Skip to content

A Scientist’s View of the Philosophy of Science

William S. Verplanck
Knoxville, Tennessee

Ernst Mayr’s book “This is Biology” is its author’s summary of the contemporary status of biology, or rather of the various disciplines within the domain of the science of biology, with thoughtful, critical review of how they have evolved. Mayr is one of the very few individuals whose credentials are so broad as to enable him to write such a book. His mastery is easily checked by a simple review of the index of this work: It is a stimulating read. Mayr has not limited support of his readers to the index; he provides them with an enlightening guide to the topics covered, as well as a glossary of terms unfamiliar to those who are not biologists. There are very few people who are in the position to describe their discipline’s relationship to other disciplines and to historical developments which have led to its present status. Mayr is one of them.

Ernst Mayr presents us with biology as a family of sciences, whose members fall into two groups. The first of these includes the evolutionary biologies, with their emphasis on historical variables. The second, the functional group, as Mayr calls it, deals with what he, like Aristotle, refers to as ‘efficient cause’. This is a Humian analysis restricting causation to contiguities or contingencies in physical space and time, the analysis that leads to reductionistic theories and research, and to the neglect of antecedents remote in time, that is, of history. Mayr identifies biology as an organicist science whose subject matter cannot be dealt with solely in physico-chemical, that is, molecular terms. The science of biology deals with highly complex organized systems, to which the physical sciences, in however advanced a stage, such as quantum mechanics, relativity, and so on, have nothing to contribute.

The title of this paper, “A Scientist’s View of the Philosophers of Science,” expresses my first reaction in reviewing this book. Mayr is both clear and convincing in demonstrating the degree to which “philosophies” of Science, and especially logical empiricism, have been based solely on the history and methods of physical sciences. Those, he asserts, are ill suited to deal with a life sciences. Kuhn he rejects in toto; Popper, he faults for his views on ‘falsification’.

A second reading led me to emphasize in a draft of this paper Mayr’s opposition to any reductionist efforts to account for the findings of the evolutionary historical family sciences in Biology. Reductionism, the attempt to account for the historical phenomena with which these biologies deal in terms of physical and chemical laws and of structure-tied function, does not work.

A third gloss demanded emphasis on the proposition that a science of behavior must be and is primarily a biological science, of the evolutionary/historical class. Hence, it must be expounded and assessed as such, rather than as a development of psychology, a science rooted in human behavior, and the “psyche” of aristotelians, rather than of the “psyche” whose functions were described by Aristotle himself.

Ernst Mayr’s critical evaluation of the biological sciences may, and should, be invaluable as vantage point for a parallel analysis of the status of the science of behavior. Those readers who are behaviorists will not fail to remark on the pertinence of his thinking to the problem of our science, to its evolution as a science as well as to its probable development in the future. This will be the case even though he or she may agree with Mayr on the impossibility of confidently predicting the course of its future development.

As you hear this summary of what might be titled “This is behaviorism,” you will recognize that it is Verplanck’s version of Mayr and as such surely reflects his own progressive and retrogressive experience in psychology, ethology, behavior, and… just plain trying to make sense of it all. He’d place more emphasis on the contributions of Aristotle than Mayr would place. (Aristotle’s are only now coming to be assimilated into our science.)

But since it is my hope that the outcome of your having heard what I have to say will be that you read Mayr’s book. Then, you can evaluate his statements with respect to what you are doing as a behaviorist, whether as a ‘scientist’ or a ‘practitioner’. Thereafter, you will be able to assess the merits, if any, of my own reading of Mayr, and of what I have derived both from him and from my own history, that is, from the evolution of my conclusions. You will be free, then, to relate what follows to Mayr’s book and agree or disagree.

Here, then, is the case for a science of behavior that is a biological science of the evolutionary/historical family, as it is identified by Mayr.
(1) The science of behavior finds its roots in the writings of the first biologist, Aristotle. The Greek word, which is best defined by a translation into Greek of the definition of the English word, behavior, the ongoing interactions of the individual with his or her environment, is the word Bios. Bios is also translated into English as “life,” “living.”

