A Concise Introduction to Logic - PART III: INDUCTIVE LOGIC - Chapter 13 and Chapter 14

PART III: INDUCTIVE LOGIC

Chapter 13: Hypothetical / Scientific Reasoning

13.1 The Hypothetical Method

Hypothetical reasoning is most immediately applied to the production of explanations. You will recall from Chapter 1 of this book that an explanation is a kind of expression that purports to shed light on some event. Many explanations are grounded in direct observation and memory. For example, suppose that you happen to be wearing a cast on your arm. You know from direct observation that you broke your arm while playing football a week earlier. Suppose somebody asks you about the cast, and you reply, “I’m wearing the cast because I broke my arm playing football.” This is an explanation. Or suppose that your mathematics instructor asks you why you failed the calculus test. You reply, “I failed the test because I didn’t study the night before.” No hypothetical reasoning is used to produce explanations of this sort.

However, it often happens that the needed explanation cannot be produced from direct observation. Suppose that you get into your car one morning and turn the key in the ignition, and the engine cranks but fails to start. Why it fails to start is a complete mystery. After all, it started perfectly the day before. Because it is impossible for you to peer into the inner workings of the engine to furnish the needed explanation, you begin by producing conjectures about why the car will not start. Perhaps the spark plugs are dirty, or the ignition coil has shorted, or the fuel pump is broken. Or perhaps someone sabotaged the car overnight. These conjectures are hypotheses, and the reasoning used to produce them is hypothetical reasoning. The hypotheses make up for the lack of direct observation in producing the needed explanation.

As a follow-up, you decide to remove one of the spark plugs and inspect its condition. You do so and ﬁnd it covered with carbon deposits. Now you think you have the answer: The car fails to start because the spark plugs are dirty. But of course this explanation could be wrong. The correct explanation could be something else, but this is the nature of all inductive reasoning. The outcome is only probable. However, having produced this explanation, you can now convert it into an inductive argument. Suppose a friend comes up and says, “I see you ran out of gas.” You reply, “No, the spark plugs are dirty, so if I replace them, the car should start.” Now you are trying to prove something to your friend, and the kind of argument used is a causal inference.

Hypothetical reasoning is used by nearly all of us in our day-to-day experience. The TV repairperson constructs hypotheses to determine why the picture appears unclear after all the ordinary solutions have been tried without success, the motorist on the freeway or turnpike reasons hypothetically to determine why the traﬃc is backed up bumper-to-bumper even though it is not yet rush hour, the physician hypothesizes about the cause of a disease prior to prescribing medicine, the teacher hypothesizes about the best way to present a complicated subject in the classroom, and the prosecuting attorney suggests hypotheses to the jury in arguing about the motive for a crime. In all of these cases the evidence is not sufficient to indicate exactly what is going on, what lies behind the scene, or what approach to take, so hypotheses are constructed to make sense of the situation and to direct future action.

Hypothetical reasoning is used most explicitly in philosophical and scientific inquiry. Every scientiﬁc theory can be viewed as a hypothesis for unifying and rationalizing events in nature. The Ptolemaic and Copernican theories about the sun and planets, Dalton’s atomic theory, Darwin’s theory of evolution, and Einstein’s theory of relativity are all hypotheses for making sense of the data of observation. The problem for the scientist is that the underlying structure of nature is hidden from view, and the data of observation by themselves are not suﬃcient to reveal this structure. In response, the scientist constructs hypotheses that provide ways of conceptualizing the data and that suggest speciﬁc questions to be answered through the design of controlled experiments.

Analogously, every philosophical system can be viewed as a grand hypothesis for interpreting the content of experience. Plato’s theory of forms, Aristotle’s theory of substance, Leibniz’s monads, and Kant’s theory about the mind are all hypotheses aimed at illuminating various aspects of experience. Just as the structure of nature is hidden from the scientist, the meaning of experience is hidden from the philosopher, and ordinary common sense will not provide the answer. In response, the philosopher constructs hypotheses that can be used to shed light on the content of experience and to provide suggestions for further analysis.

Whether it is applied in philosophy, science, or ordinary life, the hypothetical method involves four basic stages:

1. Occurrence of a problem

2. Formulating a hypothesis

3. Drawing implications from the hypothesis

4. Testing the implications

These four stages may be illustrated through the procedure used by a detective in solving a crime. Suppose that a woman has been murdered in her apartment. Initially, everything in the apartment is a potential clue: the empty wine glasses in the sink, the small container of cocaine on the coﬀee table, the automobile key found on the carpet, the strand of blond hair removed from the couch, and so on. To introduce an element of rationality into the situation, the detective formulates a hypothesis—let us say the hypothesis that the key found on the carpet ﬁts the murderer’s car.

From this hypothesis several implications can be drawn. Suppose that the key is the kind that ﬁts only late-model Cadillacs. It follows that the murderer drives a latemodel Cadillac. Furthermore, if the key is the only one the murderer owns, it follows that the car may be parked nearby. A third implication is that the murderer’s name may be on record at the local Cadillac dealership. To test these implications, the detective conducts a search of the streets in the vicinity and contacts the local Cadillac dealer for the names of recent buyers.

This example illustrates three additional points about hypotheses. The ﬁrst is that a hypothesis is not derived from the evidence to which it pertains but rather is added to the evidence by the investigator. A hypothesis is a free creation of the mind used to structure the evidence and unveil the pattern that lies beneath the surface. It may be that the detective’s hypothesis is completely false. Perhaps the key ﬁts a car that was lent to the victim for the weekend. A wide variety of other possibilities are conceivable.

The second point is that a hypothesis directs the search for evidence. Without a hypothesis for guidance, all facts are equally relevant. The mineral content of moon rocks and the temperature in the Sahara would be as relevant as the cars parked on the street outside the apartment. The hypothesis tells the investigator what to look for and what to ignore.

The third point concerns the proof of hypotheses. Let us suppose that the detective ﬁnds a late-model Cadillac parked outside the apartment building and that the key ﬁts the ignition. Such a discovery might lend credibility to the hypothesis, but it would not in any sense prove it. Concluding that a hypothesis is proven true by the discovery that one of its implications is true amounts to committing the fallacy of aﬃrming the consequent (see Section 6.6). Where H stands for a hypothesis and I for an implication, such an argument has the invalid form

Let us suppose, on the other hand, that the murderer turns himself or herself in to the police and that the only car the murderer owns or drives is a Ford. Such a fact would prove the hypothesis false because it would falsify the implication that the murderer drives a Cadillac. The argument form involved in such an inference is modus tollens:

For the hypothesis to be proved true, the car that the key ﬁts would have to be found and the owner would have to confess to the crime.

13.2 Hypothetical Reasoning: Four Examples from Science

Some of the clearest illustrations of the hypothetical method of reasoning can be found in scientific discoveries. Four examples are the discovery of radium by Pierre and Marie Curie; the discovery of the planet Neptune by Adams, Leverrier, and Galle; the discovery of atmospheric pressure by Torricelli; and Pasteur’s research concerning the spontaneous generation of life. Following is a consideration of each of these examples with special attention to the four stages of hypothetical inquiry.

Radium

In 1896 the French physicist Henri Becquerel discovered that crystals containing uranium had the power to expose photographic plates. He found that when these crystals were placed on top of an unexposed plate and left for a certain time, dark blotches appeared in their place when the plate was developed. Becquerel concluded that the crystals emitted certain rays that penetrated the opaque covering of the plates and reacted with the photosensitive material underneath. Further investigation showed that these rays were not as strong as X rays, which could be used to photograph bone structure, and so Becquerel’s interest in them lapsed.

A year later Marie Curie revived the question when she adopted it as the topic of her doctoral research at the University of Paris. In place of Becquerel’s photographic plates she substituted an electrometer, which was better suited to measuring the intensity of the rays, and she proceeded to conduct various experiments with pure uranium to determine the source of the rays that the metal emitted. When none of these experiments proved fruitful, she shifted her attention to the question of whether other metals or minerals emitted the same kind of rays as uranium. She tested hundreds of metals, compounds, and ores, but the only one that proved interesting was pitchblende, a certain ore of uranium. Because pitchblende contained uranium, she anticipated that it would emit rays; however, because it also contained a number of impurities, she expected the rays to be weaker than they were for pure uranium. Instead, they turned out to be stronger. This problem caught Madame Curie’s attention and provided the focus for her research in the months ahead.

In response to the problem, Madame Curie formulated the hypothesis that the impurities in the pitchblende somehow triggered the uranium to increase the emission of rays. One implication of this hypothesis was that mixing pure uranium with the kinds of impurities found in pitchblende would cause an increase in the emission of rays. To test this implication, Curie diluted pure uranium with various elements and measured the strength of the rays. The results were always the same: The emissions were always less than they were for pure uranium. Because of these results, she abandoned the hypothesis.

Madame Curie then formulated a second hypothesis: The intensified emissions were caused directly by some impurity in the pitchblende. The only other element besides uranium that was known to emit rays, however, was thorium, and the pitchblende that had been tested contained no thorium. Thus, an immediate implication of the hypothesis was that the increased rays were caused by an unknown element. A second implication was that this element could be separated from the pitchblende through a process of reﬁnement. At this point Marie Curie was joined by her husband, Pierre, and they began a combined eﬀort to isolate the unknown element.

