Scientific knowledge is based on verifiable evidence. By evidence we mean concrete factual observations which other observers can see, weigh, measure, count, or check for accuracy. We may think the definition too obvious to mention most of us have some awareness of the scientific method. Yet only a few centuries ago medieval scholars held long debates on how many teeth a horse had, without bothering to look into a horse's mouth to count them. At this point we raise the troublesome methodological question, 'What is a fact?' While the word looks deceptively simple, it is not easy to distinguish a fact from a widely shared illusion. Suppose we define a fact as a descriptive Statement upon which all qualified observers are in agreement. By this definition, medieval ghosts were a fact, since all medieval observers agreed that ghosts were real. There is, therefore, no way to be sure that a fact is an accurate description and not a mistaken impression. Research would be easier if facts were dependable, unshakable certainties. Since they are not, the best we can do is to recognize that a fact is a descriptive statement of reality which scientists, after careful examination and cross-checking, agree in believing to be accurate. Since science is based on verifiable evidence, science can deal only with questions about which verifiable evidence can be found. Questions like 'Is there a God?' 'What is the purpose and destiny of man?' or 'What makes a thing beautiful?' are not scientific questions because they can not be treated factually. Such questions may be terribly important, but the scientific method has not tools for handling them. Scientists can study human beliefs about God, or man's destiny, or beauty, or anything else, and they may study the personal and social consequences of such beliefs but these are studies of human behavior, with no attempt to settle the truth or error of the beliefs themselves. Science then does not have answers for everything, and many important questions are not scientific questions. The scientific method is our most reliable source of factual knowledge about human behavior. and the natural universe, but science with its dependence upon verifiable factual evidence cannot answer questions about value, or esthetics, or purpose and ultimate meaning, or supernatural phenomena. Answers to such questions must be sought in philosophy, metaphysics, or religion. Each scientific conclusion represents the most reasonable interpretation of all the available evidence—but new evidence may appear tomorrow. Therefore science has no absolute truths. An absolute truth is one which will hold true for all times, places, or circumstances. All scientific truth is tentative, subject to revision in the light of new evidence. Some scientific conclusions (e.g., that the earth is a spheroid or that innate drives are culturally conditioned) are based upon such a large and consistent body of evidence that scientists doubt that they will ever be overturned by new evidence. Yet the scientific method requires that all conclusions be open to reexamination whenever new evidence is found to challenge them. The central idea of the passage is