Fallacy

3 Assuming the Cause: "post hoc reasoning"
The trouble with the cause arises from complexity in this construction of the classes. But there is a vulgar error in talking about causal relations which comes from a false generalization about causation itself. This error takes the form of supposing that because something follows something else it is caused by it. One feature that all paired constructions of causes and effects have in common is that the effects follow the causes. This is a necessary feature, but it is by no means a sufficient one for constructing the classes of causes. If it were sufficient, everything that preceded anything else could be regarded as a cause of it. Astrology would be science instead of superstition, since there are configurations of the heavenly bodies antecedent to every event in the world. My plucking a red rose yesterday would also be a cause of my nosebleed today. When the classes of events belong to the same realm of phenomena and are closely related in time, their connection seems more plausible. If the first event is political and so is the second, the propinquity may be very convincing. Many persons who put no stock in astrology would still nod amiably if someone should say, "The death of President Roosevelt caused the cold war."
The sufficient conditions for a well-constructed class of causes are similar to those for any well-constructed generalization -- lots of representative cases and no counter instances. Putting the necessary and sufficient conditions together, we may say that one thing is a cause of another if and only if, in numerous and representative cases, every case of the one is followed by a case of the other.
Post hoc means "after this." The words refer to assuming without proof that a prior event explains a subsequent occurrence. The post hoc error is called "vulgar" since it is not only common but also ignorant. Magic, superstitions, old-wives tales, and political debate abound in instances of post hoc. The first example involves a special feature that deserves notice.
EXAMPLE COMMENT
A student for his term paper makes a study of European wars and notices that in the years preceding the outbreak of recent wars there occurred an increase of armaments among the belligerents. He states as the thesis of his paper, "Increased armaments appear as one of the major causes of European wars." The increase of armaments and the outbreak of war may both be the effects of the same cause or causes. Friction and tension between nations can result in (a) armaments and (b) war. The armament race precedes closely in time and belongs to the same realm of phenomena as the outbreak of war, but this is not a sufficient condition for saying that war is an effect of it. The possibility of a cause having several effects makes it difficult in involved situations to see, where one effect precedes another in time, that the former is not a cause of the latter. Laboratory workers attempt to isolate factors in the complex and to alter them one by one, in the hope of producing some of the effects alone. Where they succeed, experimenters know that they are dealing with linked effects. Where the events resist isolation, the experimenters accept a causal connection, the earlier standing as perhaps an intermediate cause. If it were possible to experiment with whole nations and epochs, sociologists might discover, by isolating economic, political, psychological factors, whether armament races are linked effects of antecedent causes or in fact are an intermediate cause of war.
A testimonial reads, "Relax tablets relieved my headache. You can end your headache troubles, too!" This is the crassest sort of post hoc reasoning, even assuming the sequence of events true and the opinion offered in good faith. The fact that the body rids itself of many ailments is overlooked when attention is centered on the little pink pill. No doubt the fact that our present age expects wonders from medicine explains the persistent popularity of worthless patent medicines whose patrons discover themselves "cured" after they have taken an alleged remedy. Of course, Relax pills may be effective against some headaches: we shall never establish this by such testimonials.

EXAMPLE	COMMENT
A student for his term paper makes a study of European wars and notices that in the years preceding the outbreak of recent wars there occurred an increase of armaments among the belligerents. He states as the thesis of his paper, "Increased armaments appear as one of the major causes of European wars."	The increase of armaments and the outbreak of war may both be the effects of the same cause or causes. Friction and tension between nations can result in (a) armaments and (b) war. The armament race precedes closely in time and belongs to the same realm of phenomena as the outbreak of war, but this is not a sufficient condition for saying that war is an effect of it. The possibility of a cause having several effects makes it difficult in involved situations to see, where one effect precedes another in time, that the former is not a cause of the latter. Laboratory workers attempt to isolate factors in the complex and to alter them one by one, in the hope of producing some of the effects alone. Where they succeed, experimenters know that they are dealing with linked effects. Where the events resist isolation, the experimenters accept a causal connection, the earlier standing as perhaps an intermediate cause. If it were possible to experiment with whole nations and epochs, sociologists might discover, by isolating economic, political, psychological factors, whether armament races are linked effects of antecedent causes or in fact are an intermediate cause of war.
A testimonial reads, "Relax tablets relieved my headache. You can end your headache troubles, too!"	This is the crassest sort of post hoc reasoning, even assuming the sequence of events true and the opinion offered in good faith. The fact that the body rids itself of many ailments is overlooked when attention is centered on the little pink pill. No doubt the fact that our present age expects wonders from medicine explains the persistent popularity of worthless patent medicines whose patrons discover themselves "cured" after they have taken an alleged remedy. Of course, Relax pills may be effective against some headaches: we shall never establish this by such testimonials.