WEBVTT NOTE Free Will: Two Paradoxes of Choice 1 00:00:00.000 --> 00:00:06.000 Okay, today I want to talk about two paradoxes of choice. 2 00:00:06.000 --> 00:00:10.600 I'm not sure the second one is really a paradox, but the second one is always called a paradox. 3 00:00:10.600 --> 00:00:13.900 And the first one isn't usually called a paradox, but it is one. 4 00:00:13.900 --> 00:00:17.700 So they're both paradoxes in some sense. 5 00:00:17.700 --> 00:00:26.500 And paradoxes that connect with issues about free will, about human action, and about nature of the good. 6 00:00:26.500 --> 00:00:30.100 and we'll set us up for some things that will be coming later. 7 00:00:32.100 --> 00:00:34.900 Okay, the first paradox is called Newcomb's Problem. 8 00:00:37.700 --> 00:00:40.200 A lot of different philosophers have talked about this problem, 9 00:00:40.200 --> 00:00:44.700 although they've disagreed with each other about exactly what the moral of the paradox is supposed to be 10 00:00:44.700 --> 00:00:50.700 or what, or in some sense even what the topic of it is, but I've got my line on that. 11 00:00:51.100 --> 00:00:52.700 So here's the story you're supposed to imagine. 12 00:00:52.700 --> 00:00:57.700 Imagine that you are confronted with two boxes. 13 00:00:57.700 --> 00:01:01.700 One of them is transparent and one is opaque. 14 00:01:01.700 --> 00:01:06.700 So you can see what's in the transparent one. What's in the transparent one is a thousand dollars. 15 00:01:06.700 --> 00:01:12.700 You can't see what's in the opaque one, but you are assured, and just stipulate that you believe this, 16 00:01:12.700 --> 00:01:18.700 that the box you can't see into contains either a million dollars or nothing. 17 00:01:18.700 --> 00:01:33.700 You don't know which, and you're given the following choice, you can either take the opaque box only, just the box you can't see into, or you can take both. 18 00:01:33.700 --> 00:01:45.700 Well, thus far it doesn't seem like a difficult choice. Why not take both? If you take only the opaque box, you'll get either a million dollars or nothing. 19 00:01:45.700 --> 00:01:50.700 If you take both boxes, you'll get either a million plus a thousand, or you'll get a thousand. 20 00:01:50.700 --> 00:01:55.700 Clearly, whatever happens, you're better off taking both boxes than just taking the opaque one. 21 00:01:55.700 --> 00:02:03.700 But now you're told an additional piece of information, which we stipulate again, you also believe. 22 00:02:03.700 --> 00:02:10.700 There is a reliable predictor. Some versions, it's a very, very reliable track record. 23 00:02:10.700 --> 00:02:13.700 In other versions, it's actually an infallible predictor. 24 00:02:13.700 --> 00:02:15.700 Let's go with the infallible predictor. 25 00:02:15.700 --> 00:02:18.700 I think that sharpens the paradox more. 26 00:02:18.700 --> 00:02:25.700 There's a predictor who can predict exactly what anyone will do in any choice situation. 27 00:02:25.700 --> 00:02:29.700 They're a master of psychology or they are reading your brainwaves or whatever it is. 28 00:02:29.700 --> 00:02:32.700 But this person knows what kind of person that you are. 29 00:02:32.700 --> 00:02:36.700 This person knows what choice you would make in this situation. 30 00:02:36.700 --> 00:02:44.700 And this is also the person who's in charge of deciding whether it's a million dollars or nothing that's been put into the box you can't see. 31 00:02:44.700 --> 00:02:48.700 So, I've got this thing here, I might as well use it for something. 32 00:02:48.700 --> 00:03:01.700 And just to, let's see, which is the spot on the top, all right, so just to, oh, I'm doing it the wrong way. 33 00:03:01.700 --> 00:03:17.700 But the box you can see into with, I'll do this to show you can see into it, I don't know why that shows that. 34 00:03:17.700 --> 00:03:28.700 Here's the box you can't see into, go like that to show you cannot see into it. 35 00:03:28.700 --> 00:03:41.700 And so your choice is between taking both these boxes, you have the choice of taking them both, or you have the choice of just taking this one. 36 00:03:41.700 --> 00:03:55.700 Okay, so the predictor, who knows what kind of person you are, knows what you will do in the situation as described, has either put a million dollars in this box or zero. 37 00:03:55.700 --> 00:04:09.700 What the predictor has done is, if the predictor predicts that you will choose only the opaque box, then the predictor has put a million dollars in it. 38 00:04:09.700 --> 00:04:18.700 But if the predictor predicts that you will choose both boxes, then the predictor will put nothing in the opaque box. 39 00:04:18.700 --> 00:04:25.700 So there you are facing the boxes. The predictor has already made whatever prediction she's going to make, and she's already put the money in, or not. 40 00:04:25.700 --> 00:04:33.700 She's already done whatever she's going to do with that. So there you are. You can't see into the box, but you know that it's either got the million in already, or it's got nothing. 41 00:04:33.700 --> 00:04:44.700 And now you have to make this choice. Well, now the situation is more complicated. Should you take both boxes or one? 42 00:04:44.700 --> 00:04:47.400 Well, it seems like there's a good argument for each. 43 00:04:47.400 --> 00:04:50.500 Here's the argument for taking both boxes. 44 00:04:50.500 --> 00:04:52.500 The money's already there or not. 45 00:04:52.500 --> 00:04:54.700 Nothing you can do now can change it. 46 00:04:54.700 --> 00:04:56.700 It's not as though the money's going to, you know, 47 00:04:56.700 --> 00:05:00.100 magically, quantumly vanish depending on what choice you make. 48 00:05:00.100 --> 00:05:03.600 Either the money's in there or it isn't. 49 00:05:03.600 --> 00:05:10.000 So, given that fact, if you pick, you know, 50 00:05:10.000 --> 00:05:13.800 if the money's there, then if you pick both boxes, 51 00:05:13.800 --> 00:05:21.400 If you get a million plus a thousand, whereas if you only pick the one box you just get the million. 52 00:05:21.400 --> 00:05:27.400 A million plus a thousand is better than a million, so if the money is there it's better to pick both boxes. 53 00:05:27.400 --> 00:05:34.400 If the money isn't already there, then picking both boxes gets you a thousand and picking just the one box gets you nothing. 54 00:05:34.400 --> 00:05:36.900 A thousand is definitely better than nothing. 55 00:05:36.900 --> 00:05:43.100 So it seems like whether the money is there or not, either way it seems like it makes sense you're better off choosing both boxes. 56 00:05:43.100 --> 00:05:47.100 So that's the argument for choosing both boxes. 57 00:05:47.100 --> 00:05:50.100 However, there's also an argument for choosing one box. 58 00:05:50.100 --> 00:05:53.100 The argument is this. 59 00:05:53.100 --> 00:05:56.100 The predictor is an infallible predictor. 60 00:05:56.100 --> 00:06:01.100 She's put the million in there only if you are going to choose just the one box. 61 00:06:01.100 --> 00:06:08.100 So, any possible future in which you choose just the one box 62 00:06:08.100 --> 00:06:10.100 is a future where you get the million. 63 00:06:10.100 --> 00:06:15.900 Any possible future where you choose both boxes is a future where you get only a thousand. 64 00:06:15.900 --> 00:06:23.600 So, if you really trust, and we're stipulating for the sake of argument, you do really trust that this predictor is utterly reliable. 65 00:06:23.600 --> 00:06:29.100 Then it seems you have to say that your only chance of getting the million is to pick just the one box. 66 00:06:29.100 --> 00:06:32.300 And so that's what you should do. 67 00:06:32.300 --> 00:06:39.500 And, you know, the two-boxer says, well, that doesn't make sense, because, you know, you can't affect... 68 00:06:39.500 --> 00:06:44.940 What you do now can't affect what the predictor puts in there because the predictor has already done it. 69 00:06:44.940 --> 00:06:49.660 You can't affect the past, so you should pick both boxes. 70 00:06:49.660 --> 00:06:53.980 The one-boxer says, this is actually a quote from a famous article about this, 71 00:06:53.980 --> 00:06:55.780 if you're so smart, why aren't you rich? 72 00:06:55.780 --> 00:07:01.500 In other words, if you're so smart to appeal to this idea that the money is already there, 73 00:07:01.500 --> 00:07:08.820 well still, if this really is a reliable predictor, then it's only the people that the predictor predicted would pick the one-box, 74 00:07:08.820 --> 00:07:10.740 And those are the people who do pick the one box. 75 00:07:10.740 --> 00:07:13.180 It's only the people who pick the one box that get the million. 76 00:07:13.180 --> 00:07:17.620 The people who reason your clever way and pick both boxes 77 00:07:17.620 --> 00:07:20.100 keep getting a thousand over and over. 78 00:07:20.100 --> 00:07:23.140 There's the people who follow, you know, 79 00:07:23.140 --> 00:07:25.860 who ignore the fact that the money's already either there or not 80 00:07:25.860 --> 00:07:30.780 and just, you know, just pick the one box, they get the million. 81 00:07:30.780 --> 00:07:37.300 So, which, you know, would you rather be right or rich is the way it's sometimes put. 82 00:07:37.300 --> 00:07:45.300 Well, so there's been a lot of dispute about which of these two arguments is the better argument. 83 00:07:45.300 --> 00:07:50.300 Should you accept the two box argument based on the fact that money is already either there or not? 84 00:07:50.300 --> 00:08:02.300 Or should you accept the one box argument on the basis that those who get the one box are always the ones who get the million and those who don't, don't. 85 00:08:02.300 --> 00:08:09.300 All right, my claim is that these two arguments are equally good. 86 00:08:09.300 --> 00:08:16.300 That if you accept the situation as described, if you accept that there's a predictor who is reliable in this way, 87 00:08:16.300 --> 00:08:24.300 and if you accept that she's put the money in, if she predicted you'd pick the one box, and that she hasn't, if she's predicted you wouldn't. 88 00:08:24.300 --> 00:08:31.300 If you accept all that, then I think that both arguments decisively follow. 89 00:08:31.300 --> 00:08:35.300 It is true that the money is already either there or not. 90 00:08:35.300 --> 00:08:44.300 And given that, it follows that picking both boxes can't make the money go away if it's already there. 91 00:08:44.300 --> 00:08:48.300 And picking just the one box can't make the money appear if it isn't already there. 92 00:08:48.300 --> 00:08:53.300 Therefore, the case for picking both boxes is decisive. 93 00:08:53.300 --> 00:08:58.300 On the other hand, the one box argument is definitive. 94 00:08:58.300 --> 00:09:06.300 If you accept that there is this infallible predictor and that she has done the things that are described in the story, 95 00:09:06.300 --> 00:09:12.300 then it follows that whether you get the million or not co-varies with whether you pick just the one box or not. 96 00:09:12.300 --> 00:09:18.300 And if there's something under your control that co-varies with something you really want, 97 00:09:18.300 --> 00:09:22.300 so you get the thing you want only in those cases where you make this choice, 98 00:09:22.300 --> 00:09:24.300 and you don't get the thing you want when you make the other choice, 99 00:09:24.300 --> 00:09:29.740 well then clearly it's rational you have to do the thing. So I think both 100 00:09:29.740 --> 00:09:33.580 arguments and once you accept the choice set up as described I think both 101 00:09:33.580 --> 00:09:37.020 arguments are decisive. Of course they lead to opposite results so something must 102 00:09:37.020 --> 00:09:44.660 be wrong. Well generally when you have an antinomy that is when you have arguments 103 00:09:44.660 --> 00:09:50.380 both of which seem decisive but which lead to opposite conclusions generally a 104 00:09:50.380 --> 00:09:57.380 A good strategy is to look to see if there's some common assumption behind both that's generating the problem. 105 00:09:57.380 --> 00:10:04.380 And in this case, I think the common assumption is the possibility of this situation. 106 00:10:04.380 --> 00:10:12.380 If you assume that you are in the situation as described, then you are committed to picking both boxes and you're also committed to picking just one. 107 00:10:12.380 --> 00:10:14.380 So something must be wrong. 