SCIENCE, HISTORY AND EVOLUTION Science has recently been described as humanity's greatest achievement. For most of Western intellectual history, however, science existed as a subarea of philosophy, often referred to as "experimental philosophy." Only in the past 300 years or so has science developed into a distinct academic discipline whose fruits can be seen in the spectacular rise of modern technology. Nevertheless, science is rooted in basic assumptions, e.g., for every effect there exists a cause, and in this regard is still similar to philosophy. Furthermore, science can properly answer only a limited number of questions and so complements rather than replaces philosophy. Because science has the same foundation as philosophy and is more restrictive in scope, it is a mistake to think that scientific knowledge is superior to philosophical knowledge. The rise of science can be traced to the pre-Socratic or natural philosophers of ancient Greece who combined rational thinking, observation and experimentation to explain the mysteries of the cosmos or universe. Their greatest achievement was the assumption, which guided their thinking and upon which scientific investigation is based, that the universe is orderly and knowable and so could be explained in terms of laws rather than the whim of gods. With Socrates the emphasis in Greek philosophy shifted from explaining the universe to understanding human nature and the good life. Plato, a student of Socrates, separated rationalism from empiricism, opted for the former over the latter, and replaced physics with metaphysics. In so doing, Plato led to the demise of early science. His metaphysical thinking, as well as that of his pupil Aristotle, came to dominate Western thought until science was reborn in the 17th century through the efforts of Francis Bacon, Rene Descartes and Francesco Redi. These three thinkers discovered the power of induction, deduction and experimentation respectively as ways of knowing and so freed the thinking of their day from the bondage of Aristotelian metaphysics and Church dogmatism which were interwoven in the form of scholastic philosophy. As we shall see below, the processes of induction, deduction and experimentation form the backbone of that process of discovery we call the scientific method. Today, in the minds of laypersons, science has been elevated to the status of a religion whose priests, the scientists, ply their skills in ways totally unintelligible to the masses. The rift between the two cultures of science and the humanities seems to be widening (a far cry from the days when science and philosophy were one) and must be closed by educated persons if modern society has any hope of surviving and fluorishing. Nonscientists must temper their optimism for the power of science by appreciating its limitations as a way of knowing and realizing that science cannot solve all the problems which plague modern society. Scientists, on the other hand, must come to accept full responsibility for their discoveries and not wash their hands when their work is used to the detriment of society. Definitions of Science Just what is science and why has it been so spectacularly successful in shaping Western industrial society? The answer to these questions is the focus of interest of a new breed of philosopher, the philosophers of science, who analyze what scientists do and attempt to abstract a general set of rules governing scientific inquiry. Scientists, for their part, pay little attention to what they are doing in a formal sense and attempt to explain nature and solve the problems they encounter in their quest for knowledge. There exist a number of conceptions and misconceptions about the nature of science and these will be discussed below. As we shall see there is no simple answer to the question: What is science? Science is an organized or systematized body of knowledge. This is a very common definition of science which focuses on the content of science and the factual results of scientific research. It does not, however, distinguish science from nonscience, since any logically presented body of thought can be considered science under this definition. Under this definition theology was once considered the "queen of the sciences" and the philosophical system of St. Thomas Aquinas was considered to be scientific. A telephone book is a highly organized body of information, but nobody would consider it to be a science text. Science is an authoritative, objective, dispassion- ate search for absolute truth. This concept of science is riddled with errors. First of all, despite the popular appeal of a "scientific demonstration" to sell products on TV, the validity of which rests solely on the authority of an actor dressed in a white lab coat, scientists routinely reject authority. It is the empirical evidence, not the reputation of the investigator, which decides nature, but it exists because scientists are human and not dispassion- ate in their beliefs. They are just as influenced as everyone else by bias, paradigms and prestige and often lose objectivity in the positions they hold. Many of the controversies which have punctuated the history of science were due to the subjective dispositions of scientists unwilling to relinguish their pet ideas in the face of contrary evidence. Scientists tend to magnify the results of their work through inappropriate generalization and this often brings them in conflict