the problems that we looked at last week for the regularity theory,

638
01:14:21,570 --> 01:14:26,370
said it cannot be the case that there's no more to causation than regularity.

639
01:14:26,370 --> 01:14:36,360
Instead, let's think of causation as as counterfactual dependence and the variations there in.

640
01:14:36,360 --> 01:14:40,470
Certainly Lewis saw it as a rival. Interesting, though.

641
01:14:40,470 --> 01:14:49,860
Hume didn't. I mean, Hume put forward those two explanations as if, in other words, he says and then he offers the counterfactual theory.

642
01:14:49,860 --> 01:15:01,980
So he didn't see it as a rival. An interesting this person, this philosopher, Helen Beebee, doesn't think that the CTC is a rival to RTC either.

643
01:15:01,980 --> 01:15:11,490
She says that Lewis is counterfactual theories of causation rely on the mill Ramsey Lewis view of the laws of nature.

644
01:15:11,490 --> 01:15:22,320
M. Ramsey Lewis, this is this is John Stuart Mill in his system of logic, Frank Ramsey and David Lewis himself.

645
01:15:22,320 --> 01:15:25,080
They come to this theory of the laws of nature.

646
01:15:25,080 --> 01:15:33,480
And if you remember, I spoke briefly about them last week, according to the Mill Ramsey Lewis view of the laws of nature.

647
01:15:33,480 --> 01:15:44,250
There is no more to a law of nature than the explanations generated by our best deductive systems.

648
01:15:44,250 --> 01:15:49,350
In other words, by the deductive systems that best explain our universe.

649
01:15:49,350 --> 01:15:56,220
So we make observations or observations. Tell us about correlations, about constant conjunctions.

650
01:15:56,220 --> 01:16:03,510
We those are the constant conjunctions that seem to occur irrespective of context.

651
01:16:03,510 --> 01:16:16,650
So it happens in your lab in Australia just as easily as it happens in my lab in London, etc. Those are taken out as the laws of nature that the the.

652
01:16:16,650 --> 01:16:25,140
So the best deductive system is determined by strength and simplicity in that.

653
01:16:25,140 --> 01:16:32,370
What we're looking for is the correlations, the the constant conjunctions that are as informative as possible.

654
01:16:32,370 --> 01:16:41,430
They tell us as much as possible about the world. And whilst using as little apparatus as possible, we have to introduce as little as possible.

655
01:16:41,430 --> 01:16:47,340
So you might laugh at the idea of possible worlds, but actually there are people who believe and there are a lot of people who

656
01:16:47,340 --> 01:16:53,490
believe that our best deductive systems rely on the notion of possible worlds.

657
01:16:53,490 --> 01:16:58,620
So we know that even though we're trying to use a simpler theory as possible,

658
01:16:58,620 --> 01:17:04,950
we've got to introduce possible worlds or we can't explain the world that we're actually in.

659
01:17:04,950 --> 01:17:11,670
So the mill Ramsey Lewis view of the laws of nature actually says that.

660
01:17:11,670 --> 01:17:56,090
So you might think. Pen keeps me fit all this.

661
01:17:56,090 --> 01:18:01,520
OK. Three levels. This is the level of language. This is the level of thought.

662
01:18:01,520 --> 01:18:05,510
And this is the level of, let's say, the world.

663
01:18:05,510 --> 01:18:15,080
I don't like saying that because, of course, thought and language are part of the world. But, yeah, this is the non-linguistic non thought.

664
01:18:15,080 --> 01:18:18,730
You see, these are all different.

665
01:18:18,730 --> 01:18:31,670
And you might think in exactly the same way that there's the laws of nature, which are the things that we use to explain and predict events.

666
01:18:31,670 --> 01:18:42,530
We we get our rockets to land on comets and things like that because we can do the mathematics and we can cite laws of nature and so on.

667
01:18:42,530 --> 01:18:49,190
And then underpinning that, there are natural uniformity is in our laws of nature are correct.

668
01:18:49,190 --> 01:18:54,170
If they correctly describe the natural uniformity is OK.

669
01:18:54,170 --> 01:18:58,580
In the same way that when we say the cat sat on the mat. That's true.

670
01:18:58,580 --> 01:19:05,630
Or we think the cat sat on the mat. These are true if and only if the cat is on the mat.

671
01:19:05,630 --> 01:19:13,100
So in the same way, our laws of nature are true only if the natural uniformity is that govern all worlds are as they are.

