21 Arguments VI: Inductive Arguments
I. Introduction
The last chapter introduced the distinction between deductive and inductive arguments. Deductive arguments are those whose conclusion is supposed to follow with logical necessity from the premises, while inductive arguments are those that aim to establish a conclusion as only being probably true, given the premises.
To define arguments in this way is to define them in terms of their aim or intent. That means there is a question as to whether the intention succeeds or not. One of the things that marks out inductive from deductive arguments is that there are degrees of success possible for inductive arguments. That’s because we’re reasoning about claims that involve probability and uncertainty. Deductive arguments, by contrast, are like on-off switches. Either they work logically, or they don’t.
So, let’s start this chapter by defining terms that will let us talk about successful and unsuccessful inductive arguments. For this we use the language of strength and weakness, which, just like real strength and weakness, come in degrees.
- A strong argument is an inductive argument that succeeds in having its conclusion be probably true, given the truth of the premises.
- A weak argument is an inductive argument that fails in having its conclusion be probably true, even given the truth of the premises.
With this in mind, let’s next see how we can identify inductive arguments. Then we’ll put these things together and see how we can determine when the arguments we’ve identified are strong or weak.
II. Identifying Inductive Arguments
There’s an art to knowing when you’ve got an inductive argument on your hands, and with practice you can become good at it. The general idea is to think about the reasons being offered for a claim and to ask whether someone offering those reasons is doing so in a way that says or implies that there is some uncertainty involved, and thus some degree of probability of the conclusion being true. Here are some things to look for that can help you do this.
1. Just as the parts of an argument often have indicator words that clue you in to what you’re looking at, there are certain words and phrases that often – but not always! – show up to make clear that a conclusion is being inductively rather than deductively drawn. These include the following (but I bet you can think of other words that function in a similar way once you see how these work):
Inductive Argument Indicator Words
probably
likely
there’s a good chance that
in all likelihood
tends to
2. Arguments whose conclusion is a claim about the future, where it is made based on the way things have typically happened in the past, are inductive. That’s because, even if experience provides a good basis for making predictions, the future is not fully knowable and can always surprise us.
Arguments for Claims about the Future
It rained yesterday, and it’s raining today. Therefore, it will rain tomorrow.
The Pirates did not make it to the playoffs the last three years in a row. So, they probably won’t make it to the playoffs next year either.
The sun will rise in the east tomorrow, for it has risen in the east every day for the last 4.5 billion years.
The Democrats will likely pick up seats in Congress, because midterm election years tend to favor the party that’s in the minority.
3. Generalizations are arguments that involve making a claim about a large group if things (or people) based on what is known about a small subset of that group. These are always inductive. Lots of claims in social sciences, health sciences, and the like, are supported by arguments of this kind. We’ll go into these in detail in the next chapter.
Examples of Generalizations
You have shortness of breath, coughing, and have lost your sense of smell. Those are widespread symptoms among people diagnosed with COVID-19. There’s thus a good chance that you are infected with COVID-19.
You should eat more fiber in your diet, for a study of several hundred people on a high-fiber diet showed a variety of improved health outcomes, and you should try to have good health outcomes.
A large survey of people that looked at their salaries and levels of educational achievement found that most of those with higher salaries had completed at least a four-year college degree. Therefore, people who complete a college degree tend to earn more than those who do not.
A vaccine was tested on a sample of several thousand people and it proved effective. Therefore, the vaccine will be effective for the entire population.
I met a person from the Upper Peninsula of Michigan the other day, and she had a very strange accent. I bet all those people up there talk strangely.
4. Analogies/analogical arguments are arguments in which a prediction is made about how two things (or people or groups) are likely to be similar based on known similarities between them. These are always inductive. (These will also be dealt with in more detail in the chapter after the one on generalizations.)
Examples of Analogies/Analogical Arguments
Cats have eyes that are physiologically very similar to human eyes. When mascara was put into cats eyes in a study, they became mildly inflamed. Thus, it’s likely that the same mascara will inflame human eyes if it comes into contact with them.
Dogs have four legs and like to bark. Cats have four legs. Therefore, cats probably like to bark too.
McDonalds has inexpensive burgers and fries, and it pays its workers low wages. Burger King has inexpensive burgers and fries. So I bet it pays its workers low wages too.
Note: the first of these examples of analogies is a real-world example of how animals are used to test the safety of cosmetic products. There is a factual issue about the effectiveness of such testing as well as an ethical issue about whether it is inhumane or not. The analogical argument given here presumes that such testing is an effective way to determine the truth of non-evaluative claims, such as the conclusion in this argument about mascara and its effect on human eyes. But even if such non-evaluative arguments are strong, they do not settle one way or another the issue of the ethical acceptability of such testing.
