This is part of a series on how to have better arguments with people. Most people are bad at having arguments, and I want people to be better at it. It’s a bit of a shame that the term “argument” itself comes with a bit of baggage, especially in the age of the internet. At times it feels like people don’t have arguments so much as they list excuses at one another for believing what they do. I’m here to say that it doesn’t have to be that way.

A small side note: I’m using the term ‘argument’ in two senses: the first as synonymous with a debate, and the second as the formal definition you’d see in your intro logic class. I’ll try to indicate which way I use it as I go.

Here I will talk about the concept of a logical fallacy, which is something I think a lot of people misunderstand in the general sense. I see a lot of people crying fallacy as if it’s a trump card they can play to automatically win the debate. Sometimes it’s almost as if the person who has the names and definitions of the most fallacies committed to memory has the upper hand.

Logical fallacies are presented as a sort of a zoo of phenomena, each with its own name and subspecies. They’re treated as fundamental ideas that exist on their own, and can be trifled over as such. I have a hunch that what a lot of people miss is where they actually come from, and what makes them fallacies in the first place.

I’m really horrible at chess, but bear with me as I make this analogy. I know that there are at least two ways of handling the opening moves in a game. The first is to memorize a bunch of starting positions and the best responses to them. A lot of high level play, as I understand, involves a vast knowledge bank of positions and the best moves on those positions. The other is to build a more fundamental understanding of the early game, and play from those fundamentals. Proponents of this method (like Josh Waitzkin) like to claim that it enables more creative and intuitive play.

I’m probably wrong about a lot of what I said in the preceding paragraph, but we’ll just use it as a basic analogy. Arguing is less like chess than some clichés make it out to be.

Here I want to advocate the fundamentals-based approach to understanding logical fallacies.

A logical fallacy is, in a word, a bad argument. A catalog of logical fallacies is, by extension, a list of ways in which an argument can be a bad argument. Thus, if we understand what makes a bad argument, we have everything we need to understand what makes a fallacy a fallacy. So let’s begin there.

My favorite definition of an argument is the most general one I can think of. An argument is one or more statements, one of which is designated as the conclusion, and the rest of which are designated as premises. When I say ‘statement’, I mean something that can either be true or false. This is a nice definition, since it makes no assumptions about content, nor does it make any claims about whether it needs to be any good.

Jill has a dog.
John has a cat.
---
Jake has a dog.

My examples to follow will take the above form. The conclusion is denoted by the three dashes. The above set of statements do indeed qualify as an argument. We’ve denoted Jake has a dog. as the conclusion, and the rest are premises. Each statement can either be true or false.

When we talk about an argument being good or bad, it most often has to do with its validity. An argument is valid if the truth of the premises guarantee the truth of the conclusion. Inversely, an argument is invalid if it is possible for the premises to all be true and the conclusion to be false. In other words, an invalid argument is one in which the conclusion doesn’t follow from the premises.

If Joe is a human, then Joe is a mammal.
Joe is a mammal.
---
Joe is a human.

What fallacy does the above argument commit? Who cares? It’s a bad argument. Once we understand how this argument is invalid, the fallacy to which it is attached seems redundant at best. It obscures the “why” in favor of the “what.” We cry fallacy, and the discussion is potentially derailed at the expense of the much more interesting discussion that could’ve taken place.

Why is this so important? I suppose that’s up for debate, but here’s what I think. When I say my opponent committed, say, the fallacy of affirming the consequent (which is what the above argument commits), it’s almost as if I’m doing so to avoid the legwork of understanding the argument in the first place. Crying fallacy can be a way of being dismissive. Instead, it can be more beneficial to address the argument more directly, by explaining exactly what is happening in the argument to make it a bad one. In the above instance, we can highlight the structure of the ‘if-then’ statement and demonstrate a counterexample.

This has a lot of benefits. Not only does it demonstrate an understanding of the argument we are opposing, but it also gives the opponent an opportunity to clarify their terms if we’ve misrepresented anything. Why give our opponent this ground? Because this isn’t war. A debate, at least as far as I’m concerned, should be a search for truth. This means we should be open to the possibility that we’re wrong, and even actively seek ways in which we can be wrong.

