What is the scientific method?  Everyone who sat through grade-school science class knows the answer to this question, right?: observation, hypothesis, prediction, experimentation, conclusion.  What may surprise you is that there is no such thing as the scientific method.  There are a variety of methods scientists employ in their quest to discover the truth about the natural world, none of which can be claimed as the scientific method.  Which method a scientist uses depends on what he is studying.  While the method outlined above works well for “experimental scientists” such as chemists and physicists, it doesn’t apply to “historical scientists” (i.e. those who study the past) such as paleontologists, astronomers, and evolutionary biologists.  Those in the historical sciences require a different method.

Historical scientists study the past, not the present.  They seek to discover the historical causes responsible for past events – effects that we observe in the present.  For such a task the scientific method outlined above simply won’t work.  It’s the wrong tool.  To explain the structure of the fossil record, for example, one cannot engage in experimentation.  Likewise, there is no need for making predictions since predictions address the future, not the past.  How, then, do historical scientists test their theories?

Famed Harvard paleontologist Stephen J. Gould explained that those working in the historical sciences test their theories by means of their explanatory power.  In their search for historical causes they operate according to a principle called the “method of multiple working hypotheses.”  Essentially, they postulate multiple causes that could be responsible for the observed effect, and then compare them to see which cause best explains the data in question.  This sort of approach is also called abductive reasoning, or “inference to the best explanation.”

How do scientists determine what constitutes the “best” explanation?  To establish some cause X as the best explanation one must show (1) that X is capable in principle of explaining the effect in question [causal adequacy], (2) was present and operative at the time the effect was produced [causal existence], and (3) has greater explanatory scope and explanatory power than competing hypotheses.  And if it can be shown that there is only one known cause in operation today that is capable of producing the observed effect (causal uniqueness), then scientists can be reasonably certain that that cause was responsible for producing the effect in question.

Let me illustrate how this process works.  If when I wake up, I look out my window and see that my driveway is wet, there are several possible causal explanations for this historical event: (1) the lawn sprinklers went off in the night; (2) someone washed the car with a garden hose; (3) it rained.  Those are my multiple working hypotheses.  Which can best explain the data?  If I observe that the driveway is wet but not the road, the rain hypothesis is ruled out since we would expect the road to be wet as well.  If I observe that the lawn is also dry, that eliminates the possibility of sprinklers as the cause.  Finally, I observe that the car is shiny clean.  The garden hose hypothesis best explains these observations.  But was that cause present?  Yes, I observe it lying in the driveway next to the car, so the garden hose hypothesis satisfies the causal existence criterion as well.  It follows, then, that the driveway is wet because someone washed the car with the garden hose.

Now that we understand the method historical scientists use to discover past causes, what is the best explanation for the origin of the machinery, processes, and information in the cell?  Most scientists think a valid scientific explanation must appeal to purely naturalistic processes: (1) chance, (2) necessity, (3) or a combination of both.  But what about intelligence?  Could the cause of life be an intelligent agent?  We know information routinely originates from thoughtful, conscious beings, so it seems at least possible that a designing intelligence could be responsible for the specified complexity found in the cell.  If so, intelligent agency should be included as a possible explanation for consideration as well.  Indeed, since the only known source for generating complex specified information is intelligent agents, prima facie a designing intelligence should be considered the best explanation of the data.

Many scientists will dismiss this option out-of-hand on the grounds that science cannot detect the actions of intelligent agents.  But this is not a scientific conclusion; it is a philosophical presupposition (and a false one at that, seeing that branches of science such as archaeology and forensics regularly detect the presence of intelligent agents in the natural world).  Why should a priori philosophical commitments prevent us from considering all possible causes, especially when this philosophical bias excludes what appears to be the best explanation?  If the goal of science is to discover the truth about the physical world, and the evidence points to the activity of an intelligent agent, shouldn’t we be allowed—in principle at least—to conclude that an intelligent agent is responsible?  Either the goal of science is the pursuit of truth, or it is not.  The goal of science should be to find the right answers, not the right kind of answers: those that conform to philosophical naturalism.

Historical science is a search for past causes.  Our search should not be limited to natural causes alone, but should include any and all possible causes.  We should be allowed to follow the evidence to wherever it leads, not only where naturalism allows.  As Greg Koukl has opined, “The object and domain of science should be the physical world, but its goal should be truth, not merely physical explanations.  Though science is restricted to examining physical effects, when causes are inferred, there should be no limitation.”[1]

The only cause in operation today that is known to be capable of producing specified complexity (causal adequacy) is intelligent agents, so if we find specified complex things in the past, we are rational to conclude through abductive reasoning that they have been caused by an intelligent agent.  And since intelligence is the only cause capable of producing specified complexity (causal uniqueness), it follows that an intelligent agent is the best explanation for the origin of biological information.

But wait, how do we know an intelligent agent existed at the times in question?  By logical deduction.  If there is only one known cause of a certain effect, and we observe that effect, we are justified in concluding that the cause was present at the time.  As Meyer writes, “Logically, one can infer the past existence of a cause from its effect, when the cause is known to be necessary to produce the effect in question.  If there are no other known causes—if there is only one known cause—of a given effect, then the presence of the effect points unambiguously back to the (uniquely adequate) cause.”  For example, if only volcanic eruptions can explain certain rock formations, then when we observe those rock formations in the fossil record we are justified in concluding that volanoes must have existed at that time in history.  Likewise, if biological information can only be explained in terms of intelligent agency, then the presence of biological information proves the presence of a designing intelligence in the history of life’s formation on Earth.

Consider the following parable.  Imagine that researchers in Antarctica discover certain markings on a stone, as well paintings of animals in cave buried beneath the ice.  Testing proves that the markings and paintings are 25 million years old, which is much earlier than the arrival of intelligent hominids.  Because we know that such patterns can only be produced by intelligent agents, we would be justified in concluding either that hominids evolved much sooner than we thought, or that there were other intelligent beings prior to hominids, even if we are ignorant of their identity.  The same is true of the cell.  It evidences specified complexity, and thus we would be justified in concluding that life was created by a designing intelligence, even if we do not know the identity of that designer.

Now that we have established the methods by which historical scientists do their research, and the validity of including intelligent agency in the mix of our multiple working hypotheses, we will turn our attention to the possible causes of life, beginning with the naturalistic hypotheses.  Next time….

If you haven’t already done so, read parts 1 and 2 of this series as well.

[1] Greg Koukl, Solid Ground, July/Augusts 2005 issue, page 3.