Saturday, September 11, 2004

Some Responses to Blar on the Cumulative Case Argument

Blar over at Blargblog has a very interesting weblog underway. His criticisms of my Cumulative Case post are fair-minded and stimulating. I have reproduced them here, verbatim, with some responses in blue.

Taking apart the cases for war

In a recent post at Maverick Philosopher on the cumulative case for the war in Iraq, Bill Vallicella makes at least six separate points, which range from the insightful to the dubious. I'll give my gloss on six points of his, and then I'll look at these claims in order:

1. If there are several different reasons for undertaking some course of action, that course of action may be justified by the combination of these reasons even if no single reason is sufficient to justify it.

2. The war in Iraq is justifiable based on this sort of cumulative argument. Sufficient reasons for war are provided by the combination of the the risk that Saddam had WMDs, Saddam's ties to Palestinian and perhaps other terrorists, the value of liberating the Iraqi people, the need to enforce UN resolutions if the UN failed to do so, and the value of ending ongoing hostilities like Saddam's attacks on our planes.

3. A common and fallacious debating tactic, the "Divide and Conquer" technique, is to take each of the opposing side's reasons for their belief, one by one, and show that none of them are (individually) sufficient for that belief.

4. In the run-up to the war in Iraq, many liberals made the "Divide and Conquer" fallacy. For instance, when faced with the argument that it would be good to liberate Iraq, they would respond by saying something like "if liberation is so good, then why don't we liberate North Korea?"

5. Another fallacy is to confuse reasons with one person's motives. When deciding whether to support someone's efforts to undertake some course of action, you should focus on whether that course of action was justified by the reasons for it, not on the motivation that caused that person to undertake that course of action.

6. Liberals made this Reasons vs. Motives fallacy too, arguing against the war on Iraq by claiming (for instance) that Bush was after oil or out to avenge his father.

The claim that I've labeled as Mr. Vallicella's first point is correct and worth noting. People tend to overuse the form of argument that lays out a chain of reasoning, with each step depending on all of the previous assertions. The problem is that such a chain is only as strong as its weakest link. If you happen to introduce a controversial premise or if you make a shaky jump in your reasoning, then the argument will not be convincing. Especially when dealing with practical issues, like deciding on or justifying a course of action, it's better to try to connect your conclusion to various starting points with several lines of reasoning.

However, having lots of reasons doesn't necessarily mean that they all add up to something strong. Lots of bad reasons usually don't add up to a good reason, and webs of reasoning may have weak points that many of the threads pass through. Further, it's important to figure out how all of the different arguments in favor of the conclusion combine, and how they fare when stacked up against the arguments against the conclusion or in favor of alternative conclusions, in order to see whether the conclusion is justified.

BV: My question was not whether bad reasons can add up to a good reason, but whether insufficient reasons can add up to a sufficient reason. I distinguished between a bad reason and an insufficient reason. A reason is a proposition. A bad reason is false proposition. An insufficient reason is a true proposition that does not suffice to justify the action envisaged.

Blar is right that a thorough discussion of the justification for the Iraq war has to reply to the arguments against it. That's something I did not do, and could not do given space limitations. Note also that I brought in the Iraq question only to illustrate the notion of a cumulative case; my fundamental interest is in the logic of cumulative case arguments as such.

A distinction I did not make in my post, but should have, is that between making a case for an action, and making a case for a proposition. This parallels the old distinction between practical and theoretical syllogisms. Suppose I want to establish the proposition that God exists. I give three arguments. The first is formally invalid. The second, though valid, has a false premise. The third, though both valid and possessed of true premises, is circular. It follows that each argument is bad. Do three bad arguments add up to one good argument? Of course not. Here 'Divide and Conquer' applies.

Suppose I give a sound argument for conclusion C from premises P1 and P2. Clearly one cannot refute the argument by saying that P1 taken by itself does not entail C, or that P2 taken by itself does not entail C. The premises taken singly are each insufficient to entail the conclusion; when taken together, however, they are sufficient to entail it. It is worth noting that it is not the set of premises that entails the conclusion, but the conjunction of premises. A conjunction of propositions is a proposition, but a set of propositions is not a proposition: no set can be either true or false. A proposition cannot be a set, not even if it is a set of possible worlds, pace David Lewis.

