We sometimes think of science as a body of knowledge (especially when we’re immersed in a science course that involves memorizing page after page of fact). For example we might think of the science of astronomy as everything we know about planets, stars and galaxies. But this is not quite right. Science is really the process by which we accumulated that knowledge. At its heart science is a way of learning about the world around us. (Not the only way, of course, but arguably the most effective.) The scientific method is a simple but tremendously powerful way of building an understanding of the laws which govern our universe. At its most basic, it can be described in four steps, depicted graphically in the box to the right.
1) Make observations.
2) Come up with an idea, or hypothesis, which explains the observations.
3) Develop and perform a study which tests the hypothesis.
4) Examine the results of the study, and use them to develop new hypotheses.
This is sometimes referred to as a hypotheticodeductive approach because it focuses on developing hypotheses, and then using deductive logic to see how they can be tested.
What makes this approach so powerful?
Even Albert Einstein, widely accepted as the greatest physicist of the 20th Century, was wrong sometimes.
Ideas are put to the test:
One key is that every idea is put to the test. Even brilliant scientists get things wrong occasionally. The Greek philosopher Aristotle, considered one of the greatest thinkers of all time, concluded that life could arise spontaneously from non-living material. This idea was almost universally accepted for centuries, but Louis Pasteur and other scientists helped show that it was false by testing it with experiments. Albert Einstein, one of the most respected physicists of the the 20th century, for decades argued against parts of the theory of quantum mechanics because of the essential randomness and uncertainty it showed to underlie all forms of matter at the subatomic level (this is the source of his famous quote that “God does not throw dice”). However, decades of ever more painstaking and detailed experiments have consistently supported quantum mechanics. Individual researchers often go off track, even well accepted theories can turn out to be blind alleys, but as long as scientists continue to ask questions, to look for discrepancies between theories and the data we have accumulated from experiments, and to perform new experiments, eventually we will discover our errors. This continuous process of improvement and self-correction is one of the great strengths of the scientific approach.
Constant revisions:
Scientists are perfectionists. If 100 experiments show support for our hypothesis, but a single experiment produces results inconsistent with the hypothesis, we have to reject the hypothesis and start searching for a new one. (Assuming the single experiment was well designed and executed. We might want to repeat it a few times before moving on.) However, we wouldn’t necessarily just throw the old hypothesis out and start from scratch. It may be that the original hypothesis can be revised in the light of the new findings. This may in turn lead us to ideas for new experiments. Sometimes we discover that our old hypothesis is just a special case, which only holds in some circumstances. We can then fit it into the framework of a broader theory. A case in point is Isaac Newton’s laws of motion. Hundreds of experiments over several centuries found Newtonian physics to be right on the mark. But then Einstein came along and suggested that in some conditions, such as extremely high velocity or extremely large masses, Newton’s theory would not work. Subsequent experiments produced results which agreed with Einstein rather than Newton, and our understanding of these processes had to be revised. That revision didn’t mean throwing out everything Newton had done, of course. Einstein incorporated some of Newton’s ideas into his own theory of relativity, and we can now see that Newton’s laws as a special case under the broader umbrella of relativity. Our goal is to continuously revise and improve our ideas as new observations and experimental results become available, but we can never claim to have “proven” any hypothesis beyond all doubt because we can always find conditions or circumstances in which it has not been tested. Scientists usually speak in terms of having greater or lesser amounts of confidence in an idea, based on accumulated evidence (statistics can help tell us how much confidence to have in a hypothesis, but more on this later).
Builds on past knowledge:
Another key to the success of science is that it constantly builds on previous knowledge. Above we noted that observations of the natural world are the starting point for scientific investigation. But today’s scientists are not limited to their own observations. They have access to the accumulated observations and ideas of generations of scientists who came before them. Every well-designed experiment — whatever the result — provides new insights. We observe the results, relate them to the results of other experiments, and use this to develop new hypotheses and experiments. Scientific progress is a little like a jigsaw puzzle. Each new piece we add in shows us more of the picture, but also provides some direction for adding new pieces, and extending our view even further. (Imagine how hard it would be to do a jigsaw if you just set each piece in place without considering how it fit with any others!)
Encourages creativity:
The scientific method does not put shackles on ingenuity and creativity, it simply provides a framework for applying these traits effectively. Although all scientists use the scientific method, every individual will approach problems differently, depending on field of study, aptitude, and personal interests. Some, such as chimpanzee researcher Jane Goodall, have focused primarily on observations, building up a huge body of knowledge which she and others have used to develop hypotheses about primate behaviour. Others, such as Einstein, have devoted their time mainly to developing theory, leaving it to others to design and perform the experiments needed to test the theories. Every researcher brings a unique perspective to bear on their research, and the exchange of ideas (often vigourous, occasionally heated) between scientists with very different points of view, is one of the most fruitful elements of science. That’s one reason courses in the sciences, both undergraduate and graduate, put so much emphasis on written and oral communication. The ideal scientist would combine a childlike openness to new ideas, with relentless skepticism toward untested ideas, and a ruthless rejection of ideas which fail when tested. Of course none of us meets this ideal. Training in science doesn’t make us immune to half-baked ideas, ingrained prejudices, or any other human frailty. It does, however, provide a means for weeding out the half-baked ideas and those rooted in personal bias, so that in the long haul we can accumulate a body of ideas that work (at least so far).