672
01:19:13,100 --> 01:19:19,310
But the mill Ramsey Lewis view of the laws of nature is doing away with that.

673
01:19:19,310 --> 01:19:24,200
All there is is our explanations of the world.

674
01:19:24,200 --> 01:19:34,370
So just as with the rank regularity theory of causation, you're saying that causation is the regularities themselves,

675
01:19:34,370 --> 01:19:42,810
not the necessary connexions that underpin the regularities in exactly the same way the laws of nature are the laws of nature,

676
01:19:42,810 --> 01:19:47,150
the laws of nature, the things that come out of our best deductive systems.

677
01:19:47,150 --> 01:19:59,600
Those are the laws of nature. They're not directly describing uniformity is that govern our world independently of us.

678
01:19:59,600 --> 01:20:09,890
So you can see that actually the crucial point, according to Helen Beebee, is that the counterfactual theory of causation,

679
01:20:09,890 --> 01:20:23,380
just like the regularity theory of causation, is austere, as I called it last week, austere in that it doesn't appeal to necessary connexions.

680
01:20:23,380 --> 01:20:28,340
All there is in the world are regularities not necessary.

681
01:20:28,340 --> 01:20:33,740
Connexions not not mind. Independent natural uniformity is.

682
01:20:33,740 --> 01:20:43,160
So the CTC, just like the RTC, sees causation as nothing more than lawful regularity.

683
01:20:43,160 --> 01:20:50,810
And so it cashes out those regularities in terms of chains of counterfactual dependency or influence.

684
01:20:50,810 --> 01:20:59,630
But it is in effect, just another version of the regularity theory of causation.

685
01:20:59,630 --> 01:21:08,780
And just to run through today, because I realise that it is quite difficult to take on board, we're answering the question, as we were last week.

686
01:21:08,780 --> 01:21:15,380
What is causation? And we're saying, could it be counterfactual dependence?

687
01:21:15,380 --> 01:21:19,550
Well, no, it couldn't, because there's causation without counterfactual dependence.

688
01:21:19,550 --> 01:21:27,140
That was the early pre-emption. And so on. Could it consistent chains of counterfactual dependence?

689
01:21:27,140 --> 01:21:35,150
Well, no. Because trumping and late pre-emption showed that it can't be those either.

690
01:21:35,150 --> 01:21:42,710
Could it consist in chains of stepwise influence? And as you go back, you can remind yourself what all these things are.

691
01:21:42,710 --> 01:21:49,970
And the answer to that is yes. So long as we distinguish causation and explanation.

692
01:21:49,970 --> 01:21:58,100
So we haven't found an objection to the final theory of counterfactual theory of causation,

693
01:21:58,100 --> 01:22:04,100
so long as we make that distinction between explanation and causation.

694
01:22:04,100 --> 01:22:10,970
But we have seen reason to think that the CTC is just another version of RTC.

695
01:22:10,970 --> 01:22:17,780
In other words, we've not really added more than a bit of sophistication to what we had last week.

696
01:22:17,780 --> 01:22:23,390
And next week I'm going to look at a rival to both these series.

697
01:22:23,390 --> 01:22:29,420
So a completely different take on causation.

698
01:22:29,420 --> 01:22:38,510
And what I'm doing is bringing this more up to date, because so far, I mean, I told you the CTC is although they they are still very much in play.

699
01:22:38,510 --> 01:22:42,800
They did start in 1973. So we're still, in a way, looking at history.

700
01:22:42,800 --> 01:22:46,790
So we are going to come bit more up to date next week.

701
01:22:46,790 --> 01:22:53,480
So that's it for this week. And we've got five minutes still for questions.

702
01:22:53,480 --> 01:23:02,570
Sean? I've spent the last 30, 40 years of my life delivering outcomes from causal chains.

703
01:23:02,570 --> 01:23:06,940
In effect, I'm your little dog and you put up with a single bottle at the top of it.

704
01:23:06,940 --> 01:23:12,720
Coming down the block. Exactly. Making a plan to achieve an outcome.

705
01:23:12,720 --> 01:23:24,580
And when you talked about the picking, pinching other ball and contributes to survival, I would attack a problem.

706
01:23:24,580 --> 01:23:38,280
Sorry. I would I would approach business or technical problem by questioning whether every action, every day contributed to the object.

