The Nature of Science
David S. Smith
Why Learn About Science?
It is beyond argument that our most generously educated citizens, our college graduates, are grossly undereducated when it comes to science and technology, even "ignorant" to use Ledermanís characterization. But there is no consensus at all when it comes to what to do about it, no common view of just what it is that our citizens ought and need to know about science, and no common view as to how whatever it is they ought to know can and should be taught.
With regard to what our goals should be, perhaps we can all accept something along the following lines. Our goal is to assure that our citizens know enough about science:
-so that they can tell the difference between sense and nonsense, between science and pseudoscience
-so that they can distinguish the possible from the impossible, the probable from the improbable
-so they can understand both the powers and the limits of science and technology
-so they are not at the mercy of experts-or worse, of charlatans posing as experts
-so they can be participants, not victims, in our increasingly and irreversible technological society.
David S. Saxon, Honorary Chairman of the Corporation, Massachusetts, Institute of Technology, and President, Emeritus, of the University of California, in a paper presented at the American Association for the Advancement of Science (AAAS), Washington, February 17, 1991.
TABLE OF CONTENTS
What is Science
Assumptions of Science
The Process of Science
Hypotheses, Theories, and Laws
Limitations of Science
The Practice of Science
The Nature of Science
"We live in an age of high technology". This may be the most beleaguered cliche found in the current language or perhaps even the history of the language. Every newspaper, magazine, T.V. commentator and politician produces a steady stream of high tech, Bio-tech, computer tech, and miscellaneous tech. The intriguing aspect about this unremitting public focus on technology and its significance is that it is essentially an accurate view. We do indeed live in a world of technology which is increasing in its complexity and sophistication at an exponential rate. All of this technology is ultimately based upon science. Technology is applied science. Consequently we live in a society completely dominated by science and its handmaiden technology. The great paradox is that although we live in a science dominated world the great majority of our citizens have a vague idea of what science is or what it can, or cannot do. This is fully attested to by the fact that astrologers, nutritional faddists, medical quacks, and "scientific" creationists are doing such a land office business in America of the nineties. Indeed, a Medieval mystic would probably feel quite at home with many of the notions that pass as scientific in modern America. Study after study has shown that while we are one of the most technologically advanced of the industrial nations, our population as a whole is the most scientifically ignorant of these same nations.
As a citizen of a scientifically based society it is essential to be able to tell sense from nonsense. This is especially compelling for those of you who will be pursuing careers in the health related areas. As health practitioners, you must be able to assess the scientific validity of new treatments, drugs, and procedures, as well as counsel your patients on these matters.
The purpose of this essay is to provide you with the intellectual tools necessary to assess what is truly scientific and what is not.
WHAT IS SCIENCE?
While it is traditional to begin the discussion of a subject by defining the subject, we find that there is no universally accepted definition of science. Some definitions emphasize accumulating and organizations of facts, other definitions emphasize the ability to make predictions, while still others define science as a process of accumulating knowledge. Our definition is a working or pragmatic definition: science is that branch of knowledge which attempts to explain the structure and operation of the universe in terms of natural principles which are based upon verifiable evidence.
Let us now examine this definition more closely. First we see that science attempts explanation. The simple accumulation of facts is in itself not necessarily scientific. The facts must be accumulated and organized in such a fashion as to yield an explanation or understanding of some phenomenon of nature. Second we see that science deals with the universe with all of its matter and energy including ourselves. To paraphrase the noted astronomer Carl Sagan, we are a part of the universe attempting to understand the entire universe. Third we see that science relies on natural principles of explanations. These principles are natural in that they are basic properties of the universe which can be observe or deduced by human effort. Sciences do not deal with or invoke supernatural forces as explanations. Fourth, and perhaps most significant, we see that science depends upon verifiable evidence.
Scientific explanations require evidence which can be verified by others observers. Unique, one-time occurrences or events are usually beyond scientific methodology.