I disagree with Mr. Vallicella's second claim that the war in Iraq was justified, though laying out all of my reasons for this position would take me beyond the scope of this post. One problem is that some of the reasons that he gives for war are not just insufficient but weak. What is the point, for instance, of enforcing our interpretation of a UN resolution when most of the world believes that we are acting contrary to the will of the UN?

BV: I might respond with a question of my own: What is the point of issuing a resolution if there is no will to enforce the resolution? If the U.N. resolves that Saddam shall do X, but implies by its inaction that it lacks the will to enforce the resolution, then why would the USA not be justified in enforcing the U.N.'s own resolution?

Perhaps more significantly, he ignores the two strongest reasons against the war. First, it should have been clear when we decided to go to war that a lot of people would not be happy with this. Some of these people would be Iraqis and other Arabs who might come to support or participate in terrorism.

BV: Again, I was merely illustrating the notion of a cumulative case; my intention was not to give any sort of adequate summary of all the arguments for and against the war. Indeed, I left out other reasons for the war that I might have given.

Blar is right that it could have been, and indeed was, foreseen that many people would be unhappy with the actions of the coalition forces. (And let's not forget that there was a coalition: the U.S. action was not unilateral.) But to say that we ought not topple a state sponsor of terrorism because we might provoke more terrorism is a defeatist mode of reasoning. It's like saying that I should not defend myself against an aggressor because it might cause him to become more aggressive. The correct response is to do what it takes to stop the aggressor. In the case of terrorism, the correct response is to inflict such a blow on them and the states that sponsor them that they give up. To say that that is not possible is rank defeatism.

Thus, as folks like Richard Clarke have noted, the war seemed likely to create more terrorists than it destroyed. Some of the other people who would not be happy with the war were potential allies (like Germany and the people of Spain) who could contribute to the war on terrorism in a variety of valuable ways, but might be alienated by the war in Iraq.

The second reason is the opportunity cost of the war. There are a lot of good humanitarian, anti-terrorist, and anti-proliferation things that we could be doing with over a hundred billion dollars and much of our army. Instead, we're spending our time and resources on Iraq.

BV: You need to give examples. The money I spend on food could be spent on books. But if I am starving, I will be in no mood to read any books. First things first. First eat, then read. Primum vivere deinde philosphari. Making the world safe against terrorism is job numero uno. We should spend what it takes. If I have cancer, then I'll deplete all my assets if the alternative is death.

Mr. Vallicella's third claim correctly identifies a fallacy, though I don't know how common this fallacy is. However, there is a similar form of argumentation that is not fallacious, what I might call "Divide and Debilitate". If you take the opposing side's reasons one by one, and, instead of merely showing that each reason is not sufficient for their conclusion, you show that each reason is not particularly strong, then you are arguing well, not fallaciously. Indeed, given the space limitations of almost any forum for debate, it would be almost impossible to satisfactorily address all of the opponents' arguments at once.

BV: I would need to know wehat you mean by 'strong' if that is different from 'sufficient.'

The flip side of the "cumulative case" is that it is inimical to reasonable discourse to be an endless font of arguments. If, whenever the other side challenges one of your arguments, you come back with some other argument, you create a flurry of unrelated arguments that leaves little room for comparing the strongest arguments that lie on either side of the discussion.

The fourth claim seems very distant from the actual debate over Iraq as I remember it. Though most people's reasons for their position on the war on Iraq have crystallized over time, during the run-up to that war the debate was much more of a mess. Proponents of the war tossed off a vast assortment of arguments for disarmament or regime change or hawkishness. No effort was spared in associating (at least rhetorically) the war against Iraq with the War on Terrorism. We heard a lot about aluminum tubes, a meeting in Prague, and a terrorist group in Iraq (Ansar al-Islam) which, it was rarely mentioned, happened to be outside the geographic area that Saddam controlled. There were also a wide variety of arguments against the war, some good and some bad (though I don't recall anyone making the Iraq-North Korea comparison when talking about liberation.)

