Posted from: Lipps, J.H. 1999. This is science! Pp. 3-16 in J. Scotchmoor and D.A. Springer (eds.). Evolution: Investigating the Evidence. Paleontological Society Special Publication, vol. 9.
THIS IS SCIENCE!
Jere H. Lipps
|1) Ask questions: be willing to wonder.||Start by asking "Why?"|
|2) Define the problem.||Restate the issue several different ways so it is clear.|
|3) Examine the evidence.||Ask what evidence supports or refutes the claim. Is it reliable?|
|4) Analyze assumptions and biases.||List the evidence on which each part of the argument is based. The assumptions and biases will be unsupported and should be eliminated from further consideration.|
|5) Avoid emotional reasoning.||Identify emotional influence and "gut feelings" in the arguments and exclude them.|
|6) Don't oversimplify.||Do not generalize from too little evidence.|
|7) Consider other interpretations.||Make sure alternate views are included in the discussion.|
|8) Tolerate uncertainty.||Be ready to accept tentative answers when evidence is incomplete, and new answers when further evidence warrants them.|
The last three items are particularly difficult. We all want explanations and we tend to jump to conclusions based on too little evidence. Again, an analysis of the evidence is required to determine if it is sufficient. Alternative interpretations should always be sought, even if the evidence seems compelling. Does the evidence allow for other possible interpretations? And lastly, tolerate uncertainty. No one likes uncertainty in their lives we all want, perhaps need, to know what is before us, why things happen to us, what happens when we die, and many, many more. Although difficult to do, tolerating uncertainty can be done by simply setting aside the uncertainties and, for the moment at least, accepting the uncertain.
Evidential Reasoning. In science, as in our daily lives, various claims are made by other scientists or acquaintances, TV personalities, doctors, and so on. We might even make such claims ourselves. All claims should be subjected to the analysis outlined in Table 2.
Of these points, perhaps the most critical is the last one. Any claim must be sufficient. In other words, you do not have to prove that the claim is false in order to test it, the claimant must provide sufficient proof himself. Secondly, the more extraordinary a claim, the more extraordinary the evidence must be to test it. For example, if a person claims that some herb has cured their cancer, you would be well advised to seek a good deal of further supporting evidence before risking your own life. Or if a person claims to have an extraterrestrial being in their garage, do not accept a fuzzy or even clear photograph as proof, demand a piece of the thing for further study. And lastly, the word of someone is never sufficient to establish the truth of a claim.
Table 2. Rules for evidential reasoning (Lett, 1990), or a guide to intelligent living and the scientific method (Lipps, 1999). All claims, whether scientific or not, should be subjected to these rules in order to ensure that all possibilities are considered fairly.
|Rules for Evidential Reasoning||What to do|
|1) Falsifiability||Conceive of all evidence that would prove the claim false|
|2) Logic||Argument must be sound|
|3) Comprehensiveness||Must use all the available evidence|
|4) Honesty||Evaluate evidence without self-deception|
|5) Replicability||Evidence must be repeatable|
A. Burden of proof rests on the claimant.
B. Extraordinary claims require extraordinary evidence.
C. Authority and/or testimony is always inadequate.
Judging Authority. The evaluation of authority (Point 6C in Table 2) requires special consideration. We often need to rely on other people for certain kinds of information or decisions we make in science or in our lives, simply because we cannot know enough about everything. Who can we trust to provide us with solid information? Consider the ways in Table 3 to evaluate authority, as applied to science in particular, but also in our non-scientific lives.
Even when an authority passes these tests, be aware of lapses that may reveal the degree of knowledge possessed by an expert. Well known or highly honored scientists are commonly asked to comment on subjects outside their own field of expertise. These people should be subjected to exactly the same questions as an unknown authority. Does a Noble prize winner in physics, for example, have any credibility when making pronouncements about evolution? Maybe, but the evidence and hypotheses about evolution are very far removed from the usual literature of physics. Be suspicious. In your daily life, authorities are always present making one claim or another, usually to sell something. These claimants can be evaluated rationally too.
So critical thinking, evidential reasoning and judging authority are essential to the scientific process, as well as to living an intelligent, full, and good life. Science requires more, however, and that is the development of knowledge.