Because the element was present in only the most minute quantities, separating a measurable amount from the other impurities required a great deal of eﬀort. The Curies began by grinding up some pitchblende and dissolving it in acid. Finally, after numerous stages of ﬁltration and the addition of other chemicals, they obtained a pinch of white powder. By weight, this material was found to be 900 times more radioactive than pure uranium, but since the primary component in the powder was barium, the unknown element still had not been isolated.

Rather than continue with additional stages of reﬁnement, the Curies decided to attempt a spectrographic analysis of the powder. Such analysis, they hoped, would reveal the characteristic spectrum line of the unknown element. This proposal, which amounted to a third implication of the hypothesis, was put to the test. When the powder was burned in a spectrometer, a line appeared in the ultraviolet range that was diﬀerent from that for any other element. From the combined evidence of the spectrum line and the intense radiation, the Curies announced in 1898 the discovery of a new element, which they called radium. After more processing and reﬁnement, enough of the material was ﬁnally obtained to determine its atomic weight.

Neptune

In 1781 William Herschel discovered the planet Uranus. However, the production of a table giving the motion of the new planet had to wait until the gravitational interactions between Uranus, Jupiter, and Saturn had been worked out mathematically. The latter task was accomplished by Pierre Laplace in his Mechanique Celeste, and in 1820 Alexis Bouvard used this work to construct tables for all three planets. These tables predicted the orbital motions of Jupiter and Saturn very accurately, but within a few years Uranus was found to have deviated from its predicted path. A problem thus emerged: Why did the tables work for Jupiter and Saturn but not for Uranus?

In response to this problem several astronomers entertained the hypothesis that an eighth planet existed beyond the orbit of Uranus and that the gravitational interaction between these two planets caused Uranus to deviate from its predicted position. Not until 1843, however, did John Couch Adams, a recent graduate of Cambridge, undertake the task of working out the mathematical implications of this hypothesis. After two years’ work Adams produced a table of motions and orbital elements that predicted the location of the hypothetical planet, and his computations were so accurate that if anyone with a telescope had bothered to look, they would have found the new planet within two degrees of its predicted position. Unfortunately, no one looked for it.

At about the same time that Adams completed his work on the problem, the French astronomer U. J. J. Leverrier, working independently of Adams, reported a similar set of motions and orbital elements to the French Academy of Science. The close agreement between Adams’s and Leverrier’s predictions prompted a search for the planet, but because a rather broad section of sky was swept, the planet was missed.

Finally, Leverrier sent a copy of his ﬁgures to Johann Galle at the Berlin Observatory, where a set of star charts was being prepared. It was suggested that the region corresponding to Leverrier’s computations be observed and the results matched against the charts. This was done, and a small starlike object was found that was not on the charts. The next night the same object was sighted, and it was found to have moved. The new planet was thus identiﬁed. It was named Neptune after most astronomers outside France objected to the original suggestion that it be called Leverrier.

Atmospheric Pressure

Originated by Aristotle, the principle that nature abhors a vacuum was used for centuries to explain the fact that in emptying a keg of wine an opening had to be made at the top as well as at the bottom. Because nature would not allow a vacuum to be created inside the keg, the wine would not drain from the bottom until air was let in at the top. It was thought that this principle held universally for all applications involving a vacuum, but in the sixteenth century it was found that suction pumps used to drain water from mine shafts would not work if the pump was situated over 30 feet above the water level. This caused people to wonder whether nature’s abhorrence of a vacuum, while holding true for kegs of wine, had certain limits for pumps.

In 1630 Giovanni Baliani of Genoa discovered a similar limitation in regard to siphons. When he attempted to siphon water from a reservoir over a 60-foot hill, he found that the siphon would not work. When the siphon was completely ﬁlled with water and the stoppers were removed from both ends, a vacuum seemed to be created in the uppermost parts of the pipe.

These ﬁndings were communicated to Gasparo Berti in Rome, who, around 1641, attempted to determine more scientiﬁcally whether a vacuum could actually be created. Berti designed an apparatus consisting of a spherical glass vessel attached to a pipe about 40 feet long. The apparatus was aﬃxed upright to the side of a tower and after the valve at the lower end of the pipe was closed, water was poured through the upper opening in the glass vessel. When both the pipe and the glass vessel were completely ﬁlled, the opening in the vessel was sealed and the valve at the lower end of the pipe was opened. Immediately water rushed from the bottom of the pipe, creating a vacuum in the glass vessel. This experiment crystallized a problem that had been developing for many years: If nature abhorred a vacuum, why did it tolerate the creation of one in the glass vessel? Furthermore, why did the water always descend to the same level in the pipe?

The results of Berti’s experiment were communicated to Evangelista Torricelli in Florence, who was at that time Galileo’s assistant. Galileo himself thought that the water was supported in the pipe by the power of the vacuum, but after Galileo’s death in 1642, Torricelli formulated his own hypothesis: The water was supported in the pipe by the pressure of the atmosphere. Torricelli reasoned that we live “at the bottom of an ocean of air” and that the pressure of the air pushing against the bottom of the pipe supported the water at a certain height in the pipe. A point of equilibrium was reached, he thought, when the weight of the water remaining in the pipe equaled the weight of the air pushing down from above.

From this hypothesis Torricelli derived several implications. One was that the pressure of the atmosphere would support a column of mercury about 29 inches high in a tube sealed at the top. This followed from the fact that the atmosphere supports a column of water 33 feet high, that mercury is 13.6 times as dense as water, and that 33/13.6 × 12 inches = 29 inches. A second implication was that such a tube ﬁlled with mercury could be used to measure ﬂuctuations in atmospheric pressure. This second implication won Torricelli credit for formulating the theory of the barometer. Finally, Torricelli reasoned that if such a device were conveyed to a place where the air was more rareﬁed, such as on a mountaintop, the column of mercury would descend.

The ﬁrst of these implications was tested by Torricelli’s associate, Vincenzo Viviani. Viviani obtained a 4-foot section of glass tube sealed at one end, enough mercury to completely ﬁll it, and a dish to hold more mercury. After pouring the mercury into the tube Viviani placed his thumb over the open end, inverted the tube, and placed the open end in the dish of mercury. After he released his thumb he watched the column of mercury descend to about 29 inches above the level of mercury in the dish. Thus was created the ﬁrst barometer. Its successful use in measuring atmospheric pressure came later.

The test of Torricelli’s third implication was taken up in 1647 by the French philosopher Blaise Pascal. Having received word of Torricelli’s experiments with the barometer, Pascal constructed one for himself. He readily became convinced of the correctness of Torricelli’s hypothesis, and to demonstrate its correctness in opposition to the vacuum principle, he requested that his brother-in-law, F. Perier, convey a barometer to the top of the Puy de Dôme, one of the highest mountains in Auvergne. A year later Perier was able to fulﬁll this request. He began the experiment by setting up two barometers in the monastery at the foot of the mountain. After noting that both columns of mercury rose to an identical height, he disassembled one of the barometers and instructed one of the friars to check the mercury level in the other throughout the day. Then Perier, accompanied by a group of witnesses, set oﬀ up the mountain with the other barometer. On reaching the summit, he assembled the second barometer and discovered to the amazement of all that the mercury level was more than 3 inches lower than it had been at the foot of the mountain. As a double check the barometer was taken apart and reassembled at ﬁve diﬀerent spots on the summit. Each time the results were the same.

At the midpoint of his descent Perier reassembled the barometer once again. He found that the mercury level was about midway between where it was at the bottom and at the top of the mountain. Finally, on returning to the monastery, the friar who had been watching the barometer there was questioned about what he had observed. He reported that the mercury level had not changed since early that morning when the group had departed. Pascal announced the results of this experiment to the educated world, and the announcement succeeded in abolishing the principle that nature abhors a vacuum.

Spontaneous Generation

The theory of spontaneous generation holds that living beings arise spontaneously from lifeless matter. The roots of the theory extend into ancient times. Aristotle held that worms, the larvae of bees and wasps, ticks, ﬁreﬂies, and other insects developed continually from the morning dew and from dry wood and hair. He also held that crabs and various molluscs developed from moist soil and decaying slime. Extensions of this theory prevailed throughout the Middle Ages and well into modern times. In the seventeenth century it was widely held that frogs were produced from the slime of marshes and eels from river water, and the physician Van Helmont thought that mice were produced from the action of human sweat on kernels of wheat. All one needed to do, according to Van Helmont, was toss a dirty shirt into a container of wheat, and in 21 days the container would be teeming with mice. Even Descartes and Newton accepted the theory of spontaneous generation. Descartes held that various plants and insects originated in moist earth exposed to sunlight, and Newton thought that plants were produced from emanations from the tails of comets.

The ﬁrst systematic eﬀort to abolish the belief in spontaneous generation was made by the Italian physician Francesco Redi. In response to the commonly held idea that worms were spontaneously generated in rotting meat, Redi hypothesized that the worms were caused by ﬂies. An immediate implication was that if ﬂies were kept away from the meat, the worms would not develop. To test this hypothesis Redi cut up a piece of meat and put part of it in sealed glass ﬂasks and the other part in ﬂasks open to the air. Flies were attracted to the open ﬂasks, and in a short time worms appeared; but no worms developed in the ﬂasks that were sealed.