Here are a few common situations in which crying fallacy simply doesn’t work.

Non Sequester, Non Secretariat

We’ll start with an easy one. Non sequitur is Latin for “does not follow”, and in argument speak, a non sequitur is committed when the conclusion doesn’t follow from the premises. Sound familiar? It’s another word for an invalid argument. Within the scope of this article, we can consider all logical fallacies to be subspecies of non sequitur. Out in the wild this might not necessarily be the case (since we’re only concerned with deductive reasoning here), but I can be that anyone who uses the term is simply doing so just to sound fancy.

Ad Hondomium, Ad Ho-M&M

Here’s a good rule of thumb. If someone says the phrase “ad hominem attack”, you can probably replace the entire sentence in which it occurs with “My feelings are hurt.” There’s something about that phrase that irks me — it’s almost redundant given how ad hominem is used in the context of an argument. I rarely hear an accusation of an “ad hominem attack” in which ad hominem reasoning actually took place.

Take these two arguments as an example:

Jeff is a scumbag.
---
Jeff does not own a dog.

Contrast with:

Jeff is a scumbag.
If Jeff owns a dog, then he buys dog food.
Jeff does not buy dog food.
---
Jeff does not own a dog.

These are a bit contrived, and they clearly don’t reflect any real argument that would take place, but we’ll go with them for now. Which argument uses ad hominem reasoning? The first one clearly does. The premise is a claim about a person, the truth of which does not guarantee the truth of the conclusion. There is nothing in the argument that makes any claim about scumbaggery and dog ownership, so the argument allows for the possibility of Jeff owning a dog while maintaining a state of scumbaggery.

What about the second? Clearly, the statement “Jeff is a scumbag” is irrelevant to the rest of the argument, and it is a negative statement directed toward a person, but does it commit the ad hominem fallacy? It absolutely does not, because the argument is nonetheless still valid. The weird thing about the rules of deduction is that they say nothing about the relevance of the premises to the conclusion. Therefore we can have as many premises in an argument as we want. As long as it is not possible for all those premises to be true with a false conclusion, the argument is valid.

Pennies are made of copper.
San Diego is a city in California.
Egyptians live in Egypt.
The sky is pink.
Jeff is a scumbag.
If Jeff owns a dog, then he buys dog food.
Jeff does not buy dog food.
---
Jeff does not own a dog.

We can even modify the first argument to make it valid, by adding a premise:

No scumbag owns a dog.
Jeff is a scumbag.
---
Jeff does not own a dog.

Not only does the conclusion follow from the premises, but the personal attack itself works as a relevant premise! This argument does not commit a logical fallacy, even though it clearly contains an insult, an “ad hominem attack.”

See, the ad hominem fallacy refers to a specific way of reasoning. An insult may be mean and hurt our feelings, but unless the utterer draws a conclusion directly from it, he or she is not committing the fallacy. We can even have a valid argument that contains an insult as a premise (see the last example above). This does not commit the fallacy either, but the truth of such a premise is at times difficult to prove. This is not logic’s problem.

This Begs the Question: What About Begging the Question?

This is a little off-topic as it relates to the central point of this post, but I need to get something off my chest. Question begging is not the same as question raising. When someone begins a sentence with “This begs the question,” what they really want to say “This raises the question.” They are probably not about to engage in the act of question begging.

The fallacy of “begging the question” merely refers to an argument in which the conclusion is contained in some way in the premises. For example: “Bob is great because it says so in a book that he wrote.” This is a formally valid argument, since any argument in which one of its premises is its conclusion is valid, and since the conclusion does follow from the premises, so it’s not a non sequitur. I’ll discuss informal fallacies in the “objections” section, but it’s good to understand that fallacious arguments are not necessarily invalid ones.

