PHIL 126 Lecture 8 Analogical induction

0:00 Hi everybody. Welcome to our next lecture on analogical inductions. This is the second type of inductive argument that we're going to be examining. Remember once again that whether an inductive argument is strong depends on what kind of inductive argument it is. Saw the standards for enumerative induction arguments. You need a big enough sample size and it needs to be representative enough of the target group to justify the inference of the conclusion. So it's strong if it has those features.

0:37 Analogical arguments that should just say analogical induction. I don't know why the ed is there. Okay. analogical arguments are arguments by analogy. So the idea is you say, hey, phenomenon X we all agree on. phenomenon Y shares a bunch of the properties of phenomenon X. So we should think that phenomenon Y also has this other property that X has. That's the idea. many arguments are based on analogy. We'll see today that it's many more than you might initially think. So here's the idea.

1:13 Analogical inductions. Let me actually here's here's an example. Let's do the example first. A1, humans can move about, solve mathematical equations, win chess games, and feel pain. Robots are like humans in that they can move about, solve mathematical equations, and win chess games. Therefore, it's probable that robots can also feel pain. So the idea is the idea behind this argument is that we are comparing two things humans and robots and we're listing a bunch of features that humans have a bunch of features that robots have and inferring that the features that the humans have for which we don't know whether the robots have them that the robots do have them that's the inference right so A1 doesn't talks about how humans feel pain A2 robots have all these other features in common with humans so they probably also feel That's the idea. So because two or more things are similar in se several respects, they must be similar in some other respect, some further respect.

2:21 Okay. Analogical induction has the following pattern. There should only be three things, three premises. Premise one, thing A has properties P2, P3, P4, P5, P6, however many you want. plus the property P4. In this instance, thing B has properties P1, P2, and P3, or however many you listed minus one. Therefore, thing B probably has P4. So again, arguments by analogy, like all inductive reasoning, can establish conclusions only with more or less probability.

2:58 The more similar the more similarities we consider between the two cases, the more probable the conclusion is fairly obviously. Right? So the idea is like if I list 70 ways that these two things are similar, then we should think also that they are similar in the 71st way. So arguments by analogy are in fact all over the place and are used and misused basically everywhere. We see this especially in law, science, medicine, forensics, recommendations, all of these things.

3:37 So, here's a we'll see some examples. Let's see. Here's an example. Takes a moment to set up, but here's the idea. So, say you're doing a I'll just read it. Mice or mammals have a mamalon circulatory system, have typical mamalian biochemical reactions, respond readily to high blood pressure drugs, and experience a reduction in blood cholesterol when given the new drug Z. Humans are mammals, have a mamalian circulatory system, have typ typical mamalian biochemical reactions, and respond readily to high pressure drugs.

4:13 Therefore, humans will also experience a reduction in blood cholesterol when given the new drug Z. So, medical testing on animals and in fact on humans as we'll see in a second, but on animals is analogical induction. We know that mice in particular strangely share a lot of our internal organ system. and so we're very similar on the insides to mice, which is so strange because we're so dissimilar on the outsides to mice, but whatever. So the idea is because mice and humans are so similar, we test things on mice that we don't know whether they're safe or not because we would rather give something unsafe to a mouse than to a human because it might harm the human.

5:00 And the idea here is humans are more morally weighty or something like that. So that's why we test on animals. Okay. So we see that they're similar in some respect. we infer that human beings are similar in that same respect. This is also exactly the same reasoning that's going on in human clinical drug trials. Right? So we do a drug trial of a group of a hundred humans. We test some drug on them and then when the doctor comes to you who was not in that original trial and they try to think, okay, what drug should I give you to make you better off? Well, let me look to this study that tested other people who I believe to be sufficiently similar to you.

5:47 Right? So, we regular drug trials are just arguments by analogy. Also, the reason the doctor prescribes you a drug is not because they know your body, but because they think that the way the drug interacts with your body is going to be analogous to the way the drug interacted with the bodies of the people who did the study. sort of interesting. So, analogical induction arguments have a weak spot. Part of it is because they can offer really strong support, but they can never be conclusive. And we'll talk about this. So a physician, as I just stated, might compare a patient's current set of symptoms to a disease that typically displays other similar symptoms. But since you can never conclusively or like some diseases can cause symptoms that overlap with those of another disease, we can never have a conclusive analogy.

6:56 Okay. So the idea is the basic structure. Let's do this other one. Basic structure is you identify these things that are relevant. Then you assess how strong the analogy is. So animal animals that roam in herds are susceptible to predators. Biology teaches us that the weakest members are the ones most likely to be prayed upon. From the B vantage point of natural selection, this means that the strongest members will pass on their genes, thus ensuring the overall fitness of the herd. The human brain is like an animal herd. We know that hard drugs and alcohol kill brain cells. However, like the herd, only the weakest cells are killed off. Therefore, regular use of hard drugs and alcohol successfully removes the weakest brain cells, allowing the best cells to operate more efficiently. Okay.

