Don’t Jump to Conclusions
The Role of Inferences in Making and Analyzing Arguments
Consider the following argument:
Chocolate makes you smarter.
Therefore, I should eat lots of chocolate every day.
While many of us would like this argument to be true, it just doesn’t feel like it is. Even if we grant that the first statement (chocolate makes you smarter) is true, we find it hard to take the step from that to the conclusion of eating lots of chocolate every day. Why?
It is because arguments are not just about making true statements but about connecting statements to each other, or inferences. We can think of true statements as bricks and inferences as the mortar that holds a brick wall together. Without the mortar, it is very easy for the wall to fall apart. Thus, many bad arguments don’t fail because of bad facts, but because of bad inferences.
Inferences help us in two ways. First, understanding inferences helps us establish connections within an argument. For example, we can link conclusions. Second, they help us connect the argument to other arguments. But first, let’s look at what implications and inferences are.
Defining Inference
The easiest way to think of inferences is to think of “If… then…” statements. For example, if it rains then it is wet outside. However, inferences can be more than just simple one step connections. If it rains, then you can infer more than simply that it is wet outside. When it rains, the bus might also be late, you might need to take an umbrella with you to work, and the local cafe will be more crowded as people seek to avoid the rain.
This means that inferences can be explicit, implicit, and even something that you have not thought about. No matter what, inferences are logical steps from one fact to another. There are three common types of inferences that you will encounter: deductive, inductive, and abductive inferences.
Deductive steps go from general rules to particular instances. For example, the general rule of all copper conducts electricity leads to a particular copper wire conducting electricity. The strength of this type of inference is that it logically guarantees truth, that is, if the first step (also known as premise) is true then the next step (e.g. the conclusion) has to be true as well. The end result is that you are logically compelled to believe in the conclusion if you believe in the premises in that you cannot agree with one and not agree with the other.
Inductive inferences go the opposite direction from deductive inferences as they go from particular cases to general rules. This is what science uses to create scientific theories. You take a lot of particular cases (e.g. many experiences of copper conducting electricity) and conclude that all copper conducts electricity. The important thing to note about inductive inferences is that they do not guarantee truth. Just because lots of copper conducts electricity does not mean that the next piece of copper you pick up will conduct electricity. Despite this, inductive inferences are helpful because they let us know what usually happens.
Finally, abductive inferences are inferences to the best explanation and are not about truth or probability. In other words, it goes from particular case to particular case. Contrary to what is usually said, the famous detective Sherlock Holmes uses abductive inferences to solve his cases. The best explanation for death by dog is, while possible, not that the dog is a supernatural demonic animal, but a real dog used by someone with motive to commit murder.
The best form of these three inferences is deductive as the truth of our conclusions is always going to be true. However, they are (1) hard to make and (2) often not very informative. For example, consider the following deductive argument:
Premise 1: All dogs are mammals.
Premise 2: Fido is a dog.
Conclusion: Therefore, Fido is a mammal.
By knowing that Fido is a dog, we already know that he is a mammal because being a mammal is part of the definition of what it is to be a dog. Thus, we have not really learned anything from the argument. Philosophers have argued that deductive reasoning simply ‘unpacks’ concepts rather than revealing new bits of information. Despite this, deductive inferences can be helpful because they do show us things that we should already know but have failed to realize.
Building an Argument
Now that we know what inferences are and common types of inferences, let’s look at how inferences are used to make arguments and how to evaluate them. First, inferences are used to evaluate the consistency and logic of an argument.
In every argument, there needs to be some connection between the parts or else it cannot be called an argument to begin with. Identifying the nature of the inferences between the parts ensures that there is, in fact, an argument being presented.
Back to our example of it raining and the ground being wet, there is a connection between the two. Rain — in other words, water falling from the sky — makes things wet. Therefore, it is reasonable to assume that wet ground is caused by rain. By contrast, there is no reasonable connection between rain and the grass being green. Rain alone cannot explain green grass.
Besides simply connecting parts of an argument together, evaluating inferences also ensures that the connections between the parts of an argument are correct. In other words, once we have established there is a connection between two parts of an argument, the nature of that connection also tells us the criteria by which I should be judged. If, for example, the connection is a deductive inference, then the connection must be one of moving from a general rule to a particular instance where truth will be guaranteed.
A final thing to check for within an argument is that there are no inference mistakes. Common inference mistakes are called fallacies. There are too many to go into them all one by one, but let’s look at one of the most common as an example, denying the antecedent also called the inverse fallacy.
If A then B (If it rains, then the ground is wet)
Not A (it did not rain)
Therefore, not B (therefore the ground is not wet)
While this might not sound so bad at first glance, it is an obvious mistake. It is true that if it rains then the ground is wet, but there are many ways for the ground to be wet (a sprinkler, someone pouring out a cup of water, a busted water main, etc.) besides rain. Thus, the conclusion is not an acceptable inference from the two premises.
Connecting Arguments to Life
The second aspect of inferences are that they take us outside of the topic at hand to help us evaluate it. This is to say that we do not just consider the connections between the parts of an argument but also how it connects with other parts of our lives.
An example of seeing out the full implications of a position is Jainism. Jainism is an ancient religion that claims it is wrong to harm any living thing. The followers of Jainism take this statement very seriously as they do not just avoid killing people but abstain from eating plants and animals, and even washing their bodies, as it could lead to killing microscopic life.
In other words, Jainism infers from “any living thing” that it should be taken in the most inclusive sense in that literally all living things that could be even remotely considered living should not be harmed.
Going in the other direction, the abortion debate in the US is often centered around how exclusive a concept should be. One of the many issues debated is whether to call the thing growing in a woman’s womb a ‘person’ or not. This is because calling it a person has serious implications. For instance, the US has laws charging a driver with two counts of vehicular manslaughter when a pregnant woman is killed in a car accident and gives parents tax deductions for children which includes the unborn baby of a pregnant woman.
By not using the word ‘person’ and instead calling it something like ‘fetus,’ supporters of abortion try to prevent connections (i.e. inferences) between abortion and discussions such as the ones above and others related to the killing of a person such as murder, assisted suicide, or the death penalty. This shows just how powerful implications can be to an argument.
Conclusion
In the end, good arguments are more than simply a list of true facts. A good argument is also about the connections between the various parts of an argument and how the argument fits into our larger intellectual lives. Thus, it is important to look at the reasoning, not just the conclusions, in the argument when we are evaluating it.
In fact, these connections are so important that the nature of logical connections (deductive, inductive, abductive) are a hot topic in modern philosophy. The best example of this is the problem of induction which has been raised by famous philosophers such as David Hume and Karl Popper.
The problem states that it seems impossible for any amount of experience to justify future beliefs. In other words, no amount of experience with copper warrants establishing a scientific law that all copper conducts electricity or that it will continue to do so in the future. The problem of induction then brings into question not just philosophical claims but also scientific claims as well.
Despite ongoing philosophical debates, we must connect facts to each other and to our lives if we are to build a robust intellectual life and thus, we rely heavily on inferences; making them something that cannot be overlooked.