Main Points in Chalmers, Chapter 3: Experiment In scientific research, which facts are relevant depends on the current state of development of the field of science one is working in. Acquiring the relevant facts generally requires experimental intervention. Therefore, if science is based on facts, then those facts must largely be in the form of experimental results. Experimental results are often very difficult and time-consuming to obtain. Obtaining experimental results is usually not simply a matter of straightforwardly reporting what one’s senses tell him/her. Determining whether certain experimental results are adequate may be a difficult problem requiring the scientist to judge whether a complicated experimental set-up is appropriate and to eliminate extraneous factors that could invalidate those results. Identifying possible extraneous factors and eliminating them from the experiment often means relying on some scientific theory which may turn out to be faulty. (e.g., Hertz’s experiment regarding cathode rays). Therefore, all experimental results are subject to revision. It is not enough for experimental results to be accurate. They must also be appropriate or significant for the matter being investigated. Experimental results are sometimes replaced or rejected for reasons having nothing to do with problems of human perception (e.g., errors in the experimental setup). The acceptability of experimental results is often theory-dependent (e.g., radio waves and the ether, 19th century measurement of molecular weights). Two Concerns about the Role of Experiment in Science: 1. If experimental results are always fallible and revisable, then any scientific theories or laws based on them will also be fallible and revisable. 2. If experimental results are theory-dependent, then arguments for scientific theories based on those experimental results are circular. Chalmers’s response: 1. It is possible to design experiments in which the scientific theories being tested are not the same as the theories being assumed in designing and running the experiments. 2. Although some scientific theories may be assumed in designing and conducting an experiment, the outcome of the experiment depends on the way the world is and not on which theories were assumed.