Scientific Method

Are there methods of science? And if so, methods for doing what? Most would agree that there are particular methods for observing in each of the sciences, from those employed by, say, an ethologist in observing chimpanzees in Gombe Park in Kenya to those employed by astronomers in observing the positions of celestial objects. There are methods for carrying out experiments, from properly conducting surveys to following the protocols of randomized controlled trials. There are also methods for applying mathematics to some science. And there are methods for reasoning (deductive, inductive, etc.), of the sort we employ in everyday life as well as in science, though science has methods of reasoning that are specific to it.

The history of science has commonly been accompanied by a philosophy of science, an important part of which is a theory of method—the starting point being Aristotle's works. In modern times, Francis Bacon provided in his Novum Organum (1620; its full title is Novum Organum Scientiarum) some principles of eliminative induction, René Descartes produced his Rules for the Direction of the Mind (written in Latin, probably between 1619 and 1628, and published posthumously), and Isaac Newton included in his Principia (1687) four “Rules of Reasoning in Philosophy.”

Scientific method is commonly taken to pertain to hypotheses (or theories or, more generally, the claims of some body of belief), and this will be the main focus of this entry.

Are there methods for inventing hypotheses (the context of discovery)? Are there methods for [Page 832]testing them (the context of justification)? If so, are these definitional of science? Some deny that there is such a thing as an autonomous methodology for science. The deniers include sociologists of scientific knowledge (see, e.g., the so-called Strong Program propounded by David Bloor, one of the contributors to this encyclopedia); those such as Michael Polanyi, who thought that scientific methods are tacit insights that do not have an explicit formulation; and, finally, Paul Feyerabend, who famously claimed in his book Against Method that “anything goes.” This is a position that the later Feyerabend qualified considerably when he said that each move in the game of science may be accounted for by some rule of method, but there may be no overall theory of method of the sort espoused by “rationalists” such as Karl Popper—that is, a view of method that fits all the significant moves that have been made in all the sciences.