707
01:23:38,280 --> 01:23:52,920
And obviously, in that particular case, planting a bomb is not contributing to a guy named Susie Susie.

708
01:23:52,920 --> 01:23:59,750
I would be delivering something to Susie.

709
01:23:59,750 --> 01:24:09,450
Survive the pitching outfit. Yes, a survivor.

710
01:24:09,450 --> 01:24:20,500
And in my having a rough day is evil.

711
01:24:20,500 --> 01:24:24,820
Could be mitigated by an action reaching out to you.

712
01:24:24,820 --> 01:24:32,660
So I don't have a problem if there is a fishing boat in my causal chain, it's been introduced by a risk which is now mitigation.

713
01:24:32,660 --> 01:24:38,330
So I took it out. And you were showing me the outcome isn't going to make the difference.

714
01:24:38,330 --> 01:24:44,030
That's important here. This is the thing that might make the difference. Here is what you were talking about.

715
01:24:44,030 --> 01:24:47,840
And I wish I could remember the. There is a technical term for it,

716
01:24:47,840 --> 01:25:00,320
but if you agree that Biddy's pinching the fuse is an it's a determinate and we can allow probabilistic into term of Susie's survival.

717
01:25:00,320 --> 01:25:07,310
And if you allow that Billies pinching the fuse depended counter fatuously on enemies placed in the bomb.

718
01:25:07,310 --> 01:25:11,690
Then you have a problem. I don't believe I have.

719
01:25:11,690 --> 01:25:19,370
Well, I will. Clearly, you can see that Lewis's own solution.

720
01:25:19,370 --> 01:25:26,150
He would say that, yes, we might cite that an explanation of Susie's survival, but we would never cite.

721
01:25:26,150 --> 01:25:31,040
That's an explanation, Susie. Survival. Why can't I just say to you.

722
01:25:31,040 --> 01:25:39,310
You're looking at how to manipulate an outcome. Well, you would never place a bomb as a means of manipulating the outcome.

723
01:25:39,310 --> 01:25:42,230
Susie's survival. Course she wouldn't.

724
01:25:42,230 --> 01:25:55,380
But in a context where you have Susie a bit too near a bomb, then Binnie's pinching out the fuse is a necessary condition of her survival.

725
01:25:55,380 --> 01:25:59,500
But that is because I have as a longer term desire.

726
01:25:59,500 --> 01:26:03,740
So, you know, I know that Susie will.

727
01:26:03,740 --> 01:26:11,440
I want her to survive. Hmm. If I don't know that there is such thing as a Susie and such a thing as hers.

728
01:26:11,440 --> 01:26:17,600
You know, perhaps it's a different thing. But I have a temporal overlay in this causal chain.

729
01:26:17,600 --> 01:26:21,920
You mentioned time. There's there's a flow of time. And this is something leads to something.

730
01:26:21,920 --> 01:26:25,310
Leads to something. There is a temporal flow.

731
01:26:25,310 --> 01:26:33,400
Now, interestingly enough, there are things, full anticipation functions in mathematics by atom bomb detector, which I can explain mathematically.

732
01:26:33,400 --> 01:26:38,270
I can anticipate in time something that will happen.

733
01:26:38,270 --> 01:26:44,480
So there's a you can play around it. Of course, I don't mean that common sense at all.

734
01:26:44,480 --> 01:26:48,770
But I don't have a problem with saying that there is a causal chain.

735
01:26:48,770 --> 01:26:52,250
If I do want to put it in a logical sense.

736
01:26:52,250 --> 01:27:00,770
But if there is a enemies placing a bomb, then I want to do something about it to mature and show survival.

737
01:27:00,770 --> 01:27:07,110
If there's no bomb. But you would never cite that in explanation of no.

738
01:27:07,110 --> 01:27:13,790
And you might say you wouldn't say that it is a cause of Susie's survival despite transitivity.

739
01:27:13,790 --> 01:27:19,650
But you would have to either say that causation itself isn't transitive.

740
01:27:19,650 --> 01:27:24,440
And they're all philosophers who insist on that because of problems like this.

741
01:27:24,440 --> 01:27:29,570
Two things. All you've got to find another way of saying that.

742
01:27:29,570 --> 01:27:38,060
OK, if Biddy's pinching the fuse cause Susie's survival and the enemy's placing the bomb cause Billy to pinch the fuse, then unfortunately,

743
01:27:38,060 --> 01:27:46,850
if causation is transitive, you've got to say that enemy's placing the bomb caused the disease the survival and not a cancer.