Debaters struggled to deal with the proliferation of arguments, often turning to the "Divide and Debilitate" strategy. For instance, people opposed to this war in Iraq have noted that removing Saddam was unlikely to seriously weaken Palistinian terrorism, that Saddam's attempts to attack our planes are so ineffectual as to be almost irrelevant, that there seemed to be no working relationship between Saddam and Al-Qaeda, that there wasn't strong evidence that Saddam had weapons (like nuclear bombs) that were capable of causing destruction on a massive scale, that weapons inspectors were in the country searching for any such weapons, and that using our army to forcibly end a tyranny and replace it with a better government is perhaps a risky and inefficient way to pursue humanitarian and pro-democracy goals. If these arguments are correct, then it would seem that the cumulative case does not hang together well enough to justify the war in Iraq.

BV: People on the Left seem not to appreciate the following points. First, Islamic terrorism is not identical to al-Qaeda. The latter is just one of many Islamic terrorist outfits. Second, al-Qaeda is not identical to Osama bin Laden. Thus it is quite absurd to say or imply that the whole effort is directed toward capturing or killing Osama, or that failure to do the latter shows that the effort is off track or unsuccessful. Third, it is not the events of 9/11 alone that motivate the coalition's anti-terrorist efforts. There have been many terrorists attacks in many different places for many years. The target here are Islamic terrorists whatever and whenever they strike, and the governments that support them. The total effort has many phases and many parts, with Iraq being one of them. It is silly to say that Iraq was a diversion from the war on terrorists. That's like saying that curing a man's asthma is a diversion from treating his cancer.

The fifth claim (like the other odd-numbered claims) is correct, as far as it goes. However, it misses a distinction between what we might call simple and complex courses of action. A simple action is one where almost all of the input that you put into the course of action is the decision to undertake it. If you contribute money to Organization X which helps mentor young people, then the beneficial impact of that money is in the hands of Organization X and does not depend on your motives. A complex course of action is one where you continue to exert significant influence as events unfold. If you decide to mentor a young person, that decision may turn out wonderfully or horribly depending on your continued interactions. The motives, commitment, and abilities of a person who makes a complex decision matter, because the course of action only goes well as long as he guides it well.

The final claim ignores this distinction between simple and complex courses of action. A war that involves an attempt to create a stable and democratic government in a place where there has been tyranny requires continuous decision-making and adjustments. Indeed, many of the arguments against Bush are that he has botched the creation of a stable Iraqi state through incompetence, for instance, by disbanding the Iraqi army and by putting off democratic elections.

BV: But now you are shifting the debate from arguments for the war to questions about Bush's handling of matters after the cessation of major hostilities.

Thus, even though the theoretical question about going to war in Iraq may seem like a simple yes or no question, the actual war has been highly complex.

I agree that it is unlikely that Bush was driven to war by vengeance or oil. Some people on the left did make these claims, so refuting them does not exactly involve the creation of straw men. Instead, it is more like using a straw-seeking missile to figure out which arguments to refute. There were many arguments put forth of the form "I doubt Bush's commitment & ability to bring about the state of events that I would like to see as a result of a war in Iraq." Instead of focusing on oil and vengeance, the stronger instances of this argument noted Bush's disparaging statements on nation building during the 2000 election, the administration's low-balling of the estimates of the resources needed in Iraq, the utopianism of the people advising the President who envisioned American soldiers greeted as heroes and the easy creation of a democratic Iraq, or the continuing failure to control all of Afghanistan and create a democracy there. It is the duty of someone looking for rational debate to seek out the strongest arguments on the other side, rather than to seek and destroy the weakest claims around.

BV: 9/11 changed everything, including Bush's attitude toward nation-building.

In sum, this whole question of Iraq is difficult. I am not certain I am correct, but there is a very strong case for the war, a case that liberals and leftists do not properly address. There is also a case against it, the strongest version of which is the one put forward by Pat Buchanan.