I honestly would define science as critically interpreting data. Experiments are just an optional means to more data. Observation of natural data without experimentation is still science. It's one of the reasons why I hate the term "scientific method." As if that method is the definition of science, while our most reliable science isn't a product of it. Experimentation can only yield theories, while logic can derive laws. Literally the best experimentation can ever do is create a lower class of science.