with other scientists who arrive at a different conclusion. The nature vs. nurture controversy is a good example. Ethologists studying instinctive behavior came to the conclusion that behavior was stereotyped and innate and this conclusion was quite opposite that of behavioral psychologists who studied learning and concluded that behavior was flexible and due to environmental rather than genetic influence. As it turned out both were correct and behavior is now known to be influenced by genes and the environment rather than by just one alone. Animal behavior forms a spectrum running from largely instinctive to largely learned responses to the environment. Each school studied one end of the spectrum and attempted to generalize its results. As for absolute truth, it simply does not exist in science. All scientific theories are subject to revision and so are provisional at best. The idea that science is continually progressing in gradual fashion towards some ultimate truth is also false because most of the significant advances in science have come about when ideas were totally overthrown rather than simply modified, e.g., the replacement of the Ptolemaic, geocentric universe with the Copernican, heliocentric one, and the conceptual change in geology from stationary to drifting continents. Scientific investigation is driven by what the society of the day deems important and is willing to support, e.g., star-wars and AIDS research, not by some quest for ultimate truth. Science does make progress in under- standing reality but at any given time the ideas of science are subject to change and so do not constitute absolute truth. Science is what scientists do and the scientific community decides what science is. Although vague, this description of science is probably closest to the truth of what science actually is. Earlier in this discussion I noted that scientists attempt to solve problems with little attention to what they are doing in a formal sense; it is the philosophers of science who attempt to describe what scientists do. Scientists employ a number of different approaches to problem solving and each special- ized branch of science has its own tradition, techniques and outlets for publishing results. Scientific articles are peer- reviewed, hence, what is accepted as legitimate science is determined by scientists themselves. As we shall see later on, the essence of science as viewed by both scientists and philosophers of science lies in the methodology rather than in the content of science. The true hallmark of science is hypothesis testing, i.e., any scientific explanation of nature must always be checked against reality. Nevertheless, there are many scientists included in the scientific community who rarely, if ever, develop and test hypotheses. Among these would be synthetic chemists who attempt to purify compounds, taxonomists who describe and classify species and investigators who are engaged in the description of patterns in nature. Theoretical astrophysicists (cosmologists) are still included within the scientific community even though some of their theories cannot be falsified given the conditions necessary to check their predictions. At this level the choice among competing theories, e.g., those dealing with what happened at the moment of the creation of the universe will have to be made on the simplicity and elegance of the explanation rather than on empirical evidence. In short, for cosmologists "'Beauty is truth, truth beauty,'" as Keats proclaimed in poetic form. In the long run, the nature of science cannot be described definitively because science is continously changing as scientists go about the business of explaining physical reality. At any time, however, what constitutes science is that which is accepted as such by the existing scientific community. The Scientific Method As was mentioned above, science is distinguished more by its methodology than by its content since scientific truth is relative and changing rather than fixed and absolute. Despite its shortcomings, the following criterion will serve to distinguish science from nonscience: science attempts to explain reality by constructing hypotheses that can be falsified by testing. Hypotheses or explanations which cannot be falsified by testing are nonscientific. The concept of a single, unified and universally accepted method of doing science is a myth although introductory texts emphasize the sequence of observation-induction-deduction- experimentation as the means whereby scientists develop and test hypotheses. The doing of science no doubt involves the processes of induction, deduction and experimentation, but by no means are these tools used in any rigid, set pattern by all scientists. Science is a cooperative venture which builds upon the contribution of different individuals, each of whom may use these tools in different ways and to differing extents. Some scientists are theorists who use mathematics to construct models which abstract the essentials of complex phenomena, postulate how variables interact and may lead to specific hypotheses. Other scientists are empiricists who garner facts through observation and experimentation either to discover patterns in nature or to test specific hypotheses. Some empiricists work in the laboratory under very controlled conditions and so produce precise, accurate, but often unnatural results; while others work in the field and record data in nature, although