THE DEVELOPMENT AND EVOLUTION OF

CONSTRUCTIVISM IN SCIENCE

There are various sources for the development of the ideas of constructive realism which come from the disciplines of philosophy as well as of the natural sciences. Thus, it does not stand in a particular philosophical tradition; its character and value are based upon establishing relations between very different sources and content. Some elements which served to stimulate its development include the ideas of Ludwig Wittgerstein with regard to the sciences and the experience of interdisciplinary cooperation and research in the natural sciences. Discussing and investigating the process of science surfaced the implicit, and in most cases unspoken, needs of scientists and sciences.

Another very important experience was the collapse of the philosophy of science with respect to its normative claims. On the one hand, it seems clear that reflection in the philosophy of science is deficient if it merely describes the processes of science; on the other hand, in recent decades it has been becoming increasingly unclear how normative claims can be legitimized in the context of the sciences. This uncertainty in the philosophy of science led scientists not only to feel neither understood nor in charge of the philosophy of science, but to turn to metaphysical philosophy or even to irrational ideas.

This must be seen in the context both of the development of the natural sciences in the last decades and of the inadequate response to this by modern epistemology. Modern technologies, especially the use of modern computer technology in the process of research, accelerated the trend to instrumentalism (which can be found in the basic structure of the natural sciences). That is, without having given up the claim to definite and objective knowledge, this claim has been increasingly circumvented and replaced by functional relations. This implies a great need for explanation, which, however, remained generally unspoken because it did not have direct influence upon the development of research. As the natural sciences minimize the claim to recognize the world, the area of irrational exploration (that is, areas not constructed on the basis of abstract technical or mathematical reason) continuously increases.

In many cases this has the consequence that irrational strategies of explanation (i.e. strategies not legitimized on basis of [mathematical] reason) make use of results or elements of the natural sciences. This can have unpleasant or even dangerous consequences when such explanations are used as normative. In the face of this dichotomy between contexts which can be mastered (mathematically) and other explanations, the claims of Western science are abandoned. (This, of course, could have been implicit in the structure of science for a long time--Heidegger pointed out that Western science was erroneous in principle). In response, constructive realism attempts to save scientific knowledge by a new interpretation of its procedure, understanding it as a specific form of the results of action. In this context, the instrumentalism of science is not an alien element, but the great integrator.

First, let us discuss alternative approaches in modern epistemology and the philosophy of science which manifest an increasing alienation from the natural sciences. Naturalistic, especially evolutionary, epistemologies take into explicit consideration the knowledge of the natural sciences, as Kant implicitly had done. In most cases, however, it remains unclear in these epistemologies what argumentative role is played by these disciplinary results.

We are assuming that cognitive science is a continuation of naturalistic epistemologies. It is interdisciplinary in structure, but as yet there exists no theory making clear the methods of cooperation of different disciplines. As a result it is impossible rationally to execute interdisciplinary cooperation. The effect is a strange ambivalence.

For example, the use of computers increases the possibility of hiding methodological difficulties, for it often hides the possibility of controlling the relevance and plausibility of data, whereas in the manual manipulation of data the inadequacy of the method used is clarified in discussions with other disciplines. By making complex theories explicit, computers ignore or at least are not primarily interested in a critique of the method applied. For instance in the orthodox AI approach, the question of functioning has a higher priority than the question of understanding and basic knowledge. For this reason, cognitive science often is in danger of being dominated by an instrumentalistic approach in which the higher priority is to perfect the processes of cognition, while the basic problems of cognition are completely ignored.

Another aim of this approach is to develop a methodology of cognitive science as interdisciplinary cooperation between various disciplines all taking part in the discussion of basic problems concerning, for instance, what is knowledge, what is language, what is knowledge representation, etc. An answer for these questions is very important, because, if we want to investigate the process of science, we must have at least an idea of, e.g. what we understand by knowledge or how this is represented in the disciplines, etc. The following will search for new paradigms in cognitive science, as well as in the philosophy of science.

STRUCTURE AND METHOD OF ALIENATING

INTERDISCIPLINARITY

A response to the above must be positioned, on the one hand, between methodology and ontology and, on the other hand, between a naturalistic epistemology and a theory of action. In investigating methodological rules as instructions for acting, and thus criticizing the methodologies of the single disciplines in the context of theories of action, a constructive theory must develop ontological concepts from the result of naturalistic epistemologies and also reveal the implicit ontologies of the results of single disciplines.

