Logical Fallacies

• Arguments often cite the wrong cause.
• Analogies are often weak.
• A generalisation may be weak.
• Terms may be used ambiguously.
• Circular arguments assume what they are trying to prove.
• Ad Hominem arguments attack someone personally.
• Arguments often refer to irrelevant evidence.
• Exercise

A fallacy is a logical statement that seems perfectly reasonable when you first read it, but contains a logical error that becomes apparent when you read a second time a little more carefully. There are 7 basic types of fallacy that appear on the GMAT Verbal test.

Generally, you will not need to know the technical names for these fallacies, or even be able to categorise them with 100% certainty into one of the 7 categories. Often, you will be able to see that an argument is wrong, but won't know for certain whether it is citing irrelevant evidence or a weak generalisation. It is, however, useful to be familiar with the 7 different sorts of scenario that you could encounter.

I can (almost) guarantee that you won't see a question like this:

Which type of false argument is presented here?

1. Ad Hominem attack.
2. Citing irrelevant evidence.
3. Weak analogy.
4. A generalisation is being made.
5. A circular argument

Instead, the questions would be phrased as follows:

Which of the following most weakens/strengthens the argument?
The argument is correct only if we can show that ...
Which of the following arguments has the same pattern as the one above?

etc.

1. Arguments often cite the wrong cause.

An argument may conclude that some cause produces some effect, but the cause is the wrong one. It may just turn out to be a coincidence, or possibly the cause and effect have become muddled up (e.g. concluding that having lung cancer gives a person a craving to smoke cigarettes). Here's an example:

Nearly everybody who is serving a sentence in a British prison watched more than an hour of television in childhood. Therefore, we may assume that watching television in childhood encourages criminal behaviour.

Which of the following facts, if true, would most damage the conclusion drawn?

1. Many children continue to watch television when they become adults.
2. Modern British prisons provide television viewing to the prisoners.
3. The majority of people who have not been convicted of crimes also watched a great deal of television in childhood.
4. Some people who did not watch television in childhood have been sent to prison.
5. Some television programs aimed at children contain moral messages and encourage good behaviour.

Clearly the conclusion is wrong. We may disapprove of the pervasiveness of television (personally, I regard it as "the idiot box" and watch as little as possible), but we can't conclude that it turns people into criminals. This is borne out by the fact that most people who have watched television simply don't turn into criminals - the correct answer must therefore be C.

3. A generalisation may be weak.

A generalisation involves drawing a sweeping conclusion from a small sample of cases. This may occasionally be justified. For instance, if aliens arrive on Earth, and the first few aliens off the mother ship fire lasers at you, then you are entitled to assume that the aliens are generally aggressive. However, in GMAT, it is a logical fallacy to draw a conclusion from a small sample of cases:

Of the twenty people that were interviewed on the High Street this morning, more than 75% of them believed that pensions are too low. The government clearly needs to raise pension sif it is to increase public support for its policies.

Which of the following facts, if true, highlights a potential problem in the argument above?

1. The survey was conducted outside a post office on a day when it was distributing pensions to elderly people.
2. Surveys have proved hostile to government pension policy on previous occasions.
3. The company that carried out the survey has a poor record of being able to judge public opinion.
4. A government spokesman has defended government policy as regarded pensions.
5. Pensions in some other countries are much lower than in this country.

That figure of "more than 75%" looks impressive, doesn't it! Until you notice the fact that only 20 people were asked - so "more than 75%" translates to "between 16 and 20 inclusive." Hmm! That's the major fault with this survey - you simply can't get a representative result on the basis of such a small poll. If you couple this with the fact that the survey was taken outside a post office while all these old people were coming out, having just collected their pensions, and you can't be surprised when most of the people think that pensions are too low! The correct answer must be A.

I am reminded of the (possibly apocryphal) anecdote of a government survey in some South American dictatorship several years ago that found that 100% of the people in the country approved of the death penalty. It turned out that only 10 people had been asked their opinion - and they had all been the members of a firing squad!

4. Terms may be used ambiguously.

It is easy to assume that a word or phrase is defined one way in a sentence, and then assume it is defined slightly differently in the next sentence. Here's an example:

The Sumatran rhinocerous is the oldest rhinocerous in existence. We have such a creature at London Zoo, whereas Whipsnade Zoo does not. Therefore we have an older rhinocerous than Whipsnade Zoo.

Which of the following terms is being defined in more than one way in that argument?

1. Sumatran
2. oldest
3. creature
4. existence
5. rhinocerous

Here, the definition of the term "rhinocerous" slides as we go from the first sentence to the third one. In the first sentence, the author is saying that the Sumatran species of rhinocerous is the oldest species of rhinocerous in the world. Clearly that doesn't mean that all rhinos in the world that are not Sumatran must be younger than they are, but that is what is being assumed in the third sentence. The correct answer must be E.

