….The historical sciences are related to other sciences, and all sciences share the same basic goals and methods….
Scientific Methods in Historical Sciences.
For most scientists, the main goal of science is to find truth. They want to construct theories that are true, that correspond with reality by correctly describing what really happens in nature. In a search for true theories, the main thinking tools — the generation and evaluation of theories, using observation, imagination, and logic — are described in an introduction to the simple logic of scientific method and in this brief paraphrased-summary of a central idea:
As shown in the diagram below, OBSERVATIONS (from physical experiments) are used to imaginatively generate a THEORY, which can be used with if-then logic (in a mental experiment) to make PREDICTIONS, so you can do a REALITY CHECK by comparing observations with predictions, to test whether “the way you think the world is” (in a prediction made by assuming your theory is true) corresponds to “the way the world really is” (in a corresponding observation of reality).”
Is there a scientific method? If “method” means “a single method, used in the same way by all scientists at all times,” the answer is NO, so we should not talk about The Scientific Method.Are there scientific methods? Yes. The main methods of scientific thinking, including the foundation of science — the “reality checks” made by observing reality and using logic — are used by all scientists. But details change with time and culture, and vary from one area of science to another, and from one scientist to another.
Does historical science produce reliable conclusions? This is not a useful question, because it tends to generate a yes-or-no response claiming that “YES, [all] historical science [always] produces [totally] reliable conclusions” or “NO, [all] historical science [always] produces [totally] unreliable conclusions.” Each of these extreme generalizations is wrong, because each implies the “all or none” claims inside the brackets
If we want to avoid these wrong answers, we should avoid asking a general question that encourages simplistic “all or none” thinking. Instead, we should ask specific questions about particular historical situations and claims, to encourage sophisticated “degrees of reliability” thinking. We should carefully examine the evidence-and-logic for a particular situation, and try to determine the scientifically justifiable level of confidence in the reliability of a particular claim about that situation.an application: In an effort to avoid false generalizations, in the sections below I claim only that a historical science can produce reliable conclusions [in most situations for most claims], so I think historical science usually produces reliable conclusions. I defend this claim because I think it is true, but I won’t defend a claim that seems false, a claim that every historical science always will produce reliable conclusions [in every situation for every claim]. This page closes by looking at variations in the “justifiable level of confidence” for different situations and claims.
Operation Science and Historical Science
Yes, historical science can produce reliable conclusions.
Earlier, I say that scientific methods “vary from one area of science to another.” Some variations in methods are due to differences between operation science (to study the current operation of nature, what is happening now) and historical science (to study the previous history of nature, what happened in the past). Both types of science are similar in most important ways, especially in their use of scientific logic, but there are minor differences./ Although some young-earth creationists try to contrast historical origins science with experimental empirical science (i.e., science based on observations), this is wrong because historical origins science is based on observations so it is empirical.
Although repeatable controlled experiments can be done in operation science, this is not possible for historical events. Sometimes, the limitations of historical data provide a reason for caution about conclusions. But this challenge has inspired scientists to develop methods that reduce the practical impact of data limitations, and historical sciences — in fields such as astronomy, radiometric physics, and geology — are authentically scientific.
In historical science, one way to “reduce the practical impact” is to use repeatable uncontrolled experiments to gather data. For example, other pages explain how observations of many Cepheid stars from many parts of the universe have shown that all Cepheids have similar properties, allowing them (and supernovas, which have their own consistencies) to be useful for measuring astronomical distances. These consistencies let scientists develop reliable descriptive theories, which can become explanatory theories that usually are related to (and are consistent with) explanatory theories in operation science.
Experimental Science and Observational Science
This is another useful distinction, overlapping with and similar to “operation science and historical science” in some ways, but not others.
In all historical sciences, scientists observe systems they cannot control, since they can only observe what happened the way it already happened. But in astronomy and geology, for example, they can decide “how far back in the past” to observe, by looking further out in space or deeper down in the geological column.
In non-historical sciences, scientists can do either experimental science or observational science by observing systems that are controlled or uncontrolled, respectively. In fact, in many field studies (in ecology, oceanography, anthropology,…) the goal is to observe a system “as it is naturally” with no interference due to actions by the scientists. The degree of control over a system can range from no control through partial control to total control, but for any system the result (what happens in the system) depends on nature. And when observing a set of uncontrolled systems, scientists can achieve some benefits of controlled experiments by using multi-variable statistical analysis of the systems-and-observations.
When scientists decide how they will make observations, when they decide what to observe and how, their goal can be to fill gaps in current knowledge or confirm previous observations, to resolve anomaly, distinguish between competitive theories, provide support for an argument, or impress a funding agency.