When Redi published his ﬁndings in 1668, they immediately aﬀected the theory of spontaneous generation. Within a few years, though, the microscope came into common use, and it was discovered that even though meat sealed in glass containers produced no worms, it did produce countless microorganisms. The theory of spontaneous generation was thus reawakened on the microbial level.

By the middle of the nineteenth century the theory had received considerable reﬁnement. It was thought that spontaneous generation resulted from the direct action of oxygen on lifeless organic nutrients. Oxygen was thought to be essential to the process because the technique of canning fruits and vegetables had come into practice, and people knew that boiling fruits and vegetables and sealing them in the absence of oxygen would cause them to be preserved. If they were left exposed to the air, however, microbes would develop in a short time.

One of the defenders of spontaneous generation at that time was the Englishman John Needham, an amateur biologist. Needham conducted an experiment in which flagons containing oxygen and a vegetable solution were buried in hot coals. The coals would have been expected to kill any life in the solution, but several days later the contents of the ﬂagons were alive with microbes. Needham concluded that the oxygen acting alone on the nutrient solution caused the generation of the microbes. In response to this experiment, Lazzaro Spallanzani, an Italian physiologist, conducted a similar experiment. To ensure that the nutrient solution was lifeless he boiled it for an hour. Later no microbes could be found. To this Needham objected that in boiling the solution for a full hour Spallanzani had destroyed its “vegetative force.” In addition, Needham argued, he had polluted the small amount of oxygen in the containers by the fumes and heat. Thus, it was no wonder that microbes were not spontaneously generated.

To settle the issue once and for all, the French Academy of Science oﬀered a prize for an experimental endeavor that would shed light on the question of spontaneous generation. This challenge drew Louis Pasteur into the controversy. Spontaneous generation presented a special problem for Pasteur because of his previous work with fermentation. He had discovered that fermentations, such as those involved in the production of wine and beer, required yeast; and yeast, as he also discovered, was a living organism. In view of these ﬁndings Pasteur adopted the hypothesis that life comes only from life. An immediate implication was that for life forms to develop in a sterile nutrient solution, they must ﬁrst be introduced into the solution from the outside.

It was well known that life forms did indeed develop in sterile nutrient solutions exposed to the air. To account for this Pasteur adopted the second hypothesis that life forms are carried by dust particles in the air. To test this second hypothesis Pasteur took a wad of cotton and drew air through it, trapping dust particles in the ﬁbers. Then he washed the cotton in a mixture of alcohol and examined drops of the ﬂuid under a microscope. He discovered microbes in the ﬂuid.

Returning to his ﬁrst hypothesis, Pasteur prepared a nutrient solution and boiled it in a narrow-necked ﬂask. As the solution boiled, the air in the neck of the ﬂask was forced out by water vapor, and as it cooled the water vapor was slowly replaced by sterilized air drawn through a heated platinum tube. The neck of the ﬂask was then closed oﬀ with a ﬂame and blowpipe. The contents of the ﬂask thus consisted of a sterilized nutrient solution and unpolluted sterilized air—all that was supposedly needed for the production of life. Over time, however, no life developed in the ﬂask. This experiment posed a serious threat to the theory of spontaneous generation.

Pasteur now posed the hypothesis that sterile nutrient solutions exposed to the air normally developed life forms precisely because these forms were deposited by dust particles. To test this third hypothesis Pasteur reopened the ﬂask containing the nutrient solution, and, using a special arrangement of tubes that ensured that only sterilized air would contact the solution, he deposited a piece of cotton in which dust particles had been trapped. The ﬂask was then resealed, and in due course microbes developed in the solution. This experiment proved not only that dust particles were responsible for the life but that the “vegetative force” of the nutrient solution had not been destroyed by boiling, as Needham was prone to claim.

Pasteur anticipated one further objection from the proponents of spontaneous generation: Perhaps the capacity of oxygen to generate life was destroyed by drawing it through a heated tube. To dispel any such notions Pasteur devised yet another experiment. He boiled a nutrient solution in a ﬂask with a long, narrow gooseneck. As the solution boiled, the air was forced out, and as it cooled, the air returned very slowly through the long neck, trapping the dust particles on the moist inside surface. No microbes developed in the solution. Then, after a prolonged wait, Pasteur sealed the ﬂask and shook it vigorously, dislodging the particles that had settled in the neck. In a short time the solution was alive with microbes.

When Pasteur reported these experiments to the Academy of Science in 1860, he was awarded the prize that had been oﬀered a year earlier. The experiments dealt a mortal blow to the theory of spontaneous generation, and although the theory was not abandoned immediately, by 1900 it had very little support.

13.3 The Proof of Hypotheses

The four instances of hypothetical reasoning in science that we have investigated illustrate the use of two different kinds of hypotheses. The hypotheses involved in the discovery of Neptune and radium are sometimes called empirical hypotheses, and those relating to atmospheric pressure and spontaneous generation are sometimes called theoretical hypotheses. Empirical hypotheses concern the production of some thing or the occurrence of some event that can be observed. When radium had ﬁnally been obtained as a pure metal it was something that could be seen directly, and when Neptune was ﬁnally sighted through the telescope, it, too, had been observed. Theoretical hypotheses, on the other hand, concern how something should be conceptualized. When Galileo observed the water level rising in a suction pump, he conceived it as being sucked up by the vacuum. When Torricelli observed it, however, he conceived it as being pushed up by the atmosphere. Similarly, when Needham observed life emerging in a sterile nutrient solution, he conceived it as being spontaneously generated by the action of oxygen. But when Pasteur observed it, he conceived it as being implanted there by dust particles in the air.

The distinction between empirical and theoretical hypotheses has certain difficulties, which we will turn to shortly, but it sheds some light on the problem of the veriﬁcation or conﬁrmation of hypotheses. Empirical hypotheses are for all practical purposes proved when the thing or event hypothesized is observed. Today practically all of us would agree that the hypotheses relating to radium and Neptune have been established. Theoretical hypotheses, on the other hand, are never proved but are only conﬁrmed to varying degrees. The greater the number of implications that are found to be correct, the more certain we can be of the hypothesis. If an implication is found to be incorrect, however, a theoretical hypothesis can be disproved. For example, if it should happen some day that life is produced in a test tube from inorganic materials, Pasteur’s hypothesis that life comes only from life might be considered to be disproved.

The problem with the distinction between empirical and theoretical hypotheses is that observation is theory-dependent. Consider, for example, a man and a woman watching a sunrise. The man happens to believe that the sun travels around the earth, as Ptolemy held, and the woman that the earth travels around the sun, as Copernicus and Galileo contended. As the sun rises, the man thinks that he sees the sun moving upward, while the woman thinks she sees the earth turning. The point is that all of us have a tendency to see what we think is out there to be seen. As a result, it is sometimes diﬃcult to say when something has or has not been observed.

In regard to the discovery of Neptune, the unknown planet was observed two times in 1795 by J. J. Lalande, ﬁfty-one years before it was “discovered” by Adams, Leverrier, and Galle. Lalande noted that his observations of the position of the small starlike object were discordant, so he rejected one as erroneous. Because he thought he was observing a star, he received no credit for discovering a planet. Analogous remarks extend to Galle’s observations of the planet Neptune in 1846. If Leverrier’s computations had been erroneous, Galle might have seen what was really a comet. Thus, if we can never be sure that we really see what we think we see, is it ever possible for a hypothesis to be actually proved? Perhaps it is better to interpret the proof of empirical hypotheses as a high degree of conﬁrmation.

Conversely, with theoretical hypotheses, would we want to say that Torricelli’s hypothesis relating to atmospheric pressure has not been proved? Granted, we cannot observe atmospheric pressure directly, but might we not say that we observe it instrumentally? If barometers can be regarded as extensions of our sense organs, Torricelli’s hypothesis has been proved. Another example is provided by Copernicus’s hypothesis that the earth and planets move around the sun, instead of the sun and planets around the earth, as Ptolemy hypothesized. Can we consider this theoretical hypothesis to be proved? If a motion picture camera were sent outside the solar system and pictures were taken supporting the Copernican hypothesis, would we say that these pictures constituted proof? We probably would. Thus, while the distinction between theoretical and empirical hypotheses is useful, it is more a distinction in degree than in kind.

13.4 The Tentative Acceptance of Hypotheses

A certain amount of time is required for a hypothesis to be proved or disproved. The hypotheses relating to the discovery of radium and Neptune required more than a year to prove. Theoretical hypotheses in science often take much longer, and theoretical hypotheses in philosophy may never be conﬁrmed to the satisfaction of the majority of philosophers. During the period that intervenes between the proposal of a hypothesis and its proof, conﬁrmation, or disproof, the question arises as to its tentative acceptability. Four criteria that bear on this question are (1) adequacy, (2) internal coherence, (3) external consistency, and (4) fruitfulness.

Adequacy is the extent to which a hypothesis ﬁts the facts it is intended to unify or explain. A hypothesis is said to “ﬁt” the facts when each fact can be interpreted as an instance of some idea or term in the hypothesis. For example, before the Neptune hypothesis was conﬁrmed, every ﬂuctuation in the position of Uranus could be interpreted as an instance of gravitational interaction with an unknown planet. Similarly, before Torricelli’s hypothesis was conﬁrmed, the fact that water would rise only 30 feet in suction pumps and siphons could be interpreted as an instance of equilibrium between the pressure of the water and the pressure of the atmosphere.