Science and pseudo-science:
The power of scientific enquiry, and the weight which scientific support lends to theories, have provided powerful incentive for advocates to present their ideas as scientific. Unfortunately, some individuals and organizations who present their views as science do not follow the scientific method. Although this pseudo-science is easy to spot by specialists in a given field of inquiry, it can be deceptive to those with less background in the subject. Many different varieties of pseudo-science exist (there are, of course, many ways to do science incorrectly), and so no clear and simple set of rules can be set out to uncover all forms of pseudo-science. However, there are a few simple rules of thumb that can be applied that can usually help separate real science from pseudo-science.
First, pseudo-science is rarely based on original experiments. Typically it is based on observations (often anecdotal observations), or re-interpretation of the experiments and observations of others.
Second, pseudo-science distorts the hypotheticodeductive process outlined above. In real science, hypotheses are developed on the basis of prior observations, but then tested by collecting new observations. Practitioners of pseudo-science often develop a hypothesis on the basis of a set of observations, but then use the same observations as support for the hypothesis. This creates a circular argument, since by definition the hypothesis must fit the observations it is based on. Thus pseudo-scientists present legitimate, but untested hypotheses, as well-supported theories, or even proven facts. In some cases pseudo-scientists turn the entire scientific process on its head by beginning with a conclusion they believe to be true, and then simply collecting whatever evidence they can find which supports it.
The latter approach inevitably leads to a third “red flag” of pseudo-science. Evidence which supports the desired conclusion is included, while any evidence which contradicts that conclusion is ignored. This is often called observational selection or confirmation bias, but the philosopher Francis Bacon had a simpler description: “Counting the hits and ignoring the misses.” Obviously this completely undermines the process of scientific research, in which a single well designed and executed experiment is enough to falsify any hypothesis.
A famous example of pseudo-science was a theory on the origin of the planet Venus, put forward by Immanuel Velikovsky in his 1950 work: Worlds in Collision. Velikovsky argued that Venus originated as a comet-like mass ejected from the Planet Jupiter several millennia ago. Note that at its basis, this is a legitimate scientific hypothesis. However, proponents of Velikovsky’s ideas descended into pseudo-science, by continuing to claim support for the idea despite massive evidence against it. Velikovsky also based his arguments in large part on comparative mythology (he argued that the passage of Venus near Earth produced massive upheavals which were recorded in widespread myths). At best this can be viewed as anecdotal evidence.
Historical sciences such as the study of the geological and biological history of the Earth are perhaps more prone to attracting pseudo-science because of the inherent difficulty of developing and testing new predictions (we can’t run back the clock and run through history over and over again to test our ideas), and interpretation of human myth and religion is a common theme. Another well-known example is the argument that pre-historic humans were visited by extra-terrestrials – set out by Erich von Daniken in his book Chariots of the Gods?: Unsolved Mysteries of the Past. Von Daniken’s work is classic pseudo-science. He begins with a conclusion, and then sets out all the evidence which he purports to support it. No attempt is made to develop testable predictions, and no attempt is made to look for evidence which might refute the hypothesis.
The most pervasive ongoing example of pseudo-science is scientific creationism. Many different versions of creationism exist, but the core arguments remain the same. The Biblical account of creation is accepted at face value, and evidence supporting evolution (and an age of the earth much greater than that suggested by the Bible) is explained away. It is not the purpose of this section of the Science Toolkit to counter in detail the arguments of creationism. That has been done well by many other books and websites. The goal here is simply to show, in broad terms, why creationism should not be considered science. The first reason is that the basic paradigm of creationism, that God directly shaped the earth and all living things on it, is not falsifiable. Whatever geological and biological observations we collect, the creationist can always argue that they represent the “hand of God.” To put it another way: creationism makes no testable predictions. Because of this, creationists conduct few experiments, and devote most of their time to arguing against evolution, even when they claim to be making a case for creationism (see here for an example of creationist arguments). Creationists sift through the huge scientific literature on geology and evolutionary biology, carefully collecting the few papers which they can use to argue their claims, and ignoring the thousands which flatly contradict their views.
It is worth noting in conclusion that while evolutionary biologists heap scorn on scientific creationism, no reputable scientist would claim to have proven that God does not exist or did not take a role in the creation of life. It is exactly because such ideas cannot be disproven that they stand outside the realm of science. Indeed few biologists will argue with someone who steps forward and says that, as a matter of faith, he or she does not believe in evolution. They simply agree to disagree. It is only when creationism is presented as science, and attempts are made to have it taught as science, that conflict arises. Science is not the only way to search for knowledge, but it is a powerful tool, and ideas should not be able to claim to be “scientific” unless they have passed the rigourous screening which real science entails.