108 00:10:14.380 --> 00:10:21.380 So what I suggest is what's wrong is that you cannot coherently regard yourself as being in such a situation. 109 00:10:21.380 --> 00:10:25.380 You can't regard that as a possible situation for you to be in. 110 00:10:25.380 --> 00:10:29.380 So what do these two arguments show? 111 00:10:29.380 --> 00:10:37.380 Well the two box argument shows that you can't coherently aim to bring about something that you think is already passed and settled. 112 00:10:37.380 --> 00:10:41.380 It doesn't make any sense to say... 113 00:10:41.380 --> 00:10:47.380 And that's what you're doing. If you pick just the one box, you're picking it in order to get the million. 114 00:10:47.380 --> 00:10:59.380 But since the million is in there only if you pick it, according to the assumption, 115 00:10:59.380 --> 00:11:10.380 then in effect, you are picking the one box in order to make it the case that the predictor put the money in there. 116 00:11:10.380 --> 00:11:14.380 So you're doing something now in order to make it the case that the predictor did something in the past. 117 00:11:14.380 --> 00:11:21.380 And the reason that seems to lead to a paradox is that what you do now can't change the past. 118 00:11:21.380 --> 00:11:25.380 If the predictor already put the money in, nothing you do now can prevent that. 119 00:11:25.380 --> 00:11:28.380 And if the predictor didn't put the money in, nothing you do can make that happen. 120 00:11:28.380 --> 00:11:35.380 So what the two-box argument shows is it doesn't make sense to think of yourself as aiming backward, 121 00:11:35.380 --> 00:11:39.380 as trying to bring something about that already is in the past. 122 00:11:39.380 --> 00:11:54.380 However, what the one box argument shows is that if there is something we can aim at, and it necessarily co-varies with something else we want, then we can aim at A in order to get that first thing or to get the other thing. 123 00:11:54.380 --> 00:12:08.380 If I get the million only if I pick the one box, and I can control whether I pick the one box, so I can aim at picking one box, it seems like I can aim at getting the million. 124 00:12:08.380 --> 00:12:35.380 So, I think that the moral of this is, is that you can't coherently aim at or regard yourself as aiming at something if you think that that something either is already passed and settled or co-varies necessarily with something already passed and settled. 125 00:12:35.380 --> 00:12:46.380 So that's why I think you can't aim at the predictors having done something, because that's already passed and settled. 126 00:12:46.380 --> 00:12:54.380 And therefore, you can't really, it doesn't really make sense to think of yourself as aiming at picking one box in order to get the million, 127 00:12:54.380 --> 00:13:01.380 because whether you get the million or not co-varies with whether the predictor already put it in there, which is already passed and settled. 128 00:13:01.380 --> 00:13:12.380 Now people have often said, well sure, you can't aim at something you think is already settled, or you can't choose or deliberate about something you think is already settled, if you think you know which way it's settled. 129 00:13:12.380 --> 00:13:19.380 You know, I can't decide between A and B if I think it's already settled that I'm going to do A. 130 00:13:19.380 --> 00:13:27.380 That's fairly uncontroversial. If I think that A is already settled and B isn't now ruled out as impossible, then I can't deliberate and choose between them. 131 00:13:27.380 --> 00:13:32.380 But people sometimes think that's only because we know which way it's settled. 132 00:13:32.380 --> 00:13:38.380 A lot of people think that if you think it's already settled which one's going to happen, but you don't know which, then you can still deliberate and choose. 133 00:13:38.380 --> 00:13:44.380 But I think what the Newcomb's problem suggests is, I think the moral of it is, you can't coherently aim at something, 134 00:13:44.380 --> 00:13:52.380 or you can't coherently try to make choices about something you think is already fast and settled, even if you don't know which way it is. 135 00:13:52.380 --> 00:14:00.380 As in this case, you don't know which way, it's either a million or zero, we don't know which way it is. 136 00:14:00.380 --> 00:14:06.380 You might have thought that you could therefore aim at one or the other, or choose between them, 137 00:14:06.380 --> 00:14:15.380 but I think the paradox you run into in Newcomb's problem, I want to suggest, means that you can't go ahead and do that. 138 00:14:15.380 --> 00:14:19.380 Alright, what does that moral get us? Assume that that's right. 139 00:14:19.380 --> 00:14:34.380 Assume that this shows, what I'm claiming is it shows, that you can't coherently aim at something that's either already passed and settled or else that it covariates necessarily with something already passed and settled. 140 00:14:34.380 --> 00:14:36.380 Where does that show us? 141 00:14:36.380 --> 00:14:42.380 Okay, first thing I think it shows us is it shows us a problem with certain views of motivation. 142 00:14:42.380 --> 00:14:54.380 There are some views of motivation, according to which, whenever we do something, what we're motivated by is one of our own subjective mental states, in the sense that our goal is to bring about some subjective mental state. 143 00:14:54.380 --> 00:15:01.380 This is true, for example, of psychological hedonism, where if you think that the goal of everything you do is to think you feel some kind of pleasure. 144 00:15:01.380 --> 00:15:10.380 A lot of people have thought that, thought that whatever you do, the ultimate end is to produce some feeling of pleasure or satisfaction in your mind. 145 00:15:10.380 --> 00:15:15.380 Not just that that's sometimes the goal, but that's always the goal, the only goal. 146 00:15:15.380 --> 00:15:19.380 John Stuart Mill, for example, seems to have thought something like that, 147 00:15:19.380 --> 00:15:23.380 although there are passages where maybe he's heading in a different direction, 148 00:15:23.380 --> 00:15:27.380 but at least that's on the surface, that's the most natural way to read him. 149 00:15:27.380 --> 00:15:33.380 Certainly that's the most natural way to read someone like Jerry Bentham, I think. 150 00:15:33.380 --> 00:15:38.380 Some people have thought that Mises holds this view, too. I think it's hard to tell whether he does. 151 00:15:38.380 --> 00:16:08.380 Mises often says that the point of acting is to relieve felt dissatisfaction and you could take that to mean that you've got a felt uneasiness and you could take that to mean that you've currently got some feeling of uneasiness or dissatisfaction and you want to get rid of it and so you go looking for something that will make that feeling go away and so the relief from that feeling would be the goal of our action. Sometimes Mises sounds like that's what he means and if so 152 00:16:08.380 --> 00:16:33.380 Mises talks more as though he means something purely formal, he just means that you're dissatisfied with some situation, you want to make the situation different, but not necessarily that you're trying to relieve some feeling in your mind. 153 00:16:33.380 --> 00:16:44.980 but anyway whatever Mises thought at least it's a possible view I think that the 154 00:16:44.980 --> 00:16:49.820 fact that we can't aim at something past and settled causes a problem for this 155 00:16:49.820 --> 00:16:52.940 view because think of some of the things we can aim at for example suppose that I 156 00:16:52.940 --> 00:16:59.780 buy life insurance or write a will that's an action I'm doing now what's the end 157 00:16:59.780 --> 00:17:03.780 What is my goal in buying life insurance or writing a will? 158 00:17:03.780 --> 00:17:09.780 Well, presumably it's my action as a means to my loved ones doing well after my death. 159 00:17:09.780 --> 00:17:14.780 Seems like that's the most natural way of thinking about it. 160 00:17:14.780 --> 00:17:36.780 You've got, let's take, I write my will now and it seems like that's a means to my loved ones doing well after my death. 161 00:17:36.780 --> 00:17:41.780 Seems like that's, this is the means, this is the end. 162 00:17:41.780 --> 00:17:47.780 But, is that in and of itself, or is it a means to something further? 163 00:17:47.780 --> 00:17:51.780 Well, whatever it means to it, it seems like it couldn't be a means to pleasure. 164 00:17:51.780 --> 00:17:58.780 Because we're assuming that, you know, this is after my death, I'm not around to experience the pleasure. 165 00:17:58.780 --> 00:18:08.780 Now, if you believe in an afterlife, you still don't necessarily believe that you're going to be hanging around watching your loved ones 166 00:18:08.780 --> 00:18:16.180 and getting, feeling a pleasure from seeing them do well, you know, might be off busy elsewhere. 167 00:18:16.180 --> 00:18:21.380 Anyway, people who don't believe in an afterlife also buy life insurance and write wills and so forth. 168 00:18:21.380 --> 00:18:26.280 They certainly don't think that they're going to be feeling pleasure from this in the future. 169 00:18:26.280 --> 00:18:34.180 So, writing a will can't be a means to, you know, this cannot be a means, 170 00:18:34.180 --> 00:18:39.040 My loved one's doing well after my death can't be a means to some pleasure in the future. 171 00:18:39.040 --> 00:18:45.600 Now, someone could say, well, yes, sure, but when you write the will or buy the life insurance, you get pleasure now. 172 00:18:45.600 --> 00:18:50.480 You take pleasure now in your loved one's doing well in the future. 173 00:18:50.480 --> 00:18:51.500 Well, what does that mean? 174 00:18:51.500 --> 00:19:02.740 It could mean that your loved one's doing well after your death causes you to feel pleasure now. 175 00:19:02.740 --> 00:19:06.740 And so then you'd be, but that would be a kind of backwards aiming. 176 00:19:06.740 --> 00:19:14.740 And if there's no backwards aiming, if we can't aim it at, you know, if, 177 00:19:14.740 --> 00:19:18.740 I think what Lucas's problem shows is that a means can't be later than the end. 178 00:19:18.740 --> 00:19:23.740 And so it doesn't make any sense to think of this as a means to that. 179 00:19:23.740 --> 00:19:25.740 You could say, all right, well, this drops out then. 180 00:19:25.740 --> 00:19:36.740 Well, just writing the will now causes the thought that your loved ones will do after your death, and that thought causes pleasure now. 181 00:19:36.740 --> 00:19:47.740 Well, that's true enough, but the question is, is it only the thought of your loved ones doing well after your death that you care about, 182 00:19:47.740 --> 00:19:52.740 or is it your loved ones actually doing well after your death that you care about? 183 00:19:52.740 --> 00:19:58.740 Here's a reason for thinking that it must be your loved ones actually doing well after your death and not just the thought that they will. 184 00:19:58.740 --> 00:20:01.740 Suppose that I give you a choice. 185 00:20:01.740 --> 00:20:07.740 You can either buy this life insurance policy for $100 or you can buy this magical pill for $50. 186 00:20:07.740 --> 00:20:13.740 The magical pill will make you believe that you have bought life insurance. 187 00:20:13.740 --> 00:20:19.740 So you have life insurance for $100 or the belief that you've bought life insurance for $50. 188 00:20:19.740 --> 00:20:25.740 Now, if your only reason for buying life insurance is to have the thought that your loved ones will do well after your death, 189 00:20:25.740 --> 00:20:31.740 that seems as though you'd have reason to prefer the pill to the life insurance. 190 00:20:31.740 --> 00:20:37.740 And therefore, we wouldn't be able to make sense of people if we're buying life insurance when they could buy the pill. 191 00:20:37.740 --> 00:20:44.740 You might say, well, there's a brief moment between your deciding to take the pill and your taking the pill, 192 00:20:44.740 --> 00:20:55.460 where you feel really awful, and that is, but you know, all right, you can imagine, 193 00:20:55.460 --> 00:20:59.380 you know, if that's all it is, though, you can imagine doing things to compensate for that, 194 00:20:59.380 --> 00:21:04.460 like, well, this pill not only gives you the thought, but also gives you, you know, 195 00:21:04.460 --> 00:21:06.100 five hours of intense pleasure or something. 