A hypothesis is inadequate to the extent that facts exist that the hypothesis cannot account for. The principle that nature abhors a vacuum was inadequate to explain the fact that water would rise no more than 30 feet in suction pumps and siphons. Nothing in the hypothesis could account for this fact. Similarly, Needham’s hypothesis that life is generated by the direct action of oxygen on nutrient solutions was inadequate to account for the fact that life would not develop in Pasteur’s ﬂask containing a sterilized nutrient solution and sterilized oxygen.

In scientific hypotheses a second kind of adequacy is the accuracy with which a hypothesis accounts for the data. If one hypothesis accounts for a set of data with greater accuracy than another, then that hypothesis is more adequate than the other. For example, Kepler’s hypothesis that the orbits of the planets were ellipses rather than circles, as Copernicus had hypothesized, accounted for the position of the planets with greater accuracy than the Copernican hypothesis. Similarly, Einstein’s theory of relativity accounted for the precise time of certain eclipses with greater accuracy than Newton’s theory. For these reasons Kepler’s and Einstein’s theories were more adequate than the competing theories.

Internal coherence is the extent to which the component ideas of a hypothesis are rationally interconnected. The purpose of a hypothesis is to unify and interconnect a set of data and by so doing to explain the data. Obviously, if the hypothesis itself is not internally connected, there is no way that it can interconnect the data. After Adams and Leverrier had worked out the mathematical details of the Neptune hypothesis, it exhibited a great deal of internal coherence. The hypothesis showed how all the ﬂuctuations in the position of Uranus could be rationally linked in terms of the gravitational interaction of an eighth planet. Similarly, Torricelli’s hypothesis showed how the various ﬂuid levels could be rationally interconnected in terms of the equilibrium of pressures. Internal coherence is responsible for the features of elegance and simplicity that often attract scientists to a hypothesis.

An example of incoherence in science is provided by the theoretical interpretation of light, electricity, and magnetism that prevailed during the ﬁrst half of the nineteenth century. During that period each of these phenomena was understood separately, but the interconnections between them were unknown. Toward the end of the century the English physicist James Clerk Maxwell showed how these three phenomena were interconnected in terms of his theory of the electromagnetic ﬁeld. Maxwell’s theory was thus more coherent than the ones that preceded it.

Similarly, in philosophy, Spinoza’s metaphysical theory is more internally coherent than Descartes’s. Descartes postulated the existence of two kinds of substance to account for the data of experience. He introduced extended, material substance to explain the data of the visible world, and nonextended, immaterial substance to explain the phenomena of the invisible world, including the existence and activity of the human soul. But Descartes failed to show how the two kinds of substance were interconnected. In the wake of this disconnection the famous mind-body problem arose, according to which no account could be given of how the human body acted on the mind through the process of sensation or how the mind acted on the body through the exercise of free choice. Spinoza, on the other hand, postulated only one substance to account for everything. Spinoza’s theory is thus more internally coherent than Descartes’s.

External consistency occurs when a hypothesis does not disagree with other, wellconﬁrmed hypotheses. Adams’s and Leverrier’s hypothesis of an eighth planet was perfectly consistent with the nineteenth-century theory of the solar system, and it was rendered even more attractive by the fact that the seventh planet, Uranus, had been discovered only a few years earlier. Similarly, Marie Curie’s hypothesis of the existence of a new element was consistent with Mendeleev’s periodic table and with the general hypothesis that elements could emit penetrating rays. In 1890 Mendeleev’s table had certain gaps that were expected to be ﬁlled in by the discovery of new elements, and two ray-emitting elements, thorium and uranium, had already been discovered.

The fact that a hypothesis is inconsistent with other, well-conﬁrmed hypotheses does not, however, immediately condemn it to obscurity. It often happens that a new hypothesis arises in the face of another, well-conﬁrmed hypothesis and that the two hypotheses compete for acceptance in the future. Which hypothesis will win is determined by an appeal to the other three criteria. For example, Torricelli’s hypothesis was inconsistent with the ancient hypothesis that nature abhors a vacuum, and Pasteur’s hypothesis was inconsistent with the equally ancient hypothesis of spontaneous generation. In the end, the newer hypotheses won out because they were more adequate, coherent, or fruitful than their competitors. For the same reason, the Copernican hypothesis eventually triumphed over the Ptolemaic, the theory of oxidation won out over the old phlogiston theory, and Einstein’s theory of relativity won out over Newton’s theory.

Fruitfulness is the extent to which a hypothesis suggests new ideas for future analysis and conﬁrmation. Torricelli’s hypothesis suggested the design of an instrument for measuring fluctuations in the pressure of the atmosphere. Similarly, Pasteur’s hypothesis suggested changes in the procedures used to maintain sterile conditions in hospitals. After these latter changes were implemented, the death rate from surgical operations decreased dramatically. The procedure of pasteurization, used to preserve milk, was another outgrowth of the hypothesis that life comes only from life.

Newton’s theory of universal gravitation is an example of a hypothesis that proved especially fruitful. It was originated to solve the problem of falling bodies, but it also explained such things as the ebb and ﬂow of the tides, the orbital motion of the moon and planets, and the ﬂuctuations in planetary motion caused by a planet’s interaction with other planets. Einstein’s theory of relativity is another example. It was originated to account for certain features of Maxwell’s theory of electricity and magnetism, but it ushered in the atomic age forty years later.

The factors of coherence and fruitfulness together account for the overall rationality and explanatory power of a hypothesis. Suppose, for example, that someone formulated the hypothesis that the water level in suction devices is maintained by the action of demons instead of by atmospheric pressure. Such a hypothesis would be neither coherent nor fruitful. It would not be coherent because it would not explain why the maximum water level in these devices is consistently about 30 feet, why the mercury level in barometers is much less, and why the mercury level in a barometer decreases when the instrument is carried to the top of a mountain. Do the demons decide to maintain these levels by free choice or according to some plan? Because there is no answer to this question, the hypothesis exhibits internal disconnectedness, which leaves it open to the charge of being irrational. As for the fourth criterion, the demon hypothesis is unfruitful because it suggests no new ideas that experimenters can put to the test. The hypothesis that nature abhors a vacuum is hardly any more fruitful, which accounts in part for why it was so suddenly abandoned in favor of Torricelli’s hypothesis—it simply did not lead anywhere.

In summary, for any hypothesis to receive tentative acceptance it must cover the facts it is intended to interpret and it must rationally interconnect these facts—in other words, it must be adequate and coherent. After that, it helps if the hypothesis does not conﬂict with other, well-conﬁrmed hypotheses. Finally, it is important that a hypothesis capture the imagination of the community to which it is posed. This it does by being fruitful—by suggesting interesting ideas and experiments to which members of the community can direct their attention in the years ahead.

Chapter 14: Science and Superstition

14.1 Distinguishing Between Science and Superstition

The idea that the human mind is capable of operating on diﬀerent levels in its eﬀort to comprehend reality is as old as philosophy itself. Twenty-four centuries ago Plato drew a distinction between what he called opinion and knowledge. Opinion, he said, is a kind of awareness that is uncertain, conﬁned to the particular, inexact, and subject to change, whereas knowledge is certain, universal, exact, and eternally true. Every human being starts out in life by operating on the level of opinion, and only through great struggle and eﬀort can he or she escape it and rise to the level of knowledge. This struggle is called education, and it opens the eye of the mind to realities that cannot even be imagined from the standpoint of opinion.

Today’s distinction between science and superstition is a modern equivalent of Plato’s distinction between knowledge and opinion. Everyone recognizes that science has revealed wonderful truths about the world of nature. It has put men on the moon, wiped out life-threatening diseases, and ushered in the computer age. Also, almost everyone recognizes that superstition is little better than foolishness. It leads people to fear walking under ladders, breaking mirrors, and spilling salt. Practically everyone agrees that if some claim is grounded in science, then it is probably worthy of belief, while if it is grounded in superstition, then it should probably be ignored. Where people do not agree, however, is in what constitutes science and what constitutes superstition. What one person calls science another calls superstitious nonsense.

Both science and superstition involve hypotheses, so the four criteria developed in Chapter 13 for evaluating hypotheses are relevant to the distinction between science and superstition: adequacy, internal coherence, external consistency, and fruitfulness. But the distinction between science and superstition also involves psychological and volitional elements. It involves such factors as how the observer’s subjective states inﬂuence how he sees the world, and how his needs and desires play a role in the formation of his beliefs. Accordingly, to explore the distinction between science and superstition, we must introduce criteria that include these psychological and volitional elements. The criteria suggested here are evidentiary support, objectivity, and integrity. The following account of evidentiary support encompasses adequacy and fruitfulness, and the account of integrity encompasses adequacy, internal coherence, and external consistency.

Science and superstition are, in large measure, polar opposites. Where scientiﬁc activity recognizes the importance of evidentiary support, objectivity, and integrity, superstition ignores them. Accordingly, these criteria can be used as a kind of measuring stick for sizing up the various beliefs people have about the world. To the extent that those beliefs are supported by evidence, are objective, and arise from research that reﬂects integrity, the closer they come to the ideal of science, and the more justiﬁed they are. Conversely, to the extent that our beliefs do not share in these characteristics, the closer they come to the “ideal” of superstition, and the less justiﬁed they are.