The Concept of Transfer or Translation Between Contexts

Compared to classical metaphysical theories or to this century's philosophies of science there is here neither implicit nor explicit reference to a meta-level; this is replaced by methodological action or process. This is a set of strategies having in common the transfer of one logical system of propositions from their original context into another so that the system is judged out of context. This process cannot be planned in advance, because it lacks all metatheoretical standardization. Rather it represents a game with different contexts, multiple changes between which make possible new insights, perspectives and views regarding the structure of the set of propositions. Looking at contexts in which the systems of propositions become absurd we are enabled to perceive the implicit assumptions and considerations of this system.¹

One could think, for instance, of orthodox Artificial Intelligence being dislocated in the context of the humanities. This would make clear the insufficiency of the (in most cases unreflectedly) applied methods of computer science (i.e. mathematical logic, etc.) for modeling or even better for understanding cognitive processes. One could consider the models in cognitive psychology which make use of the symbolic processing paradigm of Artificial Intelligence (e.g. J.R. Anderson).² Before such a dislocation scientists either ignore the results pretending that they are "not relevant or adequate", or they try to look for an alternative approach which, of course, is more uncomfortable than the first case.

One should distinguish three types of such dislocation or alienating interdisciplinarity (Verfremdende Interdisciplinarität) or "strangification": linguistic, ontological and pragmatic. The process of excluding contexts in which a system of statements becomes absurd is a first mode of the dislocation. This manifests the implicit assumptions and the domain of application of a system of statements without having to fall back upon metatheoretical standardization. Talk at meta-levels is not ruled out, but is seen rather as an example for strategies of dislocation.

Applying a system or set of methods of one discipline to a very different discipline is the ontological type of this process: for instance, the application of hermeneutical procedures to the natural sciences or of gestalt perception to biology at the limits of the application of quantitative methods. Conversely, the introduction of quantitative methods in the humanities does not render them move exact, but makes clear the phenomena and structures which can not be quantified.

A pragmatic form of this process is the attempt of psychology as a humanistic discipline to quantify intelligent behavior. This manifests naiveté regarding the work of the natural scientists and their methods which are too simple to quantify or even to estimate such a complex phenomenon as human intelligence. As it is not only the method, but also the scientists and their assumptions which must be analyzed, this is a third kind of alienating inter-disciplinarity which concerns the social and organizational context of scientists.

This attempt at what has been termed a constructive type of realism is not ontological; rather it is methodologically motivated. It abandons the search of traditional European epistemologies for complete insight and substitutes a methodological aim, which in turn requires meta-theoretical legitimation.

Two Types of Reality

"Avoid the mixing of `Realität' and `Wirklichkeit'!" In German there exist two words for `reality': "Wirklichkeit" (environment), the world we are living in, and which is presupposed to, our perceptions and life processes; and "Realistät" (reality), our cognitive world which is the result of a process of construction.³ In contrast to Kant's distinction between "Ding-an-sich" and human knowledge, this stresses instead the distinction between reality as the world we are living with as environment and reality as the world constructed by knowledge.

Making use of the strategies of alternating contexts, alienating interdisciplinarity offers a better understanding of reality: though environment cannot be understood, we can gain control over it by our constructions of reality, for we can understand only what we have constructed. The distinction between reality and environment does not relativize our knowledge, but avoids surrendering ourselves to the success of our constructs in the environment.

An implication is that in organizing the process of research we distinguish between that which is instrumental and whose criterion is "working well" and the construction of reality which claims to impart knowledge.

One of the greatest errors in European history was to consider normative the empirical success of constructed knowledge; such an hypostatization of empiricism can be based only in arbitrary ideological assumptions. Rather we must not lose sight of the constructive character of the empirical, which means that empirical control of theoretical constructs amounts to no more than the comparison of two different constructs in our cognitive reality. This is not to abandon the relation to the environing reality, though this is indirect. For in comparing two different constructs of reality we test not reality directly, but the extent to which it escapes the correlations of our constructs. Thus, empirical control can be a useful alternation of contexts.

EDUCATIONAL CONSEQUENCES

The above understanding of science has many consequences for styles, content and organization of education.

