NARRATOR: Thanks for joining this episode of the No. 86 lecture series, in which Professor Richard Epstein discusses Torts in the Roman Law. In Episode 2, he explains how Roman tort law differs from Anglo-American tort law, and the importance of determining causal connections. This lecture is part of a series with Professor Epstein on how this ancient legal system can provide crucial insights about modern problems. Professor Epstein is one of the most prominent legal scholars of our day. He is the inaugural Laurence A. Tisch Professor of Law at NYU School of Law, a Senior Fellow at the Hoover Institution, and Professor of Law Emeritus and a senior lecturer at the University of Chicago. As always, the Federalist Society takes no position on particular legal or public policy issues; all expressions of opinion are those of the speaker. PUBLIUS: At the end of Episode 1 on Torts, you discussed a bit about how the Romans thought about causal connections. Can you give us a fuller outline about that? How does the Roman tort law differ from our modern Anglo-American tort law? RICHARD EPSTEIN: Modern American law makes an absolutely fatal mistake, in my judgment, because it does not stall with the close cases and then start to work out. What it does is it introduces a very fuzzy notion of but-for causation, which is saying, "but for the occurrence of this, that would not have happened." Sort of a counterfactual situation, and there's no very tight physical or temporal connection that is inherent in the but-for formulation, so everybody knows that it may be a first approximation for causation, but you have to add all this stuff in. The most common candidate is foreseeability, which really doesn't do the kind of work that you need to have done on it. So, what the Romans did is they actually lucked into the right solution. Tight connections first, and then the theory of causation is how far can you extend this kind of relationship and still say that we've got ourselves these physical sorts of connection? So then, how do you actually do that? Well, the first case they give is essentially you take your hands and you strangle somebody. If somebody wants to deny that that's force, you could say it's metaphorical. And then, the other thing you could do is you could punch, but suppose now what you do is you put in your hand a sword, and you slay him. Does the introduction of an instrument essentially change the nature of the causation so that we say, you did not do it because you had this sword in your hand? And if you look at the ordinary use of transitive verbs in the English language, and somebody says, "I shot him," "I clubbed him to death," and so forth, you have exactly the same direct connection by the use of force with respect to these instrumentalities that are under the exclusive control of another individual. So, it turns out that the direct action under the stature for "occidere" applies in these particular kinds of cases. And then, things start to become just a little bit more complicated in a whole variety of situations. One of the famous illustrations that we start to use is you have somebody, and what you do is you throw him off a cliff, and he hits the water, and he doesn't die from the immediate blow, but he drowns, exhausted after desperately trying to save himself. Have you killed him by force? And the answer is, well, we don't think you've killed him by force because his thrashing around was an act that intervened, and so therefore they're not going to recover, and most people when they hear that, they think it's too clever by a hair. Essentially, the application of force is what put you in danger, but every single action that you took as the victim was in effort to try to ameliorate the force, not to try to compound its effect, right? And so, that's the way in which the situation works. Why would we want to take this thing out from underneath the stature? And so, by way of careful case exposition, they decided that that case, too, was one which was covered by occidere. Now, where does the moment of truth come with respect to these systems? Well, Romans had a peculiar pension for poisoning one another, as anybody who's ever seen I, Claudius will know. What is it that when you're dealing with poison tells you whether you're within the class of occidere or outside of it. And the first thing to know, of course, about this poison is that the dangerous potential that you have from the situation is not from the physical connection, right? Throwing some poison down your throat doesn't kill you in the same way a sword does, but the chemical interactions between what you swallow and everything else then happening is different. And so, does that change the basic rule of the situation? And the Romans thought about this for a long time, and they said, "So long as you forced it down their mouth, the digestive juices and stuff that works on this are no different than the thrashing about that you have by the man who's been thrown off the edge of the cliff, so, we're going to treat that as though it's a case of direct force, and it's covered by the stature." But this, then, invites the following kind of questions, for which you really need to have an answer. Suppose you're a little bit different under these circumstances, and instead of using brute force you decide to resort to a little bit of old-fashioned deception in order to achieve the same result. And so, what you do is you take some substance, which looks like soup or food, and you slip into it some poison. You put it in front of somebody and say, "Here is your dinner, here is your dessert." What they do is they eat or drink the stuff that's put in front of them, and they die. At that particular point, it is very difficult for somebody to say that you forced this thing down their throat, and so it's very difficult to say that this is a case of occidere. And then, what you do is you ask the next question. Okay, it's not occidere. Do we really want to let this guy off Scott-free when, in fact, the dangers of insidious behavior may, in fact, be much more serious than the dangers of the direct use of force against somebody. Why is that? Because if somebody wants to force any food down your mouth, you're going to fight like hell in order to resist it, but on the other hand, if they slip this poison to you in a deceptive fashion, then you might eat it only two hours later to wake up dead because of all the chemical effects that it had. And, again, you start asking people: is this a simple transitive verb type arrangement that