Notes on Good Scientific Practice

The Standards of Proof and the Scientific Method

In the sciences, we sometimes offhandedly talk about how we strive to 'prove' hypotheses. But what kind of proof is it that we are talking about? Is it like a mathematical proof or logical proof? Surely not. We do postulate assumptions, we use rules of reasoning and logic, and we produce conclusions; yet, we rarely seem to produce the kinds of proofs that look like a clean, 'P therefore Q. P. Thus Q', or even ones that look like the more complex logical proof pictured here on the right. Even when we do produce something of this logical form in science, we unavoidably have implicit assumptions, background supporting evidence, and a wider network of related hypotheses and arguments to fit with.

The standards of proof used in science are obviously different than the kind of standards generally accepted in mathematical proofs or logical proofs. Consider even an overly simplified sketch of a scientific method. We present hypotheses and predictions and test them with experiments and observations. We check to see if not only our conclusions, but if the whole argument, the evidence, the natural laws appealed to, and our methods of reasoning fit our best and most comprehensive understandings of the world. We publish, our work gets peer-reviewed, we respond to criticisms, we tweak our work to improve it, or are forced to abandon it if we fail to answer the criticisms. 

Thus far, the above method and standards of proof in science is obviously not as tidy as those of our "P therefore Q" proof. And it gets more complex still as we consider the differences between scientific disciplines. We may stop and ask questions of each discipline, and even of each scientific proposal, "What is it that we are testing when when we set up our experiments? The hypothesis or its assumptions?" or "How much supporting evidence, demonstration and observation is enough?" and even "What wider field or fields must this work fit?". There is no obvious correct answer to any of these questions, and it is probably the case that the "correct" answer is the one which the discipline's community reasonably agrees on. 

Dr. Benner, in his 2009 book, Life, the Universe and the Scientific Method, brings these sorts of questions into focus when assessing the relatively new scientific fields. The fields of study of paleogeneticssynthetic biology and prebiotic chemistry face unique challenges because they are often working with historical hypotheses; suddenly, a good scientific hypothesis does not seem so obviously falsifiable or testable. Furthermore, exobiology, exoplanatology and the search for the physical bases of consciousness are all fields in which the standards of proof are still in the process of being set by their scientific communities.  Finally, the interdisciplinary approach itself faces the further challenge of negotiating different standards of different scientific communities.

We might even realize that these challenges in falsifiability and testability exist in all fields of science despite agreement on methods and Standards of Proof. We offer this preliminary discussion on these issues as something to keep in mind when reading about the work on these new frontiers. Finally, although these sorts of questions concerning method and proof are still live in all scientific disciplines, this does not, and should not bar the way for their advancement. Broadly speaking, we might agree that looking for and considering carefully corroborating evidence and supporting claims has been and will be the key to progress. Thus, we want our readers to be good critical thinkers; when exploring our website, we want you to be asking yourselves what would constitute satisfactory methods and Standards of Proof.