Another thing to note is that when an argument begs the question, the inclusion of the conclusion in the premises is almost never explicit. Nobody’s going to say “I’m cool because I’m cool” with a straight face. The instance of the conclusion being used as a premise will either take the form of a rephrasing of the statement (“Your submission was rejected because it was judged as not sufficiently worthy of being accepted”), or it will be omitted altogether (take the “Bob is great” example from above) and implied from the wording. This is where the practice of searching for hidden assumptions when consuming a person’s argument.

It’s sort of a cliché to complain about the confusion surrounding the term “begging the question,” but there you go.

Straw Mayonnaise

This is probably the most common one I run into out in the wild. A straw man is a misrepresentation of the opposing argument. The motivation for committing this fallacy is to put your own argument on better rhetorical footing (not logical footing, as we’ll discuss in the next section) by demonstrating the flaws in the opposing one. A straw man is more easily set aflame than an actual man, you see.

Claims about a straw man argument are claims about an argument’s form. A true straw man, thus, is an argument that appears to have the same form as the actual one, but does not (Typically, it has different, more obvious flaws than the actual argument). If the form of the argument is the same–_even without appearing so_–then it is not a straw man argument. This is why analogies are often useful when illustrating an argument. In a good analogy, the form stays the same even with different content.

The Fallacy Fallacy Fallacy Fallacy

Let’s illustrate the straw man from above. You make an argument. I demonstrate why your argument is invalid. You claim that I’ve misrepresented your argument with a straw man. Your claim holds, but you make the additional claim that your vindication indicates that your argument is valid. Congratulations, you’ve committed the fallacy fallacy!

Here’s where we cash out on all this “crying fallacy” business. Crying fallacy does not ever affect the validity of your own position. Your knowledge of fallacies does not improve your position in any argument, especially when you deploy that knowledge offensively.

The fallacy fallacy is at the core of this post. Once you grasp it and its implications, you’ll be much better off.

Possible Objections, Addressed to My Satisfaction but Probably Not Yours

I know there are a few problems that you, dear reader, might have with what I have said here. That’s only natural; after all, I did file this under “philosophy,” and what good is philosophy without an onslaught of snarky counterarguments? Here I will try to address some of those objections.

Propositional Logic and Term Logic

I look at arguments through the modern lens of propositional logic (really, predicate logic), but I do understand that many like to make the distinction between this practice and the logic of Aristotle (known as “term logic”). I have not studied the Organon or any of its modern derivatives, so I’m choosing with this series to stick with what I understand. Nevertheless, I’m fully aware that the fallacies can be interpreted along those lines.

While this may alter the details of my discussion, I think the point of this post ultimately stands regardless of which system you choose to filter it through.

Formal and Informal Fallacies

Wikipedia (which is never wrong) goes out of its way to draw a line between formal and informal fallacies. Formal fallacies are arguments that are invalid solely on the basis of their logical form, while informal fallacies are invalid on the basis of their content (some may be formally valid). Here I’ve mostly discussed the fallacies that yield invalid arguments, but there is a whole host of arguments that are valid but nevertheless bad. Question begging, as we discussed above, is a good example of this.

Validity is merely one indicator of whether an argument is good or bad. In fact, it’s one of the more boring of these indicators. Validity is a characteristic of an argument’s structure, and an argument’s structure doesn’t give us any information about whether its contents are worth talking about.

All glubs are spilors.
Herf is a glub.
---
Herf is a spilor.

The above argument is a valid one, but it’s ultimately meaningless. “Soundness” refers to whether or not an argument’s assumptions are actually true. That is an epistemological claim, and is beyond the scope of this post.

Identifying a Logical Fallacy as a Heuristic

I’m not urging you to drop the concept of a logical fallacy altogether, as a possible objection might claim. I’m not asking you to not worry about learning what, e.g., the fallacy of the excluded middle is. What I urge is to merely think more about the structure of an argument, and make judgments about it from that basis. There are millions of ways for an argument to be a bad one, and the taxonomy of fallacies surely cannot cover them all. Furthermore, I maintain that it is almost always the case that logical fallacies (like most things) are better understood by having a grasp of their conceptual underpinnings.