Style

Traditional educational theory presented a choice between three styles or strategies: authoritarian, liberal and laissez-faire. The choice presupposed insight into human nature which, in turn, entailed extensive irrationalism or even worse: pseudo-rationalism. But such a choice between styles is neither possible nor necessary, for the style is determined by the area in which the student should be educated. In constructing scientific micro-worlds education is a type of training and, therefore an authoritarian style is most effective. Interpretation and self-interpretation can be learned only through a liberal strategy because here it is training in specific strategies; for instance in hermeneutic strategies free application is needed. For learning the central strategy of interpretation, alternation of context, an alienating interdisciplinarity, a laissez-faire style is best. By taking the lead from the specific educational area it is possible to avoid the misuse of scientific education, for its style is based on its goal, not on an ideology.

Content

A radical change in the content or goal of education is also implied. The traditional goal of education in the sciences was to develop a picture of the world in order to gain freedom in our decision making. In view of the above, however, developing a picture of the world seems meaningless, while gaining freedom in decision making appears as an irrational ideology.

On the one hand, the old conviction that science describes the world can no longer be defended. Moreover, this thesis leads to a doubling of the world and to such pseudo-questions as the ontological status of a natural law or whether natural laws existed before they were formulated. On the other hand, the notion of science as a basis for guiding freedom to a better world must be classified as an irrational ideology, for the conviction that the special sciences are able to elaborate piece by piece a picture of the world must be abandoned.

Hence, the new goals of education with respect to the sciences must be able to handle and understand scientific products for making choices between desirable and undesirable micro-worlds: or, with Paul Feyerabend⁴ we would say that students should be enabled to use science in a way that most closely agrees with the values and aims of the society they have chosen for themselves.

In view of these goals we must differentiate between training and education. The first is instruction for a specific job and is not discussed in this paper. Education is a more general term including training, but beside this leading to formation (Bildung) as developing the ability to manage life by using the available knowledge. In view of the first part of this chapter it should be noted that micro-worlds, as the immediate results of scientific activity, are not knowledge. There is need to link them up to the whole of human understanding, i.e., to the horizon of language which expresses the respective culture.

For this goal education at the high school level must be totally changed. Instead of being given a mediocre and superficial acquaintance with many scientific results (micro-worlds), students must first be trained by a few examples as to how micro-worlds are constructed with respect to the history of science, the logic of their structure, and their scientific application. But more important than this is education on how to understand the interpretations of the micro-worlds, how to modify the given interpretations, and even how to invent new ones (though this latter usually should be reserved to university education).

University education must have a totally different structure, for it is training for a job. The educational part of the study must not be understood as a mere addition to job-related training. It would be wrong to educate one to be a physicist, historian or teacher in literature simply by general philosophy of science, history of philosophy or aesthetics. For one must learn to handle and to understand the natural, human and cultural output of one's science. Thus one must learn how to discover, establish and restrict the possible applications of the products of his science. This requires interdisciplinary practice and clear rational information about the values of the society in which one lives.

Interdisciplinarity cannot be taught theoretically but requires organization. Also some theoretical information is needed for alternating contexts. Therefore the educational part, which should accompany the scientific study from the beginning, should have three sections:

a) The first section, besides history of science and, according to the circumstances some comparative theory, should contain the aims and values of one's society with regard to science. Often this can be done indirectly with helpful examples on how science works in different historical, political and cultural environments. It is important, moreover, to clarify what happens if a scientific context is implanted in a totally different society.

b) In the second section the student must take part in transferring and translating the results of one's own science to other sciences.

c) In the third section different interpretations of micro-worlds must be studied and, at least for the doctoral degree, a new interpretation of a micro-world must be developed. For this undertaking theoretical and practical experience in alternating contexts is useful.

The Reorganization of the University

The traditional university is structured according to the conviction that the sum (universitas) of the theoretical sciences delivers the knowledge which is available for mankind in the respective fields. But as most scientists have long recognized, the sum of the sciences does not generate knowledge, but rather constitutes an immense and incalculable bundle of propositions with different objects which can be connected only arbitrarily.