7:48 is this argument a good one? Should you do as many hard drugs and drink as much alcohol as you can fit into your body so that you have only the strongest brain cells left? No, obviously, right? We'll talk about why in a moment, but the thing is analogies are really easy to come up with. Like super easy. Any two things are analogous in some respect. And so analogies are really cheap. Doesn't mean they're effective, but they are cheap.

8:21 there's this famous line by David Hume, Scottish philosopher. the argument was one of the arguments in favor of the ex existence of God says something like look the universe resembles something that a human being would make and so we should think that a rational mind made the universe. And one of Hume's responses was something like, well, if you think about it, the universe also resembles a turnip in some respects. You know, like these analogies are really cheap to draw. Doesn't mean that they're always good.

9:03 Okay. So, there are some criteria that we can use to judge the strength of arguments by analogy. They are thus relevant similarities between the two things being considered, relevant dissimilarities between the two things being considered, the number of instances compared, and the diversity among the cases. As we can see, if there's more relevance, like just intuitively, if there's more relevant similarities, the stronger the argument, more relevant dissimilarities, the weaker the argument. More instances compared, the stronger the argument.

9:36 More diverse the cases, the stronger the argument. Right? That's the idea. you might like already think that these make sense and that's because you already use these criteria when people make analogies to you. Okay. So, let's consider the following case for illuminating this relevant similarities idea. So, here's an argument. In the Vietnam War, the United States had not articulated a clear rationale for fighting there and the United States lost. Likewise, in the present war, the United States has not articulated a clear rational rationale for fighting. Therefore, the United States will lose this war, too.

10:13 Intuitively, we can see what somebody's going for here, but this is if this is the only similarity between the two cases that there's not a clear rationale for fighting there. I don't think that's strong enough to think that we're going to lose this or two just on that basis. We haven't compared enough. There's these two things aren't similar in enough ways to license inferring that conclusion. But I want you to consider what the argument looks like if we start adding similarities.

10:41 So here's here's the new one. In the Vietnam War, the United States had not articulated a clear rale for fighting. There was no plan for ending the involvement of US forces. That is no exit strategy. Our military's tactics were inconsistent. And the military's view of the enemy strength was unrealistic. The United States lost the Vietnam War. Likewise, in the present war, the United States has not articulated a clear rationale for fighting. There is no exit strategy. US tactics are inconsistent and the military's view of the enemy strength is naive. Therefore, the United States will also you lose this war. We see that this argument becomes much stronger with this second version where we're considering more instances. The idea is that we think that there is something that there's something about the Vietnam War that connects all these facts together or connects the chunk of facts to the losing the Vietnam War. Something like that. We think that there's some connection between those. And so when we look at the present war and we see the same facts, if we think they were connected to losing the Vietnam war, then we should also think that those facts are connected to losing the present war.

11:55 Right? So the idea is we when we do an argument by analogy, we think that the properties are sort of cohesive in some sense. And so this brings us to the idea about relevant similarities. It's not all similarities that matter. It's only relevant ones. Now what does it mean to say that a similarity is relevant? Well, it sort of is a circular definition. A similarity is relevant to an argument by analogy if has an effect on whether the conclusion is probably true.

12:26 Right? If it if it bears on the conclusion, then it's a relevant similarity. Maybe a better way of thinking about it is if we think a a similarity is relevant. If we can tell a coherent story for how it along with the other similarities are associated with the conclusion. If we can tell a story for how they're associated with the conclusion, then or the property that's stated in the conclusion, then we say it's a relevant one. So like here's some relevant Here's some irrelevant similarities. In both wars, some combatants have green eyes. In both wars, some soldiers are taller than others. In both wars, ticket sales to movies in the US increase. Right? These things are irrelevant. If you add them, they neither strengthen nor reduce the the probability of the conclusion.

13:26 And plausibly it's because we think we think these things have nothing to do with whether you win a war or not at least in the present circumstances. Additionally, whether some property is relevant to the conclusion depends on which conclusion it is. And the same property might be relevant in some arguments but not others. So, here's one about the color of a car where it's relevant in one case and not relevant in the other.

13:58 Argument A. My Ford Fusion Hybrid and your Hummer are the same color. My vehicle averages 40 m per gallon. Therefore, your vehicle will probably average 40 m per gallon. Right? We think that the color of the car has nothing to do with the speed or the sorry, the miles per gallon of the car, the fuel efficiency. On the other hand, my Toyota Camry and your Nissan Sonata are the same color. My daughter likes the color of my car. Therefore, my daughter will probably like the color of your car, too. Right? That one, the color of the car is relevant to whether the child will like the color of the car.