All principles are base upon tangible evidence which can be observed by means of human senses or their extensions (scientific instruments such as microscopes and telescopes). Furthermore this evidence must be verifiable. The evidence must be observable by more than one person. It should in fact be observable by all people who choose to look and who are capable of observing.
As an example, some 400 years ago the Polish scientist Copernicus inferred that the earth and planets revolved around the sun and not the other way around as previously believed. Copernicus derived this idea (or explanation) from his observations of the motions of the planets and the apparent motion of the sun. We accept his basic explanation today because many other observers have confirmed his observations and agree with his conclusions. You or I do not have to accept Copernicusí ideas or observations. We are perfectly free to repeat his observations and draw our own conclusions. This is because Copernicusí explanation is scientific. It explains phenomena using natural principles, which are based upon observations (evidence) which are verifiable by others. This is the way of all real science.
ASSUMPTIONS OF SCIENCE
Science makes only one fundamental assumption and that is that the universe is governed by natural principles, which can be discovered and understood by humans. This assumption is inherent in our definition of science.
A little reflection should convince you of the necessity of such an assumption. What would be the purpose of attempting to explain why bodies are attracted to one another (gravity) if there is no reason to believe that such a principle of attraction exists. Likewise it makes no sense to search for an attraction principle if the phenomenon is deemed to be supernatural and cannot be comprehended by humans. Science must always assume that every phenomenon of nature is governed by some natural principle which, if known, is capable of being discovered by observation and logic.
THE PROCESS OF SCIENCE
How does science work? More appropriately, how do men and women go about discovering those principles which explain the organization and operation of the universe?
Observation of the universe by means of senses leads to and accumulation of facts. A fact may be defined as a truth in nature about which the great majority of observers at a given time will agree. Now if this definition of a "fact" sounds somewhat soft it is because it is. It is now considered a fact by most persons (Flat Earth Society excluded) that the earth is round. But remember that at one time most people considered it a fact that the earth is flat. We see therefore, that while facts may be immutable, peopleís concepts of them are not. In any event, science begins with these observations that we consider facts. From facts we move to explanations of the facts.
Let us again return to Copernicus. The facts that Copernicus had to deal with were the apparent motions of the sum, moon, and the planets. His explanation these facts was that these apparent motions could best be explained by the movements of the earth and other planets around the sun and the movement of the moon around the earth.
His explanation was not the only possible one, but it was the simplest one that that was consistent with the facts. This process of formulating explanations of facts is known as hypothesis formation. A hypothesis is often defined as an educated guess. It is a possible explanation of a phenomenon but necessarily the only one.
How do scientists go about forming hypothesis? There are really no set of rules. Sometimes a scientist will arrive at a hypothesis by examining all of the facts available and see what they logically suggest. At other times a hypothesis may be strictly intuitive. Hypothesis formation is the creative aspect of science. It is here that new principles are first formulated, the very beginning of explanation and understanding.
Are all hypotheses equally good? The answer here is a resounding "NO!" A good scientific hypothesis must meet two criteria.
Returning to Copernicus, we find that his hypothesis of a heliocentric (sun centered) solar system yielded predictions about the future positions of the planets on given dates. To test this hypothesis it would only be necessary to observe the predicted region of the sky on the appropriate date and see if the planet was there. Copernicus had a solid scientific hypothesis in that it agreed with all of the known facts and yielded predictions that could be tested by observation.
How one goes about testing a hypothesis depends upon the nature of the hypothesis and the predictions which logically flow from it.
There are two broad groups of hypotheses, those which are experimentally testable and those which are not experimentally testable.
Experimentally testable hypotheses lead to predictions which permit manipulations of conditions for testing. Such manipulations are called experiments. In an experiment, conditions are manipulated so that only one condition or variable (the experimental variable) will be altered. Consider the following example.