196 00:21:06.100 --> 00:21:09.580 That'll make up for those five minutes of a guilty conscience. 197 00:21:09.580 --> 00:21:11.220 It seems as though that's not the way we think. 198 00:21:11.220 --> 00:21:15.540 So we do care about things we're not gonna experience. 199 00:21:15.540 --> 00:21:20.540 We care about things that will happen after we're dead 200 00:21:23.940 --> 00:21:25.380 to our loved ones. 201 00:21:25.380 --> 00:21:30.380 So therefore, our feelings, our emotions, 202 00:21:31.020 --> 00:21:34.700 our mental states of mind can't be our only goals. 203 00:21:34.700 --> 00:21:35.780 There's some of the things we care about. 204 00:21:35.780 --> 00:21:37.680 We want to feel good, but feeling good 205 00:21:37.680 --> 00:21:39.580 can't be the only thing we care about. 206 00:21:39.580 --> 00:21:50.580 And, you know, I think that's a plausible argument in its own right, but I think it's helped by consideration of this, of this nucleus problem issue. 207 00:21:50.580 --> 00:21:58.580 Because I think what nucleus problem suggests is that it doesn't make sense to think of a means being later than the end, 208 00:21:58.580 --> 00:22:04.580 which is what it would have to be if you thought that the loved one's actually doing well after death, it causes a pleasure. 209 00:22:04.580 --> 00:22:09.220 Well, maybe it's not that the loved one's doing well in the future causes my pleasure now, 210 00:22:09.220 --> 00:22:11.700 because that would be backwards causation and that would be weird, 211 00:22:11.700 --> 00:22:17.620 but can I not take pleasure now, not just in the thought that my loved ones will do well, 212 00:22:17.620 --> 00:22:20.740 but can I take pleasure now in when my loved one's actually doing well? 213 00:22:22.020 --> 00:22:29.060 Sure, I think you can, but in that case, I mean, I think that's a perfectly good way of talking, 214 00:22:29.060 --> 00:22:33.060 but in that case, it's not a means-and relationship. I think if I take pleasure 215 00:22:33.060 --> 00:22:37.060 If we're now in something happening in the future, it's not a means end. 216 00:22:37.060 --> 00:22:41.060 The thing happening in the future is not a means to my pleasure. 217 00:22:47.060 --> 00:22:53.060 Okay, here's another upshot, I think, of Newcomb's problem. 218 00:22:54.060 --> 00:22:56.060 I think it shows that there couldn't be... 219 00:22:56.060 --> 00:23:00.060 We can't coherently regard ourselves as being causally determined. 220 00:23:00.060 --> 00:23:06.060 We can't coherently think that everything we do is the necessary result of prior causes. 221 00:23:06.060 --> 00:23:11.060 Why not? Well, suppose determinism is true. Then what's the case? 222 00:23:11.060 --> 00:23:18.060 Then it's the case that my present actions necessarily co-vary with past states of the universe. 223 00:23:18.060 --> 00:23:27.060 If causal determinism is true, then there can't be any difference in the present except as a result of some difference in the past. 224 00:23:27.060 --> 00:23:33.560 I can't do something different now from what I'm doing unless the causes acting on me are different. 225 00:23:33.560 --> 00:23:38.560 And those causes couldn't be different unless the causes acting on them were different, and so on back indefinitely. 226 00:23:38.560 --> 00:23:51.060 So if determinism is true, then if I raise my hand now, that must mean that the past causal history of the universe was one way. 227 00:23:51.060 --> 00:23:56.060 And if I hadn't raised my hand just then, then the past causal history of the universe would have had to have been the other way. 228 00:23:56.060 --> 00:24:11.060 And so, raising my hand or not raising my hand necessarily covariates, if the termism is true, with things that are already passed and settled. 229 00:24:11.060 --> 00:24:20.060 Well, could we have preferences about things that are already passed and settled? 230 00:24:20.060 --> 00:24:24.060 Sure, you could say we wish things were this way rather than that. 231 00:24:24.060 --> 00:24:44.060 Well, what the one box argument shows is if you have preferences about things that are already passed and settled and those things co-vary with things that are under your control now, then you can thereby aim at making things one way or another in the past. 232 00:24:44.060 --> 00:25:01.060 So, for example, suppose that I would like it to be the case that the past causal history of the universe is one of a certain kind that would lead up to my clapping my hands right now. 233 00:25:01.060 --> 00:25:08.060 That's what I would like. Well, then by clapping my hands right now, I can make that the case seems paradoxical. 234 00:25:08.060 --> 00:25:12.220 paradoxical, but it seems as though if I want it to be the case that that's the causal chain that existed, 235 00:25:12.220 --> 00:25:19.600 well, if I clap my hands now, then that will be the causal chain that existed. And if I don't 236 00:25:19.600 --> 00:25:24.260 clap my hands now, then that won't be. So I have a preference between these two things, and which of 237 00:25:24.260 --> 00:25:28.860 them happens, necessarily covariates with that I clap my hands now, well then by clapping my hands now, 238 00:25:28.860 --> 00:25:37.980 I could aim at bringing about a certain past causal chain. But, according to the two box argument, 239 00:25:37.980 --> 00:25:47.980 I mean that doesn't make sense, so we can form preferences about past causal chains. 240 00:25:47.980 --> 00:25:56.980 If past causal chains co-varied with our present choices, then we both could and couldn't aim at changing those past things. 241 00:25:56.980 --> 00:25:59.980 We couldn't aim at changing the past things because of the past and settled. 242 00:25:59.980 --> 00:26:04.980 But we could aim at changing those past things because they necessarily co-variate with something we can choose. 243 00:26:04.980 --> 00:26:16.980 So, once again, we have an antinomy, and I'm claiming that the common assumption behind the antinomy is the assumption that our present choices necessarily co-vary with something past and settled. 244 00:26:16.980 --> 00:26:24.980 I think we cannot coherently regard our present choices as necessarily co-varying with something that's already past and settled. 245 00:26:24.980 --> 00:26:29.980 And so, I claim that means we cannot regard ourselves as causally determined. 246 00:26:29.980 --> 00:26:43.980 Okay, so there's a, that's the first paradox I wanted to talk about and the ways in which I think that that has an upshot for free will. 247 00:26:43.980 --> 00:26:53.980 What's the second paradox? Well, the second paradox is the so-called Socratic paradox. Remember I said that I thought one of these was a paradox and one wasn't. 248 00:26:53.980 --> 00:26:59.980 wasn't. In the strict sense, a paradox is in a case where it looks like you've got equally good 249 00:26:59.980 --> 00:27:06.740 arguments for mutually contradictory conclusions, and that's why I think Newcomb's problem is a 250 00:27:06.740 --> 00:27:13.580 paradox in that sense. Socratic paradox does not obviously seem like a case of that, but there's 251 00:27:13.580 --> 00:27:17.540 just this long, long tradition of calling it the Socratic paradox, and a much broader sense of 252 00:27:17.540 --> 00:27:22.300 paradox, where paradox is just, in any case, we've got an argument not necessarily for something 253 00:27:22.300 --> 00:27:27.300 Contradictory, but for something weird, or surprising, or implausible. 254 00:27:27.300 --> 00:27:32.300 Paradox literally means contrary to opinion. 255 00:27:32.300 --> 00:27:37.300 So any time you've got something contrary to seeming, so any time you've got some argument 256 00:27:37.300 --> 00:27:42.300 whose upshot is contrary to what seems to be the case, or contrary to 257 00:27:42.300 --> 00:27:47.300 what most people would have thought, that's a paradox in this much broader, looser sense of paradox. 258 00:27:47.300 --> 00:28:17.300 Paradox. You know, the first kind of paradox can never be true. It can never really be, you know, both sound arguments for a contradictory conclusion. So whenever you've got a paradox in that strict philosophical sense, you know, something must be wrong. However, when you have a paradox where the upshot is just, you know, something surprising, that doesn't necessarily mean it's wrong. It might be. It's certainly a reason to look at it closely. But a paradox in that sense 259 00:28:17.300 --> 00:28:21.780 between something counter-intuitive, something contrary to what you expected. 260 00:28:21.780 --> 00:28:24.580 Well, lots of paradoxes are true in that sense. 261 00:28:27.060 --> 00:28:30.340 Okay, what's the Socratic paradox, which, as you might guess, 262 00:28:30.340 --> 00:28:34.340 is named because it's found in the teachings of Socrates? 263 00:28:36.180 --> 00:28:40.420 Well, choosing something expresses a preference for that thing. 264 00:28:40.420 --> 00:28:42.740 If you choose something, if it's a real choice, 265 00:28:42.740 --> 00:28:45.380 not just some sort of random twitch or something, 266 00:28:45.380 --> 00:28:46.980 but if it's really a purposeful action, 267 00:28:46.980 --> 00:28:50.980 Choosing something, your choice expresses a preference for that thing. 268 00:28:53.980 --> 00:28:58.980 Having a preference for something expresses the judgment that it's worth choosing. 269 00:28:58.980 --> 00:29:01.980 Now, that's not uncontroversial. 270 00:29:01.980 --> 00:29:07.980 You might think, well, some of your preferences are just sort of urges that don't really express any kind of judgment. 271 00:29:09.980 --> 00:29:14.980 I talked a little bit yesterday about this idea of when you're deciding what to do, 272 00:29:14.980 --> 00:29:22.340 What to do, there's a sense of what you should do that just is what you're deciding then. 273 00:29:22.340 --> 00:29:26.180 Deciding what to do is deciding what you should do in a certain sense of shit. 274 00:29:26.180 --> 00:29:32.980 So I want to claim that when you choose something, there is something like a judgment that's 275 00:29:32.980 --> 00:29:33.980 expressed there. 276 00:29:33.980 --> 00:29:38.140 If it really is an action and not just a blind urge and if it's an action of a rational 277 00:29:38.140 --> 00:30:00.140 So I think all of that is right. So the famous upshot of it is that you can't choose something you think is not worth choosing, because to choose something expresses a preference, and expressing a preference expresses a judgment that the thing is worth choosing. 278 00:30:00.140 --> 00:30:10.140 But then that seems to have the upshot that no one can knowingly or voluntarily do wrong. 279 00:30:10.140 --> 00:30:13.140 And Socrates seems to say both those things. 280 00:30:13.140 --> 00:30:18.140 And that seems hard to accept. 281 00:30:18.140 --> 00:30:26.140 So first of all, I think there's a difference between the knowingly and the voluntarily. 282 00:30:26.140 --> 00:30:39.140 And to explain the difference, let me just say what I think Plato later changes the view from the one he got from Socrates. 283 00:30:40.140 --> 00:30:52.140 Plato thinks that you can be overcome by some sort of emotion or passion, which can then make you do something contrary to your judgment of what is the best thing to do. 284 00:30:52.140 --> 00:30:56.140 And so Plato abandons the view that you can never knowingly do what's wrong. 285 00:30:56.140 --> 00:31:00.140 But Plato still thinks that you can never voluntarily do what's wrong. 286 00:31:00.140 --> 00:31:03.140 At least that's the way I interpret Plato. 287 00:31:03.140 --> 00:31:08.140 So Socrates thinks you can never knowingly or voluntarily do wrong. 288 00:31:08.140 --> 00:31:14.140 Plato thinks that you can do wrong knowingly, because you can be dragged by your passions 289 00:31:14.140 --> 00:31:18.140 and your judgment is saying, stop, stop, help, but your passion just drags you on anyway. 290 00:31:18.140 --> 00:31:23.660 Then, since you're acting contrary to your best judgment, you're not acting voluntarily, but you are acting knowingly. 291 00:31:23.660 --> 00:31:30.660 Socrates, when he strongly thinks it doesn't make sense to think that you can even knowingly do wrong. 