744
01:27:46,850 --> 01:27:51,490
Well, that that's the point you're making. There is an argument to that.

745
01:27:51,490 --> 01:27:54,680
You can say that we can't call that a cause.

746
01:27:54,680 --> 01:28:03,140
Because it was only because of it's only because of that that that causes that and these cancel each other out.

747
01:28:03,140 --> 01:28:12,230
I will for next week look up what what that is. Cause I can't remember for now because doesn't this get around to the idea that humans have to have.

748
01:28:12,230 --> 01:28:19,280
Oh, hang on. It's just coming into my mind and it's gone again.

749
01:28:19,280 --> 01:28:23,740
So I just say it just say flitted, you know, just so gone.

750
01:28:23,740 --> 01:28:32,300
OK. I think I still have it now. The problem is that we humans have to have explanations for things.

751
01:28:32,300 --> 01:28:37,120
And it all sounds fine that when you want to help Susie survive because the genes found.

752
01:28:37,120 --> 01:28:42,950
But when it gets into other areas where grandiosity of humans wants to manipulate,

753
01:28:42,950 --> 01:28:50,120
even even unconsciously, to make themselves feel good and explain something and bring logic to something.

754
01:28:50,120 --> 01:28:56,210
So without you know, in due respect to his explanation,

755
01:28:56,210 --> 01:29:03,140
this is why I'm beginning to see the merit of understanding that cause happens without any explanation.

756
01:29:03,140 --> 01:29:10,610
The course is completely independent of not the cause. Sorry, causation exists independently of us.

757
01:29:10,610 --> 01:29:13,670
I mean, not this obviously is something in place, a bloody bomb.

758
01:29:13,670 --> 01:29:19,640
It doesn't happen independently of us, but causation exists in the world independently of us.

759
01:29:19,640 --> 01:29:23,120
Explanation is something we do right.

760
01:29:23,120 --> 01:29:31,490
And in doing it, we cite cause it causes and because ah interesting causation is really only explanation.

761
01:29:31,490 --> 01:29:36,920
What we're interested in is those causes that make intelligible their effects.

762
01:29:36,920 --> 01:29:42,920
And there are lots of other causes of the effects that we want to explain, but we're not interested in them.

763
01:29:42,920 --> 01:29:48,410
If they don't make them intelligible when we're not interested in getting out of your sedergreen happiness,

764
01:29:48,410 --> 01:29:54,140
to think that explanations is led us to the worst for us.

765
01:29:54,140 --> 01:29:59,780
How much explanation has been good as opposed to how much?

766
01:29:59,780 --> 01:30:05,480
Well, I mean, Vic Einstein said explanations must come to an end.

767
01:30:05,480 --> 01:30:14,120
But that was in a different context, gone. I was just thinking about what you say about it.

768
01:30:14,120 --> 01:30:22,220
No, I didn't say anything about universes with nobody in them would be causation.

769
01:30:22,220 --> 01:30:30,460
Oh, yes. I'm sorry. Yes, I did. Even if there were no human beings, there would still be causation contests where there would still be.

770
01:30:30,460 --> 01:30:38,600
And there wouldn't be any action. But there would be causation. Causation is because we are to create these things.

771
01:30:38,600 --> 01:30:44,750
One session of time. No. This position of time just got events.

772
01:30:44,750 --> 01:30:51,320
Why do you call witnesses or no direction?

773
01:30:51,320 --> 01:30:58,340
Well, if there was a cosmic occurrence that were to wipe out all human beings.

774
01:30:58,340 --> 01:31:07,640
Now, does that mean that, for example, the London I would stop turning if it is turning over?

775
01:31:07,640 --> 01:31:13,490
No, I'm just saying it isn't isn't causation is the human concept. No explanation is.

776
01:31:13,490 --> 01:31:14,750
But causation. I do.

777
01:31:14,750 --> 01:31:25,130
Well, causation is clearly a human concept in that it wouldn't be the process of causation wouldn't have a name if we hadn't given it one,

778
01:31:25,130 --> 01:31:34,220
but it would still exist. But isn't it just a way of seeing the world which is based on the fact that we live temporal movement,

779
01:31:34,220 --> 01:31:38,390
we see things is progressing through time and changing.