Table 3. How to evaluate authority.
|Criterion||What to Consider||What to Do|
|1) Does the authority have proper credentials?||Considerable study or experience in a subject is required to become an expert in any field. Does the authority have degrees from a recognized college or university that has the faculty, libraries, and other facilities for proper education in the subject? Has the authority worked in the field for some time?||If not, don't believe him.|
|2) Does the authority have proper affiliations?||Is she identified closely with a reliable organization, such as a university, museum, government agency or corporation that practices the subject?||If not, ask how she makes a living?|
|3) Does that organization have a stake in the claims made by the claimant?||Be suspicious of anyone making claims that support the position or products of their own organization.||Seek independent evidence that the claim is correct.|
|4) Has the authority subjected her work to peer review?||In other words, have other experts evaluated the work, so that some independent assessment has been made?||If not, seek that evaluation yourself or find another authority.|
|5) Does the authority use the skills of critical thinking and evidential reasoning.||Does he use items listed in Tables 1 and 2?||If not, question him using those very skills yourself.|
|6) Is the authority a demonstrated expert in their field?||Other trustworthy people should rely on this person's expertise. Do other experts cite her conclusions?||If not, find another authority who others do rely on.|
|7) Does the authority present arguments without undue call on unsupported or untenable claims?||Does the authority present evidence sufficient that you can evaluate it?||If not, find an authority who can provide evidence supporting the claims.|
Science is a Social Activity. Scientific knowledge results from the efforts of hundreds of thousands of people over hundreds of years. Even current scientific efforts on single problems usually involve many people working in different ways all over the world. Scientific investigation is a group effort, requiring communication among all workers. In this way, all interested scientists can examine the work and conclusions of other scientists. Science is done in a social context.
The Scientific Knowledge Base is Huge. Few scientists (probably none) can speak intelligently about all aspects of science because it is so enormous. They are usually experts in only one small field of a larger discipline, just like so many other fields of human endeavor. Some may speak with authority on other aspects of science but they should be regarded with caution and further analysis. Because scientists differ in no significant respect from other people, they have the same strengths and weaknesses, and this may show in their work. All scientists are especially proud of their own ideas, for example, and they are ready to defend them. A good scientist will change her views, if enough evidence is accumulated to counter them. Some never change their minds, even when incontrovertible evidence appears. That is simply human nature, not a failure of science.
Controversy in Science. Controversy abounds in science and can be extremely helpful controversy is not a weakness. Controversy, in fact, is almost certain to develop in science because the testing of hypotheses is devised to disprove not prove them. Thus, everyone tries to disprove another's hypotheses and that scientist then may refute that disproof with yet additional evidence or presentation of alternative hypotheses. In this process of controversy, science advances. Because science is actually a group activity, each scientist builds on the work of others through criticism and support, and then contributes ideas to be used or criticized by those who follow. This is one of the great satisfactions that scientists feel their own ideas may last an eternity! But even a bad idea is often more constructive in the long run, because other scientists may sense that it is wrong and seek vigorously to correct it (Darwin, 1874. p. 606). A lingering bad idea does not indicate that science is lacking, but simply that the someone has not yet taken it to task.
Empirical and Historical Science
The sciences are not all the same either. Although the process of science remains the same, the nature of the observations may differ. These observations can be either non-historical (time independent) or historical (time-dependent) (Simpson, 1963). Sciences like physics, chemistry and much of molecular biology are largely non-historical, although they may rely on historical observations in particular instances. They deal with observations that are not expected to change with time they are time independent. An experiment done today should produce the same results as one done 10 years ago or 10 years in the future. For example, water should always flow downhill because the effect of gravity does not change with time, whether it be a billion years ago, yesterday, or today. We can expect that gravity will not change in another billion years. Sciences like astronomy, anthropology, much of biology, geology, paleontology, and evolutionary biology are largely historical, although each uses experiments commonly.
Historical sciences rely on observations or evidence (results) of phenomena that happened in the past. These results arose through a series of events, the history, and each event was contingent on previous ones. Historical scientists can only infer the causes from the results, since the results happened in the past. Many ordinary objects and processes around you have strong historical backgrounds, including your own life. You are a result of an historical process which cannot be directly observed. A good geologic example is the scientific study of the Grand Canyon. We see the Canyon today because the region has experienced a series of events in a particular order. If the events or order were different, the Grand Canyon might not exist as it is. If the metamorphism of the Vishnu Schist now at the bottom of the Canyon or the deposition of sand, mud and lime in the Paleozoic seas that once covered the area had not occurred, then the Grand Canyon would look quite different, if it would exist at all.
The historical sciences, of course, also use experimental methods. The rate of erosion of the rocks in the Grand Canyon can be determined by repeatable experiments, and thus provide additional evidence about how the Canyon came to be. By accumulating multiple observations throughout the Grand Canyon and the Colorado Plateau in which it is cut and conducting critical experiments, geologists have reconstructed a reliable model for the formation of the Grand Canyon. A similar process takes place in the study of the evolution of humans, rats, rice, or any other organism.
