Few other pathological conditions have aroused, as cancer has, the interest of so many scientific disciplines. Problems related to cancer have become of continuously increasing concern in virtually every field of medicine. In some, such as pathology, they are a major preoccupation. But in sciences other than medicine, cancer also has been receiving increased attention. One of the most urgent activities of synthetic chemistry today is the search for new compounds which might possibly be effective in the control of cancer. Physical chemistry is trying to provide new explanations about the variety of processes present in cancer. Even mathematical studies which recently have offered an interesting application of quantum theory to carcinogenesis, have found new applications in cancer.

With the rapid development of physical sciences, the medical research worker has hoped that from them might come some contribution that could help him ultimately in his difficult task. He also appears to have been anxious to take quick advantage of the progress of other disciplines for another reason, hoping that, through employing their findings and methodology, medicine in general and cancer research in particular, could be promptly changed from the empirical discipline it has been until now into a positive science. He has brought as many applications of other disciplines as possible into his study and this has led to a whole series of new methods of investigation through which interesting new information has been obtained. Yet, most of these applications have been tried chiefly because they have been at the immediate disposal of the scientist rather than because they have represented a missing link in the development of his own ideas.

The outcome has not been rewarding. Medical knowledge appears not to be sufficiently advanced to successfully utilize the avalanche of new, highly specialized information offered by the investigative methods derived from other disciplines. Basic theoretical knowledge in medicine in general, and about cancer in particular, has not yet reached the level necessary to relate and assimilate the new data. To a large extent, basic concepts about pathogenic problems are not even formulated as yet. When the medical scientist has tried to transform the new data into effective therapeutic procedures, he has failed. And the failure has made more evident how much we need basic physiopathological knowledge before we are able to take advantage of detailed data.

Meanwhile, normal development of cancer research has been hindered, side tracked from its logical course. While thousands of scientists with almost unlimited funds at their disposal are presently using the most advanced methods for the acquisition of details, almost no attempts are being made to resolve basic problems, although the cancer investigator is continuously obliged to realize the dearth of fundamental knowledge.

If we attempt to analyze this abnormal situation further, we can find indications that it may have its origin also in a distortion of the proper relationship between the two factors that, together, make for progress in research—ideas and experiment.

The experimental approach provides precise information about particular phenomena under defined conditions. The analytical method tries to investigate reality by recognizing the proper place of the various constituents of a whole, the parts being identified as such by the experimental findings. On the other hand, the conceptual method not only provides an inkling of what the completed whole will eventually look like, but also attempts to predict the properties and the relationship of the component parts.

In dealing with a highly refined and complicated subject, the analytical method by itself appears inadequate. For example, in atomic physics, the results of experiments are expressed by numbers giving the values of certain physical quantities that have been measured. In order to complete the analysis, we must simultaneously determine the numerical values of certain quantities defining the material bodies, the objects of the experiments. This is prohibitive so far as canonical coordinates by Heisenberg's uncertainty principles are concerned. With experimental knowledge somewhat curtailed, theory at present must attempt explanation.

In other areas as well, experiments present only limited numerical values pertaining to some physical quantities. Were we able to measure all quantities, we could analytically reconstruct the entire theme of the physical reality. However, when some quantities cannot be simultaneously determined, this direct reconstruction is not possible and experiments merely give an indirect approach to what we regard as "reality."

If the inadequacy of the analytical approach by itself is evident in the highly positive disciplines, such as in the physical sciences, it is even more so in biology. As Bohr and others have intimated, the conditions of uncertainty seem to be much more pronounced in biology than in physical science. The fact that experiments in biology give only fragmentary and unrelated results is not surprising; the need for a synthetic theoretical method in this field is clear.

In medicine, which is applied biology, the need for the conceptual approach is especially profound. It is true that this approach, as the sole approach, has shown its inherent weakness in the past. There was a time in the development of medicine when available data were so scarce and unreliable, and the need for ideas to provide some sort of guidance was so great, that the worker resorted to broad imagination, using it to replace almost entirely any other form of investigation.

Largely as a reaction to the high proportion of "speculations" prevalent in the early years, the experimental approach in medicine came to be emphasized. Claude Bernard, who almost single handedly was responsible for this, tried to give experimentation its rightful role. However, in ensuing decades, the relationship between theory and experimentation has been progressively distorted. An unrestrained exaggeration of the role of the experiment, the erroneous view that pure facts represent the aim of research, has led to an entirely unbalanced approach. Not only have almost any data obtained by research been considered intrinsically interesting, but obtaining them has become the sole purpose of much research. In scientific papers today, experimental data must be reported as such; any allusion to theoretical meaning is considered undesirable. Generations of scientists have been schooled to believe in the intrinsic value of the experiment. As they have applied this belief to research in biology, and as they have made unlimited use of new methods taken from other disciplines with no ideological requirement for their use, we have had more and more data and fewer ideas. Today, with great astonishment, some scientists are at last beginning to recognize not only that data alone do not generate ideas, but that science cannot progress without theory.

Ideas and experiments are integral parts of all scientific research. A balance between them is needed to assure progress. It must be understood that the function of experimentation is to guide our thinking, to help build up new concepts, and to prove their accuracy in accordance with reality. Certainly, fundamental concepts must not be mere "speculations." They should be accepted only after confirmation through experimentation. Experimentation is the necessary link between mental concept and reality. To the attempts to consider any unresolved fundamental problems in biology, one has to try to bring a rightful balance between conceptual views and experimentation.

The exaggerated importance attributed to experimentation in biological science, its use even as a substitute for ideas, has led recently to a massive attempt to solve the therapeutic problem of cancer by indiscriminate screening of chemical agents. Here, empiricism has been brought to its culmination. After tests of tens of thousands of agents, many workers are now beginning to realize that the results are almost worthless for cancer therapy in humans, that seemingly promising agents have an effectiveness limited to the conditions present in the actual animal experiments. By its impressive magnitude, the failure of indiscriminate screening, of empiricism epitomized, has begun to impel many workers to change their idea as to what must be done if the cancer problem is to be solved. A first result of this change has been a new and, this time, unbiased evaluation of just where we stand in our assault on the cancer problem. Every day more scientists are making the evaluation in their reports to the medical profession and to the public with a candidness which, only a few years ago, very few would have employed.