by Robert Day
Science can be defined as a technique used by humans to understand natural phenomena, solve physical problems and accumulate knowledge. The technique involves the application of what has become known as "the scientific method". The following description of the scientific method is adapted from a class handout I wrote with a little help from some of my students. I consider it a work in progress, since I modify it in response to student feedback each time I use it. It emphasizes aspects of the scientific method that I have often found to be either unknown to, or misconceived by non-science majors.
1) Make direct observations of physical phenomena.
Here, "observation" means any relevant data gathered using any sense(s), technique or equipment. Repeatable observations that can be mechanically or physically recorded are preferable. If no useful observations of any aspect of the phenomenon can be perceived, then the phenomenon cannot be subjected to any further scientific study and should not be considered science.
2) Use inductive logic to formulate a general hypothesis to explain the specific observations. A hypothesis is a "best guess" at explaining a set of observations. It is not supported by any evidence so it cannot yet be assumed to be true.
Forming a hypothesis may require additional assumptions based on other ("accessory") theories, or on the need for the hypothesis to remain compatible with "the scientific method". The hypothesis should be worded in such a way as to make predictions that can be tested under specific experimental conditions. This part of science is highly creative and requires speculation and flashes of insight. If no hypothesis can be formed then the validity of the assumptions may need to be re-examined. If it seems impossible to form a falsifiable hypothesis then the phenomenon cannot be subjected to any further scientific study unless more useful observations can be made; return to step one. For professional scientists, a truly vast knowledge of background information and experimentally supported accessory theories may be necessary before a plausible hypothesis can be devised.
3) Design of an experiment to test the general hypothesis, perform the experiment and gather experimental data.
The experiment should be designed to test the affect of just one variable on the phenomenon, and in such a way as to attempt to falsify the hypothesis, since logic dictates that no amount of supporting data can ever prove the hypothesis to be universally true. It is much easier to demonstrate a hypothesis to be specifically false under a given set of conditions. Experiments that yield objective, quantifiable physical data are preferable. Including appropriate controls to ensure that only one variable is tested is essential. If it is impossible to collect or imagine any way to collect data that could falsify the hypothesis, then the phenomenon cannot be subjected to any further scientific study and should not be considered science; return to step one. The philosopher Karl Popper stressed the importance of falsifiability in science and coined the term "pseudoscientific" to describe any unfalsifiable hypothesis. Evidence that comes from direct, interactive, repeatable experimentation and yields physical results is sometimes described as "empirical evidence". This is the only type of evidence acceptable to a scientist. Anecdotes ("hearsay") analogy, and references to supernatural forces or premises are not scientific evidence.
4) Apply deductive logic to the specific physical data from this experiment to see if the hypothesis is supported, needs modification, or is falsified.
5) If the hypothesis is falsified by the experimental evidence, or if it needs modification, return to step one or two.
6) If the hypothesis is not falsified and is supported by experimental evidence it becomes a theory. A theory is an experimentally supported explanation for a phenomenon.
All theories should be re-tested by other scientists using the original experiment and by their own independent experiments; return to step 3. Falsifying a new, weakly supported theory usually only requires a single piece of experimental evidence. The more experimental evidence that supports a theory (i.e., the more it resists falsification) the more widely the theory is assumed to be true. Nonetheless, it may still only take a single experiment to falsify an otherwise well supported theory. This logical asymmetry is useful because it makes it unlikely that an erroneous theory can persist. Most non-scientists do not realize that a well established theory is highly reliable precisely because a single contradictory result can destroy it. If experiments yield inconclusive or contradictory data, it may become necessary to modify the original hypothesis or examine the validity of assumptions and accessory theories. Go to step one, two or three.
For a professional scientist involved in formal research, all experimental data and conclusions must be presented to other qualified scientists so that they can be critically examined and evaluated. This process is called "peer-review" and no formal scientific study is acceptable without this because:
a) The experiment can only be usefully repeated if the original data is available for comparison.
b) Only experts in the field are knowledgeable enough to recognize and examine the validity of assumptions, accessory theories, appropriate controls and logical conclusions.
c) Publication of the work makes it available for others to use to support their own investigations.
d) Peer-review maintains a high level of accuracy and integrity in science so that the validity of published work can be assumed by its readers.
7) If the theory resists additional attempts at falsification and becomes supported by a growing body of experimental evidence then it is said to be "robust" or "well supported" but can never be considered "proven". A robust theory may be assumed to be true if this will help to produce more hypotheses or theories.
8) If a body of evidence from a variety of independent sources accumulates, and if the theory is productive in terms of new predictions and additional experimentally supported theories linking the whole field together, then these theories, taken together, are sometimes said to be "unified", or are referred to as a "prevailing paradigm". A robust theory supported by a body of evidence is the strongest assertion of truth that science can make. There are no certain facts in science.