you kill them? They said, "No, you set the poison in front of him, and he ate it." He may have caused the death, but it's not like he shot him, he kicked him, or whatever. And so, the Romans then introduced an absolutely critical distinction into their system to cover these analogous cases that are not reached by the stature, and that's the distinction between occidere and causam mortis praestare, and that mysterious Latin phrase "causam mortis praestare" means furnishing a cause of death. And if you apply it to the poison case, it seems to fit exactly like a T. And so, then what you have to do is to say, "How do we analyze these particular cases?" Then, we take a leaf from Aristotle to figure out what's going on, and we find three cases, one of which I've already mentioned to you, which the case in which you deceive somebody so they take it down, and the general rule about causal intervention is if you, a defendant, create deliberately a false impression that reduces somebody's omission, his action does not break the chain of causation because it was induced by your mistake. It would be much more difficult if it turns out that the mistake was created by a third party, and the Romans don't have a clean answer to that, nor does anybody else, but at least in this particular case it's there. The second case, which is equally apparent, is as follows: suppose I put the poison down in front of you, and you know it's poison, and I know that you know it's poison, I say, "Here's your choice. Either you take this poison, and what will happen is you will die peacefully in your sleep, or you don't take that poison and I'm gonna club you to death, and you will feel miserable for three hours before you die. Your choice." Now, we all prize freedom of choice, and generally speaking if given the choice we would take the less painful thing, but because A is less painful than B, it doesn't mean that it's a legitimate sort of choice, and so any action which is done under threat of duress by somebody else where the threat involves the use of force, again, doesn't sever the nature of the causal chain. So, what happens quite naturally in Latin, what you develop is a situation where you start looking at actions that take place subsequent to what the defendant does, right, and prior to the time of death, and you ask about whether these new, intervening act, "novus actus interveniens," inessentially override the situation. Many modernists regard this phrase as "intellectual mumbo jumbo," but if you start from the world of trespass and start working backwards like that, the intervening act is the most natural way to look at the situation, and what you realize is that some of these interventions in effect don't sever the causal connection, and other conventions do. "Well, what's the kind of connection that we have which starts to do the severance that starts to take place?" I think becomes a very fair question to ask. And here's the situation that you have to create: you must negate ignorance on the one hand, and you must negate coercion on the other hand. And so, you put the poison in front of somebody and say, "Look, here it is. It's poison. You've indicated that you'd like to kill yourself. I'm trying to help you make this decision. I'm not gonna tell you what to do, but I'm just gonna make this nice and handy. You decide." And so, generally speaking at this full knowledge and no coercion, then it turns out that the causal change comes to an end, usually. But of course, there's always going to be difficult cases. Suppose it is that you put in front of somebody, whom you know is chronically depressed and mentally impaired in a given way, and you assume that when you give them the true information, what they're going to do is essentially act on this in a perfectly irrational way, given all of their known physical and mental disabilities of which you are completely aware. And so, these then become very hard cases as to whether or not by furnishing a cause of death as somebody who has incomplete control over them. Do you treat the incomplete control in the same fashion that you teach the mistake, or you do the coercion? What you have to say about these situations is as follows: Honest people can surely differ about the ways in which you treat these marginal cases. The question is not whether or not these marginal cases give you dispute, but whether or not you could locate the margin at a particular point that is sufficiently clear, and that there are large numbers of clear cases on both sides of the line. Generally speaking, with any legal system where you have to define the difference between those consequences which are not too remote, and those consequences which are too remote. Now, furnishing a cause of death as a form of indirect causation, right, essentially presupposes the very stylized situation where you give something to somebody, but it turns out, of course, that when you're dealing with remoteness of damages, you should treat that only as a subset of a larger class of cases creating dangerous conditions under which the actions of either the individual injured party or a third party start to take place. And so, here's a variation on the particular theme, which essentially everybody seems to understand. You don't kill somebody. What you do is you put a trap in the room, and then what you do is you cover the trap over with some sod so it's indistinguishable from everything before and after, and somebody comes down the road with a horse or on foot, and what they do is they fall through the trap, and they get killed by being impaled on the spikes which have been conveniently set there as well. And now what you do is you come up and you say, "You know what, I am not responsible for this death because the actual force that was applied was by gravity when it pulled you into these stakes, and whatever happens through gravity, that's God's responsibility. That's Issac Newton's responsibility. It turns out not to be mine." Now, if you go and you look through the entire annals of the law, there's never been a legal system under sort of kind or description which has accepted that particular argument. But the question, then, is, "Why not?" And it's exactly the same problem that you had in dealing with the poison. What you did here was you created a latent effect, one which was not known to the person, and so when they stepped on that particular piece of sod, which was not supported by firm earth, essentially they acted under a system of mistake. Once they act under mistake, it turns out that