Language Problems in Reasoning

The previous essays in this series have been concerned primarily with constructing good arguments from the ground up. Without understanding the fundamental concepts, it is futile to try to engage in a serious discussion. Presumably, if you are reading this essay, you have familiarized yourself with the forms and structures of good arguments, and armed yourself with the proper tools for determining the validity of evidence, burden of proof, and cogency of both premises and conclusions. Now, we will venture into more dangerous territory. It is disturbingly easy to fall prey to faulty reasoning unless we are constantly aware of the interrelations between facts, language, and individual evalutations of the two. Those who would persuade us of the truth of their position are often skilled at the subtle manipulation of language, such that the good critical thinker must constantly hone her reasoning skills as well as her knowledge of language and symbols, and how they are perceived. For instance, it is virtually impossible for us to make many decisions unless we first make generalizations. However, we must be ever vigilant against unwarranted generalizations. We must be able to determine the validity of analogies, as this is one of the most commonly employed deceptions in invalid arguments. Thus, we must be able to examine proposed analogies for similarities, differences, and categorical distinctions. These are but two of the problems we will examine in this essay, all dealing with language.

The first problem we encounter is the fact that words have different kinds of referents. Words can describe things that materially exist, or they can describe concepts, or even other words. Furthermore, in English, many words have multiple referents, sometimes quite similar in meaning, but with important differences. Consider briefly the word "set." There are at least one hundred and nineteen commonly accepted definitions.(1) Many of these are so similar that at first it is difficult to discern the difference. For instance, "To determine or fix definitely" might seem quite similar to "To resolve or decide upon," though careful consideration will reveal that one definition explicitly involves certainty, and the other does not. Recall in the previous essay how enthymemes can be used to disguise faulty premises? How hard is it to imagine a premise that hinges on the word "set," where the unspoken implication is certainty, while the reality of the premise is not as firm? I hope you are able to see how critically important it is to make sure that if you are the positive claimant, and you are trying to construct a valid argument (as opposed to pulling some confidence man scheme), you must be meticulous and rigid in your use of language, particularly when it involves nebulous words with multiple referents. Conversely, if you are evaluating an argument, you must also consider it your duty to force the claimant to formally and conclusively define each and every term that might have any chance of being vague.

This leads us to our first rule. Nonverbal things are not homogenous. This simply means that when we are referring to something such as an emotion, an intellectual concept, or a perception, we may not assume that discreet uses of the same word are the same. Consider:
1. This creme brulee is good.
This statement refers to a perception. The dish is tasty. The texture is pleasant. It satisfies our hunger.
2. John is a good person.
This statement most likely signifies an emotion. We cannot perceive John's goodness in the same way that we perceive taste. We must evaluate John conceptually, which will lead to a feeling about his goodness. This could also apply to an intellectual concept. If we have established that goodness will be defined as conforming to a certain moral code, and we can empirically demonstrate that John conforms, we can assign him the intellectual concept of good.
Notice that even within the same sentence, the word "good" may not be assumed to mean anything. Without proper definition, the statement is vague, and could lead to invalid conclusions.

Next, we must extend this rule. Nonverbal things are not transferable between classes or definitions. This is leading us to one of the most common fallacies, that of equivocation. For now, simply remember that in a valid argument, each use of the word must have a discreet referent, and if the word is used with multiple referents, we must not presume that they are the same. Furthermore, we must be aware of the fact that even within the same argument, identical groups of words may have vastly different meanings within the context presented. Consider a Frenchman and an American Politician discussing "typical Americans." To the Frenchman, a typical American is a prudish, overly conservative boor, while the Politician considers a middle class, honest, working man to be the appropriate referent. In listening to a conversation between the two, it is easy to understand how one could completely misconstrue the words of both sides, particularly if the listener happened to have a third referent!