292 00:31:30.660 --> 00:31:36.660 If Socrates had known how Plato was going to develop his ideas after his death, I think Socrates would have said, 293 00:31:36.660 --> 00:31:40.160 wait, Plato, this notion of being dragged away by passion doesn't make sense. 294 00:31:40.160 --> 00:31:52.160 Because if you just mean that your limbs are taking over and suddenly it's like some sort of epileptic fit or something, 295 00:31:52.160 --> 00:31:56.160 that your arms and legs are suddenly doing something you don't want, well that's not an action. 296 00:31:56.160 --> 00:32:01.160 That's something happening in your body, but that's not an actual choice of yours. 297 00:32:01.160 --> 00:32:05.160 It's only a choice of yours if it expresses a judgment. 298 00:32:05.160 --> 00:32:09.160 So I think Socrates would want to cling to the stronger form of the claim. 299 00:32:09.160 --> 00:32:22.160 Now, I think that, on the one hand, I think Socrates has good reasons for these things, but on the other hand, this conclusion seems to clash with some things. 300 00:32:22.160 --> 00:32:34.160 First, it seems to clash with some obvious facts. We all know from experience, not only that we've done things that we knew we shouldn't be doing, or at least that we thought we shouldn't be doing, 301 00:32:34.160 --> 00:32:40.160 but that at the time we didn't feel as though it was just some passion completely overtaking us 302 00:32:40.160 --> 00:32:43.160 it seems as though we were choosing to do it 303 00:32:43.160 --> 00:32:46.160 so it seems as though it's a familiar feature of human experience 304 00:32:46.160 --> 00:32:50.160 that we do things that we don't really think are the best thing to do 305 00:32:50.160 --> 00:32:53.160 and yet somehow we do them 306 00:32:53.160 --> 00:32:59.160 but also the Socratic paradox seems to clash with free will 307 00:32:59.160 --> 00:33:03.160 it seems to imply causal determinism 308 00:33:03.160 --> 00:33:06.960 And I've just argued that that would not make sense. 309 00:33:06.960 --> 00:33:11.260 So, for example, it seems as though if you can never do anything 310 00:33:11.260 --> 00:33:16.360 contrary to your judgment of what's most worth doing well, 311 00:33:16.360 --> 00:33:21.160 then that means that I can't clap my hands right now 312 00:33:21.160 --> 00:33:23.760 unless I think that clapping my hands right now is the best thing to do. 313 00:33:23.760 --> 00:33:25.460 Neither I think that or I don't. 314 00:33:25.460 --> 00:33:27.060 And if I think it, then I'll do it. 315 00:33:27.060 --> 00:33:29.360 And if I think it's not, then I won't. 316 00:33:29.360 --> 00:33:35.360 And so it seems as though whatever my current beliefs and judgments are, already settles what actions I'm going to take. 317 00:33:35.360 --> 00:33:40.360 And so then it seems as though I don't have any free choice. 318 00:33:40.360 --> 00:33:48.360 So, what's the solution? 319 00:33:48.360 --> 00:33:53.360 Well, on the handout I start right off with Aristotle's solution, 320 00:33:53.360 --> 00:34:04.360 But I want to back up a bit and say a little bit about Xenophon's solution first, because Xenophon has a sort of a shorter solution. 321 00:34:04.360 --> 00:34:09.360 I'm not going to mention him. I should have his name. 322 00:34:09.360 --> 00:34:16.360 Xenophon was another pupil of Socrates, less famous than Plato. 323 00:34:16.360 --> 00:34:21.360 And here's his solution to the Socratic paradox. 324 00:34:21.360 --> 00:34:30.360 He thinks, yes, passion can overcome us and make us do something contrary to our judgment, 325 00:34:30.360 --> 00:34:36.360 but it's not as though when we act we're then doing something different from what we judge to be best. 326 00:34:36.360 --> 00:34:42.360 Zenithen thinks that what happens is that our passions confuse our judgment. 327 00:34:42.360 --> 00:34:45.360 So initially we have a clear judgment of what we ought to do. 328 00:34:45.360 --> 00:34:49.360 Then the passion comes and confuses us so that we no longer clearly see what to do. 329 00:34:49.360 --> 00:35:19.360 to do. And instead, the passion gets us to believe something else temporarily. And so temporarily, we go off and do this other thing on the basis of a different judgment. Then once the passion goes away, our judgment comes back and we say, oh, you know, I shouldn't have done that. So you have kind of a temporary amnesia or temporary confusion where you believe something different as a result of passion. Now, Xenophon's solution has certain advantages. It avoids what might seem 330 00:35:19.360 --> 00:35:24.360 It sounds like an excessive intellectualism on Socrates' part. 331 00:35:24.360 --> 00:35:30.360 A lot of people have objected that Socrates believed you always necessarily do whatever you think best. 332 00:35:30.360 --> 00:35:32.360 It makes us sound purely intellectual. 333 00:35:32.360 --> 00:35:36.360 It ignores the fact that we have emotions and passions and impulses that can influence us. 334 00:35:36.360 --> 00:35:39.360 And Xenophon's solutions solves that. 335 00:35:39.360 --> 00:35:45.360 Xenophon's solution says, no, no, you can have the Socratic paradox even if you grant that we can be influenced by our emotions. 336 00:35:45.360 --> 00:35:59.160 I think that's fine as far as it goes, but it's not good enough, I think, because although 337 00:35:59.160 --> 00:36:03.560 it may cover the fact, it may cover the cases where you're blinded by passion and just can 338 00:36:03.560 --> 00:36:07.060 no longer see what you ought to be doing. 339 00:36:07.060 --> 00:36:12.160 It does cover the cases when, as you're doing the right thing, your wrong thing, you're 340 00:36:12.160 --> 00:36:15.160 If you're saying to yourself, I know I shouldn't be doing this, 341 00:36:15.160 --> 00:36:18.160 because then it seems as though the judgment hasn't been obscured, it's still there. 342 00:36:22.160 --> 00:36:29.160 Well, I think Aristotle, although I think he's probably influenced by Xenophon's solution, 343 00:36:29.160 --> 00:36:33.160 if it is a Xenophon's solution, I mean, Xenophon may have just gotten that from Socrates, we don't know. 344 00:36:33.160 --> 00:36:41.160 But anyway, I think Aristotle builds on Xenophon's solution and then has something better. 345 00:36:41.160 --> 00:36:46.160 Aristotle distinguishes between possessing a judgment and exercising a judgment. 346 00:36:46.160 --> 00:36:53.160 First of all, this allows for the fact that you might have more than one judgment. 347 00:36:53.160 --> 00:36:58.160 I might have a judgment that I should do something and also have another judgment that I shouldn't do it. 348 00:36:58.160 --> 00:37:04.160 They could both be floating around in my mind, but which one am I acting on? 349 00:37:04.160 --> 00:37:12.160 I think this is analogous to the distinction in Austrian economics between praxeology and themology. 350 00:37:20.160 --> 00:37:24.160 This is the distinction that Mises makes. 351 00:37:24.160 --> 00:37:31.160 Praxeology has to do with the analysis of action 352 00:37:31.160 --> 00:37:39.160 and themology has to do with, more broadly, psychological causes underlying or influencing action and so forth. 353 00:37:39.160 --> 00:37:48.160 And part of Mises' point is that you can have conflicting preferences in the themological sense. 354 00:37:48.160 --> 00:37:55.660 That is to say, you know, I can have some desire for one thing and also some desire for something else incompatibly. 355 00:37:55.660 --> 00:37:58.960 You can't have conflicting desires in the praxeological sense. 356 00:37:58.960 --> 00:38:04.260 Praxeological desires are desires realized in action. 357 00:38:04.260 --> 00:38:07.160 They're the desires you're actually acting on. 358 00:38:07.160 --> 00:38:14.760 And, you know, so I might, you know, maybe I have some reason to take a drink now. 359 00:38:14.760 --> 00:38:16.160 I'm thirsty. 360 00:38:16.160 --> 00:38:18.160 I also have some reason not to take a drink now. 361 00:38:18.160 --> 00:38:20.160 It would interrupt my lecture. 362 00:38:20.160 --> 00:38:22.160 I may have a slight worry that 363 00:38:22.160 --> 00:38:24.160 someone just put poison in there, 364 00:38:24.160 --> 00:38:26.160 whatever. 365 00:38:26.160 --> 00:38:28.160 So I've got some reason to take a drink and some reason not to. 366 00:38:28.160 --> 00:38:30.160 So you could say, well, I've got some desire to 367 00:38:30.160 --> 00:38:32.160 and some desire not to. 368 00:38:32.160 --> 00:38:34.160 But when I'm actually 369 00:38:34.160 --> 00:38:36.160 drinking, you know, I can't 370 00:38:36.160 --> 00:38:38.160 enact both those desires at the same time. 371 00:38:42.160 --> 00:38:44.160 When I'm actually drinking, 372 00:38:44.160 --> 00:38:47.520 There's a preference I have that's built into the desire. 373 00:38:47.520 --> 00:38:51.280 And those are the preferences that economics deals with, 374 00:38:51.280 --> 00:38:53.800 the preferences that you actually are acting on. 375 00:38:53.800 --> 00:38:56.800 You may have, in another sense of preferences, 376 00:38:56.800 --> 00:38:58.280 various sorts of desires floating around 377 00:38:58.280 --> 00:38:59.580 that you're not acting on. 378 00:39:00.940 --> 00:39:04.280 Now, it's an interesting question whether there is some sense 379 00:39:04.280 --> 00:39:07.000 in which you can actually simultaneously act in some way 380 00:39:07.000 --> 00:39:08.120 on opposite preferences. 381 00:39:08.120 --> 00:39:11.000 And I've gone back and forth a bit on this 382 00:39:11.000 --> 00:39:13.720 and written about this, but I don't think it matters right now. 383 00:39:14.160 --> 00:39:29.160 So, Aristotle's point is that what Aristotle thinks is happening is that when you do something you know is wrong, 384 00:39:29.160 --> 00:39:34.160 what happens is you've got your better, wiser judgment that's saying don't do that. 385 00:39:34.160 --> 00:39:37.160 But you've also got another judgment that says do it. 386 00:39:37.160 --> 00:39:41.160 And he says, you know, maybe your reason is telling you don't do it and appetite is telling you do it. 387 00:39:41.160 --> 00:39:48.440 But the appetite is the judgment too. Appetite is not just some blind urge. Appetite presents, 388 00:39:48.440 --> 00:39:52.760 the thirst presents the water as choice worthy. Thirst doesn't just, you know, start moving my 389 00:39:52.760 --> 00:40:01.960 hand to the glass robotically. My thirst presents the water to me under the aspect of choice worthy. 390 00:40:01.960 --> 00:40:08.280 You know, try me, drink me, drink me. Whereas my reason says, you know, I know these people are 391 00:40:08.280 --> 00:40:20.280 So I've got both judgments, but then whichever one I act on, that's the judgment that I'm exercising, the judgment that's actually operative. 392 00:40:20.280 --> 00:40:28.280 But the judgment that I'm not acting on could still be around. I could comment on it, I could give utterance to it. 393 00:40:28.280 --> 00:40:32.280 So when you say, I shouldn't be doing this, as you go and do it, 394 00:40:32.280 --> 00:40:36.280 in fact, 395 00:40:36.280 --> 00:40:40.280 well, when you do something, saying at the same time, 396 00:40:40.280 --> 00:40:44.280 I know I shouldn't be doing this, then that shows that the 397 00:40:44.280 --> 00:40:48.280 judgment that you shouldn't be doing it is around, it's along for the ride. 398 00:40:48.280 --> 00:40:52.280 You can notice it, you can give voice to it, 399 00:40:52.280 --> 00:40:56.280 but it's not the one that you are putting into motion. 400 00:40:56.280 --> 00:41:26.280 Now, this helps somewhat with the Socratic paradox, because the reason the Socratic paradox sounds so weird is that it gives the impression, and Socrates may have thought this, I don't know whether he did or not, but certainly the way that it's often put, it gives the impression that when you act, you're always acting on this one judgment you have. You've got one judgment whether the thing is good or bad, and you act on that judgment. 401 00:41:26.280 --> 00:41:34.200 Judgment. But I think a sort of a milder way of putting it is that whenever you're acting, your 402 00:41:34.200 --> 00:41:39.080 action is expressing a judgment that's worth doing. But it doesn't mean you don't have a contrary 403 00:41:39.080 --> 00:41:43.680 judgment. The contrary judgment might be better grounded, the contrary judgment might be the one 404 00:41:43.680 --> 00:41:51.000 that you more usually have or whatever, but they could both be present. And so therefore this doesn't 405 00:41:51.000 --> 00:41:56.500 can imply that whenever you act it's somehow necessarily the result of your judgment 406 00:41:56.500 --> 00:41:59.500 because you might have contrary judgments you can pick which one to act on. 407 00:42:01.500 --> 00:42:09.500 A second point is that some people thought that praxeology itself implies determinism. 