The second common move in arguments that's supposed to solve the skeptical regress problem is guarding. The basic idea of guarding is very simple. It's just making the premises of your argument weaker so that there will be fewer ways in which your opponents can raise trouble for them, or show that they're false. So, here's an example. Suppose someone says, we ought not to build any new nuclear power plants because they'll explode. Now, wait a minute. How do you know that any of them really will explode? How can you tell that in the future? You don't even know what kinds of standards they're going to be built to. So, someone can object to that argument by saying, you're not justified in asserting that premise that these new nuclear power plants will explode. So, how do you stop that problem? Well, you simply say, we ought not to build any new nuclear power plants because some of them might explode or even weaker because I believe that some of them might explode. Now, we've got a premise. I believe that some of the new nuclear power plants might explode. That's going to be hard to deny. Okay, what's an opponent going to do? Show that you don't really believe it? Show that it's not true that they might explode, that there's no possibility that they'll explode? Or that any of them will explode? You're not claiming all of them will, just some of them. So, by guarding the premise in this way, you make it more likely to be true and less subject to objection, and that's what guarding does. It enables you to start an argument with that premise if other people agree that it's true because it's so weak. You get them to agree to share your assumption by weakening your premises. Now, of course, what happens now is someone says, wait a minute. They might explode. Sure, and the sun might not come up tomorrow. All kinds of things might not happen or might happen. Might's too weak to establish that we shouldn't have new nuclear power plants. So, what you're suggesting is not just that it might happen, but that it's likely to happen. And, if the chance of it happening really is so slight as to be negligible, then you weaken the premise too much. And it's not going to follow that we shouldn't have nuclear power plants. So, the issue's going to come down to, is the risk of them exploding or some of them exploding enough to justify the conclusion that we ought not to have them? And that's going to depend on exactly how much risk there is. And by looking at the guarding term at the might term, and questioning it, and saying, can we replace might with probably or something like that? Then, we're going to have a better handle on how to asses the argument. So, when you see someone using guarding like this, you need to ask, why did they put in the guard? And, have they put in too much guarding that has weakened it so much, that the conclusion no longer follows? So, the general trick of guarding is to weaken the premise so it's going to be harder to deny. And that's how your argument gets going. But there are at least three different ways to do this, okay? One is the extent, the other is probability, and the other is mental. First, guarding by extent. We need a new alcohol officer on our campus because all students drink too much. Well, that's clearly false. Not all students drink too much. Most students drink too much. Well, not most. Maybe not most. Many students drink too much. Okay, many, too many because it's too much. Some students drink too much. Notice that you can guard or weaken the claim, the premise, from all students drink too much to most students drink too much to many students drink too much to some students drink too much. And as you move down that scale, the premise gets harder and harder to deny. So, whether this counts as guarding depends on the expectation. If you expect the claim that most or all students drink too much, then it's guarding to say many or certainly to say some. But if you don't expect the many or most or all, then to say some is simply to say that you're talking about some. So, it's guarding when you weaken it beyond what would otherwise be expected in the context. This standard is going to be hard to apply because it might be difficult to say what the expectations are of the different people involved in a conversation. But, that's what guarding is. It's not guarding every time you use the word mini or most. It's only guarding when you're expecting all in the person, instead merely claimed mini. The second kind of guarding concerns probability. Some people would say it's absolutely certain that O. J. Simpson killed his wife. And other people said, well, he probably killed his wife. Or it's likely that he killed his wife. And others will say, there's a chance he killed his wife. And others will say, he might have killed his wife. So, when you change from it's certain to it's probable, or it's likely, or there's a chance, or he might of, again, you're moving down a continuum. And the further you move down that continuum, the easier it is to defend your premise, because you're claiming less. And the question is going to be whether you've done it too much. If you simply say, he might've killed his wife, therefore he ought to be convicted. That's clearly wrong. If you say he probably killed his wife, so he ought to be convicted. That might be wrong too if there's a strong burden on the prosecution to prove guilt beyond a reasonable doubt. But you don't want to require that it's certain that he killed his wife, because then you'll never be able to convict any criminals. So, you need something like almost certain. You know, which is a bit of a fudge word, or beyond a reasonable doubt. In any case, however you assess whether or not he should have been convicted depending on how you assess how likely it was that he really did what he was accused of. Apart from all those questions, the point here is simply that in arguments in general, you can make the argument more defensible by weakening the premise so there are fewer ways to show that the premise is wrong. And that's the second type of guarding. Now, the third kind of guarding we can call mental because it has to do with the mental state of the person asserting the premise. You might say, well, I know that the President is 50 years old. But you might say, I believe that the President is 50 years old. You might say, I tend to believe I or I'm inclined to believe that the president is 50 years old. And there's another continuum as you move from knowledge to belief to inclination to believe. Again, you are making the premise weaker and weaker which makes it harder and harder to question, or deny, or doubt that premise. So, you avoided a problem for your argument, and potentially this is a way to stop the skeptical regress.