In written debate, there are multiple ways of distinguishing between usages. First, a subscript may be used with a legend provided in the introduction, or prior to the presentation of the proposition. When speaking of things that change over time, dates may be used in the same way. For instance, I might discuss current science(1800) as opposed to current science(1950). So long as I include the dates, there will be no confusion as to the referent. Quotes are also a good way to separate one usage from another. This, of course, is only useful when there are no more than two uses. For instance, I might discuss magic as the practice of sleight of hand and illusion, and then contrast it with "magic," the practice of casting spells and hexes. Sadly, another viable option is the use of the hyphen, as in the word "religio-cultural" for a discussion of the effects of both religion and culture on a particular subject. I say sadly simply out of personal taste, for I tire of reading essays in which every concept is hyphenated, leading to the perception of over-wordiness. (Ironic use of hyphen intended.)

FALLACIES*

Finally, we get to every debater's favorite topic: Fallacies. In short, a fallacy, as defined by Richard Whately, is "any unsound mode of arguing, which appears to demand our conviction, and to be decisive of the question in hand, when in fairness it is not."(2) The next few points are extremely important, and must be the cornerstone of any debate. Remember the essay on burden of proof, positive versus negative claims, and presumption? With that in mind, commit the following truths to memory, and if you apply them consistently and thoroughly to every argument you hear, you will have conquered perhaps the most formidable obstacle to good critical thinking. Note that I don't claim to have produced a complete list of fallacies, but I believe it to be thorough enough for virtually all arguments that the average person will encounter.

* ANY premise that contains a fallacy is invalid, and renders any argument which includes it ABSOLUTELY USELESS in debate or critical thinking.
* ANY argument that contains either a fallacious premise or conclusion is invalid, and is ABSOLUTELY USELESS in debate or critical thinking.
* If the fallacy is contained in the positive claimant's argument, the negative claimant can be considered the winner in a debate unless the positive claimant can eliminate the fallacy and present a valid argument.
* If the fallacy is contained in the negative claimant's argument, this does not IN ANY WAY remove the obligation of the positive claimant to produce a valid, cogent argument for his position. Despite any fallacies in the negative claimant's argument, the presumption is still ALWAYS THE STATUS QUO.

Types of Fallacies
Unsupported Assertion: Essentially, this is any assertion that is not backed by properly supported evidence. This is often called a naked assertion. Refer back to the essay on evaluating evidence to recall guidelines for establishing the weight and validity of evidence. Also, be aware of the possibility of counter-evidence before accepting less than definitive evidence. I will give you a wonderful example from my own life. In a recent letter to the editor, the writer was decrying the evils of alcohol, and recounted a long list of ills supposedly being wrought on our city by the consumption of alcohol by local college students. Among the supposed correlations were violent crime, robbery, assault, and oddly enough, serial murders. As there were no citations to support this assertion, I looked up statistics for both states within the U.S. and for other countries as well. I discovered that there is actually no statistical correlation between per capita alcohol consumption and the rate of these crimes in the same area. While the writer's letter was emotionally moving, and might potentially be convincing to those who were not moved to check out the truth of his assertion, the facts are against him, and his conclusion, namely that alcohol ads should be removed from TV, is immediately disregarded until such a time as he can either produce counter evidence, or another valid argument in support of his position. (Clever readers will note that his conclusion was not derived from the argument, so there were two causes of its invalidity!)

Undistributed middle. If you recall from the previous essay, distribution is simply the inclusion of a categorical in middle term of a syllogism. This fallacy can lead to false analogies.
"Apples are a kind of fruit. Pears are a kind of fruit. Therefore, apples are a kind of fruit." (Thanks, REVLyle!)
Notice that in this example, the middle term, (Apples are a kind of fruit) is not expressed as a universal or an absolute. Therefore, the conclusion is invalid. The major premise more properly reads, "Apples are among the group of things known as fruit."

False analogy. Remember that an analogy is only valid if the two items are in the same category, AND the conclusion in question is relevant to the category. For instance, consider the following:
Apples are fruit a kind of fruit with seeds. Pears are a kind of fruit with seeds.
Apples are red. Therefore pears are red.
While it is true that apples and pears are analogous with regard to having seeds, the presence of seeds has no bearing on the exterior color of the fruit. Indeed, we can think of many seeded fruits that are not red. (Note that while some pears are, in fact, red, the partial truth of the conclusion has no bearing on the validity of the argument!)

