The Scientific Method
It is important to know something about the way in which new theories are produced. To illustrate this process, sometimes called the scientific method, let us follow the thoughts and actions of a scientist who happens upon the box pictured in Figure 4. The box is apparently constructed from some heavy metal and has three rather stiff ropes protruding from the holes labeled A, B, and C. Her curiosity aroused, our scientist attempts to open the box to discover its purpose and inner workings. She soon discovers, however, that she lacks the necessary tools and must be satisfied with observations made from outside the box.
Figure 4. A curious box.
An initial tug on rope C produces no apparent movement of ropes A and B. This observation triggers the thought that the box may contain three unconnected, independent ropes. This mental visualization is our scientist's first hypothesis of the nature of the interior of the box.
Now, she reasons, if this hypothesis is correct, each rope should move to its limit without affecting (moving) any of the other ropes. In order to test her hypothesis, she pulls each rope and observes the effect of that action on the other ropes. Rope C appears to move independently of the other, but the effects of ropes A and B on each other can be expressed as a mathematical law. If x represents the distance traversed by rope A, and y the distance traversed by rope B, the law becomes
x = 2y
Because of the specific interdependency of ropes A and B, the scientist now proposes the model pictured below in Figure 5. In this model, rope A is wound around a drum attached to an axle; rope B is wound around the axle itself. The circumference of the drum is greater than the circumference of the axle (in fact, 2 times as great), and rope A is therefore played out at a greater rate than rope B. Rope C remains unattached and independent.
Figure 5. A model of the interior of the box.
This model provides an explanation for all of the experimental data, and it also permits the formulation of new questions and predictions. For example: Can rope C be withdrawn from the box completely, or is it held inside by a knot at the end? Are the ends of ropes A and B attached to the drum and the axle? If the model is correct, then when rope B is withdrawn to its limit, rope A may not have reached its limit, and (assuming that the ropes are attached) a further pull on rope A may wind rope B back into the box. These questions suggest additional experiments which might never have been conceived without the help of the model.
While our scientist has spent a considerable amount of energy investigating an almost trivial problem, her approach to the problem contains many of the features of the "scientific method": experimental observation that leads to the formulation of a law, a hypothesis that leads to a model or theory, and the subsequent use of the model to design new experiments. Each hypothesis leads to a model, which may be discarded after additional experiments are performed, or, if the experiments are all consistent with the model, the model is retained until contradictory evidence is obtained.
Because of the central role of models, it is important to be cognizant of a number of their characteristics. First, the scientist usually draws on her own experiences in fashioning a theory. In our example, it might be suggested that the ropes are controlled by elves residing in the box, but our scientist has never seen an elf, nor does she believe in the existence of such creatures. On the other hand, she has seen mechanical devices such as winches that employ ropes on drums, and she has seen a spool of thread. Many models, designed to account for the behavior of matter so small that it has never been seen, are based on the behavior of macroscopic bodies, such as billiard balls, which lie within the realm of everyone's experience.
On the other hand, some models are mathematical and abstract in nature. For example, the mathematical nature of the contemporary model of the electron makes many of its features difficult to visualize. Indeed, some scientists feel that the most significant scientific discoveries occur within the realm of mathematics.
It is also important to realize that a given set of experiments and observations can usually be explained by more than one model. Our scientist could have developed a model based on gears rather than drums, and in fact there are a number of alternate models that will satisfactorily account for the behavior of the ropes. As data and observations accumulate, one of a set of equally good models may become more satisfactory than the rest, or the choice of model may be based on considerations of simplicity or symmetry or usefulness.
Finally, the fact that models may not, and very likely do not, correspond to reality cannot be overemphasized. Since the box cannot be opened, the scientist will probably never know if the box really does contain a drum and axle. When the model is intended as a picture or visualization of matter at the sub microscopic, molecular level, the problem is even more acute. Atoms cannot possibly be either billiard balls or mathematical abstractions, nor is it likely that atoms behave like billiard balls. And yet, the billiard ball model of atoms is at the heart of the determination of the structure of the nucleic acids DNA and RNA, the revelation of the genetic code, and all of its biological implications. Thus, while the correspondence between the model and reality may not be very high, the benefits of the model, the development of new experiments, the discovery of new laws of nature, and so forth - may be very great indeed.
Reference: C. Yoder, O. Retterer, M. Thomsen, and K. Hess, Interactive Chemistry, Mosby Year-Book, 1997.