5. Circular arguments assume what they are trying to prove.

Circular arguments take the form "Fact A proves fact B. Fact B proves fact A." The thing that you are trying to prove in the first place is being used as a piece of evidence to prove itself. This is best illustrated with an example:

Shakespeare was the world's greatest playwright, which means that he must have written the world's greatest plays. The fact that he wrote the greatest plays in the world only goes to prove that he is the world's greatest playwright.

Which of the following statements best matches the argument shown above?

1. Dr. Smith is the only member of the twenty-strong English department who thinks that Byron wrote greater poetry than Keats. Therefore Byron was not as good a poet as Keats was.
2. A miscarriage of justice occurs when an innocent person is sent to prison. However, there are no miscarriages of justice because people in prison are guilty. Otherwise they would not have been sent to prison.
3. The presence of enlargened white corpuscles in the blood would indicate that the patient must be suffering from disease X. However, his blood sample contains no enlargened white corpuscles, so he cannot be suffering from disease X.
4. Blair denies having committed the robbery, but has implicated Brown. Brown on the other hand, claims that he was elsewhere at the time, and points the finger squarely back at Blair.
5. I believe that the presence of the tartar sauce helps to accentuate the flavour of the other ingredients, but they in turn release chemicals which tend to neutralise the tartar sauce. Therefore the dish is much less spicy than pure tartar sauce would be.

You can see that argument about Shakespeare whirling round on itself, can't you. It can only prove that he is the greatest playwright by pointing out that he wrote the greatest plays. However, this is only true if he turns out to be the world's greatest playwright, which is what we are trying to prove in the first place.

The option that matches this one is B. People in prison must be guilty. How do we know this? Because innocent people are not sent to prison. How do we know this? Because the only people in prison are guilty. This is a circular argument - and one which is patently wrong. You only have to study the cases of the Guildford Four, the Birmingham Six and the Arsenal Nil to know that!

Option D contains some of the elements of a circular argument, but there is no logical error in two men protesting their innocence and implicating each other.

6. Ad Hominem arguments attack individuals personally.

Ad Hominem is Latin for "to the man" and refers to arguments that attack a person because of his or her personal characteristics, rather than attacking the logic of his or her argument:

Professor Tully has a general background in Artificial Intelligence but not in Neural Networks specifically. Therefore he is not a suitable candidate to organise this year's conference on Neural Networks.

This argument is weak because it

1. assumes there is no-one more suitable than Professor Tully to organise the conference.
2. ignores the possibility that Professor Tully may know something about Neural Networks.
3. confuses Professor Tully's specialist knowledge with his ability to organise conferences.
4. decries Professor Tully's background in Artificial Intelligence.
5. implies that Professor Tully's knowledge is too broad to be useful.

Presumably, the most important skill to have when organising the Neural Networks conference is the ability to organise conferences. The argument stated says that the reason that the professor is not suitable to organise the conference is that his speciality does match that in the conference itself. The option that refers to this comparison is option C.

Any argument that attacks the messenger, not the message, is an Ad Hominem attack.

7. Arguments often refer to irrelevant evidence

Another word for "refer to" is "cite", so you may see this type of fallacy referred to as "citing irrelevant evidence." Either way, it means that a statement is given and we are told that it must be true because of a certain piece of evidence, even though that evidence has little to do with the arguments in question.

We may predict that earthquakes in England are comparatively rare. After all, there have been no recorded earthquakes in the neighbouring countries of Scotland and Wales.

This argument only becomes logical if we could show that

1. the climate in England is similar to the climate in both Scotland and Wales.
2. Wales and Scotland have on average the same number of earthquakes a year.
3. Scotland has fewer than ten earthquakes a year.
4. other countries which are close to England, such as Ireland or France, also have a low risk of earthquakes.
5. countries that share a border also share the same risk of suffering earthquakes.

When you look at the options, the absurdity of the original argument becomes apparent. We gain nothing by counting the number of earthquakes in Wales and Scotland - or in France and Ireland either, for that matter - and the argument certainly has no connection with climate (weather). The only thing that makes the argument a good one would be to show that countries which are near each other have the same risk of earthquakes. While I grant that this is sometimes the case (earthquakes do tend to occur in certain "zones" around the world), it doesn't necessarily follow.

You will note that you don't have to be an expert on earthquakes or plate tectonics. You just have to be able to spot the hole in the argument and the necessary plug that would fill it. In this case, you don't even have to believe the argument - it isn't true in this case! However, if we could show that option E were true, then the argument would become sound.

---

Go back