CHAPTER 12
ARGUMENT ANALYSIS
12.1 WHAT IS A GOOD ARGUMENT?
We have discussed different aspects of arguments. It is time to consider how they help us explain what a good argument is. Intuitively, a good argument is one in which the premises provide good reasons for the conclusion. This is of course quite vague. Let us try to make it more precise.
Condition 1: The premises are true or highly plausible
The premises of a good argument must be known to be true, or they have to be at least highly probable. This criterion should be rather obvious. We have no reason to accept an argument if the premises are false or are unlikely to be true.
Condition 2: The argument is deductively valid or inductively strong
Deductively valid arguments are of course valuable. Valid arguments cannot lead us from true premises to false conclusions. But we have seen that inductively strong (and hence invalid) arguments play an equally important role in reasoning. It will be too restrictive if we demand that all good arguments must be valid. We would have to give us most of our scientific knowledge. Note that if a good argument is either valid or inductively strong, this implies that an inductively weak argument can never be a good argument.
Condition 3: The premises are not question begging
The first two conditions are still not sufficient for a good argument. Consider this circular argument, where the conclusion appears as a premise:
This is surely a bad argument since no independent reason has been given to show why oatmeal is healthy. However, the argument is actually sound. First, the premise is indeed true, because oatmeal has lots of fibre and can lower blood cholesterol. Furthermore, the argument is valid. Since the premise is the same as the conclusion, it is impossible for the conclusion to be false when the premise is true! This example shows that not all sound arguments are good argument. To deal with this problem, we should require that the premises of a good argument cannot be question-begging—that is, they should not assume what the argument is trying to establish. (See also page 176 for further discussion.)
Condition 4: The premises are all relevant to the conclusion
Consider this argument:
This is presumably a non-question-begging, sound argument. If we accept the premises, we ought to accept the conclusion. Yet there is something wrong with the argument—namely, that the third premise is irrelevant to the conclusion, even though it is true. If we remove this particular premise, it does not affect the strength of the argument at all. The extra premise is a distraction and liable to create confusion, and it fails to provide a good reason for the conclusion. Bearing this in mind, we should require that a good argument does not contain any irrelevant premises.
Summary: Definition of a good argument
1. The premises are true or highly plausible.
2. The argument is deductively valid or inductively strong.
3. The premises are not question-begging.
4. All the premises are relevant to the conclusion.
12.2 FOUR WAYS TO ATTACK AN ARGUMENT
Now that we know what a good argument is, what should we do when we come across an argument that is not good? It is important not to rely on your gut feeling and just dismiss the argument. See if you can think of one or more reasons why the argument should be rejected. In general, there are four main ways to attack an argument: two direct methods and two indirect ones:
1. Direct method 1: Attack the premises. If you can show that an argument relies on at least one implausible premise, that is a good way of showing that the argument is not good enough. But sometimes you do not have to go all the way to show that a premise is false. You might argue that there is simply not enough evidence to show that the premise is true. This falls short of arguing that the premise is false, but it passes the burden of proof to the opponent. But remember, just because an argument has a false premise, it does not follow that the conclusion is false!
2. Direct method 2: Attack the reasoning. Even if the premises are all very plausible, you need to check whether the reasoning of the argument is acceptable. The argument might be invalid or inductively weak, or question begging.
3. Indirect method 1: Attack the argument indirectly by attacking the conclusion. If you can show that the conclusion of an argument is false, this implies that there must be something wrong with the argument. This strategy of refuting an argument is useful when it is difficult to evaluate an argument directly, perhaps because it is too long or convoluted. Of course, this strategy does not really explain what is wrong with the argument.
4. Indirect method 2: Give an analogous argument that is obviously bad. The idea is to compare the original argument with another argument. If the new argument is obviously bad, and it has the same structure as the original one, then the original one is likely to be a bad argument as well. This is a good strategy to use when it is difficult to see what is wrong with an argument, or your opponent refuses to admit that the argument is no good.
As an example, consider this argument:
Capital punishment is wrong because it is always possible to punish an innocent person by mistake.