Extended Analogy: Assuming that two specific items are analogous because they share an unrelated category.
The KKK are part of a fringe movement, separate from mainstream society.
Rosa Parks was part of a fringe movement, separate from mainstream society.
Because Rosa parks was admirable for being part of a fringe movement, so the KKK is admirable.

Non causa pro causa: This is a cause-effect assertion without connection.
I had a dream that it would rain today.
It rained today.
Therefore, my dreams caused the rain.

Post hoc ergo propter hoc: Assuming that something is a cause of an event simply because it preceded it.

Non Sequitur: Literally, "It does not follow." This is when the conclusion cannot be logically drawn from the premises.
My aunt is a good cook.
My aunt enjoys shopping for cookbooks.
Therefore, my cousin is a good cook.

Affirmation of the consequent. Consider the following:
A causes B.
B is true.
Therefore, A is true.
While this appears to be a reasonable argument, it is not necessarily true. Consider the possibility that B may have multiple, mutually exclusive causes. I will not be going into truth tables in this series, but if you do a web search, you can easily find one. Suffice it to say that a truth table demonstrates why this is a fallacy in the symbols of logic.

Denial of the antecedent is essentially the opposite of Affirmation of the Consequent:
A causes B.
A is false.
Therefore, B is false.
Again, this is not necessarily true, and is invalid as a syllogism.

Fallacy of Accident: This is when an argument moves from a general rule to a particular case. Again, we can refer to the lack of distribution to understand why this is fallacious.
"Practically everyone knows that the world is a sphere. Therefore, that chick from The View knows that the world is a sphere."

Converse Fallacy of Accident: This is the opposite of a Fallacy of accident, and involves arguing from a small number of cases, or from a special case, to a general rule.
"My sister lost her virginity at 28. Therefore, most women lose their virginity at 28."

Irrelevant Conclusion: Just as the name implies, this is a fallacy in which the point proven doesn't contribute to the proposition at hand.
"Alcoholics Anonymous is a very well respected organization. Therefore, AA is the best way to quit drinking."

Begging the Question: Related to circular reasoning, this fallacy consists of demonstrating a conclusion through a premise that assumes the conclusion. Forgive me for going against my word, but I'm going to use a theist argument here because it is simply the best known example of begging the question in the Western world.
"The Bible is inerrant. The Bible states that it is true. Therefore, the Bible has to be true."

Circular Reasoning: Very similar to begging the question, this fallacy assumes a premise that is actually the conclusion. Note that circular reasoning is actually logically valid, for it is a tautology, equivalent to saying A=A. Unfortunately, tautologies are not demonstrations of material truth, and cannot contribute anything useful to a debate unless the truth of A is established by some other means.

Many Questions: Also known as a loaded question. This is a question that contains a hidden presumption, and the only choices made available dictate the answer, based on that unsupported assumption.
"When you beat your wife, did you use a club or your fists?"
Clearly, there is no way to answer the statement as asked without implicitly agreeing to the unspoken assertion that I did beat my wife. This wording does not allow me to answer that I did not beat her.

Equivocation: This is the use of one word with two or more meanings. In a syllogism, this creates four terms instead of three.
People who volunteer have a good spirit about them.
Demonic spirits are in possession of people who do bad things.
Therefore, people who volunteer are not possessed by demonic spirits.

Fallacy of Composition: Asserting that because individual parts of a whole have a quality, that the whole must also have that quality.
All of the materials used to build this ship are lightweight.
Therefore, this ship weighs very little.

Fallacy of Division: The opposite of the Fallacy of Composition. This assumes that a property of a whole must also apply to the parts.
Human beings are made of atoms.
Humans are sentient.
Therefore, atoms are sentient.

Ad hoc: related to Post hoc ergo propter hoc, this is an after the fact explanation for something. Note that explanations are not arguments!
Astrology is an accurate predictor of personality traits.
Here is a study that demonstrates no link between star sign and personality traits.
Well, astrology is too subtle to be measured, but it's accurate on a more personal level.

