Here, for example, is Dr. Hawking about those troublesome extra dimensions required by string theory but apparently unavailable for parking cars. "I must say that personally, I have been reluctant to believe in extra dimensions," he writes on Page 54 of the new book. "But as I am a positivist, the question `Do extra dimensions really exist?' has no meaning. All one can ask is whether mathematical models with extra dimensions provide a good description of the universe."
In other words, if the experiments come out right, it doesn't matter. This could be considered jarring if you cling to the notion that science is the search for a reality that is deeper than the measurements on a laboratory table. But, quantum theory and relativity have taught us, science is about what can be observed and measured or it is about nothing at all. In science, as in democracy, there is no hidden secret knowledge, all that counts is on the table, observable and falsifiable. All else is metaphysics.
Interesting that in a discussion including quantum theory, with its uncertainty principle, that 'science is about what can be observed and measured'.
And if science is to be limited to what is 'observable and falsifiable', what happens to inference? In fact, the quote from Hawking is based on the opposite point. Existence of extra dimensions is meaningless as long as the mathematical models work. That's a pure argument for the value of inference. Unless, of course, you hold that whatever works mathematically in one model is the single possible correct answer. Actually, I think, hypotheses need experimental validation. The experimental results can sometimes be considered validation only by using another layer of inference.
Any discussion about the evolutionary relationships of dinosaurs is inference about inaccessible, unrepeatable events. Alternatives must be considered possible, if less likely under selected criteria.
Anyone want to proffer a definition of science that puts inference and observation in their correct places?