A drug is being tested as a possible headache cure. If this drug is effective then you would predict that people suffering from headaches who receive it will have their headaches disappear. Assume that 1000 headache sufferers receive this drug and 900 of them report that their headache disappeared within the hour. Would this unequivocally support the hypothesis that the drug cures headaches? The answer is no, because in this test we have no real way of knowing whether it was the drug or some other factor which caused the headaches to disappear. What is needed is an identical group of headache sufferers who do not receive the drug for comparison. This comparison control group should be identical in every respect except for the experimental variable (the drug in this case). A proper experimental design would be to take these 2000 headache sufferers and randomly divide them into two groups of 1000 each. One group (the experimental ) would receive the drug while the second group (the control) would receive a non-active tablet which is termed a placebo. In such tests with humans we must always give a placebo as humans will frequently begin to feel better (or at least differently) if they think they are supposed to. In this experiment both groups will think that they received the real drug. Now if 900 of the experimental group should report relief of their headache while only 250 of the control group reported relief we would probably judge the drug to be effective and the hypothesis confirmed. We have therefore experimentally confirmed the major prediction of the hypothesis (that being that the drug would cure headaches).
It should be noted that in experiments with humans it is usual to administer the drug and placebo in such a manner that the person administering the experiment does not know which is which and therefore cannot influence the experiment by comments or attitudes. This kind of test where neither the subjects nor the administrator know which are experimental and which are controlled is know as a double blind test and is the type utilized for evaluating drugs, vaccines, and other procedures designed for use by humans. This is then the way which we can test our first category of hypotheses.
In double blind experiments the person performing the experiment will not know which group is the experimental one and which is the control.
Hypotheses that deal with subjects and phenomena that cannot be manipulated are not testable by means of experiments. Testing and verification of these hypotheses must be based upon predictions which can be verified or refuted by observation in the natural world. The kind of reasoning involved would proceed like this. "If this hypothesis is correct, then we would expect to find, observe, detect, etc." Copernicus, hypothesis would be tested in this way. Obviously we cannot experimentally manipulate the orbits of planets but we can make predictions about the future positions of the planets in those orbits.
Another group of similar hypotheses are those which are historical in nature. These hypotheses explain events which have occurred in the past. Geologists frequently utilize these types of hypotheses in explaining the history of the earth. For example when one descends into the Grand Canyon a succession of layers of rock are passed. Geologists explain this layering effect by developing the hypothesis that these layers represent sediment which was deposited over immense periods of time and which formed rock layers that became exposed by the erosion of the river in cutting the canyon. If this is a true hypothesis then we would predict that we would find similar layering where other rock layers that have been exposed (we do). We would also anticipate that deep earth core samples through sedimentary rock would show similar layering, (they do). Finally it would be predicted that these same layering processes would be occurring today in appropriate areas such as river deltas and ocean floors (they are).
It may therefore be seen that experimentation is not the only valid way to test hypotheses although experimentation is usually a much simpler and quicker way when possible.
When is a hypothesis considered to be proven true? It has been realized by scientists in this century that it is really impossible to ever completely prove any scientific hypothesis unequivocally. This is due to the nature of the scientific process itself. Just because the predictions of a hypothesis have always proven true in the past does not mean that they will continue to be true in the future. It is always possible that at some future date a prediction may prove false. Until every possible predictive case has been tested we will not know if the hypothesis is universally true. As it is rarely possible to test every possible prediction of a hypothesis then it is impossible to prove it 100%.
Most scientists seek probabilities of truth. Based upon the number of predictions which have been verified a hypothesis may be considered as having a low or high probability of being true. The probability that the sum will rise in the east tomorrow morning is extremely high based upon the verified evidence accumulated. It is remotely possible that it could rise in the west or perhaps not rise at all.
