Module: Argument analysis
Quote of the page
We do not learn from experience... we learn from reflecting on experience.
- John Dewey
In this tutorial we shall discuss what a good argument is. The concept of a good argument is of course quite vague. So what we are trying to do here is to give it a somewhat more precise definition. To begin with, make sure that you know what a sound argument is.
This means that if we have an argument with one or more false premises, then it is not a good argument. The reason for this condition is that we want a good argument to be one that can convince us to accept the conclusion. Unless the premises of an argument are all true, we would have no reason to accept to accept its conclusion.
Is validity a necessary condition for a good argument? Certainly many good arguments are valid. Example:
All whales are mammals.
All mammals are warm-blooded.
So all whales are warm-blooded.
But it is not true that good arguments must be valid. We often accept arguments as good, even though they are not valid. Example:
No baby in the past has ever been able to understand quantum physics.
Kitty is going to have a baby soon.
So Kitty's baby is not going to be able to understand quantum physics.
This is surely a good argument, but it is not valid. It is true that no baby in the past has ever been able to understand quantum physics. But it does not follow logically that Kitty's baby will not be able to do so. To see that the argument is not valid, note that it is not logically impossible for Kitty's baby to have exceptional brain development so that the baby can talk and learn and understand quantum physics while still being a baby. Extremely unlikely to be sure, but not logically impossible, and this is enough to show that the argument is not valid. But because such possibilities are rather unlikely, we still think that the true premises strongly support the conclusion and so we still think that the argument is a good one.
In other words, a good argument need not be valid. But presumably if it is not valid it must be inductively strong. If an argument is inductively weak, then it cannot be a good argument since the premises do not provide good reasons for accepting the conclusion.
For more information about inductive strength, see the previous tutorial.
Notice that criteria #1 and #2 are not sufficient for a good argument. First of all, we certainly don't want to say that circular arguments are good arguments, even if they happen to be sound. Suppose someone offers the following argument:
It is going to rain tomorrow. Therefore, it is going to rain tomorrow.
So far we think that a good argument must (1) have true premises, and (2) be valid or inductively strong. Are these conditions sufficient? The answer is no. Consider this example:
Smoking is bad for your health.
Therefore smoking is bad for your health.
This argument is actually sound. The premise is true, and the argument is valid, because the conclusion does follow from the premise! But as an argument surely it is a terrible argument. This is a circular argument where the conclusion also appears as a premise. It is of course not a good argument, because it does not provide independent reasons for supporting the conclusion. So we say that it begs the question.
Here is another example of an argument that begs the question :
Since Mary would not lie to her best friend, and Mary told me that I am indeed her best friend, I must really be Mary's best friend.
Whether this argument is circular depends on your definition of a "circular argument". Some people might not consider this a circular argument in that the conclusion does not appear explicitly as a premise. However, the argument still begs the question and so is not a good argument.
Here, plausibility is a matter of having good reasons for believing that the premises are true. As for relevance, this is the requirement that the the subject matter of the premises must be related to that of the conclusion. Why do we need this additional criterion? The reason is that claims and theories can happen to be true even though nobody has got any evidence that they are true. If the premises of an argument happen to be true but there is no evidence indicating that they are, the argument is not going to be pursuasive in convincing people that the conclusion is correct. A good argument, on the other hand, is an argument that a rational person should accept, so a good argument should satisfy the additional criterion mentioned.
So, here is our final definition of a good argument :
A good argument is an argument that is either valid or strong, and with plausible premises that are true, do not beg the question, and are relevant to the conclusion.
Now that you know what a good argument is, you should be able to explain why these claims are mistaken. Many people who are not good at critical thinking often make these mistakes :
- "The conclusion of this argument is true, so some or all the premises are true."
- "One or more premises of this argument are false, so the conclusion is false."
- "Since the conclusion of the argument is false, all its premises are false."
- "The conclusion of this argument does not follow from the premises. So it must be false."
Answer the following questions.
These are some arguments (or just premises) that students have given to support the idea that there is nothing morally wrong with eating meat. Discuss and evaluate these arguments carefully. Think about whether the premises are true, and whether they support the conclusion that it is morally acceptable to eat meat.
This section is a more abstract and difficult. You can skip this if you want.
One interesting but somewhat difficult issue about the definition of a good argument concerns the first requirement that a good argument must have true premises. One might argue that this requirement is too stringent, because we seem to accept many arguments as good arguments, even if we are not completely certain that the premises are true. Or perhaps we had good reasons for the premises, even if it turns out later that we were wrong.
As an example, suppose your friend told you that she is going camping for the whole weekend. She is a trustworthy friend and you have no reason to doubt her. So you accept the following argument as a good argument:
Amie will be camping this weekend. So she will not be able to come to my party.
But suppose the camping trip got cancelled at the last minute, and so Amie came to the party after all. What then should we say about the argument here? Was it a good argument? Surely you were justified in believing the premise, and so someone might argue that it is wrong to require that a good argument must have true premises. It is enough if the premises are highly justified (of course the other conditions must be satisfied as well.)
If we take this position, this implies that when we discover that the camping trip has been cancelled, we are no longer justified in believing the premise, and so at that point the argument ceases to be a good argument.
Here we prefer a different way of describing the situation. We want to say that although in the beginning we had good reasons to think that the argument is a good one, later on we discover that it wasn't a good argument to begin with. In other words, the argument doesn't change from being a good argument to a bad argument. It is just that we change our mind about whether the argument is a good one in light of new information. We think there are are reasons for preferring this way of describing the situation, and it is quite a natural way of speaking.
So there are actually two ways to use the term "good argument". We have adopted one usage here and it is fine if you want to use it differently. We think the ordinary meaning of the term is not precise enough to dictate a particular usage. What is important is to know very clearly how you are using it and what the consequences are as a result.