their particular actions are charged back to the defendant who made the pit and not charged to the plaintiff alone. So, even though the setting on the road and the setting with the poison look in some sense to be wildly diverse, the fact that there's a latent effect which causes harm when the ordinary actions of the plaintiff essentially is involved is in fact good reason to say that the causal connection is not going to be severed. PUBLIUS: The obvious follow-up question is whether there are cases where the causal connection will be severed? Can you give us an example? RICHARD EPSTEIN: Here, of course, is easy to find those particular cases if you keep your eye on the distinction between a latent or hidden effect and a patent or an obvious effect. So, just change the situation that somebody puts a pit in the road, and there's some spikes there. It's the middle of broad daylight, and what somebody does is they come along and they see this particular pit. One of the things they can say is, "I have a right to use this highway, and if you create this obstruction and dangers, I'm gonna use it anyhow, and if anything happens to me then I'm going to be able to sue you for the full damages," and there's nobody in their right mind who thinks that people are allowed to do that or that anybody would want to do that. So, in the event of the perfectly obvious effect, the response is either you stop and turn around, or what you do is you go around the thing. Does this mean that you're entitled to no damages? Well, the correct answer is it's not a different form of liability, and that's the liability that is associated with blocking somebody's right of way, but if you think about it again for another second, and you could take ten steps to the left, and then go forward on your path without having to worry about anything. Do we really want to bring a cause of action for the ten second delay that is otherwise done? "De minimis non curat lex." If it turns out, however, when they put this blockade, what you have to do is to change routes and you have to go somewhere else and take a canoe or extra expenses, then, in effect, what happens is you can recover the cost of mitigation associated with avoiding the hazard, but since those costs are vastly lower than getting yourself impaled upon that, you can not go headlong into the particular situation, and so what we then do is we start to analyze the choices that somebody makes when confronted with a known difficulty to see whether or not they have or have not acted in accordance with the basic rules. To give a famous English case from about 1885 or so, which kind of illustrates this pattern, is you have somebody who owns a ship, and the ship is badly disabled when it turns out it's hit by somebody else, and the Captain, with his basically impaired vessel, incomplete control over the steering mechanism, the loss of his logs and his other devices for navigation, is trying to make his way to port, and he runs ashore on some kind of a shoal. And the question is, "Can the guy who rammed into the ship say, 'look, it's not my fault because it was he who beeps the ship, not me.'" And the answer to that was given by the English cases is you can't possibly do this, and why is that? Because it's the same thing that we talked about in the Roman situation, and that is so long as the fellow who's sailing the ship is under an impairment because he doesn't have his full control over his vessel and the full information on where he goes, he is going to be fully protected so long as we know he's trying to avoid future harm rather than to create it, and so the recovery is going to be allowed. Now, does this mean that recovery is indefinite? The answer is, of course not. And what of the situation? One is he's got an impaired boat, and he sees an obstacle in front of him, and he says, "Oh, goody," and he decides to sail the ship into trouble. At that particular point, since he had the capacity to avoid the harm and deliberately chooses to encounter it, it's going to be on his side of the ledger, not somebody else. So, we have one form of causal intervention, which ends the chain of causation. And then, the second situation is he makes it safely to port, and what somebody then does is fix the rigging on the boat and restores the logs, and he now goes out a week later than he otherwise would've done, and because he's a week later he gets hit by some kind of a storm which wouldn't have happened if he'd gone out a week earlier. Can he recover for that? And the answer is almost always, no. And the reason is if he would've come early, he would've had the same probability of being injured by a storm. So, what we do is we substitute in one probability of, say, .01, for another probability of .01, and it turns out that we just don't want to give you insurance against that probability in less there is "an increased risk or hazard of some damage taking place." So, the standard illustration of that makes perfectly good sense. What you do is you derail somebody from a train, and now it's dark and they have to walk back through the woods, and they're attacked by a stranger. It turns out there's an increased risk of hazard until you reach a position of safety, so you who created the risk are responsible, and you're responsible even though there's a third person who's obviously wrong because they've committed a deliberate trespass against your person, but once you reach the point of safety, namely the house to which you are headed, and something happens to you there, unless there's still an increased risk or hazard, which is highly likely, then you're not there. So the point about the Roman system is that it not only creates situations where you know what the liability is, but it also creates very clear rules which indicate this far and no further, so that causation doesn't have this infinite capability of spreading throughout the world and bringing everything down in its way. NARRATOR: Thank you for listening to this episode in the Roman Law unit of the No. 86 lecture series. The spirit of debate of our Founding Fathers animates all of the No. 86 content, encouraging discussion and critical reflection relative to how each subject is widely understood and taught in law schools and among law students. Subscribe to the No. 86 Lecture series on your favorite podcast platform to have each episode delivered the moment it’s released. You can also go to fedsoc.org/no86 for more lectures and videos on Property, Contracts, and the Common Law. Thanks for listening. See you in class!