Science is best defined as a method for attempting to gain knowledge about the natural world that entails making testable predictions, making observations and performing experiments to test them, forming provisional explanations based on those observations and experiments, and continually testing those provisional explanations through further observations and experiments.

As the philosopher Karl Popper articulated in The Logic of Scientific Discovery and elsewhere, science functions most effectively when data or observations are gathered for the purpose of contradicting rather than for the purpose of bolstering proposed explanations of natural phenomena.[1] In this understanding of science, every proposed explanation of the natural world is taken as provisionally valid and the ultimate purpose of performing experiments and making observations is to look for examples that falsify one’s proposed explanations, not to look for examples that confirm them. A proposed explanation of natural phenomena must be capable of being falsified by further experiment or observation to be scientific under these criteria. Now, to claim that all science is always carried out based on the standard of falsifiability would be simplistic and probably even naïve. But falsifiability is an ideal demarcation of genuine science from every other attempt to gain knowledge. Without this demarcation, science would lose much of its rigor.

Because all scientific propositions are synthetic propositions, the three elements necessary for a synthetic proposition to be apparently justified and apparently true—(1) positive evidence, (2) consistency with already held apparently justified apparently true beliefs, and (3) intersubjective confirmation—are also necessary for scientific propositions to be apparently justified and apparently true.

As in any other field, positive evidence is necessary for scientific claims to be valid. In science, positive evidence can consist of either direct sensory observations of natural phenomena or observations of the measurements and readings of machines and instruments. Normally, such observations are only made after a hypothesis has been formed. Hypotheses are often based on mathematics or deductive logic rather than observation, but they can only become provisional explanations, and belief in such explanations can only become apparently justified and apparently true, if repeated observations which do not falsify those explanations are made.

Any purported scientific claim must also be consistent with already held apparently justified apparently true beliefs about the laws and attributes of existence. Ordinarily, this means a scientific claim must fit within whatever paradigm (current set of norms, ideas, and practices) exists in a given area of scientific inquiry.[2] However, if a purported scientific claim is not consistent with the current paradigm, this does not necessarily mean the claim is invalid. In some cases, it simply means that the current model needs to be tweaked or even replaced with a better one. At times, modifications in the prevailing paradigm are minor and do not constitute fundamental theoretical shifts, as when genetics became part of the basic framework of the theory of evolution under the influence of Mendel and others. In these situations, the basic framework stays intact, but a significant new element is added. In contrast, paradigm shifts are major changes in prevailing theory that occur when mounting data and observations require that the prevailing scientific consensus theory be modified, specifically when the old paradigm cannot adequately account for new data and observations. For example, a paradigm shift occurred when Einstein’s theories of relativity and multiple physicists’ development of quantum mechanics provided a more complete picture of the physical universe than classical mechanics. Even though classical mechanics remains valid as a highly accurate approximation of the physical laws that prevail in circumstances other than those involving very strong gravitational fields, very fast speeds, and very small objects, it is still true that with relativity and quantum mechanics, our fundamental understanding of the basic laws of the physical universe was completely altered. Of course, people had known about the existence of gravity and electromagnetic radiation prior to the development of relativity and quantum mechanics, but the new theories provided a much more complete understanding of these phenomena, and the latter eventually led to awareness of two other fundamental forces, the strong and weak nuclear forces. But all these cases have one thing in common: in all of them, the scientists involved made these claims because they believed them to be consistent with the apparent laws and attributes of existence.

In science, intersubjective confirmation is called peer review. This means that other experts, or people qualified by education and experience to comment in each field of study because of their awareness of the methodologies, results of research, and common interpretations or explanations of the evidence in that field, provide feedback on the methods, evidence, and proposed explanations in that field. An indispensable element of peer review is open discussion of contrary viewpoints by qualified experts about the same evidence using reason and appeals to that evidence. If the voices of some qualified experts are dismissed or stifled because they do not agree with popular or official/government-sanctioned interpretations of the evidence, this undermines the process of peer review. Of course, as with any form of intersubjective confirmation, scientists who engage in peer review can have biases and personal idiosyncrasies that prevent them from being fully reliable for the purpose of confirming or disconfirming other people’s experiences. However, without peer review, there would be far more false positives and false negatives and far more ideas that are inconsistent with current understanding, which would undermine the rigor that makes science as reliable as it is.

Despite its rigor and the truly dazzling technological breakthroughs it has helped foster in the modern era—e.g., electrical lighting, vaccines and cures for many diseases, and improvements in the efficiency of food production, as well as more harmful technologies like weapons of mass destruction—the scientific method is not the only means of arriving at apparently justified apparently true beliefs. While science is the most proven and reliable means for forming beliefs based on experience, this does not make it the only valid means. Indeed, most everyday apparently justified apparently true beliefs are not based on the scientific method. Many are not falsifiable at all. Many experiences cannot, by their very nature, lead to proposed laws and principles that are publicly testable or testable in a controlled and systematic way. Virtually all wholly private experiences fall into this category. Such experiences are not suited for data collection/observations by multiple individuals operating under a common set of principles and assumptions, and thus cannot be intersubjectively confirmed. This does not, however, mean that such experiences do not reflect some aspect of the world or of human consciousness that is consistent from person to person and era to era.

[1] Popper, Karl. The Logic of Scientific Discovery. 2nd English ed. New York: Routledge, 2002.

[2] See Thomas Kuhn’s 1962 book The Structure of Scientific Revolutions.