To be more efficient (so they’ll use less of their valuable time and resources) or for systems that cannot be physically observed, scientists can run thought experiments by asking “if we do this, what might happen and what would we learn?”
An important practical aspect of science is experimental design, which is introduced as a foundation for the simplicity of science and is analyzed in my overview of scientific method and detailed examination.
Prediction & Retroduction, Mechanism & Agency. Prediction and Retroduction: Because theory-based inferences are usually called predictions, the non-scientific meaning of “prediction” (focusing on the “pre”) can lead to the mistaken impression that in science a logical inference must be made before an event occurs. But in historical science the timing of prediction is not a cause for concern, because a scientific prediction (which is a logical if-then inference where one if-condition is to assume the theory is true: if this theory and if this system, then we will observe _) can be logically valid when it’s made after an event has occurred. In fact, a theory-based inference can be logically valid even after observations are known, in a logical process of retroduction where the goal is to find a theory whose logical if-then predictions will match the known observations. In historical science, the main goal is to describe (and maybe explain) what did happen, not to predict what will happen. But some predictions are possible in some historical situations, when these predictions are allowed by the types and amounts of contingency operating, and our knowledge about the contingencies.
Mechanism and Agency: In some historical situations, only undirected natural process is involved, so a mechanistic explanatory theory can provide an adequate description and explanation. In other situations, “what happens” depends on the decisions and actions of an agent. This introduces an element of unpredictability, but a historical detective using scientific reasoning (in forensics, psychology, sociology, anthropology, archaeology, geology, astronomy, or history) only has to determine what did occur, not predict what will occur, in a descriptive theory. And in an agency explanatory theory, proposing that “agent action was involved” is the scientific conclusion of a historical detective.
Unobservable Causes of Observable Effects. Can scientists logically infer the existence of things they cannot observe? Yes, if an unobservable cause produces observable effects. This cause-and-effect principle is used in operation science. For example, even though electrons and ideas cannot be observed, modern theories propose electrons (in chemistry) and ideas (in psychology). Why? Because our observations are explained in the most satisfactory way — using “if… then…” hypothetico-deductive logic asking “if this theory is true, then what should we expect to see?” — by theories proposing the existence of unobservable causes (electrons and ideas) that produce the effects we observe.
Similarly, in historical science we can logically infer the existence of causes we did not observe, if these unobserved causes produced effects we can observe. Therefore, when skeptics ask “Were you there? Did you see it?”, they are ignoring the principle that scientific logic depends mainly on observable effects, not observable causes. Because of this principle, even if an event or process was not directly observed, a plausible scientific theory can propose that the event or process did occur.
Miracles in Operation Science and Historical Science. In operation science, we have no reason to suspect that God is miraculously changing the results of everyday experiments. But even if there were occasional miracles, the scientific practice of demanding reproducible results (and being able to test the reproducibility by repeating the same experiment) would negate the effects of occasional miracles, so the ultimate conclusions of operation science would not be affected.
In historical science, an occasional miracle could have significant consequences in history. These consequences could be important when we’re trying to develop an accurate historical science that finds the truth of what actually happened in history. For example, IF life would not emerge from nonlife by natural chemical evolution but God wanted life so He miraculously created it and then let it evolve by neo-Darwinian mechanisms, a science that ignores this possibility (because it is restricted by methodological naturalism) would not be able to find the truth.
An Invitation to Examine the EvidenceEarlier in the page, I say: “Sometimes the limitations of historical data provide a reason for caution about conclusions.”But we usually have reasons to be confident about the conclusions of historical science. Therefore, “we should carefully examine the evidence-and-logic for a particular situation, and try to determine the scientifically justifiable level of confidence in the reliability of a particular claim about that situation.” When doing this we should also consider “the big picture” and the logical principle of multiple independent confirmations when the evidence-and-logic in a wide variety of situations leads to the same conclusion, and this occurs when we examine the abundant evidence for an old earth and old universe.
Postmodern radical relativists — ranging from young-earth creationists who are selective postmodern skeptics (they challenge the credibility of historical science but not operation science) to comprehensive postmodern skeptics (who challenge the credibility of all science and most other scholarship) claim that in historical science the evidence is almost always inadequate, so the conclusions of scientists must be determined by their nonscientific beliefs. Most scholars, including myself and other members of ASA, think these relativists are exaggerating the logical difficulties, making mountains out of molehills, and historical sciences have a solid foundation — the logical evaluation of empirical evidence. That provides a reliable way to learn about the history of nature and the fascinating world created by God. We encourage you to explore ORIGIN EVIDENCE where you can carefully examine a wide range of historical situations and scientific claims, so you can determine the levels of confidence that you think each claim deserves when it is evaluated based on the evidence-and-logic of science.