Aristotle’s works, when translated into modern English, show him to be a behaviorist, an operationalist, and a contextualist (Randall, J.H. Jr (1959) “Aristotle”). Mayr independently puts it this way, “Much of Aristotle’s discussion [e.g. of ‘eidos’ and the like] becomes remarkably modern if one inserts modern terms to replace obsolete sixteenth and seventeenth century vocabulary” (Mayr, E., (1976) “Evolution and the Diversity of Life”).

Aristotle was able to set forth principles in the biological sciences that remain critical, despite his total lack of the scientific prostheses, the instruments, that we can now employ. He did this almost two millennia before the development of scientific methodology through the last five centuries.

(2) Behavior is defined as the ongoing interaction of the individual with his or her environment. The ongoing processes that are observed are transient and leave only records, whether archaeological artifacts or cumulative curves. Behaviors cannot be stored in museums, or obtained as samples drawn from warehouses. The science of behavior is necessarily a historical one. In accounts of ‘causality’, explanation cannot be expected to derive from the physical concept of contiguities in time and space, which work satisfactorily in the physical sciences and are well illustrated by the instantaneous rates, dx/dt of calculus, but by what analyses parallel to Aristotle’s ‘four causes’: material, formal, efficient and final. A remarkable parallel to these recently emerged in Tinbergen’s statement of the four very different kinds of questions that the behaviorists called ethologists must answer in their research.

(3) The science of behavior studies the behavior of the whole individual, and not of specific structures of the individual. In Mayr’s terms, it is organistic–its subject matter is the behavior of the intact individual, not of truncated ones, or of specific body parts.

(4) The environment with which the individual interacts must be described and understood in the terms of and in the language of common-sense realism, of tables, chairs and things that are good to eat, and not of molecular structures nor of the particles, of quantum physics, nor of impalpable phenomena. To paraphrase Kantor, shoppers encounter no metaphysical, epistemological, or ontological questions when they choose a steak in the supermarket. (This is the war that Gestalt’s theorist won, even as they lost every reductionistic theoretical battle.)

(5) The science of behavior must be Baconian in its collection of descriptive empirical data. The adequacy of such descriptions is assured by the use of the number system–statements in quantitative rather than qualitative, that is, verbal terms. But such data must then be classified, putting likes with likes, with due note of the differences that remain between items that have been classified together. From such progressive classification, generalization can be adduced, and descriptive statements become “abstract” generalizations–ones that “explain” or “enable understanding.”

Such flexible systems, rather than rigid ones as dictated by “rigorous” methodology, are, in Mayr’s view, the key to scientific advance and explanation. First naturalistic observation, then the experimental observation made possible by the first classification, then the puzzlement: how? and why? of discrepant results. These lead to guesses, conjectures, “working hypotheses,” the identification of what needs to be done next, thus checking out the guess or guesses, enabling acceptance (for the time being) of the guess that works best, the pragmatic answer.

Such systematic categorization, such classification, is an essential step in the development of fruitful theory. Could Darwin have developed evolutionary theory if he had not had at hand, some when he boarded the Beagle, and some on his return, the biological classification of Linnaeus, the geological classification of Lyell, and the demographic and economic data of Malthus and Smith?

The science of behavior must then be descriptive and classificatory. It must further be historical and probabilistic. The qualification “historical” emphasizes that behaving is a process, an ongoing sequence of events ever-changing, whose description will lean upon the description of events not only immediately antecedent to the behavior observed, but also on more remote antecedents. It is probabilistic in that the conditions required for precise chronological predictions (rather than logical predictions), whose probability of verification reach 1.00 , are seldom if ever achieved. Variability in control of the pertinent behavioral variables will preclude. Mayr notes that “there are many branches of science in which prediction plays a very subordinate role, and in many of the non-applied sciences the question of control never comes up.”