408 00:42:09.500 --> 00:42:13.500 Because they thought, well look everything we do is motivated. 409 00:42:14.000 --> 00:42:18.500 We don't do any random actions. If they're random they're not actions. 410 00:42:18.500 --> 00:42:21.100 Everything we do is the result of our desires. 411 00:42:21.100 --> 00:42:24.500 You do it because you desire that more than you desire anything else. 412 00:42:24.500 --> 00:42:28.300 Our actions express our strongest preference. 413 00:42:28.300 --> 00:42:33.900 You might think that just straightforwardly implies some sort of psychological determinism. 414 00:42:33.900 --> 00:42:40.300 But this very point that both Aristotle and Mises stress in different contexts 415 00:42:40.300 --> 00:42:48.100 is that the desires we're talking about are desires that are built into the action. 416 00:42:48.100 --> 00:42:51.300 They're not some prior cause of the action. 417 00:42:51.300 --> 00:42:54.900 It's not as though at time one, I've got a desire, 418 00:42:54.900 --> 00:42:57.580 and then at time two, my limbs start moving. 419 00:42:57.580 --> 00:43:01.340 And the reason that counts as being motivated by that desire 420 00:43:01.340 --> 00:43:05.940 is the desire somehow kicked it off, the way that you knock over one domino. 421 00:43:05.940 --> 00:43:10.900 And if a domino, 10 dominoes down, knocks it, falls over, 422 00:43:10.900 --> 00:43:12.940 it's because it was caused by this earlier one. 423 00:43:12.940 --> 00:43:15.780 Now, when you're acting, the desire is present right then. 424 00:43:15.780 --> 00:43:17.300 The desire is built into the action. 425 00:43:17.300 --> 00:43:22.660 Desire is an aspect of the action. Desire acting on is an aspect of the action. 426 00:43:22.660 --> 00:43:29.140 Now, of course, it ordinarily isn't going to make much sense for you to act on some desire now 427 00:43:29.140 --> 00:43:33.140 unless you had that desire earlier. So you probably have to have the desire earlier, 428 00:43:33.140 --> 00:43:37.860 at least in most cases. You have to have the desire earlier in order to be able to act on it now. 429 00:43:37.860 --> 00:43:46.340 But still, what makes my action of drinking be motivated by thirst 430 00:43:46.340 --> 00:43:51.860 It's not that I had thirst earlier than I drank. That's not the decisive thing. 431 00:43:51.860 --> 00:43:59.500 What makes it is that in drinking, I'm choosing it as a thirst quencher rather than as something else. 432 00:43:59.500 --> 00:44:04.660 I'm thinking of the following example. Suppose that you're out in the middle of the Sahara Desert 433 00:44:04.660 --> 00:44:08.860 and you've been lost for days, you had no water, you're desperate. 434 00:44:08.860 --> 00:44:12.740 Suddenly someone comes along and offers you some water and says, 435 00:44:12.740 --> 00:44:16.340 I'll pay you $50 to drink this water. 436 00:44:16.340 --> 00:44:18.540 Well, now, you've got two reasons to drink it. 437 00:44:18.540 --> 00:44:22.740 One is you're dying of thirst, and the other reason is it's kind of nice to have $50. 438 00:44:22.740 --> 00:44:29.340 But at that moment, probably, unless you're very cool-headed, 439 00:44:29.340 --> 00:44:35.340 probably the $50 is not even something you're going to pay attention to at that moment. 440 00:44:35.340 --> 00:44:38.240 You're just going to grab the water. 441 00:44:38.240 --> 00:44:41.940 You're going to choose it as a thirst quencher, not as a money-bringer. 442 00:44:41.940 --> 00:44:46.100 And you could, if you had more cool-headed, you might choose it as both, but probably not. 443 00:44:46.100 --> 00:44:53.380 But anyway, which one you do it for is a matter of something that's built into the action. 444 00:44:53.380 --> 00:44:58.820 It's not as though first you have a desire over here and then some mechanical motion of limbs follows later. 445 00:44:58.820 --> 00:45:06.260 That's not what action is like. The desire is actually simultaneous with the action. 446 00:45:06.260 --> 00:45:09.620 And that means that although it may be true that you always act on your strongest motive, 447 00:45:09.620 --> 00:45:15.060 It doesn't mean that I'm always act now on whatever my strongest motive was a second ago. 448 00:45:15.060 --> 00:45:18.980 I might not have decided. We might sort of decide what our strongest motive is going to be. 449 00:45:18.980 --> 00:45:24.420 I might have a lot of different motives. And whatever one I'm acting on at the moment 450 00:45:25.060 --> 00:45:27.620 is the dominant one. But it's not clear that it's being the dominant one. 451 00:45:27.620 --> 00:45:31.460 It's just a separate fact from the fact that it's the one I'm acting on. 452 00:45:31.460 --> 00:45:56.460 Now, while this isn't in the notes, another thing that might be worth mentioning is, Mises seems to have thought that praxeology implies determinism for a different reason, not just because you might think that our actions are caused by our desires. 453 00:45:56.460 --> 00:46:07.960 But Mises said that, one way he put it, is that the concepts of cause and effect are implied in action. 454 00:46:07.960 --> 00:46:15.960 Because whenever I choose something as a means to an end, I must assume that the means is causally related to the end. 455 00:46:15.960 --> 00:46:27.660 It wouldn't make any sense for me to choose raising this glass to my lips as a way of quenching my thirst, 456 00:46:27.660 --> 00:46:31.460 unless I thought that raising the glass to my lips would cause the quenching of thirst. 457 00:46:31.460 --> 00:46:42.960 If I thought there was a purely random relationship between raising the glass to my lips and quenching my thirst, 458 00:46:42.960 --> 00:46:59.960 Well, I think that's true enough, but the question is whether cause and effect is the same thing as cause of determinism. 459 00:46:59.960 --> 00:47:07.960 And I don't think so. I think that we can have cases of cause and effect that aren't deterministic. 460 00:47:07.960 --> 00:47:16.960 And let me borrow an example from Elizabeth Anscombe, I believe it was Anscombe I saw this in. 461 00:47:16.960 --> 00:47:25.960 Suppose that, you know, grant for the sake of argument that quantum physics is really indeterministic. 462 00:47:25.960 --> 00:47:27.960 It's not just a matter of our knowledge, but it really is. 463 00:47:27.960 --> 00:47:30.960 Alright, so let's say I've got a chunk of radioactive material. 464 00:47:30.960 --> 00:47:36.960 If physics is indeterministic at the quantum level, then what that means is that this chunk of radioactive material 465 00:47:36.960 --> 00:47:45.960 is emitting radioactive particles at a rate that's in principle unpredictable. 466 00:47:45.960 --> 00:47:49.960 You know, just radioactive particles zap out every once in a while. 467 00:47:49.960 --> 00:47:54.960 Now, you know, we have a pretty good idea of what the probabilities are. 468 00:47:54.960 --> 00:47:57.960 We have a pretty good basis for saying, you know, within the next second, 469 00:47:57.960 --> 00:48:05.960 you know, it'll spew out, you know, approximately 10 particles in the next second or something like that. 470 00:48:05.960 --> 00:48:13.960 But we cannot predict in principle. It's not settled in principle exactly how many or exactly what points during the second will happen. 471 00:48:13.960 --> 00:48:23.960 Alright, so here's Anscombe's example. I take this piece of radioactive material and I hook it up to a Geiger counter, which will then click. 472 00:48:23.960 --> 00:48:31.960 Every time that the radioactive material sends out a radioactive particle, the Geiger counter will click, registering that it's received that particle. 473 00:48:31.960 --> 00:48:42.960 And therefore the clicking of this will be random. We know that it's likely that there will be about 10 clicks a second, but we can't predict exactly when the click will come. 474 00:48:42.960 --> 00:48:48.960 Now I take this Geiger counter with the radioactive material attached to it and hook the whole thing up to a bomb. 475 00:48:48.960 --> 00:49:07.960 The bomb. And I rigged the bomb so that the bomb will go off if a click comes between, you know, one-fifteenth and one-sixteenth of the way into the second. 476 00:49:07.960 --> 00:49:17.960 If between one-fifteenth of a second and one-sixteenth of a second, the Geiger counter clicks once or more, then the bomb will go off, otherwise not. 477 00:49:18.960 --> 00:49:20.960 Then I stick this whole thing under your bed. 478 00:49:21.960 --> 00:49:36.960 And it just so happens that the radioactive material does send out a particle at the right time, it hits the Geiger counter, the Geiger counter clicks, the bomb goes off, and you are blown to bits. 479 00:49:37.960 --> 00:49:43.960 And so, you know, I'm taken to court and accused of killing you, and I say, wait a second. 480 00:49:43.960 --> 00:49:56.960 I didn't cause this person's death because the causal chain leading from me to the person's death had some randomness in it. 481 00:49:56.960 --> 00:50:09.960 There was this random thing where I couldn't guarantee whether the Gregor Conner were quick at the crucial moment or not. 482 00:50:09.960 --> 00:50:14.960 That was purely random. And so you can't accuse me of causing the death. 483 00:50:14.960 --> 00:50:19.960 Maybe there was a good probability that it would cause your death. 484 00:50:19.960 --> 00:50:25.360 You know, maybe there was, you know, you can, you know, you can make it if it was one chance in ten, 485 00:50:25.360 --> 00:50:28.860 or you can make it if it was nine chances in ten, you know, spin the story however you want, 486 00:50:28.860 --> 00:50:32.060 just as long as it's a fairly decent probability that this would kill you. 487 00:50:32.060 --> 00:50:34.360 I mean, if it's one in a million, then you might think, all right, I'm off the hook. 488 00:50:34.360 --> 00:50:40.360 But it's like, if it's one in ten, you're inclined to think that, you know, I'm guilty of killing you. 489 00:50:40.360 --> 00:50:42.960 And, you know, they might say, no, I haven't caused your death. 490 00:50:42.960 --> 00:50:46.760 Well, what Asuka wants to say is, it's plausible to say that I have caused your death. 491 00:50:46.760 --> 00:51:00.260 And so she wants to say that, therefore, the fact that there is some indeterminism in the course of the causality doesn't change the fact that it is causality. 492 00:51:00.260 --> 00:51:09.560 X causes Y, you know, if I increase the probability of something, and then the upshot is that it happens, 493 00:51:09.560 --> 00:51:15.360 it seems as if I've made a causal contribution to it, even if ultimately nothing necessitated it. 494 00:51:15.360 --> 00:51:26.920 So, I think Mises is right that we can't view the world in terms of means and ends unless we view it in terms of there being causal regularities. 495 00:51:26.920 --> 00:51:32.560 If it's just sort of random what will happen after I raise the glass to my lips, just sort of completely unpredictable, 496 00:51:32.560 --> 00:51:39.960 then I couldn't coherently regard raising the glass to my lips as a means to quenching my thirst. 497 00:51:39.960 --> 00:51:46.520 But if I think that it's overwhelmingly probable that raising the glass tinted lips will quench my thirst, 498 00:51:46.520 --> 00:51:53.800 and if therefore by raising the glass tinted lips I can render it overwhelmingly probable that my thirst will be quenched, 499 00:51:53.800 --> 00:52:04.640 then I can aim at it, since, after all, our actions generally don't guarantee that the result is going to happen anyway. 500 00:52:04.640 --> 00:52:10.640 Okay, I think what's a further upshot of all this? 501 00:52:10.640 --> 00:52:19.640 Well, one thing I think it shows is that our patterns of preference cannot be innate. 502 00:52:19.640 --> 00:52:20.640 Why is that? 503 00:52:20.640 --> 00:52:24.640 Well, let's first talk about the preferences in the praxeological sense. 504 00:52:24.640 --> 00:52:27.640 That is, the preferences that are embodied in action. 