these data are often difficult to interpret due to the variability and complexity of the natural environment. Solutions to significant scientific problems take time and the application of all three approaches: model building, field study and laboratory work. Thus, the so-called scientific method does not exist other than as an ideal abstraction of the work of many different scientists; but there do exist certain ingredients that contribute to the methodology of science and these will be described and evaluated below. Components of scientific methodology Data collection. Many consider the starting point of any scientific investigation to be data collection. The random collection of facts, however, without any guiding hypothesis is not only tedious, it is most often fruitless. Data are usually collected to test specific hypotheses, not to formulate them; but, when no prior information about a phenomenon exists, data are collected to discover patterns or regularities in nature which then are explained by hypotheses. The inducto-deductive approach to science does indeed start with data collection for the purpose of finding a pattern. The pattern is recognized inductively by extending limited observations to make a generalized statement. The hypothetico- deductive approach to science, on the other hand, starts with hypothesis building to explain already recognized patterns. Both of these approaches then use data collection for testing hypotheses. Induction. Perhaps the greatest misconception about science concerns the role of induction in scientific methodology. Induction is the logical process of arguing from the particular to the general and so allows one to make a general statement based on a limited number of observations. For example, suppose that, after trapping mice in several different woodland lots, I notice that all the mice I have captured are brown. Through induction, I could then argue that all mice are brown. Such inductive generalizations describe broad patterns in nature which then require some hypothesis to explain them. Many think that induction leads to hypothesis construction; it doesn't except for some very limited theories which explain nothing and are only generalized patterns, e.g., the cell theory. Induction produces patterns, not hypotheses. The so-called "laws" of physics and chemistry are really inductive generalizations of patterns so regular in nature that they follow precise, mathematical relationships. These laws are not, however, explanations of why the patterns obtain - such explanations are called hypotheses or theories (see below). The proper role of induction in scientific methodology is to describe patterns which then will be explained through hypotheses (inducto-deductive version), or to lend confidence to hypotheses via the verification of predictions (both inducto- deductive and hypothetico-deductive versions). Induction is not the process whereby hypotheses are formulated. Hypotheses. Hypotheses are tentative explanations of why patterns occur, i.e., they are statements of the mechanistic processes which produce patterns in nature. Nobody really knows how hypotheses are formed; they result from luck or individual, creative genius. No formula exists for producing them. To be scientific, however, hypotheses must be stated in such a way that they can be falsified. The criterion of falsifiability, which characterizes scientific hypotheses and provides a basis for distinguishing between scientific and nonscientific hypotheses, was introduced in 1934 by Sir Karl Popper (a reknowned philosopher of science) in his book, The Logic of Scientific Discovery. According to this criterion, to be scientific an hypothesis must be tested empirically against nature and accepted only if such testing cannot falsify it. The hypothesis that God is the creator of nature is nonscientific because there is no way of falsifying it through testing, even though it appears to be definitionally true and logically consistent with the philosophical propositions that every effect has a cause and nature is an effect, not a cause. Note also that the aim of a scientist is to falsify, not to prove hypotheses. Proof exists only in mathematical theorems, not in scientific methodology. Hence, hypotheses which have not been falsified through testing are accepted; but such acceptance results only in provisional (as opposed to absolute) truth because scientific hypotheses still are capable of falsification and might be falsified in the future. Remember that the discarded theories of the past were accepted as scientific truths in their day! Deduction. Hypotheses are tested through their predictions, which are produced by the process of deduction. Deduction is the logical process of arguing from the general to the particular. As general explanations, hypotheses can be extended to unknown phenomena which are predicted from the hypothesis using deductive reasoning. In science the term prediction means the statement of a fact not known to exist at the time the statement is made, but one which must exist if the hypothesis is true. The hypothesis is tested by attempting to find the predicted fact either through observation or experimentation. Don't equate the term "prediction" with the statement of a future event (the usual meaning of the term prediction) because a scientific prediction can be either a prediction in the sense of a future event or a postdiction (the statement of