Writing As The Key To Competence – Part 3: How To Make & Break Arguments

In my previous article, I discussed how writing can develop your critical thinking skills…something that’s fundamental to thriving in the 4IR because machines can’t do this very well (yet). After unpacking the different components of what it means to be a critical thinker, I wanted to dive more deeply into an understanding of what distinguishes a good from a bad argument.

But, to come to grips with this, you will need a quick lesson in argument structure and development. So, it might a get a bit technical, but you need to grasp this to be able to assess and attack arguments with skill. Here we go…

Arguments can be divided into two parts – the premises and the conclusion. The premises are the reasons or propositions that together support the truth of the conclusion. The premises take the form of declarative sentences that are either true or false. And when we say that an argument is “good”, we mean that the premises are true, and the conclusion follows from the premises – i.e. they are related in the right way.

Now, here’s where it gets a bit tricky…There are different kinds of arguments – deductive, inductive, and abductive arguments. It’s important to understand how these differ and what it takes to make them persuasive. So, let’s jump right in with the first one.

Digging Into Deductive Arguments

A deductive argument is one that focuses on logical structure. If the premises are true, then the conclusion is necessarily true because the logical structure guarantees this. Here’s a simple example of a deductive argument…

1. If Paul lives in Johannesburg, then he lives in South Africa.
2. Paul lives in Johannesburg.
3. Therefore, Paul lives in South Africa.

The premises of the argument are expressed in the first and second sentences, while the conclusion is expressed in the third sentence. To see that this argument is deductively valid, we need to identify its logical structure. We do this by replacing each proposition with a variable, like so:

1. If A, then B
2. A
3. Therefore, B

The “if…then…” logical connective is what we call a hypothetical, and once we’ve isolated the logical structure, we can test whether it is the kind of structure that will guarantee the truth of the conclusion (if the premises are true). We test this by seeing whether we can come up with a counterexample. In other words, is there any content you can plug into these variables where the conclusion does NOT follow?

To cut to the chase, the answer is ‘no’. Any content that follows this structure will guarantee the truth of the conclusion (again, if the premises are true). But what would happen if we inverted the propositions? Would this argument also be deductively valid? Let’s test it:

1. If Paul lives in South Africa, then he lives in Johannesburg.
2. Paul lives in South Africa.
3. Therefore, Paul lives in Johannesburg.

The logical structure now looks like this:

1. If B, then A
2. B
3. Therefore, A

But clearly the conclusion doesn’t follow from these premises because Paul could live in South Africa but live in Cape Town rather than Johannesburg. This means the logical structure does not yield a deductively valid argument. We say an argument is deductively sound when the logical structure is deductively valid AND the premises are all true. This would be an unbeatable argument!

There are many different logical structures we can test and a number of them have names because they are so common. The first argument structure we looked at in this article is called Modus Ponens and the second structure (which is not deductively valid) is called the Fallacy of Affirming The Consequent.

In attacking these kinds of arguments, you have to either show that the logical structure is not deductively valid, even if it may seem that way, or you have to show that one of the premises isn’t true. This is how you successfully dismantle these kinds of arguments…

Now let’s turn to inductive arguments.

Investigating Inductive Arguments

While deductive arguments offer certainty, inductive arguments can only secure probability. These are the kinds of arguments that are used in science, where we can only ever be more or less sure about something, but never 100% certain. In other words, even if the premises are true, the conclusion doesn’t necessarily follow. That’s because empirical data always leaves room for error.

The drawback of deductive arguments is that the conclusion can’t move beyond the information contained in the premises as its bound by tight logical structures. But inductive arguments can do this because they draw generalizations based on the evidence at hand. You can enhance your chances of being right about the conclusion by having a sufficiently representative sample size, as well as a sample size that’s big enough.

Here’s an example. Let’s say we conduct a survey to find something out about a specific population. The inductive strength of the argument goes up the greater our sample size and goes down the smaller it is. Obviously, this is because a generalization is more probable the more people you’ve surveyed. So, if we are investigating something specific about a town of 3000 people and our sample size is 1000, this would make the generalization we draw much more probable than if we had only surveyed 50 people.

The biggest mistakes people can make when it comes to inductive arguments include the Hasty Generalization Fallacy, where the sample size is too small or qualitatively unrepresentative, or the False Cause Fallacy, where the person attributes causation regarding two phenomena when there is only correlation.

So, when it comes to undermining inductive arguments, you should look at the nature of the sample used or show that there is only correlation if causation has been attributed.

Last one…abductive arguments.

About Abductive Arguments

The final type of argument is an abductive argument. This is where we draw an inference to the best explanation based on a set of observations. Essentially, we’re isolating the best explanation for what we’re observing. Here’s an example to illustrate.

A “good” explanation is one that delivers substantive predictions which could turn out to be false. This means that we have a way of determining how good the explanation is by looking at its predictive success. You should also consider more than one explanation to make sure you’ve isolated the one with greater predictive powers.

We can tell whether an observation fits better with one explanation over another by applying the Surprise Principle. It says that “An observation O strongly supports H₁over H₂if both the following conditions are satisfied, but not otherwise. (1) If H₁ were true, O is to be expected; and (2) if H₂ were true, O would have been unexpected.”

The error in reasoning that is likely with these kinds of arguments is when the person argues for an explanation just because they can’t think of any rival theories. This is known as “The Only Game In Town” Fallacy. You can undermine an abductive argument by coming up with a better rival explanation that has equally strong predictive success or (even better) greater predictive success.

And there you have it! These are the different arguments you can build and break. In the next article, I will go through many different errors in reasoning (called fallacies). Once you have a decent grasp of different fallacies, you’ll be able to critique and undermine other people’s arguments, building your critical thinking muscles like a champion.

This will give you the superpower of problem solving, making you a real force to contend with!