Being scientists can never hope to completely prove a hypothesis true they rely instead upon attempting to falsify it. While numerous verified predictions will not prove a hypothesis to be true, several false predictions will quickly cast doubt upon its accuracy. A scientific hypothesis must be capable of being falsified, that is, yielding predictions which are capable of being shown to be false. No such false predictions may ever be found if the hypothesis is true. This concept is so important to the operations of science that it bears repeating. In science it is never possible to prove that any idea is absolutely correct, but it is possible to prove that an idea is not correct. This is the absolute core of science. Scientific explanations must by the nature of the process always be tentative and subject to revision as new facts are discovered. Consider the following example. A woman has studied crows for several years. She has observed that all crows are always black, and based upon her observations she forms the hypothesis is the prediction that any future crow will also be black. For the next ten years this scientist and other scientists examine thousands of crows and invariably find that they are always black. Then one day an albino (white) crow is found. The hypothesis has been falsified, or has it? Certainly an important generalization should not be discarded simply because one or a few exceptions are found. The great majority of people could spend their entire life observing crows without seeing a white one. It would be better to modify the hypothesis. Perhaps we would say that the overwhelming majority of crows are black or more precisely all crows are black with the exception of rare albino members which occur at a frequency of 1 per 100,000 birds. If on the other hand we found that about 50% of all crows were white we would have to reject the hypothesis as being false. This example demonstrates another important property of science: not only are scientific explanations tentative, they are also self-correcting. As new data are discovered hypotheses are modified. Consequently as science progresses we get an ever more accurate view of the universe but any and all scientific explanation must be considered tentative and subject to revision.
Certain scientific hypotheses are considered more probable than others because they have been exhaustively tested by a large number of scientists and have been found to be true in every case. Copernicusí hypothesis is one such example and today we consider such highly substantiated hypotheses as his to be true beyond a reasonable doubt.
HYPOTHESIS, THEORIES, FACTS, AND LAWS
These four terms are almost universally misunderstood by the lay public. How many times have you heard someone say "thatís only a theory." The average person equates theory with hypothesis, and idea without much evidence, and only one of several equally probable explanations. Most people perceive a progression of accuracy ranging from theory to law and then to fact. This view is quite erroneous. To a scientist, a fact is an observation about nature while a theory is a series of statements, which explains facts. To be considered a scientific theory an explanation must have a large body of facts supporting it, but theories never become facts, theories always explain facts. A hypothesis which has been tested a number of times by a number of different people and has always proved true may be accorded the status of a theory. The demarcation between hypothesis and theory is not at all clear cut.
Laws are often presented in elementary texts as theories which are universally accepted. This is somewhat erroneous as most laws relate to specific observations which occur each and every time a specific event occurs. Laws are generally much more restricted than are theories and often are derived from general theories. In one of Newtonís laws we find that for every action there is an equal and opposite reaction. A rocket engine thrust backward and the rocked moves forward. Note that the law governs a very restricted and predictable event. It is not at all accurate to conclude that laws are more accurate or correct view of the force than does Newtonís law of gravity.
The differences between hypothesis, theories, laws, and facts are frequently quite arbitrary and may be used differently by different scientists. Regardless of what we term a scientific explanation, it must be testable by observation and measurement.
No experiment you perform, however brilliantly conceived and executed, will satisfy more than 5% of the people concerned.
Highly probable scientific explanations will usually be accepted by the majority of reputable scientists.
LIMITATIONS OF SCIENCE
The current standards of living which we enjoy and the present understanding of the universe which we have developed are testimony to the immense power of science. In a span of a few thousand years we have moved out of caves and gone to the moon. It was only about 400 years ago that William Harvey discovered the circulation of blood but today we transplant genes between species. Such a staggering progression may suggest that science have no limits. This is not correct.
The great power of science lies in explanation by means of natural principles which can be verified, but these natural and verifiable aspects also set the limits of science because they permit science to only deal with what can be observed and measured. Science cannot deal with areas which are not subject to objective observation. This eliminates all areas which are based upon value systems developed by human beings. Science cannot deal with the religious or supernatural because by definition these do not provide observable predictions which can be verified or refuted. Neither can science deal with philosophy, ethics, or esthetics because these are human endeavors and not governed by natural laws. Such endeavors are belief systems which are not based upon observation and experiment. This does not mean that such endeavors are somehow inferior. On the contrary, ethics, philosophy, esthetics, and religion are some of the areas which provide life with its greatest riches. These human values are what truly set us apart from the remainder of the living world and provide that uniqueness we term human. They are not subject to scientific measurement and therefore cannot be evaluated scientifically.