505 00:52:27.640 --> 00:52:31.640 The preferences in some sense only exist insofar as we're acting on them at that moment. 506 00:52:31.640 --> 00:52:38.760 Well, if we necessarily have to regard our actions as free, and I claim we do, because 507 00:52:38.760 --> 00:52:42.560 otherwise we have to regard things we can choose as co-varying with things that are 508 00:52:42.560 --> 00:52:48.680 settled, and I'm claiming that that's incoherent, if we regard our actions as free, you know, 509 00:52:48.680 --> 00:52:57.200 we're free, you know, so I'm free either to, you know, to drink the water or not, well, 510 00:52:57.200 --> 00:53:04.200 And drinking the water embodies, if I drank the water it would embody a preference for clenching my thirst. 511 00:53:04.200 --> 00:53:10.200 If I didn't drink the water it would embody a preference for not being poisoned by all you guys. 512 00:53:10.200 --> 00:53:14.200 Well then I can control which of those preferences exist in the praxeological sense, 513 00:53:14.200 --> 00:53:17.200 where the preference exists only so long as it's in the action. 514 00:53:17.200 --> 00:53:21.200 I can determine which of those preferences I have just by choosing whether to drink or not. 515 00:53:21.200 --> 00:53:28.200 So, in that sense, it can't already be settled what my preferences are going to be. 516 00:53:28.200 --> 00:53:37.200 So, if I've got free will, then I can control those preferences. 517 00:53:37.200 --> 00:53:39.200 But what are preferences in the themological sense? 518 00:53:39.200 --> 00:53:42.200 In the sense that we're not talking about preferences embodied in action, 519 00:53:42.200 --> 00:53:48.200 but just the various desires and impulses that you happen to have. 520 00:53:48.200 --> 00:54:08.200 Well, those preferences exist when you're not acting on them but I want to claim that even though the preferences you have that you don't act on don't depend on any specific actions. 521 00:54:08.200 --> 00:54:16.200 I think they do depend on patterns of action generally. Here's why. 522 00:54:16.200 --> 00:54:28.200 I think part of what it means for you to have a certain psychological state has to have something to do with how you actually apply it. 523 00:54:28.200 --> 00:54:32.200 And here I want to borrow a message from an example from Wittgenstein. 524 00:54:32.200 --> 00:54:41.200 Wittgenstein gives this example of this guy who is allegedly planning something and you're wondering what he's planning. 525 00:54:41.200 --> 00:54:49.440 I'm planning. You say, well, no one can know what he's planning because you can't look into his head and see his plans. 526 00:54:49.440 --> 00:54:56.520 So Wittgenstein asks, all right, suppose he wrote his plans down. Then would you know his plans? 527 00:54:56.520 --> 00:55:00.120 Well, suppose he writes all his plans down but he doesn't act on them. 528 00:55:00.120 --> 00:55:04.100 He writes down, you know, I'm planning to do this, I'm planning to do that, I'm planning the next thing, 529 00:55:04.100 --> 00:55:07.720 and he just sits there, doesn't do any of those things. 530 00:55:07.720 --> 00:55:13.720 It's not just occasionally, you know, when you plan something, it doesn't do it, but it's sort of systematically. 531 00:55:13.720 --> 00:55:17.720 He never does any of these things on this list. 532 00:55:17.720 --> 00:55:22.720 Well, it seems like we'd have a hard time saying that these really are his plans. 533 00:55:22.720 --> 00:55:27.720 They might look like plans, maybe they have the word plans written in big red letters at the top, 534 00:55:27.720 --> 00:55:33.720 but who knows what they are, but maybe there are things he's fantasizing about, 535 00:55:33.720 --> 00:55:38.720 If he's advertising about, maybe he's writing some sort of fiction or something, whatever it is, 536 00:55:38.720 --> 00:55:41.720 it seems like they're not his plans if he never acts on them. 537 00:55:41.720 --> 00:55:45.720 If he only just sometimes doesn't act on them, you think, oh well, we know how that is. 538 00:55:45.720 --> 00:55:50.720 But if he just never acts on any of them, or hardly ever, you just think, oh, it's not plans going on. 539 00:55:50.720 --> 00:55:54.720 All right, so Wittgenstein says, all right, now suppose you could peer into his head, 540 00:55:54.720 --> 00:55:57.720 and there you see the plans written in his mind. 541 00:55:57.720 --> 00:56:02.720 Also, big red letters saying plans at the top, but once again he doesn't act on them. 542 00:56:02.720 --> 00:56:05.720 How does peering into his mind, how does that help? 543 00:56:05.720 --> 00:56:13.720 I think part of Wittgenstein's moral there is that a mental state doesn't count as a plan unless you actually act on it. 544 00:56:13.720 --> 00:56:17.720 Now again, you don't have to act on it all the time. 545 00:56:17.720 --> 00:56:20.720 You don't have to act on it consistently. 546 00:56:20.720 --> 00:56:23.720 But if you just never act on it, it just doesn't count as a plan. 547 00:56:23.720 --> 00:56:30.720 And more broadly, you have some sort of desire. 548 00:56:30.720 --> 00:56:33.720 There are lots of desires that we're not acting on all the time, 549 00:56:33.720 --> 00:56:37.720 but if you have some sort of desire and you never act on it, 550 00:56:37.720 --> 00:56:40.720 then it's doubtful that you really have that desire. 551 00:56:40.720 --> 00:56:44.720 Now, of course, what acting on a desire means can be tricky. 552 00:56:44.720 --> 00:56:48.720 There are cases where you desire to do some particular action, 553 00:56:48.720 --> 00:56:50.720 and you don't ever do that action, 554 00:56:50.720 --> 00:56:56.720 but nevertheless you might do something else that somehow shows that you have that desire. 555 00:56:56.720 --> 00:57:12.720 So, probably would be too strong to say that, you know, so, you know, maybe I want to kill my father and I never actually kill my father, but nevertheless the fact that, you know, other things I do show that I would like to kill my father or something like that. 556 00:57:12.720 --> 00:57:35.720 So we can allow for there being desires you never act on, but nevertheless some will have to express themselves in their actions, just as you might have plans you never acted on, but you see either you're trying to act on them or you're doing somehow or other it has to express itself in action in order even to count as a desire in the themological sense. 557 00:57:35.720 --> 00:57:48.720 So even tendencies or impulses are various things we have. If we have free will, we don't have to act on them, but you could say, well, you've always got them, even if you don't act on them. 558 00:57:48.720 --> 00:58:01.720 But if you have enough control of your actions, you can avoid ever acting on those tendencies. And if you consistently don't act on the tendencies, then it seems to weaken the case for saying that you've got them anymore. 559 00:58:01.720 --> 00:58:06.720 All right, a couple of morals of those. 560 00:58:06.720 --> 00:58:13.720 First moral, you might think that certain facts about your preferences are genetically determined. 561 00:58:13.720 --> 00:58:21.720 People thought, for example, that someone's degree of time preference or their degree of risk aversion or something like that could be innate. 562 00:58:21.720 --> 00:58:28.720 I'm going to argue that we can't consistently think that those are innate 563 00:58:28.720 --> 00:58:39.720 because those are preferences that are embodied in action and if we have free will, we can control our actions. 564 00:58:39.720 --> 00:58:45.720 If we can control the specific actions, then ultimately we can control the overall pattern of actions. 565 00:58:45.720 --> 00:58:47.720 Someone might say, well, maybe not. 566 00:58:47.720 --> 00:58:59.720 Maybe not. Maybe, you know, suppose that any, you know, suppose for example that, you know, let's say that being an alcoholic means that, well I'll simplify it a bit. 567 00:58:59.720 --> 00:59:07.720 Being alcoholic means that you, you know, that you usually don't turn down a drink. That's not really what being alcoholic means, but let's simplify it. 568 00:59:07.720 --> 00:59:11.720 Being an alcoholic, part of what it means I'll say is you usually will not turn down a drink. 569 00:59:11.720 --> 00:59:16.720 You might say, well you can turn down any particular drink. You have free will to turn down any particular drink. 570 00:59:16.720 --> 00:59:23.720 But you can't consistently turn them all down, because you're just stuck with this preference that you will accept most drinks. 571 00:59:23.720 --> 00:59:28.720 So although you have control of your individual actions, you don't have control over the overall pattern. 572 00:59:28.720 --> 00:59:33.720 But it seems to me that if you have control of your individual actions, you do have control over the overall pattern. 573 00:59:33.720 --> 00:59:39.720 You know, because, you know, suppose that, you know, suppose that you're going to be offered ten drinks today, 574 00:59:39.720 --> 00:59:45.720 and suppose that you've turned down the first four, exercising your free will. 575 00:59:45.720 --> 01:00:05.720 Well, you still have free choice about the next one. You can deliberate whether to accept it or not, and your deliberation is going to be, you have to regard the choice as not already settled in order to deliberate about it, so you have to regard yourself as free to make that choice. 576 01:00:05.720 --> 01:00:13.720 You're free to make that choice, and you're free to turn down the fifth drink, you're free to turn down the sixth drink, and so on, and so you could, in principle, turn them all down. 577 01:00:13.720 --> 01:00:17.720 In which case, what happens to this alleged pattern that you're stuck with? 578 01:00:17.720 --> 01:00:25.720 So I think that if you're free to choose individual actions, then you also have freedom to choose your overall pattern of actions. 579 01:00:25.720 --> 01:00:38.720 And if you're free to do that, then if you do that consistently enough, then you don't really count as being stuck with a particular pattern of preferences. 580 01:00:38.720 --> 01:00:44.600 at Time Preference, Risk Aversion, Alcoholism, or anything else. 581 01:00:48.840 --> 01:00:55.920 Second moral is that one crucial feature of Aristotelian ethics, the notion of habituation, 582 01:00:55.920 --> 01:00:59.200 is, I'm going to argue, a priori. 583 01:00:59.200 --> 01:01:01.040 Here's why. 584 01:01:01.040 --> 01:01:08.400 Aristotle thinks that it becomes easier and easier to be virtuous because the more often 585 01:01:08.400 --> 01:01:11.880 When you do virtuous things, the easier it becomes to do it next time. 586 01:01:11.880 --> 01:01:15.560 You get, you know, your impulses to do the virtuous thing get stronger, your impulses 587 01:01:15.560 --> 01:01:17.040 to do the under-virtuous thing gets weaker. 588 01:01:17.040 --> 01:01:21.240 Of course, you also thought it's true the other way, in the case of bad habits. 589 01:01:21.240 --> 01:01:25.400 The more often you do the wrong thing, you can start strengthening your impulses to do 590 01:01:25.400 --> 01:01:28.080 the wrong thing and weakening your impulses to do the right thing. 591 01:01:28.080 --> 01:01:33.040 And so, that situation is a crucial part of versatility in ethics. 592 01:01:33.040 --> 01:01:35.200 You might think it's a sort of crucial part of ordinary common sense. 593 01:01:35.200 --> 01:01:39.200 But you might think, well, that just happens to be a psychological fact about us. 594 01:01:39.200 --> 01:01:42.200 You might think, well, we might have turned out to be different. 595 01:01:42.200 --> 01:01:47.200 We might have turned out to be the sorts of beings for whom habituation just doesn't apply, 596 01:01:47.200 --> 01:01:53.200 where no matter how often you do something, your impulses to the opposite are never any weaker. 597 01:01:53.200 --> 01:02:01.200 But I think that these free will considerations suggest that it's not just a happenstance. 598 01:02:01.200 --> 01:02:04.200 It's not just an accidental psychological fact of us. 599 01:02:04.200 --> 01:02:09.200 I think that any rational agent would have to be a rational agent to whom habituation applies. 600 01:02:09.200 --> 01:02:15.200 Or to put it another way, only beings capable of habituation could be rational agents. 