14:37 So, the same thing, two cars being the same color, is an irrelevant difference in one case and a relevant difference in the other case. Okay, so that is relevant similarities. Let's move on to relevant differences or relevant dissimilarities. Generally speaking, the more relevant dissimilarities or danalogies there are between the things being compared, the less probable the con conclusion is. Dissimilarities, in other words, weaken arguments by analogy.

15:10 Consider the argument up here about the drug right here. Consider this argument. What if we discovered that cholesterol-lowering drugs that work in mice almost never work in humans? Well, then we would have good reason to think that this conclusion doesn't work. There's a relevant dissimilarity between the two cases, which is that cholesterol-lowering drugs that work in mice never work in humans.

15:42 I think I wrote that on here. Yeah. So, pointing out dissimilarities in analogical induction is probably the most common way to undermine the argument, right? Sometimes finding one dissimilarity, like in this cholesterol-lowering drugs case, is enough to show that the argument should be rejected. Other times, we can debate about whether the dissimilarity is a relevant dissimilarity or not. We'll do that plenty. Okay. The third thing to consider is the number of instances compared. Basically, the greater the number of instances that show relevant similarities, the stronger the overall argument. So, think back to this case of the Vietnam War and the current war that we're in. Imagine that there was not just one instance of a war in America in recent American history where all of these things were true, but suppose there were five wars in recent American history where all five of those things were true.

16:42 The argument would be greatly strengthened because it showed that those cluster of properties occurred not only once but five independent times. What is the likelihood that this case will be different than the others? probably lower than 50%. Of course, coincidences happen. Maybe these properties aren't clustered in the way we think and the argu and the conclusion might be false. Maybe the United States will not win the will not lose the war.

17:20 That's okay, right? We're we're considering inductive arguments. The conclusions don't have to be certain. They just need to be probable. It just needs to be reasonable to believe them. Even if they turn out to be false, it's totally reasonable to believe things that happen to be false. You just can't know about that. You know what I mean? Okay. And finally, the final criterion is diversity among cases. So, dissimilarities between the things being compared weaken an argument by analogy. Right? These dissimilarities are supposed to show us that the things being compared aren't like super strongly analogous to one another.

18:03 We've also seen many cases that if there are many cases of similarities, this can strengthen the argument too. All the different wars. However, in this criterion, the final one, there's a different thing going on here. We are considering the diversity among the cases that were consider that that are being put forward as part of the analogy. If the cases are really really different from one another but all have the set of properties that you care about then that gives us good reason to think that anytime we see that cluster of properties we also see the final property. So here's here's a version of the argument. I know it's a lot just take it nice and slow. You'll get the pattern. You don't have to hold a lot in your head. So in the 1990s, US senator, a Republican from Virginia, was chairman of the Commerce Committee, had very close ties to Corporation X, had previously worked for Corporation X before coming into office, and was found to have been taking bribes from Corporation X.

19:06 In the 1980s, another US senator, this time a Democrat from Texas, was chairman of the Commerce Committee, had very close ties to Corporation X, had previously worked for Corporation X before coming into office, and was found to have been taking bribes from Corporation X. You see the pattern? Number three, in the 1970s, another US senator, this time an independent from Arkansas with strong religious values, which chairman of the Commerce Committee, had very close ties to Corporation X, had previously worked for Corporation X before coming into office and was found to have been taking bribes from Corporation X.

19:38 Premise four. Now, the newly elected Senator Jones, this one a Democrat from New York with strong support from labor unions, is chairman of the Commerce Committee, has very close ties to Corporation X and has previously worked for Corporation X before coming into office. Therefore, Senator Jones will probably take bribes from Corporation X. So, here we see a bunch of similarities in question.

20:09 Right. senators who are chairmans of chair people of the commerce committee really close ties to this corporation previously worked for the corporation or were found to have been taking bribes from this corporation. And we actually see what makes this strong, especially strong is that the cases are really diverse despite those handful of similarities. They're diverse in the following respects that we sometimes take to be relevant.

20:40 What era it was in. You've got people from 70s, 80s, 90s, and presumably, let's say, 2020s. You've got senators of different political parties, Republican, Democrat, independent, and you've got senators from different areas, New York, Arkansas, Texas, Virginia. This combination of things suggests that the similarities are not merely accidental. They're not they don't just happen to cluster because they cluster in really different arenas.

21:17 And so we think that they're strongly linked so that anytime you see the cluster of four or five properties, you're very likely to see the last one as well. Okay. So the last thing I want to do is I want to talk about analogical inductions for moral claims. So this is a very common way to argue in morality is to argue with cases. The idea is you consider a case in isolation and then you see in what ways that case is analogous to real life situations that we care about the ethics of and we try to figure out what to do in the real life ethical scenario on the basis of thinking about these cases because we think they have all the same relevant similarities is the idea. That's the idea. So consider a case in isolation.