Unfortunately many people do not appreciate this fact and we often see various individuals or groups of individuals attempting to justify an ethical, religious, or even political position scientifically. A good example is the current debate over abortion. It is not at all uncommon to see one side or the other of this issue stating that "scientific evidence clearly shows a fetus to be a person (or not a person) at conception." The truth is that no such evidence exists or can exists or can exist because "person" in the sense of an individual protected by the constitution is a legal-ethical definition. Science can describe what happens during embryonic development but when an embryo becomes a "person" depends upon oneís value system and is not scientifically testable. Abortion is therefore an ethical issue and not a scientific one.
The limitations of science are very obvious in the area of esthetics. It is impossible to demonstrate scientifically that one painting is better art than another. What constitutes good art is a matter of appreciation and is culturally derived. Paintings that were critically acclaimed in the Victorian period are now considered to be trite and maudlin while those considered to be monstrosities by the Victorians are now hailed as powerful works or art.
Another example of scientific limits comes from the area of religion. Most religious beliefs are based upon the existence of one or more supernatural beings. Being they are supernatural they immediately fall outside of the realm of science which can only deal with the natural and tangible. It is impossible to scientifically prove or disprove the existence of God. He/She is spiritual and not observable or measurable. Whether God exists or not is largely a matter of faith (of lack of) on the part of each individual.
The three previous examples clearly show that science cannot make decisions in ethics, esthetics, or religion. These are all culturally derived values and can only be viewed from the value system of the culture in which they flourish.
The Practice of Science
The previous discussion has focused on the theoretical nature of science. The question now is how science actually done in a practical sense? How do new scientific discoveries become known? A fairly standard procedure has evolved over the years for the practice of science.
It begins with the individual scientist who has an idea (hypothesis). The scientist sets about to research and test this idea. This may involve experiments and/or observations of the natural world. Once the scientist is satisfied that she or he has discovered new knowledge, the research will be presented to other scientists for their evaluation. This almost always involved the publication of the research scientific journal.
There are tens of thousands of scientific journals world wide. Some are general, publishing research from a number of different scientific fields. The American journal Science and the British journal Nature are examples of general journals. Others tend to be highly specialized, such as the Journal of Immunology which only publishes research dealing with the science of body defense, immunology. All referred journals have a panel of subject experts known as referees. When a scientist submits his/ her research to the journal, the paper containing the research will be given to several of the referees who will evaluate it terms of its scientific merit. Very often it will be sent back to the scientist for revision. Once the referees are satisfied that the research has been conducted in a competent manner, it will be published in the journal. Then all interested scientists may read the paper, evaluate the research, and repeat it if deemed necessary. If after vigorous evaluation and testing by a number of scientists, the reported research is sustained, then the concepts will become part of the common knowledge of science.
The vigorous and rigorous evaluation that research published in refereed journals receives from scientists insures that inferior quality research will be exposed hypothesis with little factual support will be discarded.
Pseudoscience means false science. It is something that claims to be scientific but when carefully examined is found to be lacking in some or all of the commonly accepted attributes of science. Today as never before in history we are bombarded by all types of charlatans and hucksters peddling some types of nonsense which they term scientific. One of the reasons we see so much pseodoscience sit ha science in our society enjoys great prestige. Therefore to say something is "scientific" or "scientifically proven" is to give it an air of legitimacy and accuracy. The modern day "snake oil" peddler frequently cloaks himself in a mantle of self proclaimed science.
At this point you should know enough about real science to be able to distinguish it from pseudoscience, but sometimes it is difficult to do, especially if you are not familiar with the products and of concepts begin promoted. Fortunately pseudoscience frequently (but not always) possess readily identifiable characteristics. You should strongly suspect pseudoscience if the proponents of a position, product, concept, or claim exhibit one or more of the following characteristics:
If an experiment requires "n" ingredient, there will always be "n-1" ingredients in stock.
Pseudoscience often appears to have the appearance of real science, but upon closer inspection it is usually found to be in violation of acceptable scientific methodology.