Note, however, that to say a belief is justiﬁed is not to say it is true in any absolute sense. As we saw in Chapter 13, all beliefs that arise from science are tentative at best. But such beliefs are the best ones we can have for now. Also, to say that a belief is not justiﬁed is not to say it is absolutely false. It is quite possible that a belief grounded in superstition today could tomorrow be grounded in science. But such a belief is not worthy of assent today. An analogy can be found in rolling dice. No sensible person would bet even money that a pair of dice will come up “snake eyes” on the next roll, even though he realizes that tomorrow it might be discovered that the dice were loaded in favor of this outcome.

14.2 Evidentiary Support

In the preceding chapter we saw that hypotheses in themselves are mere conjectures, and before they are believed they should be supported by evidence. This rule applies equally to the hypotheses that underlie science and to those that underlie superstition. This rule is strictly obeyed in science, but it is often ignored in the realm of superstition. For example, in the sixteenth century Copernicus formulated the hypothesis that the sun is the center of our planetary system and that the earth revolves around the sun—in opposition to the prevailing Ptolemaic hypothesis, which put the earth at the center. In the years that followed, the telescope was invented, and thousands of subsequent observations conﬁrmed the Copernican hypothesis and disconﬁrmed the Ptolemaic hypothesis. Without these observations, the Copernican hypothesis would never have been adopted.

In contrast, consider the superstition that breaking a mirror will bring bad luck. No evidence has ever been collected in support of this hypothesis. No tests have been conducted and no experiments performed. Possibly on one occasion or other someone did break a mirror and later lost money on the stock exchange or was injured in a car accident, but surely it would be unreasonable to believe that breaking a mirror caused the loss or the accident. Such reasoning constitutes a classic case of the post hoc ergo propter hoc (false cause) fallacy. But in spite of the lack of evidence, many people believe the mirror hypothesis.

Before inquiring further into the need for evidence, however, we must ﬁrst investigate what counts as evidence. Does the testimony of authorities count as evidence? What about ancient authorities? Does the Bible count as evidence? The answer is that scientiﬁc hypotheses are about the natural world, so only observations of the natural world count as evidence. Every scientiﬁc experiment is a question the experimenter asks of the world, and the result of that experiment is nature’s reply. The problem with the testimony of authority is that we have no certain knowledge that the authority is correct in his or her assessment. The same holds true of the Bible. We have no way of knowing whether what the Bible says about the natural world is true. If someone should reply that the Bible is divinely inspired, then the obvious reply is, How do we know that? Do we have any observational evidence for it? Appeals to authority ﬁgures and the Bible amount to passing the evidentiary buck.

Another kind of evidence that is considered unreliable is anecdotal evidence. Suppose that you have cancer and a friend advises you that eating garlic can cure it. You decide to take this advice, and after eating a clove of garlic every day for a year, the cancer goes into remission. Did the garlic cure the cancer? Evidence of this sort is called anecdotal, and science usually rejects it. The trouble with anecdotal evidence is that it is too isolated to establish any causal connection. Thus, the garlic evidence ignores the thousands of people with cancer who have eaten garlic and have not been cured, and it ignores the thousands of people who have experienced spontaneous remission of cancer and have not eaten any garlic. Also, there is no way to turn the clock back and try the experiment again.

One of the key features of scientiﬁc evidence gathering is that an experiment be replicable under controlled conditions. This means that the experiment must be repeatable by diﬀerent scientists at diﬀerent times and places. Replicability helps ensure that the outcome of the experiment did not result from anything peculiar to one certain experimenter operating at a single place and time. Also, the controlled conditions are designed to eliminate the inﬂuence of extraneous factors. Perhaps, in reference to the garlic example, the cure was eﬀected not by the garlic but by something else that was eaten, or by any one of a thousand other factors that occurred during this time, or by any combination of these factors.

The evidence oﬀered in support of superstitious hypotheses is rarely replicable, and when it is, the outcome almost always fails to support the hypothesis. For example, the belief in ghosts is usually supported by what one or more individuals claim to have seen on some unique occasion. This occasion can never be repeated. And the belief in psychic phenomena such as extrasensory perception is sometimes claimed to be supported by experiments involving Zener cards: cards imprinted with crosses, circles, wavy lines, stars, and squares whose image an observer might “transmit” to a psychic receiver. But when these experiments have been repeated under carefully controlled conditions, the outcome has never been other than what would have been expected to occur through mere chance.

Another defect found in superstitious hypotheses is that they are often framed so vaguely that it is virtually impossible to provide any kind of unequivocal conﬁrmation. For example, according to feng shui (pronounced fung-shway), an ancient Chinese system of magic, bad luck travels in straight lines, whereas good luck does not. As a result, one invites bad luck by living in a house or apartment that has two (or, what is worse, three) doors lined up in a row. But what, exactly, is the meaning of bad luck? What is interpreted as bad luck today may turn out to be good luck tomorrow. If a person loses $1,000 in the stock market today, that may lead her to be more cautious in the future, and that increased caution may save $10,000 later on.

In contrast, the hypotheses of science are often framed in the language of mathematics, or they can at least be translated into some mathematical expression. This fact provides for extremely accurate conﬁrmations and is largely responsible for the extraordinary success science has enjoyed during the past 500 years. For example, in 1802 the French chemist Joseph Louis Gay-Lussac formulated the hypothesis that if the temperature is raised one degree Celsius in a closed container of gas—any gas—the pressure of the gas will increase by 0.3663 percent. The hypothesis has been tested thousands of times by chemists and students in chemistry labs, and it has been found to be correct.

Closely related to the problem of vagueness is the breadth with which a hypothesis is framed. If a hypothesis is framed so broadly and comprehensively that even contradictory evidence serves to conﬁrm it, then the hypothesis is not really conﬁrmed by anything. Suppose, for example, that a health care practitioner should invent a hypothesis involving diet: Practicing this diet is guaranteed to make you feel great, but before it has this eﬀect it may make you feel either rotten or the same as usual. After following this diet for six months you report that you feel the same as before. The practitioner replies that your experience conﬁrms the hypothesis, because this is what the diet is supposed to do. On the other hand, suppose that after six months you feel great, or perhaps rotten. Again the practitioner will report that your experience conﬁrms the hypothesis. Hypotheses of this sort are not genuinely scientiﬁc.

In 1919 the philosopher Karl Popper discovered this very problem concerning hypotheses. In response, he argued that any genuinely scientiﬁc hypothesis must be framed narrowly enough so that it forbids certain things from happening. In other words, the hypothesis must be falsiﬁable. In the years following its announcement, many philosophers criticized Popper’s falsiﬁability criterion because, strictly speaking, hypotheses are rarely susceptible of being disproved. But, as we saw in Chapter 13, hypotheses can be disconfirmed (or rendered less plausible). Thus, we can retain Popper’s basic insight by requiring that any genuinely scientiﬁc hypothesis be disconfirmable. This means that the hypothesis must be framed narrowly enough so that it is possible for evidence to count against it. Newton’s gravitational hypothesis, for example, satisﬁes this criterion because the discovery of two large bodies that failed to attract each other would tend to disconﬁrm the hypothesis. But the dietary hypothesis just mentioned fails the disconﬁrmability criterion because no outcome could ever count against it.

A problem closely associated with excessively broad hypotheses arises in connection with what are called ad hoc modiﬁcations of hypotheses. For an example, suppose that you are a sociologist conducting research into alcoholism. You formulate a hypothesis that alcoholism is caused by cultural factors that present alcohol consumption in a favorable light. When you gather evidence to support this hypothesis, however, you ﬁnd that relatively few people who come from such cultures are alcoholics. Thus, you modify the hypothesis to say that alcoholism is caused by cultural factors but only when a genetic predisposition exists. But then you ﬁnd that many alcoholics drink to ease the pain of depression and other psychological problems. Thus, you modify the hypothesis once again to take this fact into account. Further research shows that parental drinking patterns play a role, so you add another modiﬁcation. These changes are called ad hoc (“to this”) modiﬁcations because they are introduced purely to cover some problem or anomaly that was not recognized when the hypothesis was first framed.

The problem with ad hoc modiﬁcations is that their purpose is to shore up a failure of evidentiary support in the original hypothesis. As more and more modiﬁcations are added, the hypothesis becomes self-supporting; it becomes a mere description of the phenomenon it is supposed to explain. For example, suppose that we introduce a certain hypothesis h to explain the occurrence of a certain phenomenon x among a group of entities A, B, C, D, E. As ad hoc modiﬁcations are added, we ﬁnd that A has x because of some unique attribute a, B has x because of b, and so on. In the end our hypothesis states that anyone who has attributes a, b, c, d, e exhibits x. But the set of attributes a, b, c, d, e is simply a description of A, B, C, D, E. If we should ask why entity A has x, the answer is that A has x because of a, where a is just a unique something that A has. Applying this analysis to the alcoholism hypothesis, if we ask why a certain person (let us call him Smith) is an alcoholic, the answer is that Smith is an alcoholic because he has a certain attribute s that causes him to be an alcoholic. The explanation is vacuous.

Another problem with ad hoc modiﬁcations is that they result in hypotheses that are so complicated that applying them becomes diﬃcult. Science has always favored simplicity over complexity. Given two hypotheses that explain the same phenomenon, the simpler of the two is always the preferable one. In part this preference is aesthetic. The simpler hypothesis is more “beautiful” than the more complex one. But the preference for simplicity also results from the application of what has been called “Ockham’s razor.” This is a principle, introduced by the fourteenth-century philosopher William of Ockham, that holds that theoretical entities are not to be multiplied needlessly. Why settle for a complicated theory when a simpler one works equally well? Besides, the simpler one is easier to apply.

