Posts Tagged ‘ problem of induction ’

4. Two kinds of argument

There are many ways premises can support their claims. In this post we’re going to look at two common forms widely used in academic and popular writing: inductive generalisations and arguments from authority. To begin, first read Sarah Graham’s short article “Effects of Smoking May Be Passed Down Through Generations”.

Look at the structure of each paragraph in Graham’s article. Paragraph 1 claims that the dangers of smoking for pregnant women may be greater than previously believed. The support for this claim is that research has found that the grandchildren of women who smoked when they were pregnant are more likely (double, in fact) to have asthma.

By the lights of logic, this is an informal but very common way of arguing: the writer tries to persuade you of the argument’s truth by citing an authority that you should believe: recent research. Hence, this kind of argument is called ‘Argument from authority’.

Other arguments from authority are

1. Jesus was the Son of God. How do I know? It says so in the Bible.
2. Darwin’s principle of natural selection is true for sure. It must be because it is accepted by modern science.
3. Pam is cheating on her boyfriend. How do I know? She told me so.
4. An argument’s definitely not the same as an explanation. I don’t know why, but that’s what it says in this text book.
5. Wittgenstein famously said that ‘the world is all that is the case’.
Therefore, we know that anything that isn’t the case in not part of the world.

Arguments from authority are only as good as the reliability of the authority being quoted. As an independent and critical thinker, you should never wholly accept a claim just on the basis of an argument from authority. If you do, you have to accept that others might disagree with you simply because they do not trust that authority. Good critical thinkers must judge an argument for themselves, and not believe it just because others tell them to.

Paragraph 2 of Graham’s article claims that if a woman smokes while she is pregnant, both her children and grandchildren may be more likely to have asthma as a result. The support for this claim comes from three premises:

Premise 1:

children of women who smoked while pregnant were 1.5 times as likely to develop asthma as the offspring of nonsmokers were.

Premise 2:

If both the mother and grandmother smoked during pregnancy, the risk increased to 2.6 times that of children of nonsmokers.

Premise 3:

Most surprising, even when a mother did not smoke while she was pregnant her child had nearly double the risk of developing asthma as a child from a smoke-free home if her mother had smoked during pregnancy.

This form of argument is known as ‘Argument from induction’. Inductive arguments are based on evidence, observation and past experience. Most scientific arguments are inductive. If the premises are true, you have good reason to believe the claim, but arguments from induction are never certain – they are only probable (look at the language used in the claim: ‘may be more likely’). It is possible that further evidence could undermine the argument (just check the history of science to see!).

Some philosophers have been so upset by this observation that they refuse to accept any argument based on induction. Referring to the problem of induction, the 18th Century Scottish philosopher David Hume famously complained that the past was not a reliable guide to the future.

While inductive reasoning cannot rule out the logical possibility that things might change, it remains true that, in practice, we could neither live our daily lives nor do science if we did not place regular faith in inductive generalisations.

Other arguments from induction are

1. Every swan we have ever seen is white. Therefore, all swans are white.
2. The sun always comes up in the morning. Therefore, the sun will come up tomorrow morning.

You will notice that 1 and 2, both contain universal statements. These are statements that have the logical form ‘All As are B’ (like “all swans are white” in 1), or which can easily be parsed into that form (as “The sun always comes up in the morning” can be rephrased as “All mornings are mornings with sunrises”).

Most universal statements are inductive generalisations, meaning we think that they are true because we have observed instances of them many times. Nonetheless, be careful because some universal statements are not generalisations based on evidence. Consider, for example,

‘All unicorns have only one horn.’

This universal statement is not an inductive generalisation. It is a semantic tautology; that is, a statement that is true by definition: the meanings of the terms make the proposition as a whole necessarily true (you may remember we saw another kind of tautology in post 1: What is an argument).

Other universal statements like this are

‘Every bachelor is an unmarried man.’
‘No object can be coloured both red and green all over at the same time.’ (this is a logical impossibility ruled out by the meanings of “coloured all over” and “at the same time”).

For some universal statements, it remains unclear whether they are semantic tautologies or inductive generalisations. For instance, are the universal statements

‘Every event has a cause.’ and  ‘Nothing can be in two places at the same time.’

inductive generalisations, or simply true because of the meaning of the terms in the sentence? Philosophers and quantum physicists seem unable to agree, but for now it is enough that we recognise that universal statements usually express either inductive generalisations or semantic tautologies.

• premises can provide different kinds of support for a claim. Two kinds of support are ‘argument from authority’ and ‘argument from induction’
• universal statements have the form ‘All As are B’. Some universal statements are true because of the meaning of ‘A’ and ‘B’, while others are inductive generalisations based on past experience

Try Exercise 4 to test your understanding of this post, or continue reading

%d bloggers like this: