Logical Fallacies in Experimental Research

Inasmuch as a given researcher, more often than not, bases his own research on the published research of others in the field, he needs to understand the way inferences are drawn. And when it is his turn to publish his own research, he should make every effort to make his inferences easily understood. In either event, he should watch out for some pitfalls in thought, as well as in writ­ten or spoken language, that have been identified by logicians and are referred to as fallacies. Following is a brief review of the important ones.

1. Fallacies of Ambiguity

As with definitions (dealt with earlier in Chapter 2), ambiguity can mislead the reader or listener, often unknowingly and some­times intentionally. A few variations follow:

1. Equivocation refers to using the same word (or phrase) within a statement with two (or more) dif­ferent contextual meanings.

Example: Waiting for leaves to fall, I have hard work waiting this fall, which I expect leaves me tired.

2. Amphibole is the name given to the fallacy where the meaning of a statement is rendered indetermi­nate because of the loose or awkward way of com­bining the words.

Example 1: The dog fought hard to save the puppy, but it died.

Example 2: The political party in this election that is very corrupt, is likely to be chosen. (Which is corrupt, the political party or the election?)

3. The fallacy of accent is committed when, depend­ing on which word (or group of words) is accented, we are led to different meanings.

Example: In summer, storing chemicals in the attic is not a good idea. (Try accenting “in sum­mer,” “storing,” and “chemicals,” alone, each time.)

4. The fallacy of composition is said to have been com­mitted when the properties of the parts are taken obviously to manifest as properties of the whole.

Example: Cotton is expected to be a “light” material, but a mattress made of cotton can be heavy.

5. The fallacy of division is the reverse of the above, wherein the property of the whole is taken obvi­ously to manifest as a property of the parts.

Example: Americans are tall as a group, but there are some who are short.

2. Fallacies of Irrelevance

Relevance in either drawing inferences or advancing arguments based on given premises is not as obvious in practical affairs as we normally assume. In research, because of the nature of the disci­pline, irrelevance is found less than in practical affairs, but delib­erate effort is still necessary to keep it in check. Some major fallacies that arise because of irrelevant connections between pre­mises and conclusions are reviewed below. Latin names, which are traditionally attached to fallacies, are also mentioned in parentheses.

1. Appeal to inappropriate authority (ad verecundium): Making reference to opinions or judgments of authorities (or experts) within a given field of study is common, and indeed helpful. Such refer­ence should be confined to the particular field of study. An expert in an unrelated field of study should not be cited as an authority. For example, a worker in chemistry appealing to the opinion of Bernard Shaw or a researcher in tennis quoting Einstein is inappropriate.
2. Argument against the person (ad hominem): When, in the process of disputing (or disagreeing) with an inference or a judgment, the author is attacked with the purpose of discrediting his work, the fal­lacy is known as ad hominem, meaning “against the person.” In courts of law, political campaigns, and religious propaganda, the ad hominem fallacy is generously used, linked with such criteria as “credi­bility,” “honesty” and “trustworthiness.” There should be no room for such in scientific research.

3. Other Miscellaneous Fallacies

1. Argument based on ignorance (ad ignorantiam): When a proposition is claimed to be true because it has not been proved false, or when a proposition is claimed to be false because it has not been proved true, this fallacy is committed. A pure believer and an uncompromising agnostic, arguing on whether God exists or not, may both use this fallacy to score some points. But in anything claiming to be scientific, its use is undesirable.
2. Begging the question (petitio principii): Also known as circular reasoning, this fallacy is committed when we have assumed what we set out to prove.

Example: I want to prove that the Ten Com­mandments, as given in the Bible, are the words of God.

A skeptic’s questions: “How do you know?”

My answer: “It says so in the Bible.”

A less serious version of this fallacy is our assumption that the laws of nature will operate in the future as they operate at present, which is known as the principle of induction. Philosopher David Hume takes strong exception to this, point­ing out that using the fact that we have seen some­thing happen during the past up to the present as proof that it will happen in the future amounts to a circular argument. We can hope that the princi­ple of induction in nature may prevail, but we cannot prove it.

3. Fallacies of accident and reverse accident: These have particular significance in experimental research because both are related to generalization, and both serve as reminders that generalizations can be hasty and that they may, and quite often do, have exceptions. Witnessing widespread affluence in the United States, a casual visitor may be led to the inference that there are no “homeless people” there, even though we know there are homeless people, some of whom may starve or freeze to death in winter. This is an example of applying generalization to particular cases; it is known as the fallacy of accident. Similarly, witnessing one or two slums in Mumbai, a foreigner might think that all people in India live in wretched condi­tions. This is an example of carelessly generalizing from a particular case and is known as the fallacy of reverse accident.
4. Fallacies of appealing to emotion (adpopulum), pity (ad misericordium), and force (ad baculum): An orator’s speech to fellow countrymen at times of war usually contains several examples of ad popu- lum. The language of the defense attorney plead­ing for lenience in sentencing a criminal may contain many ad misericordiam “arguments.” And the government of practically every country, pur­suing the principle of “Might is Right,” engages in ad baculum, which may find expression in the lan­guage of political leaders. Fortunately for science, these fallacies pose no problems because they are little used.
5. Fallacy of complex questions: This refers to asking a question in such a way that the conclusion, often an accusation, is hidden in the question.

Example: A father to his teenage son: “Have you stopped drinking yet?”

Again, fortunately, no experimental researcher is likely to ask, or to be asked, such questions.

Source: Srinagesh K (2005), The Principles of Experimental Research, Butterworth-Heinemann; 1st edition.