One consequence of Ockham’s razor is that naturalistic explanations are preferable, at least at the outset, to supernatural ones. For example, over the years there have been numerous reports of statues depicting Jesus or the Virgin Mary that are said to weep tears of water, oil, or blood. After viewing these statues, thousands of faithful come away convinced that they have seen a miracle. However, Ockham’s razor says that we should look for naturalistic explanations ﬁrst. Perhaps the tears came from condensation or from rainwater that may have leached into the statue from small surface cracks. Another possible explanation is pranksters. Scientists have shown how the appearance of tears can be replicated by drilling a small hole in a statue’s head, pouring a liquid into the hole, and then scratching away a bit of lacquer or paint from the corner of the eyes. The liquid will percolate down through the porous substance of the statue and escape through the scratches.

Another, closely related consequence of Ockham’s Razor is that explanations based on known realities are preferred to those based on the fantastic or the bizarre. For example, there is a region in Peru, high in the Andes mountains, known as the Nazca Desert. The ﬂoor of that desert features large drawings of geometrical ﬁgures and animals such as monkeys, hummingbirds, and lizards. The drawings were produced more than a thousand years ago by scraping away the iron-oxide-coated pebbles that cover the land, but given their huge size, they can be recognized as coherent ﬁgures only from a position high above. One accepted explanation is that the Nazca Indians produced the drawings using simple tools and surveying equipment to impress the rain gods that dwell in the clouds. However, the best-selling author Erich von Däniken proposes that the lines represent landing strips for alien spacecraft, and it was those aliens who guided the Indians from overhead in producing the drawings. Given its farfetched character, this explanation would seem to violate Ockham’s razor.

Returning to the question of evidentiary support, one of the surest ways to know that our hypotheses are supported by evidence is that they lead to predictions that turn out to be true. Each true prediction represents a pillar of support for the hypothesis. But some predictions are better than others, and the best ones are those that reveal ways of viewing the world that would never have been dreamed of apart from the hypothesis. If a hypothesis leads to predictions of this sort, and if those predictions are conﬁrmed by evidence, then the hypothesis has earned a very special kind of support. Such a hypothesis reveals hidden truths about nature that would never have been recognized without it.

A classic example of a prediction of this kind resulted from the hypothesis underlying Einstein’s general theory of relativity. One of the consequences predicted by this hypothesis is that light is aﬀected by gravity. In particular, the hypothesis predicted that a light ray coming from a star and passing by the sun would be bent in the direction of the sun. As a result, the position of the star with respect to other stars would appear to be diﬀerent from what it was usually observed to be. Of course, testing such a prediction under normal circumstances would be impossible, because the light of the sun is so bright that it completely blocks out the light from stars. But it could be tested during a solar eclipse. Such an opportunity arose on May 29, 1919, and scientists took advantage of it. The prediction turned out to be true, and as a result Einstein’s theory was quickly adopted. Within a few years the theory led to the discovery of atomic energy.

Hypotheses that yield striking, novel predictions are largely responsible for progress in science. And it is precisely these kinds of predictions, argues the philosopher Imre Lakatos, that distinguish science from pseudoscience. Of course, not every scientiﬁc hypothesis leads to such startling predictions as Einstein’s, but this kind of hypothesis may at least be integrally connected to broader, umbrella hypotheses that have led to such predictions. In contrast, the hypotheses underlying astrology have been around for twenty-seven centuries, and they have produced not a single startling prediction that has been veriﬁed and not a single new insight into the course of human events. They have produced no master plan for future civilization and no hint about future discoveries in physics or medicine. This lack of progress over centuries is one reason that the philosopher Paul Thagard named astrology a pseudoscience.

14.3 Objectivity

Our beliefs about the world are objective to the extent that they are unaﬀected by conditions peculiar to the experiencing subject. Such conditions can be either motivational or observational. For example, a belief that is motivated by the emotions of the experiencing subject and that exists for the primary purpose of satisfying those emotions tends to lack objectivity. Also, a belief that is grounded in observations peculiar to the experiencing subject, such as visual hallucinations, lacks objectivity. Even though objectivity is an ideal that can never be completely attained, practically everyone would agree that beliefs are more trustworthy if their content is not distorted by the experiencing subject. The scientist constantly strives to avoid such distortions, but the superstitious mind either revels in them or, in the more tragic cases, succumbs to them.

All superstitions exist at least in part to satisfy the emotional needs of the experiencing subject. The chief emotions that give rise to superstitious beliefs are fear and anxiety, and they are often reinforced by a disposition to fantasy and mental laziness. Much of the fear and anxiety is generated by the fact that everyone dies. Death can come suddenly, as in a freeway accident, a fall from a roof, or an avalanche, or it can come as a result of cancer, heart failure, or stroke. Short of death, everyone is subject to injury with its attendant pain, and most people at some time experience the mental suﬀering that accompanies rejection, loneliness, and failure.

People have little control over these facts of life, and to relieve the anxiety they produce, many resort to charms and amulets, the rosary beads dangling from the rearview mirror or the scapular or medal worn around the neck. If nothing else will protect us from the terrors of life, perhaps these objects will. After all, science has failed to conquer disease and death, and it oﬀers to the believer nothing but tentative truths that may change tomorrow. To the person facing an uncertain future, dejection, or loneliness, it may seem more reasonable to dial up the Psychic Friends Network and buy a bit of immediate consolation than to trust in science.

A second element in the human condition that generates anxiety is freedom and the responsibility freedom entails. The idea that you, and you only, are in charge of your destiny can be an extremely frightening idea. Many people recoil from the thought and seek refuge in a leader or guru. They turn all their power of critical thinking over to this leader and blindly follow his or her instructions to the last detail. When the leader orders them to believe any form of nonsense, no matter how silly, they do so. The belief or practice ordered by the leader, they are told, is essential to their protection. But, following such orders can sometimes lead to tragedy, as it did in the Jonestown massacre in 1978 and the Heaven’s Gate suicides in 1997.

A disposition to magical ways of thinking and mental laziness greatly facilitates the ﬂight to superstition. Many people, if not most, are fascinated by the mysterious, the arcane, and the occult, and some would rather believe an explanation clothed in magic than they would a scientiﬁcally grounded one. The psychologists Barry F. Singer and Victor A. Benassi performed a series of experiments on their students in which People are they had a magician pose as a “psychic” and perform demonstrations of psychic feats. fascinated by the Before the demonstrations began, the students were told repeatedly in the clearest lanmysterious. guage that the magician was only pretending to be a psychic, and that what they were about to witness was really a series of conjurer’s tricks. Nevertheless, in spite of these warnings, a majority of students concluded, in one experiment after another, that the magician was really a psychic. Furthermore, many concluded that the magician was an agent of Satan.

The disposition toward the magical and the fantastic is greatly reinforced by the media, particularly television and motion pictures. The media are slavishly subservient to the entertainment desires of their audience, so, given a widespread fascination with the magical, the media issue a constant stream of movies, miniseries, and “news” stories devoted to that subject. These programs touch everything from vampires and disembodied spirits to irrational conspiracies and the intervention of angels. This persistent attention to the fantastical increases the public’s acceptance of superstitious explanations whenever realistic ones are not readily available, or even in the face of realistic explanations.

A disposition to mental laziness also assists in the formation of superstitious beliefs. It is, in fact, extremely diﬃcult to ensure that one’s beliefs are supported by evidence and that they pass the test of internal coherence. Sloppy logic is so easy it is no wonder people resort to it. Most of the informal fallacies treated in Chapter 3 can arise from Superstitions are sloppy thinking. After old Mrs. Chadwicke hobbled past the church, lightning struck supported by sloppy the steeple and burned the church to the ground. Obviously old Mrs. Chadwicke is a thinking. witch (false cause). Furthermore, old Mrs. Chadwicke wears a black cape and a black hood. It must be the case that all witches wear such clothing (hasty generalization). And of course witches exist because everybody in the village believes in them (appeal to the people).

Another kind of sloppy thinking involves an appeal to what might be called false coherence. A farmer discovers that one of his cows has been killed. At the same time the farmer happens to read a story in a local tabloid saying that a satanic cult is operating in the vicinity. The cult practices its rites on the thirteenth day of each month. The cow was killed on the thirteenth. Thus, the farmer concludes that the cow was killed by Satan worshippers. This line of thinking involves many loose ends, but that rarely deters people from drawing a conclusion. Becoming a clear, critical thinker is one of the primary goals of education, but unfortunately becoming educated is no less of a struggle for students today than it was for students in Plato’s day.

Thus far we have focused on emotions and dispositions in the experiencing subject that lead to superstitious beliefs. We now turn to some of the many ways that our observation of the world can be distorted. Such distortions consitute avenues in which conditions peculiar to the experiencing subject enter into the content of observation. When such distorted observations are combined with the emotions and dispositions mentioned earlier, superstitious beliefs are likely to arise. The distorted observations can occur in the same person who has the emotions and dispositions or they can be conveyed secondhand. In either case, the combination leads to superstition.

