Philosopy of Science seeks to understand the nature and justification of scientific knowledge. Since it is difficult to distinguish science from non-science, there are legitimate arguments about the boundaries between science and non-science. This is known as the problem of demarcation. There is however, a set of core precepts that have broad consensus among philosophers of science and within the scientific community on what constitutes scientific knowledge. For example, it is generally agreed that scientific hypotheses and theories must be capable of being independently tested and verified by other scientists in order to become accepted by the scientific community.
There are different schools of thought in philosophy of science. The most popular position is empiricism, which claims that knowledge is created by a process involving observation and that scientific theories are the result of generalizations from such observations. Empiricism generally encompasses inductivism, a position that tries to explain the way general theories can be justified by the finite number of observations humans can make and the hence finite amount of empirical evidence available to confirm scientific theories. This is necessary because the number of predictions those theories make is infinite, which means that they cannot be known from the finite amount of evidence using deductive logic only. Many versions of empiricism exist, with the predominant ones being bayesianism and the hypothetico-deductive method.
Empiricism has stood in contrast to rationalism, the position originally associated with Descartes, which holds that knowledge is created by the human intellect, not by observation. A significant twentieth century version of rationalism is critical rationalism, first defined by Austrian-British philosopher Karl Popper. Popper rejected the way that empiricism describes the connection between theory and observation. He claimed that theories are not generated by observation, but that observation is made in the light of theories and that the only way a theory can be affected by observation is when it comes in conflict with it. Popper proposed falsifiability as the landmark of scientific theories, and falsification as the empirical method to replace verifiability and induction by purely deductive notions. Popper further claimed that there is only one universal method in science, and that this method is not specific to science: The negative method of criticism, trial and error. It covers all products of the human mind, including science, mathematics, philosophy, and art.
Another approach, instrumentalism, colloquially termed "shut up and calculate", emphasizes the utility of theories as instruments for explaining and predicting phenomena. It claims that scientific theories are black boxes with only their input (initial conditions) and output (predictions) being relevant. Consequences, notions and logical structure of the theories are claimed to be something that should simply be ignored and that scientists shouldn't make a fuss about.
Finally, another approach often cited in debates of scientific skepticism against controversial movements like creationism, is methodological naturalism. Its main point is that a difference between natural and supernatural explanations should be made, and that science should be restricted methodologically to natural explanations. That the restriction is merely methodological (rather than ontological) means that science should not consider supernatural explanations itself, but should not claim them to be wrong either. Instead, supernatural explanations should be left a matter of personal belief outside the scope of science. Methodological naturalism maintains that proper science requires strict adherence to empirical study and independent verification as a process for properly developing and evaluating explanations for observable phenomena. The absence of these standards, arguments from authority, biased observational studies and other common fallacies are frequently cited by supporters of methodological naturalism as criteria for the dubious claims they criticize not to be true science.