Historical science has a greater margin of error most of the time than non-historical, experimental science because scientists cannot repeat each event and must view only the results of those events through a filter of time. That margin should not, however, be mistaken for a lack of knowledge. We understand the formation of the Grand Canyon in all aspects, but not in every detail. In evolutionary biology, so-called missing links are details, not evidence that destroys the theory. Just because you may not have any information about your great great grandmother does not invalidate her existence nor that she is a part of your history. Paleontology and evolutionary biology are largely historical sciences that reveal the broad patterns, and very commonly even the detailed patterns, of evolutionary history. Gaps do exist in that record, just as there are likely to be gaps in your own family's historical record, but that does not invalidate or make the science less substantive.
Scientists and teachers tout the "scientific method" (Table 4) as the way we do science; it appears dull and agonizing. This erroneous formalization of the scientific thinking process makes people fear science. Few scientists actually work that way. Instead, they get excited, hopeful, interested, intrigued, and puzzled. Their ideas come to them in the shower, on the freeway, while playing baseball with their kids, as well as in the laboratory, office or library. Good science is a creative process without rules on how to do it.
Table 4. "The Scientific Method" is commonly taught as the way science is done. It is, however, more truthfully the way science is presented to others. Most scientists are creative people, who use their imaginations, personal experiences, wonder, joy, and excitement to develop their scientific ideas. Later, they reorder all of this creativity into the method, so others will understand them. Most science only gets to the "Test Hypotheses" stage.
Elevate to Theory
Elevate to Laws
Alternative Ways of Doing Science
How, then, do scientists sort out all this creativity to make order and understanding possible? They use various ways. More than a hundred years ago, T.C. Chamberlain laid out three ways of doing science (Table 5) (Chamberlain, 1897). Chamberlain and later authors (Platt, 1964) made clear that the third method, that of multiple working hypotheses, was the most efficient and preferred way to proceed.
Scientists do three fundamental activities. They gather data, develop hypotheses, and test hypotheses, not necessarily in that order. Within these activities, individual scientists may work differently. Chamberlain (1897) and Platt (1964) carefully pointed out the strengths and weaknesses of each method. Data gathering is simply the accumulation of observations experimental, observational, or mathematical. Hypothesis development requires that it be consistent with all known data, that it fit logically within other accepted hypotheses in science, that it be testable, and that it have predictive power. Hypotheses must be testable within the abilities of science; otherwise they remain simply an idea without use. They must also predict that certain phenomena will occur if certain experiments or observations are undertaken. A test must attempt to disprove the hypothesis since proof in science cannot be attained. The more critical tests that an hypothesis passes, the more confidence we can have in it.
The concept of a scientific hypothesis differs significantly from the common use of "theory" in non-scientific language. "Theory" here means an ad hoc idea which does not possess the characteristics of a good hypothesis. The common usage of "theory" should not be confused with the scientific use of "hypothesis" or "theory."
Table 5. Ways of doing science, according to Chamberlain (1897).
|Type of Approach||What is Done||Good Points||Bad Points|
|1) Description of phenomena||Describes what is observed.||Provides necessary data.||Fails to provide explanation.|
|2) Single Hypothesis method to account for phenomenon.||Provides a single, testable hypothesis.||Scientist may devote much energy and creativity to "proving" or supporting her/his idea||Other hypotheses may equally or better explain the phenomenon. Scientist may be blinded to these other hypotheses.|
|3) Method of Multiple Working Hypotheses, all worked nearly simultaneously.||Provides several to many possible hypotheses for testing with the available data.||Many ideas are tested with the same or little additional effort, and new hypotheses can be generated as they develop. Scientist is not wed to a favorite idea, hence is less biased.||Energy may be dissipated among many hypotheses.|
Descriptive Science. Descriptive science is important and nearly all scientists engage in it. It provides the fundamental data surrounding any scientific problem. However, simply answering the "what" questions, without understanding the "why" questions, is unsatisfactory, although necessary. Description of observations is essential to the understanding of processes, for a scientist must know what he is trying to explain. But mere description by itself does little to explain what the processes are that have taken place to give rise to the observations. Observations in science include an experiment, say in chemistry, and of objects in context, say fossils in a rock layer. An experiment is simply a single observation on a par with an observation of those fossils in rock. Both observations take on significance when they can be made multiple times, thus confirming their generality. The results of an experiment can be readily measured with precision and commonly repeated, but it is nevertheless just another fact, as are the fossils (for which precise measurements and repeated observations are common), supporting or disproving a scientific hypothesis. Although many scientists must stop, at least at times, with description, most proceed to one of the next steps.