One well-documented phenonemon that inﬂuences our observation of our own bodily states is the so-called placebo eﬀect. Recall that a placebo is any kind of “medicine” or procedure that provides no medicinal or therapeutic beneﬁt by itself but that can eﬀect a cure when the patient is told that it has such beneﬁt. For example, patients with knee pain have been told that an operation will cure them, and after they undergo a minor incision that, by itself, has no therapeutic eﬀect, the pain often disappears. Also, patients who suﬀer from nervous tension or depression have been told that a little colored pill (which consists of nothing but sugar) will cure them, and after they take the pill, the tension or depression disappears. Obviously in these cases it is not the placebo alone that eﬀects the cure but the placebo together with the suggestion implanted in the patients’ minds by their doctors.

Another well-documented effect that influences our observation of the world around us is called pareidolia. This is the eﬀect by which we can look at clouds, smoke, or the textured coatings on walls and ceilings and see animals, faces, trees, and so on. We project familiar visual images onto vague, relatively formless sensory stimuli and “see” that image as if it were really there. Pareidolia is responsible for a good deal of religious superstition. For example, a few years ago a woman making burritos saw the face of Jesus in the skillet burns on a tortilla. She built a shrine to house the tortilla, and thousands of believers came to pray before it. More recently, a grilled cheese sandwich that believers thought bore the image of the Virgin Mary sold on eBay for $28,000. But incidents of this sort are not conﬁned to the Christian world. Followers of Islam have seen the Arabic words for “Allah” and “Mohamed” in such media as ﬁsh scales, egg shells, and lamb’s wool.

Closely related to pareidolia is the concept of the perceptual set, where “set” refers to our tendency to perceive events and objects in a way that our prior experience has led us to expect. The idea of perceptual set is a product of Gestalt psychology, according to which perceiving is a kind of problem solving. When we are confronted with a problem, such as ﬁnding the solution to a riddle or puzzle, we enter into a state of mental incubation in which potential solutions are turned over in our minds. This state is followed by a ﬂash of insight (assuming we are able to solve the puzzle), after which the solution seems obvious. When we consider the puzzle at a later time, the solution leaps into our minds. Such a solution is called a gestalt, from the German, for form or conﬁguration. Analogously, every act of perception involves solving the puzzle of organizing sensory stimuli into meaningful patterns. Each such pattern is a perceptual gestalt, or set, and once such a set is formed, it serves to guide the processing of future perceptions. As a result, we perceive what we expect to perceive.

In 1949 the psychologists Jerome S. Bruner and Leo J. Postman performed a famous experiment in which subjects were shown replicas of ordinary playing cards—but some of the cards had been altered by reversing their color. For example, in some groups of cards, the three of hearts was black and the six of spades was red. Of twenty-eight subjects, twenty-seven initially saw the altered cards as normal ones. One subject identiﬁed the black three of hearts as a three of spades on forty-four successive showings. This experiment clearly shows that we perceive what we expect to perceive and, indeed, this fact is familiar to everyone. For example, we expect to receive a phone call, and while taking a shower we think we hear the phone ringing, only to be told by someone in the other room that the phone did not ring. Or, while driving, we might approach a red octagonal sign that reads ST_P (our view of the sign being partly blocked by a tree branch between the S and the P). However, we bring the car to a stop, because we perceived the sign to read STOP. In fact, what our sense of vision received was three consonants (S, T, P), meaningless until processed through perception.

Yet another factor that influences our sense of vision is the autokinetic effect. According to this eﬀect a small, stationary light surrounded by darkness will often seem to move. One can prove the existence of this eﬀect for oneself by looking at a bright star on a dark night, or by observing a small stationary point of light in a dark room. Psychologists speculate that the autokinetic eﬀect results from small, involuntary motions of the eyeball of the observer, and they have shown that the eﬀect is enhanced by the reports of other observers. If someone standing nearby says that she just saw the object move, others will often conﬁrm this report. The autokinetic eﬀect is thought to be responsible for many claims of UFO sightings.

Hallucinations of various sorts can also distort the content of perception. Two kinds of hallucination that aﬀect many people in the drowsy moments between sleep and wakefulness are hypnagogic and hypnopompic hallucinations (Hines, 1988, 61–62). The former occur just before drifting oﬀ to sleep, when the brain’s alpha waves are switching to theta waves, and the latter occur just before awakening. During these moments the subject may experience extremely vivid, emotionally charged images that seem very real. These hallucinations are thought to be responsible for the ghosts and other appearances that people sometimes see in bedrooms.

Collective hallucinations are another kind of perceptual distortion that can occur in large crowds of people. Before such hallucinations can happen, the crowd must be brought to a heightened emotional state, which may be brought on by the expectation of seeing something important or miraculous. An occurrence of this sort may have happened on October 13, 1917, when some 70,000 people gathered in the village of Fatima in Portugal expecting to see a miraculous sign from heaven. At midday, one of the children who was supposedly in contact with the Virgin Mary cried out to the people to look at the sun. They did so, whereupon thousands saw the sun swirl amid the clouds and plunge toward the earth. Of course, if the sun had actually moved, it would have triggered seismograph readings all over the globe. Also, many of the people there did not see anything unusual, but their reports were discounted. Nevertheless, even to this day many of the faithful take this observation of the swirling sun as evidence of a miracle.

Finally, the operation of memory can distort the way we recall our observations. Human memory is not like the process whereby a computer recalls information from its hard drive with total accuracy. Rather, it is a creative process susceptible to many inﬂuences. When images are recalled from human memory, they are retrieved in bits and pieces. The brain then ﬁlls in the gaps through a process called confabulation. The brain naturally and unconsciously tries to produce a coherent account of what happened, but precisely how the gaps are ﬁlled in depends on such things as one’s feelings at the time of recall, other people’s suggestions about the event recalled, and one’s own successive reports of what happened. Given that memory recall is selective to begin with and that many details are inevitably left out, the ﬁnal picture recalled may range anywhere from a fairly accurate representation to a total fabrication.

These eﬀects represent only a few of the ways the subjective state of the observer can inﬂuence human observation and memory. To avoid such distortions, scientiﬁc inquiry restricts human observation to circumstances in which known aberrations of perception and recall are least likely to occur. In the natural sciences, much if not most observation occurs through instruments, such as volt meters, Geiger counters, and telescopes, the behavior of which is well known and highly predictable. The results are then recorded on relatively permanent media such as photographic ﬁlm, magnetic tape, or computer drives. In the social sciences, techniques such as double blind sampling and statistical analysis of data insulate the observer from the outcome of the experiment. Such procedures provide considerable assurance that the data are not distorted by the subjective state of the experimenter.

14.4 Integrity

Our eﬀorts to understand the world in which we live have integrity to the extent that they involve honesty in gathering and presenting evidence and honest, logical thinking in responding to theoretical problems that develop along the way. Most forms of superstition involve elements of dishonesty in gathering evidence or a failure of logic in responding to theoretical problems. Such failures of logic can be found in the lack of response by the community of practitioners to problems involving the adequacy, coherence, or external consistency of the hypotheses related to their practices.

The most severe lack of integrity arises when the evidence is faked. One of the more striking examples of faked evidence is found in the case of the Israeli entertainer Uri Geller. Beginning in the early 1970s, Geller presented himself in numerous venues throughout the world as a psychic who could perform marvelous feats such as bending spoons, keys, nails, and other metal objects through the sheer power of his mind. Such objects would appear to bend when he merely stroked them with his ﬁnger, or even without his touching them at all. Scientists were called in to witness these feats, and many came away convinced of their authenticity. But in fact Geller was just a clever trickster who duped his audiences. Geller’s trickery was exposed in large measure by the magician James Randi.

After watching videotapes of Geller’s performances, Randi discovered how Geller performed his tricks, and in no time he was able to perform every one of them himself. Sometimes Geller would prepare a spoon or key beforehand by bending it back and forth several times to the point where it was nearly ready to break. Later, by merely stroking it gently, he could cause it to double over. On other occasions Geller, or his accomplices, would use sleight-of-hand maneuvers to substitute bent objects in the place of straight ones. In yet another trick, Geller claimed to be able to deﬂect a compass needle by merely concentrating his attention on it. As he would wave his hands over the compass, the needle would spin—and his hands had been thoroughly examined earlier for hidden magnets. But Geller had concealed a powerful magnet in his mouth, and as he bent over the compass, the needle would spin in tune with his head gyrations.

For another example of faked evidence, let us look at ﬁre-walking. Practitioners of this art claim that their self-help seminars can alter a person’s body chemistry so as to allow him or her to walk barefoot over a bed of glowing coals without being burned. One of the leading gurus of this business is Tony Robbins of the Robbins Research Institute. Robbins uses what he calls “neurolinguistic programming” to cure all sorts of physical and psychological ailments, from irrational fears and impotence to drug addiction and tumors. As proof of the eﬃcacy of this technique, he invites those who have taken his seminar to engage in a ﬁre-walk. By merely believing they will avoid burning their feet, he tells them, they will survive the ordeal unharmed.