601 01:02:15.200 --> 01:02:21.200 Here's why. If it's really true that by changing our actions we change our tendencies, 602 01:02:21.200 --> 01:02:27.200 because you don't count as having a tendency unless you act on it often enough, in some sense of act on it, 603 01:02:27.200 --> 01:02:34.200 And that suggests that it must be true that by changing our actions, we can change our tendencies. 604 01:02:36.200 --> 01:02:39.200 Okay, that's all about that. Yes. 605 01:02:40.200 --> 01:02:49.200 I agree that we can't change the past, but it seems to me that it doesn't fall from that, that we can't affect the past. 606 01:02:49.200 --> 01:02:58.200 Suppose that the predictor works through direct pre-submissions, and let's say that the predictor 607 01:02:58.200 --> 01:03:07.200 either puts the money in the box or left it empty 24 hours ago, then it seems to me that by your choice now, 608 01:03:07.200 --> 01:03:17.200 you can either bring it about that the predictor puts the money in there 24 hours ago, or bring it about that you left it empty. 609 01:03:17.200 --> 01:03:30.200 And so it seems to me then that the two boxers claim that in the relevant sense the money either there or it's not isn't true. 610 01:03:30.200 --> 01:03:40.200 You said that it's passed and settled, and I agree that it's passed, but it's not clear to me that it's settled. 611 01:03:40.200 --> 01:03:51.200 The predictor putting the money in there or not is temporally first, chronologically first, 612 01:03:51.200 --> 01:03:55.200 but it seems to me that your choice is logically first. 613 01:03:55.200 --> 01:04:01.200 Now, given that the predictor putting it in there or not is temporally first, 614 01:04:01.200 --> 01:04:07.200 obviously there's some sense in which the money is either there or it's not. 615 01:04:07.200 --> 01:04:14.560 So let's suppose that the million dollars have been there, but it seems to me that you're 616 01:04:14.560 --> 01:04:24.480 still free to take only the second box, and if you do, it's not that the money will vanish, 617 01:04:24.480 --> 01:04:33.360 rather if you do, the money will never have been there, that you might say, well, given 618 01:04:33.360 --> 01:05:02.360 So the question is this, suppose that the predictor rather than predicting on the basis of an analysis of my mental state at the time of the prediction actually has precognitive powers and peers into the future. 619 01:05:02.360 --> 01:05:12.360 So that in some sense the predictor's decision is caused by my future choice. 620 01:05:12.360 --> 01:05:20.360 My choice in the future is seen by the predictor and that causes the predictor to think a certain way and to make a choice. 621 01:05:20.360 --> 01:05:23.360 So you actually have a kind of backward causation or something like that. 622 01:05:23.360 --> 01:05:28.360 So the question is whether the money is already there. 623 01:05:28.360 --> 01:05:46.360 Well, the money being there is temporarily prior to my choice, but the suggestion would be that my choice is somehow logically or causally or something prior to the money being there, because if I make a different choice, that will make it the case that the predictor all along made a different choice. 624 01:05:46.360 --> 01:05:53.360 Now this very issue got debated by the late scholastics. 625 01:05:53.360 --> 01:05:58.360 There was a big dispute about whether God could change the past, 626 01:05:58.360 --> 01:06:05.360 or whether God is looking into the future and seeing things and so forth. 627 01:06:05.360 --> 01:06:15.360 And this whole dispute about, you know, Occam said something like this in order to water down the doctrine of predestination 628 01:06:15.360 --> 01:06:36.440 nation by saying, well, you can, you know, God predestines only those that he sees really are predestined to heaven, only those who sees really are going to merit salvation and predestined to hell, only those who are, who foresees are really going to merit damnation. 629 01:06:36.440 --> 01:06:42.840 Now, I don't think that the notion of backward causation makes sense. 630 01:06:42.840 --> 01:06:49.740 I don't think that the notion of literal precognition of the way the Medievalists put it was knowing 631 01:06:49.740 --> 01:06:54.480 the future in itself as opposed to knowing the future in its causes, that is, not inferring 632 01:06:54.480 --> 01:07:00.200 the future from its causes for someone directly seeing some future thing that is undetermined 633 01:07:00.200 --> 01:07:01.200 at the time of the event. 634 01:07:01.200 --> 01:07:11.200 I don't think that makes sense, because if, well think for a moment about the attitude of the predictor. 635 01:07:11.200 --> 01:07:21.200 The predictor is looking into the future to see, to decide what choice he wants to make. 636 01:07:21.200 --> 01:07:33.040 But, you know, so the predictor is saying, I'm going to stick the million in there only if I look and see you choosing just the one box. 637 01:07:33.040 --> 01:07:40.240 But of course, if I'm the predictor and I look into the future and I see you choosing just one box, presumably I see you getting the million or I see you not getting the million. 638 01:07:40.240 --> 01:07:45.240 Because it seems like I don't have any choice to make there either. It seems as though that's already settled too. 639 01:07:45.240 --> 01:07:50.480 But what are we supposed to picture? 640 01:07:50.480 --> 01:07:53.040 We say, all right, so here's the sequence of events. 641 01:07:53.040 --> 01:07:57.880 Suppose that I were to make the opposite choice. 642 01:07:57.880 --> 01:08:01.880 So the money is already there. 643 01:08:01.880 --> 01:08:05.800 I decide to choose both boxes. 644 01:08:05.800 --> 01:08:08.160 You want to say that doesn't make the money go away. 645 01:08:08.160 --> 01:08:11.040 It just makes it the case that the money was never there. 646 01:08:11.040 --> 01:08:14.240 So you've got the sequence of events. 647 01:08:14.240 --> 01:08:21.240 The money is there, I choose both boxes, and now the money not only is no longer there, but was never there. 648 01:08:21.240 --> 01:08:25.240 It seems like it's an inconsistent sequence of events. 649 01:08:25.240 --> 01:08:32.240 It says that it has the money being there at time t, and then it has the money also not having been there at time t. 650 01:08:32.240 --> 01:08:36.240 And you can say, well, the first one goes away, but is the first one going away something that happens? 651 01:08:36.240 --> 01:08:39.240 It seems like the first one going away never happened. 652 01:08:39.240 --> 01:08:57.240 I agree that that's incoherent, but if the million dollars is in the second box, that shows that you won't make the choice of choosing only the second box, but not that you can't. 653 01:08:57.240 --> 01:09:04.240 So there is no possible world in which the million dollars is there and you take the second box. 654 01:09:04.240 --> 01:09:21.240 and there's no possible world in which the second box is empty and there's no possible world in which the second box is empty and you take only the second box. 655 01:09:21.240 --> 01:09:29.240 But your choice determines which of these two possible worlds is actual. 656 01:09:29.240 --> 01:09:37.240 Here's my problem with that, though, is that one of them is already actual. 657 01:09:37.240 --> 01:09:39.240 Temporarily. 658 01:09:39.240 --> 01:09:43.240 Yeah, I don't know what the non-temporal sense of actual is. I guess that's the way. 659 01:09:43.240 --> 01:09:49.240 One of them is already actual. And once it's actual, it can't change from being actual to being non-actual. 660 01:09:49.240 --> 01:09:52.240 We agree on that. So one of them is already actual. 661 01:09:52.240 --> 01:09:58.240 But nevertheless, I'm supposed to think of myself as having the power to make the other one have been actual. 662 01:09:58.240 --> 01:10:08.240 But it seems as though that implies that there's a possible route from the actual world to the other one. 663 01:10:08.240 --> 01:10:26.240 But it seems like there isn't. Because if I'm in this one, if I'm already in this one, I guess I think that one way of putting it is that I think that once something happens, it is settled. 664 01:10:26.240 --> 01:10:34.240 And you want to say, well, there's some sense of settled in which it's not settled yet. 665 01:10:34.240 --> 01:10:43.240 This is a very old dispute. It goes back to Aristotle and the Stoics and Diodorus Cronus and all these guys. 666 01:10:43.240 --> 01:10:50.240 But yeah, certainly there are moves for your side to make. I don't think I've blown you out of the water. 667 01:10:50.240 --> 01:10:54.240 But I think I've got more moves to do. David. 668 01:11:54.240 --> 01:12:21.240 Okay, so the question is this, why isn't this the case, that the fact that the money is already there or not means that it's not rational for us to rely on, or we can't coherently rely on or something like that. 669 01:12:21.240 --> 01:12:30.740 This fact that there's this, you know, un-varying connection between whether you get the money and whether the predictor made the prediction of the past, 670 01:12:30.740 --> 01:12:35.240 why isn't it just so that we can't rely on that information? 671 01:12:35.240 --> 01:12:43.240 And therefore, rather than saying that both arguments are right given the assumptions, and therefore beating up on the assumptions, 672 01:12:43.240 --> 01:12:50.740 why not just say, no, the assumptions are all right, but what this shows is that you have to reject the argument for taking just the one box. 673 01:12:50.740 --> 01:13:01.300 Well, I guess because if it still seems to me, it seems to me a sort of a 674 01:13:01.300 --> 01:13:08.500 praxeological principle, that if I can choose between A and B and A and B 675 01:13:08.500 --> 01:13:13.900 co-vary with these other things C or D, so that I get A, you know, so any situation 676 01:13:13.900 --> 01:13:17.740 where I get A, I get C, and any situation where I get B, I get D, and I have some 677 01:13:17.740 --> 01:13:32.740 If I can choose between these two things and they co-variate these other two things I care about, then how can I not treat these as means to those? 678 01:13:32.740 --> 01:13:43.740 So it does seem to me as though they're both praxeologically irresistible, given the setup. 679 01:14:17.740 --> 01:14:28.740 Well, the question is, we need to have an argument against someone who says, 680 01:14:28.740 --> 01:14:35.740 well, if you're going through a deliberative process, trying to think about which one of these to get, 681 01:14:35.740 --> 01:14:41.740 and if the deliberative process then issues and you're getting this one, they're going to say, why isn't that a choice? 682 01:14:41.740 --> 01:15:11.740 And I think, you know, this is an argument for saying, you know, an argument against that view, but I think, you know, you do need an argument, because otherwise someone could just say, well you shouldn't, you know, you can't just define choice in such a way that causal, determined choice doesn't count as choice, because after all, there could be a case where there are two things you want, and you think about, you know, I'll avoid using words like deliberate and choice, you think about which one you want most, 683 01:15:11.740 --> 01:15:17.140 After thinking about it, it seems to you that this is the one that would most satisfy your preferences. 684 01:15:17.140 --> 01:15:20.140 And then you, on the basis of that, you then do that. 685 01:15:20.140 --> 01:15:23.840 You can see why a determinist is going to say, why not call that choice? What's wrong with that? 686 01:15:23.840 --> 01:15:30.140 And so I think that you need some kind of argument to show that there's something wrong with that. 687 01:15:30.140 --> 01:15:36.940 Because otherwise, they'll just say, well, you're cheating by defining choice in such a way that 688 01:15:36.940 --> 01:15:40.340 you're just defining the determinist choice out of existence. 689 01:15:41.740 --> 01:15:51.740 You may know something about this. 690 01:16:11.740 --> 01:16:23.740 I mean, it seems like you can have, well, I guess there's two things, you can have this position, but you can never manifest it, right? 691 01:16:23.740 --> 01:16:38.740 Putting in standard cases of fragile things that are safely packed away so they never kind of risk being broken, that sort of thing. 