a past event unknown at the time the statement is made). Suppose that, based on observation of a comet's appearance in 1941, 1962 and 1983, an astronomer hypothesizes an orbit for this comet requiring 21 years to complete one full revolution around the sun. This hypothesis would yield two predictions: (a) that the comet will return in the year 2004, which is a literal prediction, and (b) that the comet did appear in the year 1920, which is a postdiction. If both predictions are verified, does this prove the hypothesis? The answer is no; verified predictions only support the hypothesis but don't prove it because (1) future observations could falsify it, e.g., if the comet was recorded in the year 1911 or returns in the year 2020 or (2) alternative hypotheses might be made with the same predicted 21 year cycle. Thus, the unique feature of science is that its hypotheses can be falsified through testing and the role of the scientist is to reject hypotheses. Hypotheses that cannot be rejected must be accepted as provisionally true. The basis of hypothesis testing is through predictions generated from the hypothesis by a process of deduction. Experimentation. An experiment is simply a way of testing hypotheses by finding out if specific predictions are true or false. In a more restricted sense an experiment is a manipulation of the conditions which result in some phenomenon, whereas an observation is simply a witnessing of the phenomenon. The most effective type of experiment is a controlled experiment wherein all factors that can influence a phenomenon are kept constant except one which is allowed to vary. In this way the effect on the phenomenon under study of the variable can be determined precisely and accurately. The controlled experiment is erroneously considered by many to be the hallmark of science, but in actuality it is hypothesis testing (through either observation or experimentation) which distinguishes science from nonscience. Theories. An hypothesis which has broad explanatory power and has survived repeated testing increases in level of certitude and eventually becomes elevated to the status of a theory. The difference between the terms "hypothesis" and "theory" is inexact but refers to the scope of the explanation and the level of certitude attached to it as a consequence of repeated testing. The greater the number of specific predictions which are verified through experiment, the higher is the level of certitude attached to the hypothesis. But note that this higher level of certitude obtains only because of inductive reasoning: arguing from particulars (the number of verified predictions) to the general (the validity of the hypothesis). Thus, in scientific methodology induction plays a crucial role in acceptance of hypotheses or theories as well as its role in recognizing patterns in the inducto-deductive approach to science. What happens when one specific prediction of a theory is falsified? Does this mean that a long accepted explanation of how nature operates must be rejected? The answer is no! The theory will be retained as long as it functions in providing a reasonable interpretation of nature and continues to generate new information by being the guiding light behind research projects. Theories, therefore, are paradigms for interpreting information and generating new information through questions raised by the theory itself. A theory is not just the explanation of a phenomenon, it is a way of looking at the world and so is exempt from naive falsification, i.e., rejection based on the experimental falsification of one or two of its predictions. Of course, if more and more predictions are falsified as time goes on, the usefulness of the theory diminishes and it may be replaced by a new theory which provides a better paradigm for interpreting nature. Theory replacement, according to the philosopher Thomas Kuhn, is the most exciting event in science and the one which accounts for the success of science over the long run. Note again that the replacement of one theory by another does not necessarily mean gradual progression towards some form of absolute truth. In many instances the new theory doesn't build on the one it replaces, it actually destroys it. Accepted theories represent the body of current scientific truth and they are the means whereby we make sense out of the world, interpret new information and expand the frontiers of knowledge through research effort. The Problem of History The physical sciences (physics and chemistry) have a distinct advantage over biology and the social sciences in that the phenomena studied in the physical sciences are independent of history. The law of gravity applies for all time (at least after a fraction of a second after the "Big Bang") and so the behavior of an object with regard to gravity is not influenced by its recent past history. The same cannot be said of phenomena studied in the other sciences due to the effect on an individual's behavior brought about by both its developmental and evolutionary history, i.e., its ontogeny and phylogeny. The time-free dimension of physical science allows extensive use of experimentation since the phenomena under study are influenced by processes which are ongoing, hence, subject to manipulation. All causes studied in the physical sciences are proximate; but, due to evolutionary history, the life and human sciences must consider ultimate causes. Consequently, predictions