A robust prevailing paradigm is considered to be as close to "truth" as science is likely to get and usually requires the appearance of significant new observations or experimental evidence before its validity can be rejected. Occasionally though, a well established theory may be challenged or falsified by a single, reproducible, elegant experiment yielding a logical an unavoidable conclusion. (The inability of an individual or group to reject a theory in the face of powerful falsifying evidence is sometimes evidence of "dogmatism". A "dogma" is an idea that is so engrained in society that it is accepted as true, even in the face of contradictory or falsifying evidence.)
9) Contrary to what some non-scientists believe, a scientific law is not a theory that has been proven true, since that is impossible. A law is merely a descriptive term that can be applied to all known instances of an observation. For example F = MA. However, a law is different from a theory in that it does not attempt to explain anything. Laws merely describe known physical observations. Theories attempt to explain observations and predict entirely new ones.
Notes on the origin, usefulness and identification of scientific explanations to explain phenomena:
Intelligence is a far more improbable proposition than life itself. We can infer this from the fact that life on Earth apparently evolved spontaneously a relatively short time after the formation of the Earth, yet the appearance of technological intelligence took almost half the projected life-span of our solar system. Early humans knew nothing of the scientific method and used problem solving strategies that drew heavily on association by trial and error, tradition (perpetuated by mimicry, language or instinct), authoritarian decree, emotional responses, myth, anthropomorphism, animism and analogy. Many of these strategies are still used by humans today with varying degrees of success. As human knowledge accumulated, it became clear to some that these strategies by themselves were unsatisfactory and that those cultures which relied exclusively on them suffered a competitive disadvantage. Slowly, forced by competition and necessity, logic and the scientific method began to evolve. Science did not spring from the human mind in its present form because of some innate, obvious truth. It is not easily practiced by most humans since it involves considerable discipline of thought and the ability to recognize and separate scientific evidences and assumptions from unscientific ones. In its current form it has been found useful for explaining a wide variety of natural phenomena. A comparison of early and more recent scientific writings shows us that humans are continuing to refine and improve the scientific method and that we are gradually learning to enforce it's central principles more stringently. This may be one reason why scientific knowledge accumulates ever-more rapidly yet requires major revision less frequently.
The scientific method's greatest strengths are its use of logic and its drawing on the powerful principles of precedent, and uniformitarianism (i.e. The assumption that natural laws stay pretty much constant through time and space and that experimentally supported theories can be drawn upon as assumptions.) Science's other great strength and most novel characteristic compared to other attempts to understand reality is the idea that no scientifically constructed idea can ever be considered finally proven, hence all scientific ideas are subject to perpetual extension, modification or rejection at any time if falsifying data is found. This requires an admission of earlier error, which has proven to be one of the hardest skills for the practitioners of science to master.
Science leaves many questions unanswered, or even totally unasked if they cannot conform to the requirements of the scientific method. This is contrary to human nature since human intellect has evolved to be uncomfortable leaving any question unanswered. As an aid to survival, humans tend to use small fragments of dubious evidence to make snap decisions about the nature of reality and stick to them doggedly, even in the face of contradictory evidence. In the struggle for survival, a bad decision is often better than no decision at all. This might explain why so many humans respond to scientific ideas and premises so cynically. Most instinctively can't accept that some ideas can be "relatively true" yet unprovable, while other "important" questions ("is there life after death" ?) must remain totally unasked by science.
Notice that in the above description of the scientific method there are many scenarios under which a field of study can change from scientific to one that can no longer be "studied scientifically" i.e. can no longer be considered science. One example is any hypothesis presented as immutable fact by its advocates. Experiments are pointless unless falsification or modification is accepted as a possibility, so this type of hypothesis cannot be subjected to the scientific method and is not science.
Since it is difficult or impossible to design any experiment that can falsify the existence of supernatural or unobservable phenomenon, it follows that any hypothesis that includes these is also, in practice, not subject to examination by the scientific method and should also not be considered science. The inability of non-scientists to fully understand science may be one reason why science and the supernatural often seem to be associated in misconceptions about science. To the non-scientist, science, magic and occult or futuristic pseudoscience seem to all fall into the general category of "weird things I don't understand", and appear therefore to be somehow related.
The key characteristic of pseudoscientific concepts (concepts with questionable scientific validity) is the inability imagine any experiment that will falsify the concept, the absence of supporting empirical evidence and, to a lesser extent the absence of a peer -reviewed description of the concept in a scientific forum, such as a respected scientific journal. Without these, scientists will not recognize a concept as a true scientific theory.
The following are examples of "theories" that a scientist would discount as pseudoscientific (i.e., useless) if not accompanied by a body of empirical evidence:
Explanations that include scientific jargon or technical terminology, even if used appropriately. Such explanations are sometimes called "scientistic" and should be rejected no matter how plausible they sound if there is no supporting empirical evidence.
Explanations that draw heavily on analogy. Modern science uses analogy only as a tool for explanation, never as an argument to support a hypothesis. Analogies have no logical supportive value because events which appear superficially similar are not necessarily related at all. Reasoning by analogy is one of the most common problem- solving strategies used by humans, yet it often produces explanations for phenomena that are wrong or arguments that are fallacious.