The Single Hypothesis Method. A single idea, model or hypothesis to explain a set of observations is commonly developed by scientists. A large number of scientists stop with the development of an hypothesis that accounts for their observations. However, the method of using a single hypotheses is fraught with many pitfalls. First, a scientist with a single hypothesis is a knight in armor. He is forced to defend his idea because it is the only one that he possesses. Second, data that do not fit the hypothesis are easy to ignore because there is no other place to use it. Thus, the data collected tend to support the hypothesis, yet the best supported hypothesis can still fail on a single critical observation. Third, a scientist with a single hypothesis has her ego at stake, and thus resists counter hypotheses made by other scientists. Because scientists are like other people with regard to their egos, this resistance to alternate hypotheses results in a loss of objectivity, and sometimes bitterness may ensue and controversy abound when others try to disprove the hypothesis.
The Method of Multiple Working Hypotheses. A better way for scientists to work, and many do, is to use Chamberlain's "Method of Multiple Working Hypotheses" (Chamberlain, 1897). In this method, a scientist thinks of all the possible hypotheses that might account for his or her observations, and then goes on to test each one. In this way, his ego is attached not to a single hypothesis, but to the development and testing of all of them (Platt, 1964 ). She takes pride not in a single hypothesis but in an array of them. She has an easier time using all the data, for many hypotheses are available to apply them to. Furthermore, single tests can be devised for each hypothesis and those that fail can be eliminated rapidly, thus increasing the efficiency of scientific progress. For those hypotheses that survive all conceivable tests, they can be arranged in some order of more probable to less probable based on the degree of support generated for each from the data, and the predictive value of the hypothesis. This arrangement then provides a research plan for other or future workers who may have additional data or newer techniques or instruments. In applied situations where decisions must be made on incomplete information, for example your health or environmental issues, a scientist left with two or three hypotheses can still make an informed decision based on which one is best supported by the available evidence. Science is thus more efficient when the method of "multiple working hypotheses" is fully utilized.
To show in a simple way, how each might be applied to a hypothetical study, assume that three investigators want to know why a certain butterfly species congregates in certain kinds of trees. A descriptive scientist might spend the rest of his life counting butterflies in numerous trees, and know only that each tree has some capacity for butterflies without understanding why. While that might be important, it is not an efficient methodology. A second scientist might have a single favorite hypothesis, say that the butterflies need certain nectar from the trees' flowers and that is why the butterflies congregate there. This scientist will work hard and long to find ways to prove his assertion, for he has no alternatives except to try to prove his idea. A third scientist, using the method of multiple working hypotheses, will acquire some data, develop a number of hypotheses to account for those observations, then design tests that will disprove (not prove) each hypothesis. She may find a single tree someplace that contains an anomalous number of butterflies, thus suggesting that the first scientist has wasted a good deal of his life trying to understand numbers in trees. She may find that some trees with butterflies have no nectar, thus disproving the favorite hypothesis of the second scientist with a single, well understood observation. A single failed test means that hypothesis can be discarded. In this process, additional data are collected that can be brought to bear on the remaining hypotheses, as well as the development of yet additional ones. A large number of possibilities are dealt with efficiently and quickly. The third scientist moves through all previously proposed hypotheses and those she developed herself, eliminating some and accumulating evidence to support those remaining. In the end, since no hypothesis can ever be "proved", she has several that are supported in one degree or another. The scientist can then develop further tests, or communicate the work to others for them to consider. Science has thus progressed quickly, efficiently, and a future research plan has been laid out, and with these has come greater understanding of the phenomenon, although not necessarily the "truth". We may never know when truth is attained, for to prove scientific hypotheses is impossible.
Science Is Not Based on Faith, But on Hypotheses Testing. Occasionally someone asserts that science is based on faith. This is simply not true. Someone knowledgeable in the scientific process accepts the proclamations of other scientists outside their own field of expertise because we know that the scientific process yields testable hypotheses subject to elimination. The proclamation may not be correct but we accept it anyway, knowing that it is subject to scrutiny and correction by other scientists more expert in that particular field. As a paleontologist, for instance, I tentatively accept Einstein's Theory of Relativity, not on the basis of faith, as some people would assert, but on the knowledge that it has been tested, will continue to be tested, and that it has withstood those tests. Faith is not involved at all. I am ready to change my mind as soon as I understand from other authorities in physics that Einstein's theory has failed a test.