We might attack the argument using the four methods as follows:
- Attack the premises: Is it always possible that an innocent person is executed by mistake? It might be argued that in some crimes there were many independent witnesses. Perhaps the criminal was apprehended right away at the crime scene, and the whole crime was recorded on surveillance video. There is therefore little doubt that the person being caught is guilty.
- Attack the reasoning: Even if mistakes are always possible, this is just one consideration and it does not immediately follow that capital punishment is wrong. Maybe there are many other considerations in support of capital punishment. We need to balance these factors before deciding whether capital punishment is acceptable or not.
- Attack the conclusion: Punishment should be proportional to the crime. Capital punishment is not wrong because this is what justice requires in the case of hideous crimes.
- Give an analogous argument that is obviously bad: With imprisonment, it is also possible to punish an innocent person by mistake. But it would be absurd to stop sending people to jail because of this.
Of course, there is a lot more we can say about capital punishment. The responses just given might not be very convincing, and you need not agree with any of them. They serve only to illustrate the fact that many arguments can be attacked in more than one way.
12.3 ARGUMENT ANALYSIS: CHECKLIST
The analysis of argument is one of the most basic parts of critical thinking. To sum up, there are three main steps:
1. Clarify the argument.
2. Evaluate the argument.
3. Think about further relevant issues.
We have already discussed many aspects of this process. The table below lists some of the main tasks involved. You can use this as a checklist when you want to analyze an argument systematically. It would be good to internalize these steps as part of your natural thinking habit:
| Step | Tasks and questions |
| 1. Clarify the argument | Identify premises and conclusion. |
| Clarify the keywords. | |
| Simplify the argument using your own words. | |
| Draw an argument map. | |
| 2. Evaluate the argument | Is the argument a good one? |
| Are the premises plausible? | |
| Is the argument valid or inductively strong? | |
| Any fallacy in the argument? (See Chapter 19.) | |
| Any reason to think that the conclusion is false? | |
| Any obvious counterexample? | |
| 3. Explore further issues | How good is the argument overall? |
| How important is the argument? | |
| Is the conclusion surprising? | |
| Can the argument be repaired or improved? | |
| Are there other arguments with similar conclusions? | |
| What about arguments with the opposite conclusion? | |
| Can the argument be applied elsewhere? | |
| Any further information that might be relevant? |
EXERCISES
12.1 Are these statements true or false?
a) Not all sound arguments are good arguments.
b) Some inductively strong arguments are not good arguments.
c) If a good argument is invalid, it also cannot be inductively weak.
d) A good argument that is not inductively strong must be valid.
e) A sound argument can contain irrelevant premises.
f) If an argument is inductively weak, it cannot be a good argument.
g) If an argument is not good, it is either inductively weak or unsound.
12.2 Which of these arguments are question begging?
a) I like anything that is sweet. Chocolate ice cream is sweet. So I like chocolate ice cream.
b) I like chocolate ice cream because chocolate ice cream is what I like.
c) I like chocolate ice cream best, because it is my favorite ice cream.
d) I ordered chocolate ice cream, because it is my favorite ice cream, and I always order my favorite ice cream.
12.3 Apply the direct and indirect methods to criticize these arguments:
a) Cloning animals or human beings is unnatural, so it is wrong and we should not do it.
b) We should not trust scientists because they keep on changing their theories. Today they say that this is true. Tomorrow they come up with a different theory and say something else.
c) It is useless to punish students because they will always make mistakes.
12.4 Evaluate these arguments and in particular pay attention to any hidden assumptions.
a) Low taxation is desirable because it has lots of benefits. It is good for business and investment, and the citizens are likely to be happier.
b) The universe could not have existed forever. Radioactive material will decay until they cease to become radioactive. Since the universe still contains radioactive material, there must have been a beginning. Otherwise all the radioactive material will have long been gone already.
c) Indira donated $2 million to the Democratic Party and only $500,000 to the Liberal Party. So she is probably going to vote for the Democratic Party.