produced by physical scientists can be very precise and most often are in the form of mathematical equations, whereas those in the other sciences are less exact and verbally stated. This has given rise to the distinction between "hard" and "soft" sciences with the physical sciences classified as "hard" and those which must consider history classified as "soft." Since history deals with the past, historical hypotheses cannot be tested experimentally because the past is over, not ongoing. Historical hypotheses rely more on postdiction and observation than on prediction (in the literal sense) and experimentation for testing, but this makes them no less scientific because historical hypotheses can be falsified. Biology and the social sciences may be "soft" relative to the physical sciences, but they still are sciences because their hypotheses can be tested. For biologists, the present is the product of the past and cannot always be studied apart from its history. Whereas the universe is some 15-20 billion years old with the units studied by physical scientists appearing soon after the Big Bang, the units studied by biologists are far more recent in origin. As far as we know, life is unique to the planet Earth, which formed about 4.5 billion years ago, and the first evidence of life exists in rocks dated 3.5 billion years ago. The great diversity of life seen in the fossil record is even more recent and begins about 600 million years ago. Each individual is the product of an historical process, i.e., development, and so the behavior of an individual cannot be fully understood apart from a knowledge of its ontogeny. Likewise, phylogeny or species history is an important consideration in understanding the limits to the potential of each individual. We are what we are because we evolved from primates. Had we evolved from some other mammalian group, e.g., carnivores, our appearance and behavior would have been quite different. To what extent can the past be studied using scientific methodology? The events of history are largely unique and unrepeatable, therefore, they cannot be studied experimentally. How is it possible to approach history methodologically so that the historical aspect of biology can be studied scientifically? The following methodological approaches are used by evolutionary scientists to reconstruct the past and so create falsifiable hypotheses about historical phenomena. The fossil record The only direct window to the past is to be found in the fossil and archaeological records. The fossil record documents change in the species composition of the Earth over time, but it is notoriously incomplete. In addition it reveals pattern not process and so can be interpreted in various ways to support different hypotheses regarding the process behind the pattern. Radiometric dating techniques provide an accurate means of estimating the age of fossils and archaeological artifacts and so provide a time frame so that history can be observed directly. The principle of uniformitarianism This method for understanding the past is based on an assumption: the processes which shaped the past are ongoing and so can be studied today. First formulated by the Scottish geologist James Hutton (1726-1797), the principle of uniformitarianism was brought to prominence by Charles Lyell (1797-1875) - the geologist who was a close friend and important influence over Charles Darwin. According to this principle, processes such as plate tectonics, volcanic explosion and natural selection are uniform over time so that past events can be explained as the result of contemporary processes acting over long time periods. It also allows us to study past processes experimentally because they are still operating today. Thus, in a limited sense through this principle, the past can be studied experimentally. The opposing view that geological processes were unique and acted in an irregular fashion is called catastrophism. This position was held by Georges Cuvier (1769- 1832) who was a creationist hailed as the father of comparative anatomy and paleontology. The battle between catastrophism and uniformitarianism was one of the major confrontations in pre- Darwinian science and still is around today as we will see later in our discussion of gradualism vs punctuationalism. Historical scientists use this principle to suggest that physiological and ecological adaptations of present species also characterized extinct species. Consequently, what we learn from studying living species can inform us about the biology of extinct species, even though we cannot observe directly the physiology and ecology of species known only from fossils. For example, we can make statements about past climates by comparing fossil fauna and flora with existing species. Contemporary species vary in their ability to tolerate temperature so that some are restricted today to warm regions, others to colder ones. Fossil relatives of each can be used to estimate the nature of past climatic conditions. The conclusion that a certain period of the past was warm because the fossil species represented in that period have the same appearance as contemporary species which are restricted to warm climates rests, of course, on the assumption that morphological indicators of warm climates today were also indicators of warm climates in the past. Also, growth rates of trees vary with temperature and moisture and this variation is documented in the width of