No True Scotsman: This is actually a variant of Ad hoc, in which definitions are changed to exclude conflicting evidence.
No Scotsman drinks Kentucky Bourbon
My uncle is a Scotsman, and he drinks Kentucky Bourbon.
Well, no TRUE Scotsman drinks Kentucky Bourbon.

Red Herring: Introduction of unrelated material designed to divert from the main argument.
There is a direct correlation between school attendance and good grades.
Yes, but there are mothers who have to work and can't take their children to school!

Reification: Treating a concept as an existing material object.
You can't put "love" on the table so that we can see it. Therefore, it does not exist.

Shifting the Burden of Proof: This has been covered several times in this series. Any time the burden of proof is shifted from the positive claimant to the negative claimant, it is a fallacious argument. (This could also be considered a red herring!)
You can't prove that Santa Clause doesn't exist. Therefore, he does.

Strawman: An intentional misrepresentation of the opponent's argument:
All democrats believe that we should raise taxes to pay for minority women to have as many children as they want. This is proof that democrats are irrational.

Ad hominem: This is the tactic of attacking the person who is arguing a point, rather than their argument.
Al Gore is a loser. Therefore, global warming doesn't exist.

Tu quoque: A special form of ad hominem in which the fact that both sides have done a particular thing is pointed out as proof against an unrelated argument.
You cheated on your test. Therefore, your grade is invalid.
So what? You cheated, too.
(Note that this has nothing to do with whether or not the grade is invalid.)

Bifurcation: Presenting only two possibilities when there are actually more. This is related to the loaded question.

Fallacy Fallacy: This has been mentioned before. Just because a premise is fallacious, the conclusion is not necessarily false. (NOTE: The argument IS invalid!)

Argumentum ad nauseum: Literally, arguing until the opposition is sick of arguing, hoping they will concede the point. Repeating the same argument over and over without addressing counter arguments.

Special Pleading: This is where an exception to a rule is asserted without proof that such an exception is warranted.
"It's right that Paris Hilton was let out of prison early. After all, she's a celebrity."

Next, I'm going to lump a great many fallacies into one group, and discuss them generally, while making special note of the more common ones. This group is known as the "Appeal" group. In all cases, a certain thing is used as evidence for a conclusion when in fact it is not relevant to the conclusion. The most common of these are appeal to authority, appeal to numbers, and appeal to ignorance. When we appeal to authority, we are asserting that because a person who is respected says a thing is so, that it is so. While authority can be used as evidence, it is only valid if the authority can demonstrate the evidence by which he arrived at his conclusion. In other words, his presence in the argument is irrelevant. Appeal to numbers is commonly used in arguments such as "Seventy five percent of the population believes that our government is trustworthy. Therefore, the government is trustworthy." Clearly, large numbers of people have often been wrong about a great many things.

Appeal to ignorance is a favorite among mystics, theists, and conspiracy theorists. Essentially, the lack of an explanation is cited as proof for their explanation. People who use this fallacy will often say, "Lack of evidence is not evidence of lack." In other words, they are trying to positively prove their own position solely through the lack of definitive proof of another position.

Other forms of appeal: Appeal to antiquity (age of a belief), Appeal to current thought (It is new, therefore it is true.) Appeal to force (might makes right); Appeal to money (Those with money are likely to be correct.); Appeal to poverty (those who are poor are more likely to be honest); Appeal to emotion.

At this point, I am unsure whether or not I will write any more essays in this series. It is possible that I will devote an essay to refutation, but at present, I am unsure of how relevant it would be in this forum. In any case, I hope that these essays are helpful to you, and that you will be able to use the skills I've discussed to better your own critical thinking skills, and perhaps rid yourself of any harmful false beliefs you might have.

Thanks for reading.

1. http://dictionary.reference.com/browse/set?r=75
2. Richard Whately, Elements of Logic, James Munroe and Co., Boston, 1848, p. 143

* Special thanks to Mathew at infidels.org for his list of fallacies at http://www.infidels.org/library/modern/mathew/logic.html#antecedent