It follows that the definition of this science in terms of the prediction and control of behavior are, at best, unrealistic and certainly misleading.

(6) Like other sciences, the science of behavior has necessarily developed its own methods and concepts, many of them unique. Our methods of experimental observations, derived from the observation and categorization of the interactions of environment with individual in the absence of experimental manipulation, parallel them.

These fall into three classes, based on the procedures carried out by the experimental investigator, as well as on parallel events observed in natural, uncontrived circumstances. In the first, an activity of the individual is either forced or precluded, and the consequent changes in behavior measured. In the second, the environment (or parts of it) is manipulated, and the change in the individual’s activity measured (in nature, lightning happens). In the third, a change in the environment is produced by an activity of the individual. These are the paradigmatic operations–response, stimulus and dependence operations respectively. Each controls specific processes, and in both ‘nature’ and the laboratory, complex combinations of these occur. The “three term contingency” will be recognized as the combination of a stimulus with a dependence operation.

More specifically, behavior science has developed its own methods, and its own concepts, and will be cautious in choosing names for the concepts dictated by the product of its methods. Words chosen from the vocabulary of philosophy, the lexically defined terminology of the vernacular, may lead to faulty communication. Words from the latter may be introduced into the behaviorist’s vocabulary only when data enable them to be defined by quantitatively stated observational data.

This link between conceptual terminology and methodology is a point that is not explicitly stated or emphasized by Mayr, but it is explicitly stated in both Kantor and in Wittgenstein. (Wittgenstein’s name is absent from Mayr’s index–a singular omission.) Each Science develops its own methods and a language dictated by those methods. Each thus comes to be isolated from others. There is no assurance that the languages of the various sciences will necessarily enable eventual translation from one to another, and hence produce the “Unified Sciences” dreamed by Carnap.

(7) From the above, it follows that the key to both the methods and concepts of behavior science is measurement. Analysis of the measures of behavior that are made demonstrates that the science of behavior is independent of those other sciences which systematically rely on the measurement system of the physical sciences. These include psychophysics and ergonomics, but not the evolutionary biological science of behavior.

(8) “Reductionism”, the attempt to provide an account of the findings of one science in terms of another, more basic science, is a viewpoint that is present in psychology, but is alien to the historical/evolutionary biological sciences. Mayr addresses such efforts explicitly, pointing out that such viewpoints preclude the investigation of many problems, using behavioral ones as his example. To others of the biological sciences, such efforts are merely irrelevant.

In the science of behavior, the early behaviorist expected to “explain” behavior in terms of physiology. But Watson, in his later writings, ignored the limits set by such aspirations. At about the same time, Kantor explicitly opposed such efforts, as did the logical behaviorist, Wittgenstein, somewhat later. Even as Wittgenstein’s were published, the thoroughly reductionistic behavioral theory of Clark Hull fell flat, and Skinner wrote “Are theories of learning necessary.”

Since then, James J. Gibson rejected, on the basis of behavioral observations, all reductionistic accounts of the individual’s interactions with his or her environment. That Gibson was a behaviorist, as are his followers, is undeniable, however closeted and/or self-hating they may be. His writings will inevitably prove translatable into the language of behavior in the foreseeable future.

Finally, W.R. Uttal, in his book “Toward a new behaviorism: the case against perceptual reductionism,” makes an about-face, rejecting reductionistic thinking, whether physical/physiological or cognitive/mentalistic. He correctly identifies the science of behavior as implicitly non-reductionistic.

As behavior scientists, we must recognize that reductionism has not worked and join Ernst Mayr in his assertions that it cannot work, as he gives the reasons why.

The study of “light-waves” and “sound waves” “belongs” to physics. The study of the anatomy and functions of the brain is the subject matter of neuroanatomy and of neurophysiology.

The science of behavior states for the “brain scientists” problems to be investigated.

(9) In the science of behavior the primary subject matter is the systematic exploration of the consequences of the paradigmatic operations for the ensuing behavior of the individual, and in “applied” behavior science, the development of procedures for using those methods to produce specific changes in the behavior of individuals.