Some hypotheses and theories may not yet have been subjected to the critical tests, either because no other scientist has gotten interested in the problem, or the technology or proper discoveries have not yet developed. I am always on the watch for claims by physicists that might change the acceptance of Einstein's theory. In this sense, science is self-correcting and errors are eliminated. They may not be eliminated immediately, but eventually they will be. In rapidly moving fields, like molecular biology, corrections come every day; in other less active fields, say sponge systematics, corrections may take decades.
Evolution is the scientific study of how organisms bacteria, archaea, protists, fungi, animals and plants came to be. Evolution relies on the same sort of scientific processes and evidence as do all other fields of science. The methods and processes of evolutionary biology do not differ in any significant way from those used in any other science. Evolutionary biologists use critical thinking, evidential reasoning, judgment of authority, hypotheses development, data gathering, and hypotheses testing just like all other scientists. Evolutionary biologists practice nonhistorical or historical science or both. All biologists contribute to evolution in one way or another. Some use experiments, for example with fruit flies, while others observe the sequence of fossilized organisms in the rock record. Some rely heavily on the other sciences for complementary data, for example the evolutionary paleontologist is beholden to physicists for the understanding of dating by radioactive decay of various rocks.
Darwin (Darwin, 1859) brought together a large number of observations and developed the hypothesis that the environment acted to select individual variants in a population that had particular characteristics. He did not know what caused the variation, such as eye color, growth rates, or height, but he determined that if a particular character were consistently selected by the environment through successful breeding, in the same way that dog breeders selected characters they desired in their dogs, then changes would ensue in future generations. This hypothesis he called "natural selection". As he and others gathered more evidence and added additional hypotheses in the following decades, his hypothesis became strengthened and called a theory. Evolution is a simple elegant theory in its basics, yet it rapidly becomes complex. This complexity leads to other hypotheses to account for aspects of the overall theory. For example, punctuated equilibrium the idea that species are in some kind of evolutionary stasis for long periods of geologic time and then rapidly evolve into another species was presented as an alternative to Darwin's idea that evolution was a gradual affair with changes taking place slowly and evenly over a long time. Punctuated equilibrium enhanced Darwin's theory by clarifying how evolution took place through time it did not disprove the theory as some religious zealots have claimed. Today, many biologists and paleontologists around the world are working to better understand and test Darwin's theory in all its details. Some of these biologists confirm Darwin's ideas, some develop more critical details, and all constantly test the theory, but the theory of evolution has yet to be disproved in spite of all this effort.
Science is a way to understand our world and universe through repeated reliable observations and hypothesis development and testing. The observations must be repeatable by others besides the originator. From this, we conclude that nature is independent of a single observer or set of observers. Nature behaves similarly under similar conditions. It does not differ from observer to observer, as long as the conditions are identical. Thus, science is reliable; observations and tests are not dependent on the kind of people who practice it, what they might feel about it, or what their previous experiences are. Anyone using the same techniques and procedures will have similar or identical results. As far as we can tell, the world and universe are real.
Science requires certain skills the ability to think critically, to reason evidentially, and to evaluate authority. Using these skills, scientists can then gather data, develop hypotheses to account for the data, and test the hypotheses. The most efficient method of advancing science is the "Method of Multiple Working Hypotheses" that entails simultaneous and continuous development and testing of a number of hypotheses. Hypotheses and theories in science are never proven, only disproven. In this sense, a final, a true answer may never be attained, but the supporting evidence, resistance to disproof, and the logical fit with other scientific knowledge provides differing degrees of confidence. By evaluating the degree of confidence, scientists can make valuable decisions about scientific issues and design new technologies even if an answer is not known with certainty.
In order to do this well, scientists must be excited, creative, and innovative. The best scientists are the most creative scientists. They let their minds wander, putting various ideas and bits of data together to find possible solutions to a particular puzzle about the universe. Look at the pictures from the Hubble Space Telescope, the images from the Martian Pathfinder, or the bones of a dinosaur partially excavated from the ground. Imagine the thrill of waiting for an experiment to run to completion, of studying a whale breaching at the ocean's surface, a comet hitting a planet, or a snake eating a rat. These are each different and attract different kinds of people, but they can all be good science. This is excitement, and it is addictive! In a good way, and often for life. A good life. Science is fun! And the good news is: Almost anyone can learn and practice the skills of science and improve their lives.