Fallacy Book: Fallacies





Before moving on to the fallacies, it is necessary to have a brief discussion about how the term “fallacy” is used. Unfortunately for those who prefer to see fallacies as logical errors, people do use the term “fallacy” to refer to a factual error. For example, someone might say “a lot of people think that Google created Android from scratch, but that is a fallacy. Google based Android on Linux.”  While thinking that Android was created from scratch would be an error, it is an error about the facts, rather than an error in reasoning. If someone said “Android is a bad operating system. After all, many creepy geeks use it”, then this would be an error in reasoning (this is an example of an Ad Hominem fallacy). Even if many creepy geeks use Android, this does not prove that the operating system is bad. While both are mistakes, they are two different types of mistakes. One is an error about the facts, the other an error in reasoning.

To see the distinction, think about balancing a checkbook. I can make a mistake by doing the math incorrectly (which would be an error in reasoning), and I can make a mistake by entering the wrong amount for a check (factual error).

To use another analogy, think about cooking. One way I could screw up a meal is by cooking badly. This would be like an error in logic or reasoning. Another way is that I could use the wrong (or bad) ingredients. That would be like making a factual error. It is one thing to get the facts wrong (factual error) and quite another to reason badly about them. This is why I avoid referring to untrue claims as fallacies; I just call them false claims. I will not, however, say that people are wrong to use the term this way.

In logic classes, a fallacy is often said to be an argument in which the premises fail to provide adequate logical support for the conclusion.  This type of fallacy can be called a structural fallacy: the defect lies in the construction (logic) of the argument. A deductive fallacy, which is an invalid argument, is a paradigm example of a structural fallacy. Whether an argument is valid or invalid is an objective matter and can be tested by various means, such as truth tables and proofs. Deductive fallacies are often called formal fallacies.

Inductive fallacies are usually described as less formal than deductive fallacies and are often called informal fallacies. An inductive fallacy is a weak inductive argument in which the premises do not provided adequate support for the conclusion. In some cases, an inductive fallacy will be a structural fallacy: the defect lies in the structure of the argument itself and any argument with that structure will also be a fallacy. Ad Hominem fallacies are examples of structural fallacies: they are bad arguments because the pattern of reasoning is bad.

In other cases, an inductive fallacy occurs due to the argument being weak because it does not meet the standards for that argument type. In such fallacies, the structure of the argument is not the problem. For example, the Hasty Generalization fallacy and a strong Inductive Generalization are structurally identical. In both cases, the inference is that because X% of the sample of Ys are Z, it follows that X% of all Ys are Z (I am leaving out the margin of error here).  A Hasty Generalization is simply an Inductive Generalization made from a sample that is too small. Determining that a Hasty Generalization has occurred requires examining the sample, looking at the pattern of reasoning will not reveal the mistake. This sort of fallacy could be called a criteria fallacy or a fallacy of standards. I would use the term “standard fallacy” but that would be confusing.

While deductive arguments can be evaluated with certainty, this is not true for inductive arguments. There are objective standards, but assessing inductive reasoning is bit like judging figure skating: very good and very bad cases are easy to spot, but in between cases can be reasonably disputed. While inductive fallacies are often categorized as informal fallacies, this term also refers to a broader category of errors in reasoning.

People also use the term “fallacy” to refer to bad reasoning that does not involve explicit errors in logic. Rhetorical devices are sometimes classified in this way as are various persuasive techniques and strategies. These methods, devices and tactics are commonly used to try to substitute persuasion for argumentation or to “win” an argument through deceptive means. For example, the Red Herring (introducing a new issue to distract from the original issue) is not an argument but is an avoidance tactic often used in argumentation. These could be called method fallacies. They are not bad arguments, but this is because they are not arguments at all (in the philosophical sense). No doubt there are other uses of the term that I have not mentioned, and language always evolves.

When reading through the fallacies, be sure to keep in mind that there is no official governing body for fallacies. While there are traditions and common practices, you will see other names and different definitions for these fallacies and many of these are correct. Just keep an eye out for those who try to redefine various fallacies for nefarious purposes.

It is also a good idea to keep in mind that not all things that look like fallacies are fallacies. In some cases, they are not fallacies because they are not involved in reasoning at all. In other cases, an argument that might be mistaken for a fallacy could be a non-fallacious. I have taken care to make note of common occurrences of this sort in the fallacy descriptions. As a final point, there are far more fallacies than are listed in this book (or any book). So, something that does not match a named fallacy might still be a fallacy.


A Factual Error

Portland, Maine is the capital of the United States.


Valid Deductive Argument

Premise 1: If Bill is a cat, then Bill is a mammal.

Premise 2: Bill is a cat.

Conclusion: Bill is a mammal.


Extended Deductive Argument

Argument1 Premise 1: If pornography has a detrimental effect on one’s character, it would be best to regard it as harmful.

Argument 1 Premise 2: Pornography has a detrimental effect on one’s character.

Argument 1 Conclusion: It would be best to regard pornography as harmful.

Argument 2 Premise 1: If it is best to regard something as harmful, then the government should protect people from it.

Argument 2 Premise 2: It would be best to regard pornography as harmful (conclusion of argument 1).

Argument 2 Conclusion: The government should protect people from pornography.


Deductive Fallacy (Invalid Argument, Affirming the Consequent)

Premise 1: If Portland is the capital of Maine, then it is in Maine.

Premise 2: Portland is in Maine.

Conclusion: Portland is the capital of Maine.

(Portland is in Maine, but Augusta is the capital.)


A Strong Inductive Argument

Premise 1: 70-80% of humans have brown eyes.

Premise 2: Sally is a human.

Conclusion: Sally has brown eyes.


Inductive Structural Fallacy (Circumstantial ad hominem)

Premise 1: Dave supports the tax reduction for businesses and says it will be good for everyone, but he owns a business.

Conclusion: Dave must be wrong about the tax reduction.


Inductive Criteria Fallacy (Hasty Generalization)

Premise 1: Having just arrived in Ohio, I saw one white squirrel.

Conclusion: All Ohio squirrels are white.


Method Fallacy (Red Herring)

Reporter: “Senator, while you claim to want to address economic inequality, you have repeatedly used your insider knowledge to enrich yourself on the stock market. What is your response to your critics?”

Senator Smith: “I must remind you that I am working hard to pass laws to address climate change.”

Originally appeared on A Philosopher’s Blog Read More