22:08 You can think about it on its own. It's not messy. and then you infer by analogy how things are in the real world. So in moral philosophy, this property P is often what's called a normative property. You don't need to know exactly what normative means, but something like morally right, morally wrong, something along those lines. Okay, I want to give an example. We'll talk about another example, but this one I don't want to try to resolve. I just want to present it to you as something to consider. It's about immigration and whether the state meaning a country has a right to restrict immigration to migrants.

22:51 So we're going to consider a case and try try to argue that it's analogous to another case. So Marvin wants to travel to a town where employers would would hire him, landlords would rent to him, and customers would buy his labor or products. Right? So Marvin's trying to go to this town to visit the marketplace where he would get employment. He would contribute to the landlord's pockets and customers would buy his labor products, right? He'd be a good marketplace addition.

23:28 Sam, Uncle Sam, threatens Marvin with force to stop him from entering. Says, "No, you are not allowed to come into our marketplace, to our town." Now, suppose these are all the relevant moral facts. There's nothing else. Marvin's not dangerous. They're not overcrowded. Nothing like that. Just suppose that that is the situation. Would you think that it would be morally okay for Sam to threaten Marvin with force? Shoot him if he comes in, throw him out if he comes in to stop him from entering.

24:09 I think a lot of people would think if you're really thinking about this case, you don't have a good reason to kick Marvin out that no, it would be wrong to block Marvin from coming in. Why? because it blocks this peaceful voluntary exchange of goods and services that we think would make everybody better off. So we think it wrongs Marvin to be blocked from entering this town for no good reason.

24:41 Okay. The person who proposes this argument, philosopher named Michael Huer, thinks that this case is actually very very analogous to cases of migration. So here's his alternative case. A peaceful migrant wants to enter a country where one employees would hire them, two, landlords would rent to them and three people would trade with them. The state threatens the migrant with force or penalizes people who try to help the migrant to stop entry.

25:16 So these cases, these two scenarios are analogous on all the relevant facts minus the final one, the bottom one, right? Employers would want to hire them, landlords would rent to them, people would buy their labor or products, and the state threatens them with force. So if we think that it's wrong morally wrong in the in the marketplace scenario and they're analogous in every morally relevant respect then we should think that it's morally wrong in the migration case as well.

25:55 So if you like this argument then you think blocking people from oh sorry let me put that a different way. The argument is an argument by analogy. There's only one way, two way, one way to show that an argument by analogy is not strong and that's to show that the analogy doesn't work. Are there relevant dissimilarities between these two cases?

26:29 I guess here's the Yeah. What are the are these similar? So, we'll go the other way first. Here's some similarities between the cases. Person wants to access willing traders. Coercion prevents movement. Third parties are willing to contract. And and the result is that there's a peaceful exchange blocked. Are these similarities relevant to whether it's morally right or wrong? Sure. Seems so, right? Blocking peaceful exchange among consenting adults seems to be relevant to whether it's right or wrong.

27:06 So there's from this argument there are three conclusions if you cannot restrict if you cannot find a disanalogy between the cases. Conclusion A is that restricting entry is a rights violation against the migrant. They have a right to freedom of movement. Conclusion B restricting entry is impermissible all things considered. That's really strong. or C restricting entry is primaascashia philosophers's term meaning like at first glance or without any good reason otherwise that's sort of what it means restricting entry is wrong unless we can find some reason why it might be right in some case probably the conclusion that the author wants us to take is C or that's probably the most defensible conclusion from So if you want to reject C then you need to find a relevant dissimilarity between the cases. Here's some candidate ones. I just want to throw them out. I don't want to solve the case for anybody.

28:15 so here are candidate ways of of undermining the analogy. Each one of these needs an argument that it is actually relevant to the moral property we care about. So jurisdiction or authority. a state has special standing to set the terms of entry that individuals lack. You need to make an argument for that and how it actually bears on this case. Another way out is to say that with public goods that affects things whereas letting somebody into a marketplace affects things doesn't affect things. So large-scale migration can affect welfare, state finance, infrastructure or political stability.

28:54 Third, security crime or war screening and ex and exclusion may be necessary to prevent serious harm. And finally, democratic self-determination. Citizens have a claim to shape membership and policy and that can sometimes override the wishes of people to move freely. Now, you might think that some of these are pretty good and some of them are pretty bad. That's probably the case. but notice that for a lot of these what they restrict what they permit is some instances of migration but not all of them. You can't restrict all of them just by using these like the second one.

29:39 Okay, that's what we got for the second kind of inductive argument. Arguments by analogy I believe. Yep, that's it. Godspeed.