rings observed in cross sections of their trunks. By correlating the ring pattern observed in cross sections of living trees with weather records of the recent past, climatologists can determine how temperature and rainfall affect the growth of rings. Paleoclimatologists, then, use this information to reconstruct past climatic regimes by studying the ring structure of fossil trees. Again, any conclusion drawn from such studies assumes that trees responded in the past to climatic factors in the same way they respond today in producing growth rings. Since past processes can never be known for sure, the principle of uniformitarianism is not a scientific hypothesis; it cannot be falsified. Nevertheless, the more success historical scientists have in piecing together a picture of the past using this principle, the more confident they become regarding the validity of this assumption. One application of this principle which is less valid is the idea that rates of change produced by processes have remained constant or uniform over time. This is certainly not true for rates of volcanic action or sedimentation; hence, attempts to date fossils by means of extrapolation from existing rates is not valid. One exception to this constancy of rates assumption is the rate of radioactive decay based on theories of physics whereby geological strata are aged. Radioactive isotopes (uranium 236, rubidium 87 and potassium 40) decay to form a product (lead 206, strontium 87 and argon 40, respectively) at a constant rate measured as the isotope's half- life. After one half-life only 1/2 of the original amount of the isotope will remain; after two half-lives only 1/4 will remain; after three half-lives only 1/8 will remain, and so on. From a knowledge of the amount of isotope trapped in a rock when it was formed (rocks must be formed instantaneously, hence only igneous rocks formed by volcanic eruption can be dated accurately - sedimentary rocks in which most fossils are found cannot be dated because they form very slowly) and its half- life, scientists can estimate the age of the rock by a method called radiometric dating. Different isotopes have different half-lives and the method of radiometric dating used for each isotope has its own unique assumptions. Nevertheless, that the result is independent of the peculiarities of each isotope is suggested by the agreement in dating that results when different methods are applied to the same rock. Constancy of rates is claimed for macromolecule (DNA and protein) change due to mutation - the constant ticking of a so- called molecular clock. This idea is very controversial today but molecular clocks have been used to estimate the timing of divergence between apes and humans which we will discuss later in the course. The principle of imperfection This principle, called the panda principle by Stephen Jay Gould, enables one to reconstruct the past history of a trait which is imperfectly designed to achieve its function. The panda's thumb, to cite Gould"s example, is not a true thumb; rather, it is a modification of a bone in the wrist. Consequently, the panda is not related to primates, which have true, flexible thumbs, but apparently evolved from carnivores whose digits are basically the same as those of the panda. Perfectly designed structures which are highly adapted to the environment in which the species possessing them lives, e.g., the vertebrate eye, mask the role of history in their origin and so provide little evidence for evolution. To the contrary, perfectly designed structures suggest a designer and so seem to offer more evidence for creation than for evolution. The angiosperms, or flowering plants, exploded in diversity soon after they evolved and are characterized by the possession of a unique structural feature: flowers. How did flowers evolve from nonflowering plants? A detailed examination of floral structure reveals that flowers are nothing but modified leaves and therefore evolved from nonreproductive parts of these plants. Despite the detailed adaptation of flowers in many angiosperm species to attract insect pollinators, the existence of less well-adapted flowers in other species betrays their evolutionary origin. The principle of sequences Finally, the phylogenetic history of traits or species can be revealed by studying existing variation and arranging this variation into a series of stages which suggests the temporal pathway involved in their evolution. This method, which I will call the principle of sequences, is based on the idea that rates of evolution are unequal so that there can exist side by side at any cross section of time all of the stages in an ongoing process which proceeds sequentially from simple to complex (or from primitive to advanced). Alternatively, this method can be used for sequences which occur at equal rates, but which begin at different times. Thus, according to this principle, we can learn something about the nature of the past by studying existing diversity and ordering it into a structural sequence which suggests a temporal sequence. The principle of sequences is used primarily to hypothesize the phylogeny of existing species. An example of this method is the reconstruction of the evolution of the balloon fly's curious mating habit. Male balloon flies court females by constructing a hollow ball of silk which they present to a potential mate who then accepts the male's advances and mates with him. The evolutionary history of this unique behavior has been suggested by a detailed examination of the various mating behaviors of other species in the same family of dance flies, the Empiidae, to which the balloon fly belongs. When the entire array of mating behavior variation in this family is classified in a sequence of simple to complex, a pattern emerges which suggests the possible phylogeny of the balloon fly. This sequence is described as follows. 