The six characteristics presented are not the only ones associated with pseudoscience but they seem to occur most often. It is sometimes very difficult to recognize pseudoscience unless one is an expert in the area. For example sometimes pseudoscience advocates will cite experiments that prove their point. Upon further inspection one may find that the experiments lacked controls, used adequate or biased samples, or have never been repeated by any reputable scientist. However at the time of the claim it may be difficult to ascertain that kind of information. Under such conditions you should always examine the credentials of the individuals making claims. Are they truly experts in the area which they advocate? Do they engage in real research in the appropriate area and is this research submitted to reputable, refereed, scientific journals for the review by other experts? The possession of a Nobel prize in physics does not give an individual any expertise in IQ testing. Remember, the ability to sound scientific is no the same as being scientific.
We will now examine four areas of pseudoscience which are currently in vogue. These are quack cures, nutritional fads, astrology, and "scientific" creationism.
Until recently advocates of nutritional supplements could make virtually any unsubstantiated claim they wanted to because these supplements were considered foods and not drugs. Recently the Food and Drug Administration (FDA) has attempted to regulate the unsubstantiated claims made by the sellers and advocates of these supplements, congress delayed the implementation of the new regulations. Whether they will ever go into effect is problematic at this time.
Finally we have the "natural phenomena". The local health food stores may sell "natural vitamin C" at a much higher price than the "inferior" synthetic variety. Actually Vitamin C (ascorbic acid) is the same chemical substance regardless of its origin and has the same effect in the body. No nutrientís valve is enhanced simply because it was extracted from a pile of dirt!!
All reputable nutritionists agree that normal healthy people need eat only a balanced diet obtainable at any supermarket to maintain their health, vigor, and vitality. Members of the health related community have an obligation to be able to recognize genuine scientific progress in terms of nutrition from pseudoscientific quackery, and to inform their patients accordingly.
Creationism was until fairly recently not classified as pseudoscience because it was completely religious and made no scientific claims. In the early 1960ís a group of creationists began calling themselves "scientific" creationists and proposed that the creation events as spelled out in Genesis could be sustained scientifically with out reference to the Bible. Unfortunately for their beliefs, the available evidence complete refutes their thesis. Never-the-less their views in one guise or another, largely in the hope of introducing their religion views into the public school curriculum.
Creationism as a religious belief canít be promoted in the schools because of the constitutional prohibitions involving church and state. However, if it could be established as a scientific thesis, then it could be taught as an alternative to evolutionary theory, and this is what "scientific" creationism is all about.
"Scientific" creationists propose the their "model" is as scientifically valid as the voluntary concept and therefore should be given equal time of fairness and in the spirit of scientific open mind. This is an appealing view to the public which believes in fair play and the right of all ideas to be given equal time. Scientific theories rely on observable evidence and therefore do not require consideration of ideas which have been refuted by that evidence, no matter how much we may wish them to be true. Creationists have never presented any scientific acceptable evidence n support of their position. The bulk of their "research" has been confined to attacking legitimate science and pointing out the flaws that they perceive in evolunatry theory.
In their view evolutary biology and all other science (chemistry, physics, geology, astronomy) which support it is flawed an therefore their preconceived position must be true. The great scientific flaw in "scientific" creationism is not the lack of evidence to support it or even the great weight of evidence against it. Rather it is the impossibility of scientifically testing it. The fundamental assertion (belief) of creationism is that all living things, (plus the earth and universe) were created by a supernatural creator using UNKNOWN PROCESSES that cannot be verified or refuted by scientific methods. Therefore according to the scientific creationists themselves, origins can never really be known or understood. As this completely takes creationism beyond the realm of science (it contradicts the fundamental assumptions of science) the term "scientific" creationism is an oxymoron , a self contradiction.
The previous examples demonstrate that pseudoscience generally attars a thesis which (1) has not been scientifically tested (2) has been tested and found false (3) is not capable of being scientifically tested. It is critically important that you as a member of a society which is so highly dependent upon science and societies, be able to recognize the difference between real science and pseudoscience.