It is said there is a scientific method, i.e., a pattern of the way scientific researchers go about their work. There are diagrams of the process, much like the one shown above. The strange thing is that I was never taught this method in any science class through my Ph.D. Such diagrams may represent how concepts progress as science advances, but they do not represent how discovery actually comes about.

One problem is that it suggests that if one just follows the diagram, new science will be created. But there is no checklist of steps that scientists use and that will automatically produce scientific knowledge. I don’t know any scientist who consults a picture or postulation of the scientific method. I’ve not used it or instructed my graduate students in its application.

Another problem with the method diagram is that it appears to be an infinite loop. One can enter the loop at any point, but frequently, it begins with an observation of some behavior that raises a question in the investigator’s mind. This leads to some hypothesis about the phenomenon and some prediction about how it will respond when tested. The experiment produces more data, which leads to a refinement or revision of the original hypothesis, and so on.

Each time around, the data become more exact and/or more relevant and the theory becomes more refined. But we look in vain for any final outcome. The only product shown in this diagram is a theory which will be revised or replaced on each circumlocution of the circle. Yet we know there are discoveries that have endured, formulas that will always work within their limits, and we’ve also seen how the experiment step in the cycle provides a basis for that fact.

If a model is needed, I propose the one above. It recognizes the law and explanation aspects of a theory and the different loop of steps involved in their developments. A figure-8 pattern between the two loops, with empirical observation in the center, shows how the lobes interact in practice. Both sides of this process have products that are new models or rationalizations of accepted laws or clearer, more precise, or more broadly applicable statements of existing ones. And finally, the steps have verbs rather than just nouns, implying that humans are involved in conceiving each step.

Scientific knowledge often proceeds from the particular to the general. We interpret a specific event, then seek to generalize our interpretation of it to other occasions of that phenomenon. But it can also be the other way around. A phenomenon can be empirically characterized, and a law established before being satisfactorily understood.

The central place of experimentation and observation in this diagram of the scientific method underscores the degree to which scientists care about empirical evidence and will change their explanations in the light of fresh evidence. And it is just this insistence on verification through observation that is often cited as the basis for credibility in scientific results.

It Starts with a Question

If a model for investigation is not included in a scientist’s curriculum, what is taught about the process in graduate school? Each mentor has a different approach which their students carry on and modify. What I learned is that one does not investigate a subject, but rather one seeks to answer a question. For my thesis work, it was why platinum electrodes did not work as well as mercury droplets for electroanalysis. Newton wondered why the rate of fall of heavy objects did not depend on their mass. Kekule was stuck trying to work out a structure for the molecule, benzene C6H6 for which no linear or branched arrangement would work.

Reasonable research questions can range in their ultimate significance from a predictable extension of a known process to a transformative new concept. “Will an increase in temperature speed up the process?” or, “Why does the photoemission of electrons depend on the color of the impinging light rather than its intensity?” An answer to the former question may help improve a specific synthesis or analysis. The answer to the latter question transformed our understanding of light and won Einstein the Nobel prize.

Recognizing and formulating questions whose answers will be the most exciting or pertinent is a key factor in successful research. And, as Edelson[i] suggests, “Can you ask the question in such a way as to facilitate the answer?” Decisions regarding the means of observation, the set-up of the experiments, and the distillation of the raw data into the form of a formula all depend on the creativity, knowledge, and experience of the scientist(s) involved.

Thomas Kuhn,[ii] a historian and philosopher of science who began his career as a physicist, refers to areas of scientific knowledge as paradigms. He distinguishes research that adds to existing paradigms from that which creates new ones. Questions that are logical extensions of understood phenomena are important. And there is often great creativity in the methods employed in finding the answers. In fact, most research falls into this category. It can be very valuable work, bringing new concepts to the point of usefulness.

Bold, “outside the box” questions that lead to new paradigms are riskier to pursue, but when they succeed, they have the greatest effect on how science is understood and used. Therefore, recognizing and formulating questions whose answers will be the most exciting or pertinent is a key factor in determining scientific success. This is true whether one is extending a paradigm or creating a new one.

Conceiving exceptional questions is more than a skill; it requires curiosity, knowledge, insight, and imagination. A list to which some might add “luck”. Oliver Sacks[iii] said it beautifully. “Science sometimes sees itself as impersonal, as “pure thought,” independent of its historical and human origins. It is often taught as if this were the case. But science is a human enterprise through and through.”

[i] 1994 interview in The New Yorker

[ii] Kuhn, Thomas S., The Structure of Scientific Revolutions, University of Chicago Press, 3rd edition. 1996. First published in 1962.

[iii] Oliver Sacks, Everything in its Place, p. 39.