Explanations supported only by untested stories or accounts, "hunches", feelings or speculations. Descriptions of events with vague or unknown origins should be ignored and even one's own eye-witness accounts or recollections should not necessarily be considered evidence of anything because the human eye and brain are not accurate, objective recording instruments. Such evidence is referred to as "anecdotal". Unfortunately, anecdotal evidence is widely accepted as useful by the general population.
Any explanation that relies exclusively on qualitative descriptions as supportive evidence. Although this kind of study can be useful they are often rejected by scientists because they tend to be subjective, difficult to reliably reproduce and less likely to yield specific, testable predictions.
Finally, for serious professional science, any hypothesis that could be, but has not yet been subjected to the scientific method or any theory that is supported by a body evidence but has not yet been subjected to official peer review is not usually considered reliable.
Point and Counterpoint; Common Competing Ideas About Science Held By Scientists and Non-Scientists.
¥ Scientists tend to think of science as.....
* Non-scientists tend to think of science as.....
¥ A technique useful for solving problems, either as part of a formal research program or in any aspect of one's life.
* An obscure hobby involving test tubes and complex electrical equipment with no practical use.
¥ Generally interesting but scientists are realistic about tedious elements and limited rewards.
* Generally boring. Some have naive ideas about science being entirely glamorous and filled only with adventure and daily exciting discoveries.
¥ Mainly quantitative.
* Mainly qualitative, the quantitative parts cannot be understood by normal people.
¥ Progresses in small steps using highly focused, discrete questions within definite sub-disciplines. Rarely considers truly grandiose issues.
* Science is a continuous blend of facts that cannot be separated into discrete questions. Usually considers broad, grandiose issues.
¥ A part of daily life. Used routinely by individuals to solve trivial problems.
* No connection to daily life accept possibly the invention of new products. Not used routinely by individuals.
¥ Easy to understand, anyone can learn it.
* Utterly incomprehensible to normal people.
¥ Conducted by normal people with ordinary lives and altruistic intentions.
* Conducted by old Austrian men with frizzy gray hair and thick accents who often lose their spectacles and occasionally declare that "this could mean the end of civilization as we know it". Female scientists are aloof and sexless. Most scientists are eccentric or insane and usually have wicked or misled ulterior motives.
¥ Influenced by politics, career goals, economics and practical consideration of which questions are likely to yield useful data and continued funding.
* Isolated from politics, not driven by career goals, economic or practical concerns.
¥ Characterized by a distinct career path linked extensively to academic institutions. Practitioners have hierarchic organization and rank like any other institution.
* Doesn't have a distinct career path or organizational hierarchy. Most non-scientists are not familiar with the mechanics of academia and underestimate the role of academic institutions in science.
¥ Performed only by qualified specialists in accredited institutions.
* Performed by social misfits in their castle or basement. Since they are unfamiliar with scientific research careers, many non-scientists overestimate the role of commercial, governmental or amateur research and do not weigh the importance of formal academic training and credentials heavily enough.
¥ Offers the hope of eventually solving most of humanity's problems.
* Responsible for causing most of humanity's problems.
¥ Performed by sleek, sexy, individuals who are attractive because of their rational nature and great intellect.
* Performed by geeks who are repellent because of their social ineptitude, poor fashion sense, cold-fish demeanor and know-it-all attitude.
¥ Performed by individuals who are familiar with the modern scientific method and its limitations.
* Not performed using any fixed "method". Non-scientists typically have vague ideas that science involves "experiments" or "research" without really understanding these terms. Most non-scientists think that the object of experiments is to definitively "prove" things to be true and have little or no understanding of empiricism.
¥ A body of theories supported by experimental evidence, subject to constant revision, modification and extension.
* A collection of dry, immutable facts that are "discovered" and "proven to be true".
¥ Made up of hypotheses, which are mere speculation, theories, which are supported by experimental evidence and paradigms or laws, which are strongly supported by a reliable body of interconnected experimental evidence.
* Made up of hypotheses and theories, which are mere speculation, and "scientific facts", which have been "scientifically proven" true.
¥ Cannot be applied to the study of the supernatural. Not related to the supernatural.
* Can be applied to the useful study of the supernatural. Similar to magic in that both are incomprehensible, outside normal experience, performed by a caste of individuals with special knowledge, and used to achieve dubious or mischievous goals. (Hence, Frankenstein's monster and Dracula appear in movies together.)
¥ Not in direct conflict with religion, although some see the two as incompatible. Others prefer not to think about it too much.
* In direct conflict with religion, most prefer not to think about it too much.
¥ Often social. Usually involves collaboration, free exchange of findings and peer review of all work.
* Rarely social. Scientists usually work alone and are protective of their findings.
¥ Cynical and critical of all unfamiliar statements and research. Examine statements and evidence closely for flaws, regardless of source.
* Tend to accept unfamiliar statements and research findings without critical appraisal of evidence, especially if information is thought to originate from a scientific or authoritative source.
¥ Does not involve looking in awe at unusual phenomena and talking about how neat they are.
* Usually involves looking in awe at unusual phenomena and talking about how neat they are.