692 01:16:38.740 --> 01:17:07.740 But, I mean, I'm not sure I understand the inference from the fact that you can make it a case that you can never manifest a disposition to act in such and such a way, to the claim that therefore you can make it a case that you don't have a disposition to act in such and such a way. 693 01:17:07.740 --> 01:17:20.740 Okay. So the question is, a thing can have a disposition that it never manifests. A glass can be fragile and it never breaks. 694 01:17:20.740 --> 01:17:25.740 In fact, you could have thousands and thousands of glasses all lined up and none of them ever breaks, but they're all fragile. 695 01:17:25.740 --> 01:17:31.740 They're being fragile doesn't mean that they break from time to time or they break frequently or anything. 696 01:17:31.740 --> 01:17:41.740 It just means that they would break if struck with some defined force, or sugar is soluble in water. 697 01:17:41.740 --> 01:17:47.740 That doesn't mean that any given bit of sugar ever will dissolve in water. 698 01:17:47.740 --> 01:17:50.740 It could spend its entire career never encountering any water. 699 01:17:50.740 --> 01:17:54.740 It just means it would dissolve in water if confronted with it. 700 01:17:54.740 --> 01:18:06.740 So why do I want to say that we can't have desires or tendencies or impulses or, you know, middle stuff 701 01:18:06.740 --> 01:18:13.740 without it being somehow manifested in action, you know, often enough? 702 01:18:13.740 --> 01:18:18.740 And I don't want to make sort of a strong claim that, you know, you never have a desire 703 01:18:18.740 --> 01:18:22.740 without actually enacting that desire. I want to say something fuzzier than that. 704 01:18:22.740 --> 01:18:32.740 The question is, why treat desires as different from things like fragility? 705 01:18:32.740 --> 01:18:42.740 Well, I think the reason is I've sort of been convinced by Wittgenstein 706 01:18:42.740 --> 01:18:49.740 that there's a kind of intimate connection between thought and action 707 01:18:49.740 --> 01:19:19.740 The disposition of a kind that is different from these physical cases, the disposition to act isn't just a disposition to engage in certain mechanical motions, if it were, you could have a robot that would be disposed, you could have a robot that was disposed to, you could have a robot that was disposed to engage in a certain kind of behavior in certain kinds of situations, but it's never placed in any of those situations, 708 01:19:19.740 --> 01:19:23.700 Program such that if you were to, then the actions would flow forth, and so it makes 709 01:19:23.700 --> 01:19:30.380 perfect sense to say that that robot is disposed to exhibit this behavior even if it never 710 01:19:30.380 --> 01:19:49.140 actually exhibits it. But an action isn't just a behavior, the action implies, the action 711 01:19:49.140 --> 01:19:56.140 Action implies the desire, so you can't have an action isn't just sort of the behavioral side of things, the action actually includes the desire. 712 01:19:56.140 --> 01:19:59.140 Likewise on the other side, the desire includes the action. 713 01:19:59.140 --> 01:20:10.140 In the case of a robot that's programmed to do certain things, you can point into its circuitry and identify where the programming is somewhat apart from the actual acting. 714 01:20:10.140 --> 01:20:19.140 But the example I was giving with the plans is supposed to show that nothing is going to count as the plans unless they're acted on. 715 01:20:19.140 --> 01:20:21.140 You could say, well, look, here's the thing in the brain. 716 01:20:21.140 --> 01:20:23.140 What is the thing in the brain? 717 01:20:23.140 --> 01:20:29.140 I think the thing in the brain might show that you maybe have some sort of predisposition to some sort of mechanical behavior. 718 01:20:29.140 --> 01:20:34.140 But I don't think the brain, just pointing to the brain, can show you have a predisposition to the action. 719 01:20:34.140 --> 01:20:45.140 Because, you know, the brain is just a physical thing and anything you can infer from it is just going to be sort of more kickings around, but it doesn't really show sort of the meaning side of it. 720 01:20:45.140 --> 01:21:01.940 So I want to say that in some sense the desire and the expression of it are sort of internally related 721 01:21:01.940 --> 01:21:09.540 in a way that the things in the brain and subsequent muscular motions are externally related to each other. 722 01:21:09.540 --> 01:21:15.540 You can grasp one without conceptually grasping the other, unless the other two are twisted together. 723 01:21:39.540 --> 01:21:49.540 You have to have any particular number of actual manifestations of what a business is for. 724 01:21:49.540 --> 01:21:55.540 You can't have business that develops a business that doesn't manifest in any business at all. 725 01:21:55.540 --> 01:22:03.540 People have plans, they don't actually carry out. They abandon the plan for whatever. 726 01:22:03.540 --> 01:22:11.540 or whatever, but for the calendar plan, there has to be some kind of difference between having it as a plan and merely tripling down out of work. 727 01:22:14.540 --> 01:22:21.540 But then I guess what I wonder is, if you're sort of making it a matter of just not having felt this position, 728 01:22:21.540 --> 01:22:38.540 and not having unmanifested individualism, then one of the questions you can ask, whether it's the case that one of the differences that could be relevant is the difference is not actually a physical fault. 729 01:22:38.540 --> 01:23:01.540 It seems like some of the things that we cite, that we cite as positions of interest to them, like fear or pain, or things like that, maybe things that are manifested in involuntary characteristics, right? 730 01:23:01.540 --> 01:23:28.540 So if someone has a theory that you have to emanate tendency towards calories and things like that, there's plenty of things that you could do to, there's plenty of things that could indicate, you know, 731 01:23:31.540 --> 01:23:49.540 The continuing presence of fear, even if the person isn't acting in such a way as to act on the disposition in front of them. 732 01:23:49.540 --> 01:23:54.540 And so, you know, there's just, you know, a few things like that. 733 01:23:56.540 --> 01:23:59.540 Arguing is all good action, but then there's the struggle. 734 01:24:01.540 --> 01:24:06.540 The thing is, that would undermine the claim that they're actually a catalyst, right? 735 01:24:06.540 --> 01:24:10.540 Because they're not acting on just the things we've spoken on. 736 01:24:15.540 --> 01:24:18.540 But it wouldn't undermine the claim 737 01:24:18.540 --> 01:24:38.100 Question for the audience. 738 01:24:38.100 --> 01:24:44.260 Okay, so the question is, I've allowed this fuzziness, as I called it, that it's not the 739 01:24:44.260 --> 01:24:57.260 It's not the case that every desire or every mental state you have, you have to actually act on, it's just that you can't have any that aren't somehow, you know, expressible. 740 01:24:57.260 --> 01:25:09.260 And so then the question is, well then, what if some of these ways that they're expressed are things that you don't have control over? 741 01:25:09.260 --> 01:25:16.260 So, for example, if I have extreme fear, you know, that might be expressed in chosen actions like running away, 742 01:25:16.260 --> 01:25:24.260 but it might just be expressed by, you know, shaking and trembling and blinching and things that are not under my control, 743 01:25:24.260 --> 01:25:30.260 and therefore there's no guarantee that habituation would be able to get rid of those. 744 01:25:30.260 --> 01:25:45.260 Yeah, I think I agree. I didn't want to say that we have sovereign control over all the contents of our minds so that we can get rid of any psychological state through habituation. 745 01:25:45.260 --> 01:26:02.260 But by habituation, we can, any states that we act on, because what Aristotle is concerned with primarily for virtue is the states that we act on, the states we express through our actions. 746 01:26:02.260 --> 01:26:10.260 I want to say that the states we express through our actions, to the extent that they are expressible through actions, to that extent they are controllable. 747 01:26:10.260 --> 01:26:26.260 But so that to the extent that my extreme fear makes me run away, that aspect of it is controllable. 748 01:26:26.260 --> 01:26:30.260 I can get myself to be sort of less likely to run away. 749 01:26:30.260 --> 01:26:37.260 But it might not be the case that I can get rid of the trembling and the flinching. 750 01:26:37.260 --> 01:26:42.260 Maybe I can, maybe I can't, but I'm not claiming that there's sort of an a priori argument that can get rid of that. 751 01:26:51.260 --> 01:26:55.260 You're saying that rational agents must be habituated to a priori? 752 01:26:58.260 --> 01:27:05.260 Yeah, the question is, am I saying that any rational agent must a priori be one of whom habituation is true? 753 01:27:05.260 --> 01:27:15.260 I have this habit of making bold claims and I can't get rid of it. 754 01:27:35.260 --> 01:27:46.260 And the only point I can give you is because these kind of reviews have been going back on to the initial plan and so it's hard to actually try to figure out something in the multiple questions. 755 01:27:46.260 --> 01:28:01.260 But this story has many other greater implications, like for example, the paradox that I'm not saying, and I'll say it's more cultural, but I've always heard the story of the death of Harvard, 756 01:28:01.260 --> 01:28:16.140 ______ 757 01:28:16.140 --> 01:28:24.020 Okay so this is a question about one of the, you know, one of the suggested 758 01:28:24.020 --> 01:28:28.100 readings I had. So one of the suggested readings in the background was this 759 01:28:28.100 --> 01:28:34.540 This dialogue, platonic dialogue, or probably platonic dialogue with Hipparchus, which I 760 01:28:34.540 --> 01:28:43.180 didn't say anything about, but it's about the Socratic paradox in effect, but the question 761 01:28:43.180 --> 01:28:50.260 is about this long digression in the middle about the historic character Hipparchus, who 762 01:28:50.260 --> 01:28:54.540 was a ruler of Athens, who was assassinated, and so the question is twofold. 763 01:28:54.540 --> 01:29:01.400 One, does the long digression in the dialogue shed any light on the rest of the dialogue? 764 01:29:01.400 --> 01:29:07.900 And second, most accounts of the death of Prakas present it as the glorious assassination 765 01:29:07.900 --> 01:29:12.980 of an evil tyrant, and this dialogue presents it as the tragic assassination of a benevolent 766 01:29:12.980 --> 01:29:13.980 ruler. 767 01:29:13.980 --> 01:29:16.820 For the first question, I've got nothing to say. 768 01:29:16.820 --> 01:29:22.460 I have never been able to figure out what that digression is doing there, and I haven't 769 01:29:22.460 --> 01:29:30.460 I spent a lot of time thinking about it, and maybe if I did I would come up with something wise to say about it, but I got nothing on that at the moment. 770 01:29:30.460 --> 01:29:45.460 On the other question, well, whether a particular assassination was a glorious deed or a wicked one depends on the political perspective of the author. 771 01:29:45.460 --> 01:29:51.060 Plato was politically rather on the conservative side. He identified more with the aristocratic 772 01:29:51.060 --> 01:29:59.420 side, and so he was more sympathetic to Hipparchus. A lot of the anti-Hipparchus things were written 773 01:29:59.420 --> 01:30:06.060 by people who were more sympathetic to the democratic factions. What the right story 774 01:30:06.060 --> 01:30:11.900 is about Hipparchus I couldn't say. My guess would be that Hipparchus was no saint and 775 01:30:11.900 --> 01:30:22.900 And the people who killed him were no saints either, but exactly the details that they disagree on, I don't know. 776 01:30:22.900 --> 01:30:29.900 Are there other representatives who talked to you about what I'm saying? 777 01:30:29.900 --> 01:30:38.900 Probably. I can't think of any offhand, but I can't remember whether Aristotle talks about the Constitution of Athens or if so what he says. 778 01:30:38.900 --> 01:30:44.900 But I can imagine him saying something negative about it, but less negative than Plato. I wouldn't be surprised if it went like that. 779 01:30:44.900 --> 01:30:48.900 Okay, I think we are out of time. See you this afternoon.