1. In some species mating is quite haphazard with the males rushing up to females, who sometimes mistake them for prey species instead of potential mates and eat them. 2. Other species avoid this potential hazard by offering females a prey item and then mating with her while she is consuming her meal. 3. More advanced species wrap the prey item in a silken net before presenting it to the female, while in others the net itself becomes more important as a signal of courtship because it is presented to females either empty or with a piece of vegetation inside. The balloon fly's reproductive tactic represents the culmination of this sequence in which the net is a care- fully woven ball which the female accepts as a species recognition cue. The four methods explained above enable evolutionists to develop hypotheses regarding past events. The major problem history poses for science is that past events and processes cannot be studied experimentally. Biologists cannot go into the lab and attempt to produce an amphibian from a fish, although this transition did occur in the past. Critics of historical study argue that the past cannot be studied using the scientific method; but, this criticism stems from a very narrow view of scientific methodology - one based on physics as a model. Physics studies the proximate causes of phenomena which act at the present time and can be manipulated experimentally. Historical sciences study ultimate causes which acted in the past and hypotheses regarding ultimate causes can be tested through use of postdiction and observation. Prediction and postdiction The testing of hypotheses requires the making of predictions which are then checked empirically. Hypotheses dealing with proximate causes can produce predictions (in the literal sense) by stating the future result of an experimental manipulation in either the field or the lab. For example, if factor A is believed to be an important component in the process which produces phenomenon B, a scientist could predict that if factor A is modified, a specific variation in phenomenon B will occur. The hypothesis that factor A is important in producing phenomenon B can then be tested by performing the experiment to see if the predicted result will occur. Obviously, literal predictions cannot be used to test hypotheses regarding the past because scientists cannot experiment with the past. Historical scientists test their hypotheses by making postdictions, i.e., statements about events which have already happened, rather than predictions. But note that postdictions constitute a type of scientific prediction because a scientific prediction is defined as a statement not known to be true at the time the statement is made, but which must be true if the hypothesis is true. Thus, a scientific prediction can be either a prediction (in the literal sense) or a post- diction since in either case the statement is about something not known to be true when the statement is made. Postdictions are checked by observation rather than experimentation. For example, suppose that a paleontologist observes in successive sedimentary strata two closely related fossil species. The one (species B) in the higher layer possesses several characters which are advanced over those found in the other (species A). The paleontologist might hypothesize that the primitive species (species A) evolved gradually into species B and so predict (= postdict) that in strata between species A and B there must exist a third species which is structurally intermediate between the two. If discovered, this intermediary species would support the hypothesis. Hence, postdictions are scientific predictions used to test historical hypotheses, and they are checked through observation rather than experimentation. Darwin was well aware that his theory of evolution could not be tested experimentally and so offered a different criterion for testing his theory. Darwin drew upon Whewell's concept of a consilience of inductions as a test of historical hypotheses, i.e., to what extent can the hypothesis or theory explain a large amount of information obtained from different, often unrelated, fields of inquiry? This information, when explained satisfactorily by the theory, feeds back to support the theory in an inductive argument - the greater the explanatory power of the theory, the more likely it is to be true, or at least useful in explaining reality (see previous discussion of scientific theories). With the methodology described above in mind, let us now take a closer look at evolution in terms of patterns which can be interpreted through the paradigm of evolution. The phenomena described under these headings constitute an argument for evolution because together they form a "consilience of inductions." Each topic by itself has weaknesses as evidence for evolution, but taken together they form a chain of diverse links which argues persuasively for evolution as the reason for species origin and against the creationist explanation.