The demarcation problem is the philosophical problem of determining what types of hypotheses should be considered scientific and what types should be considered pseudoscientific or non-scientific. It also concerns itself with the ongoing struggle between science and religion, in particular the question about which elements of religious doctrine can and should be subjected to scientific scrutiny. This is one of the central topics of the philosophy of science, and it has never been fully resolved. In general, though, a hypothesis must be falsifiable, parsimonious, consistent, and reproducible to be scientific.
 Methodological naturalism
The assumption of methodological naturalism is arguably the most basic and important foundation of the scientific method. While it does not explicitly reject the existence of the supernatural - a position associated with philosophical naturalism - it limits the applicability of science to the natural world and the observable laws shaping it. In a nutshell, this means that an explanation that has to resort to a supernatural cause, like Intelligent design, must be considered unscientific.
The Viennese philosophers who introduced the positivist paradigm effectively laid the groundwork for the modern philosophy of science and one of its most important strands of thought. The early Positivists favored a rather strict approach to the demarcation and strongly affirmed the empirical nature of science, meaning that questions that cannot be empirically verified or falsified are irrelevant to scientific thought. This obviously set science in stark contrast to religion, but also to the philosophical schools in the vein of classical rationalist tradition that emphasized pure thought.
 Karl Popper
In his book The Logic of Scientific Discovery, Karl Popper proposed the idea that scientific hypotheses must be falsifiable; unfalsifiable hypotheses should be considered pseudoscience. Popper's emphasis on falsifiability changed the way scientists viewed the demarcation problem, and his impact on philosophy of science was enormous.
 Later attempts to resolve the demarcation problem
 Kuhnian paradigms
Thomas Kuhn coined the concept of a paradigm shift, in which scientists come to a consensus on certain relevant assumptions, facts, or methodologies. Any research done in a pre-paradigm state is considered non-scientific.
 Lakatos' research programs
Imre Lakatos combined elements of Popper and Kuhn's philosophies with his concept of research programs. Programs that succeed at predicting novel facts are scientific, while ones that fail ultimately lapse into pseudoscience.
The concept of Non-overlapping magisteria is a relatively recent attempt at proposing a clear demarcation between science and religion. It explicitly restricts science to its naturalistic foundations, meaning that no conclusions about supernatural phenomena like gods may be drawn from within the confines of science. This idea has come under heavy criticism for ignoring the blatantly irrational nature of modern-day fundamentalism, and adherents of this theological orientation have unfortunately not paid the same respect to science.
 Rejection of the problem
Some philosophers have rejected the idea of the demarcation problem, such as Larry Laudan. Others like Susan Haack, while not rejecting the problem wholesale, argue that a misleading emphasis has been placed on the problem that results in getting bogged down in arguments over definitions rather than evidence.
 See also
- Massimo Pigliucci: Nonsense on Stilts: How to Tell Science from Bunk. Chicago: University of Chicago Press, 2010.
- Demarcation: Popper, Kuhn, and Lakatos
- Science and Pseudoscience, Stanford Encyclopedia of Philosophy