780
01:31:38,390 --> 01:31:46,130
Whereas if you have no individuals witnessing the changing things through time, then nothing would be changing.

781
01:31:46,130 --> 01:31:55,460
Everything will just be as it is. We would have no thing leading to something else because this time you're actually witnessing.

782
01:31:55,460 --> 01:32:00,760
Well, the question of whether time is real is a very big philosophical question.

783
01:32:00,760 --> 01:32:10,250
And we might think the time isn't real. And if time isn't real, then then there can't be temporal successions that are causation.

784
01:32:10,250 --> 01:32:18,020
But I do think that's a different question. And I completely accept that without human beings, there'd be no explanation.

785
01:32:18,020 --> 01:32:24,950
But I don't think I accept that without humans, there'd been a I mean, the big bang caused the universe.

786
01:32:24,950 --> 01:32:32,420
There were no humans around at the time. It's true that we couldn't know that the Big Bang caused the universe.

787
01:32:32,420 --> 01:32:35,540
Were it not for various theories and so on.

788
01:32:35,540 --> 01:32:39,200
But but I think that it existed without us.

789
01:32:39,200 --> 01:32:44,000
Seems to me to be. But then I'm an invert.

790
01:32:44,000 --> 01:32:51,960
This idea of causation is something that's very imposed on the universe.

791
01:32:51,960 --> 01:33:03,740
Okay. I just want. We accept the statement that explanation is always going to be, yes, they probably can,

792
01:33:03,740 --> 01:33:07,580
because there can be legal explanations and there could be other types of explanation.

793
01:33:07,580 --> 01:33:17,750
But Ann Lewis would actually say that all explanation is causal because even legal explanations will cite causes.

794
01:33:17,750 --> 01:33:23,150
So it's your intention to kill that made it murder.

795
01:33:23,150 --> 01:33:30,470
So it is the fact that you caused the death of somebody and intended to cause that death that makes it murder, etc.

796
01:33:30,470 --> 01:33:38,330
So Lewis would definitely say there are no explanations that are not causal or that are not.

797
01:33:38,330 --> 01:33:48,770
Such that we cite a cause in them. Well, if you look at sort of like visual perception, it's similar to white and it can be something like analogy.

798
01:33:48,770 --> 01:33:57,200
It could be a way of thinking that is not based on explanations would be something else outside of causality.

799
01:33:57,200 --> 01:34:00,870
Okay. I mean, it certainly does look as if there could be.

800
01:34:00,870 --> 01:34:05,720
But Lewis would say no, there are no explanations that are not causal.

801
01:34:05,720 --> 01:34:13,270
And one more question then. I think we'd better finish the problem with the speaker between the explanation and causation.

802
01:34:13,270 --> 01:34:17,450
I mean, I decide to throw a brick and break a window.

803
01:34:17,450 --> 01:34:22,990
Then that's an explanation why we never wrote me in the sidebar.

804
01:34:22,990 --> 01:34:26,090
Doesn't that have to be a break in the chain of causation?

805
01:34:26,090 --> 01:34:32,780
Because otherwise, I mean, my decision is simply a result of other causes, and I'm not really there at all.

806
01:34:32,780 --> 01:34:37,880
Well, actually, I mean, that that is exactly exactly the problem of free will.

807
01:34:37,880 --> 01:34:44,060
I mean, if if your decision is not itself a caused event,

808
01:34:44,060 --> 01:34:52,280
then it's it's not true that we can go back in the day and find events that cause other events, etc.

809
01:34:52,280 --> 01:35:00,110
But but I mean, the question of whether reasons are causes is something that I'll look at at some length.

810
01:35:00,110 --> 01:35:05,540
It could be a decision during Wunderbar. It would actually have to be on course.

811
01:35:05,540 --> 01:35:10,250
Well, not. No, no, no. There are lots of philosophers who think that decisions are caused as well.

812
01:35:10,250 --> 01:35:17,890
They're just part of the causal chain. But you're right. There are other philosophers who think that if it's caused, it's not a decision at all.

813
01:35:17,890 --> 01:35:23,090
Yeah. Okay, let's finish there and we'll start again next week.

814
01:35:23,090 --> 01:35:26,610
Looking at the rival theory. So a completely different type of thing.

815
01:35:26,610 --> 01:35:30,960
Or at least we'll look at the argument for the existence of a rival theory.

816
01:35:30,960 --> 01:35:38,881
Then the following week, we'll look at the rival theory.