A second major concern is the behavior repertory of the individual and of his or her environment. Remarkably, these have been all but ignored by behavior scientists.

The behavior repertories of individuals, whether men or mice, and the historical (biological) events that determine them can not be neglected. The failure of Hullian theory vis a vis Tolmanian theory, both behaviorists, is readily attributable to the great differences between the behavior repertories of the individual rats in their respective colonies, dependent on the environments in which the repertories developed.

The behavior and behavior repertories of environments remain all but unexplored by behaviorists, evidently, leaving the problem to ecologists.

Three major components can be identified in the behavior repertories of individuals:

First is the species-characteristic repertory: those behaviors which have through evolution coevolved with the physical structures and function of the individual, which include the structure of brain, nervous systems and glands of the species of the individual. As the Brelands showed many years ago, extrapolation of behavioral findings from species to species must be carried out cautiously.

Second is the community-characteristic repertory: those behaviors that the individual shares with other members of the groups or communities–the social environments–within and with which he or she behaves. Study of these community-characteristic behaviors have seldom been investigated by behavior scientists, who leave them to sociologists, economists, linguists, and other “social scientists,” each specializing in investigating one or another subset of the community-characteristic repertories of the human.

Third is the individual behavior repertory: those behaviors that are the unique product of each individual’s specific interactions with each of the successive environments with which he or she has behaved, beginning with the first kicks at the uterine wall, and the first rock music the mother listens to as the cochlea and auditory system of the fetus develops.

As participants in a biological science, and more especially an evolutionary/historical one, we have done remarkably little work on the organization of behavior repertories, neglecting to recognize, it seems, the relationship of such organization to such obsolescent concepts as “emotion,” “self,” and many other concepts and concept-names encountered in non-behavioral psychology.

For a start, consider the relationship to one another of three major components identified by Freud as id, superego and ego. (Note that he shared with Skinner the “black-box” approach to physiological participants.)

* * * *

In sum, by putting like with likes, Mayr’s views and our own, the science of behavior is a biological science. It has achieved major results, but has much left to do. Mayr’s book, in many ways so friendly and so relevant to the emerging science of behavior, may stimulate investigative research in new areas.

So, as we may begin to extend our investigations, close reading of “This is Biology” has much to offer. Among other advantages, it implicitly and sometimes explicitly identifies some pitfalls we may encounter. Let me summarize:

First is reductionism, the attempt, even as we observe and investigate behavior, to account for it in terms of objects or events that are not themselves the behavior of the individual, e.g., the structure of the brain.

Second: We must not become preoccupied with that part of human behavior that is termed “verbal.” Such preoccupation is the Cognitive Fallacy, the assumption that human verbal behavior opens a high road to the description and explanation of behavior. Verbal behavior (especially when it is conversing with oneself) is only one part of the human repertory. Pavlov, having reductionistically accounted for what he called “the first signal system,” proceeded to study “the second signal system,” that is say, language. This led Pavlov into the field we now call Psychiatry.

Third: There is always the danger of inadvertently ‘writing’ and following general theories, based on limited sets of observation. Presently, too many behavior analysts have shown such a tendency in their use of the concept ‘reinforcement’. They use the term reinforcement to refer to any increase in response-measures. They then proceed to assume that if a response increases in measure, it has been reinforced by an environmental consequence of that behavior, and not by any other environmental operation. This necessarily leads them to hunt for, and thus invent “reinforcing stimuli” that are fictitious. Reinforcement-by-consequence is only one of the operations that strengthen a behavior or produces a new one.

Finally, we must avoid limiting our studies to permutations and combinations of a limited set of operations, and then fail to consider the full implications of the concepts to which those have led us. We have far too often failed to extend the basic methodological and conceptual systems to give accounts of broader areas of behaving.

But all this may be no more than the complaints of a geriatric curmudgeon.

Creative Commons License is licensed by Admin under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.