The truth is that anyone, whether or not he has taken the seminar and regardless of what he believes, can, under controlled conditions, walk across burning coals and escape unharmed. The physicist Bernard J. Leikind proved this, at least to his own satisfaction, when he showed up at a Robbins seminar in the fall of 1984 (Frazier, 1991, 182–193). Even though he had not attended the sessions and he declined to think cool thoughts as per the instructions of the attendants, he found that he could perform the ﬁre-walk without even getting singed. He explained his success by noting certain basic laws of physics. In spite of their high temperature, wood coals contain a very low quantity of heat, and they conduct heat very poorly. Also, the foot is in contact with the coals for only a second at a time, thus allowing only a small quantity of heat energy to ﬂow to the foot. As a result, the feet of ﬁre-walkers rarely sustain injury (or at least serious injury).

For a third example of faked evidence we need look no further than the thousands of fortune tellers, palm readers, and mentalists who use the art of “cold reading” to divine all sorts of amazing truths about their clients’ lives. Most people who engage the services of these “readers” do so because they have problems concerning love, health, or ﬁnances. The reader knows this and often begins the reading with a ﬂattering spiel that is tailored to ﬁt practically everybody. This recital is intended to put the client at ease and condition him or her to open up to the reader. All the while the reader is taking in every detail: the client’s age, sex, weight, posture, speech patterns, grammar, eye contact, build, hands, clothing (style, age, neatness, and cost), hairstyle, jewelry, and whatever the client might be holding or carrying (books, car keys, etc.). All of these provide clues to the personality, intelligence, line of work, socioeconomic status, religion, education, and political aﬃliation of the client.

The reader uses this information to formulate hypotheses that are then presented to the client in the form of subtle questions. Depending on the client’s reactions—facial expression, eye motion, pupil dilation, gestures—the reader can often tell if he or she is on the right track. Once the reader hits on something close to home, the client will usually react in amazement and begin revealing more personal details. After appropriate intervals, the reader will then rephrase this information in a diﬀerent sequence and feed it back to the client, to the client’s ever-increasing amazement. The client then provides even more details, which the reader weaves together with everything else learned to that point. The use of a crystal ball, satin cape, or tarot cards combined with a polished sense of conﬁdence convey to the client that the reader can literally read the client’s mind.

If the deceptive techniques of the magician who pretends to be a psychic, the neurolinguistic programmer, and the cold reader are accepted at face value, they appear to constitute evidence that really supports the hypotheses underlying these activities. But faking the evidence is not the only way in which the practitioners of superstition lack integrity. The other way concerns the reaction of the community of practitioners to problems that arise in connection with the adequacy, coherence, and external consistency of those hypotheses.

Such problems arise in connection with scientiﬁc hypotheses no less often than they do with superstitious ones. When they arise in science, the community of scientists shifts to what the philosopher Thomas Kuhn calls a puzzle-solving mode, and scientists work on them with great persistence until the problems are solved. This puzzle-solving activity occupies the attention of the vast majority of scientists for the greatest part of their careers, and it constitutes what Kuhn calls “normal science.” Furthermore, it is precisely this puzzle-solving character of normal science, Kuhn argues, that distinguishes science from pseudoscience.

For example, after the Copernican hypothesis was introduced, a problem turned up in connection with stellar parallax. If, as the hypothesis held, the earth travels around the sun, then, in the course of its orbit, the farthest stars should appear to shift in position with respect to the nearer ones. An analogous phenomenon can be observed as you change your position in a room. The distant lamp, which originally appeared to the left of the chair in the foreground, now appears to the right of it. In the case of the stars, however, no parallax could be observed. The explanation given at the time was that the stars were too far away for any parallax to be detectable. Nevertheless, stellar parallax constituted an adequacy problem that the community of astronomers regarded as a puzzle, and they worked on it for 300 years. Eventually more-powerful telescopes were produced that did indeed detect a change in position of the stars as the earth orbited the sun.

In contrast, when an astrological prediction fails to materialize, the community of astrologers never sets to work to ﬁgure out what went wrong. Astrologers never recheck the location and birth time of the client or the exact position of the planets at the time of his birth. They merely charge forward and issue more predictions. Similarly, when the bumps on a person’s head fail to indicate essential features of that person’s personality, or when the lines on his palm fail to reveal features of his life, the community of phrenologists and the community of palm readers never try to account for the failures. They just ignore them and move on to the next batch of clients. Such a response reveals a lack of integrity on the part of these practitioners toward their respective hypotheses. Something is clearly wrong with the hypotheses or with the measurements, but no one cares enough to do anything about it.

A similar response occurs in connection with coherence problems. Most superstitions involve serious incoherencies, many of which arise from the lack of known causal connections. For example, if astrology claims that the planets inﬂuence our lives, then there must be some causal connection between the planets and individual humans. But what could this connection be? Is it gravity? If so, then astrologers need to show how inﬁnitesimally small gravitational ﬂuctuations can aﬀect people’s lives. On the other hand, if some other causal inﬂuence is at work, the astrologers need to pin it down. What kind of laws govern it? Is it an inverse square law, like the law of gravity, or some other kind of law? Analogously, if the lines on a person’s palms indicate something about the person’s life, then what form of causality is at work here? Do the lines inﬂuence the life, or is it the other way around? And what laws does this form of causality obey?

Any absence of a causal connection is a defect in coherence, because it signals the lack of a connection between ideas functioning in a hypothesis. However, such a lack of coherence need not be fatal to a hypothesis. Physicians from the time of Hippocrates knew that willow leaves, which contain the essential ingredient of aspirin, had the power to relieve pain, but they failed to understand the causal connection until recently. What distinguishes the biomedical community from the community of astrologers lies in their respective reactions to such problems. The members of the biomedical community recognized the aspirin problem as a puzzle, and they worked on it until they found the solution, but members of the astrological community are unconcerned with identifying the causal mechanism by which the planets inﬂuence human lives. Similarly, members of the community of palm readers and members of the community of phrenologists care nothing about identifying the essential causal connections implied by their respective hypotheses.

An even more serious problem is posed by hypotheses that are inconsistent with established theories or laws. A case in point may be found in claims made by promoters of the Transcendental Meditation (TM) movement. The practice of TM was popularized in the 1960s by the late Maharishi Mahesh Yogi, and since then it has attracted thousands of adherents. It consists in the silent repetition of a mantra, which induces a mental state similar to self-hypnosis. For many who have tried it, the beneﬁts are mental and physical relaxation leading to a sense of rejuvenation. But with further instruction in TM (at considerable cost to the student), longer and deeper trances can be induced that, the Maharishi claimed, allow the meditator to levitate—to hover in the air without any physical support. Thousands of disciples, he claimed, have learned how to do this, and he released photographs that purported to verify this claim. But of course if levitation actually occurs, it constitutes a violation of, or a suspension of, the law of gravity.

The inconsistency of the Maharishi hypothesis with such a well-conﬁrmed theory as the law of gravity is probably suﬃcient reason to assign it to the category of superstition. But the reaction of the community of TM practitioners to this inconsistency leaves little room for doubt. In 1971 the Maharishi bought the grounds and buildings of what was formerly Parsons College in Fairfield, Iowa, and he converted the site into Maharishi International University. The university then became the home of the International Center for Scientiﬁc Research, which, one would think, would be the perfect forum for investigating levitation. Given the availability of scores of alleged levitators, the “scientists” in residence could conduct in-depth studies into this phenomenon. Their ﬁndings could provide the basis for interplanetary space travel, to say nothing of what they might do for safer airplanes. However, from its inception, the International Center has conducted not a shred of research on levitation. No experiments have been performed and no scholarly papers have been written. This response is inconceivable for any bona ﬁde center of scientiﬁc research.

14.5 Concluding Remarks

Distinguishing science from superstition is no idle preoccupation of armchair philosophers, as some have suggested, but an issue vital to the future of civilization. In Stalinist Russia responsible scientists were shipped off to the gulag for refusing to knuckle under to the state’s ideas as to what was scientiﬁc. And in the United States, court battles have been fought over what counts as science for curriculum reform in the public schools. Also, the attempt to distinguish science from superstition has longstanding roots in the history of philosophy. It can be taken as a modern equivalent of the same question Plato asked long ago; numerous philosophers since then have addressed the question from their own perspectives.

On the foregoing pages we have examined some features that are characteristic of scientiﬁc inquiry and some contrasting features that are characteristic of superstition. The purpose of this exposition has not been to provide the suﬃcient and necessary conditions for a bright demarcation line between science and superstition. Rather, the purpose has been the more modest one of setting forth a group of family resemblances that a fair-minded inquirer may use in rendering a judgment that a set of beliefs is more probably scientiﬁc or more probably superstitious.

To the extent that a set of beliefs rests on hypotheses that are coherent, precisely tailored, narrowly formulated, supported by genuine evidence, and productive of new insights, these beliefs can be considered scientiﬁcally grounded. This judgment is reinforced by the conscientious response of the scientiﬁc community to problems that develop concerning the adequacy, coherence, and external consistency of those hypotheses. But to the extent that a set of beliefs rests on hypotheses that are incoherent, inconsistent with well-established theories, vague, overly broad, motivated by emotional needs, and supported by evidence that fails to be trustworthy, as well as leading to no new insights, then those beliefs tend to be superstitious. Such a judgment is reinforced by a reaction of oblivious unconcern on the part of the community of practitioners to problems that arise in connection with the adequacy, coherence, and external consistency of those hypotheses.

posted on 2025-06-21 21:54 ZhangZhihuiAAA 阅读(50) 评论(0) 收藏举报

刷新页面返回顶部


博客园 © 2004-2025 浙公网安备 33010602011771号浙ICP备2021040463号-3

导航