What does one do in such a puzzling situation? If nothing is done then the huge science/industry will go its own way, organing itself without consciousness so that all communication, beyond technical cooperation, will be lost. Powerful linguistic systems as parts of single sciences produce new linguistic products which escape the overview of the scientists. For the formalized languages follow the fundamental character of languages as self-organizing and creating new linguistic systems.

Whereas ordinary language always refers these new systems to everyday practice, in formalized language this direct and continuing touch with the human action is lost. Linguistic formalization generalizes a specific type or aspects of human activity and by abstraction loses touch with other types or aspects of human action. Therefore, for formal language, in contrast to ordinary language, the modification and renewal of the respective language--activity--relations is blocked, for if these relations are changed the respective formal language loses its instrumental character. This would be to move from the field of technology to that of poetry understood as an opening of one's the relations in one's language--activity--relation. Thus, ordinary a computer-program could become a poem.

To control this development, first we must be clear that in this case every general effort must fail. Therefore the contemporary university organization, with its generalized structures and restriction to merely abstract cooperation with the society, must fail.

In this situation some modified version of the Platonic principle that kings and philosophers should be one could be formulated: the scientific specialists must be representatives of society with respect to their special science, and the scientific community must become aware that it can influence scientific development only by being involved in the efforts of alternative is either to be ruled by an uncontrolled development of the standard of contemporary life as well as the level of therefore the loss of a wide range of possible human activities. (The ecological movement is correct in its diagnosis of the problem, but very often wrong in its remedies.)

We must abandon the old metaphysical conviction that there is a final and general knowledge. Removing irrational hopes and looking at how science really works, we must constitute a new concept of scientific knowledge as intellectual freedom with respect to functioning mechanisms.

As a type of definition, every word is important here. "Intellectual" means that the man is able to describe functioning mechanisms in his own way. Different choices can be made for didactic reasons, reasons of beauty, and so on. As understood by Hegel, "freedom" has a substantive meaning, i.e. among all possible choices only those are free which change the mechanisms. Therefore to be expert in handling the mechanisms is a precondition of intellectual freedom and in consequence of scientific knowledge.

This may seem a very restrictive definition of scientific knowledge, but not if we keep in mind that we cannot aspire to any generality or totality for scientific knowledge. For all totalities and generalities regarding scientific results or procedures are arbitrarily (not freely) elected, and therefore cannot be considered to be scientific knowledge.

A correct formulation for this concept of scientific knowledge would be that to have understood a few micro-worlds is to have grasped the possibilities of scientific knowledge, which means to have entered into scientific knowledge. Scientific knowledge is a specific type of human activity, taught, developed and increased for over 200 years in Europe. It cannot be learned in totality because it is an open system. On the contrary, it suffices to be instructed by examples in order to understand, that is, to preform this type of activity.

What are the consequences for university reorganization? Students, researchers and teachers must acquire the possibility to learn free choices regarding single micro-worlds. In view of this, they must have the opportunity to learn the different strategies of resituating a micro-world, which strategies can be related to different operational (not classificatory) levels of understanding micro-worlds.

One level is logical. The structure of a micro-world can be replaced by another structure which fulfills at least the same goals. Of course, this requires scientific ingenuity and in this sense Albert Einstein understood classical mechanics better than did Sir Isaac Newton himself. But serious critics too have a better insight, even if they are not able to replace the criticized micro-world by a new one. Of course, these criticisms cannot be judged until a respective micro-world is invented, and in this sense Leibniz and Mach had a good understanding of the Newtonian micro-world.

A second level is linguistic: the level of translation, by which, for example, Immanuel Kant obtained his understanding of the mechanics.

A third level is that of application, where understanding can be gained from failure as well as success. Examples here are the use of mechanics not only for engineering, but for physiological and psychological explanation. In experiencing applicability, as in recognizing inapplicability, freedom is gained regarding the micro-world. Other levels are related to society, culture in general, poetry and so on.

What specifically does this mean for university structures? The traditional divisions in the universities should be replaced by units with a small staff: a chief who could be in the place of the traditional "professor", one or two assistants, and one or two secretaries. These units originate from the appointment of the professor, who is appointed to a general discipline rather than to a precisely delineated subdivision. He should have restricted teaching obligations in order to leave opportunity to establish research programs.

Beside these units there should be a large pool of young "researchers", with the duty to cooperate in a research program of their choice. Under specific conditions they could become leaders of research programs, which would be a first step toward becoming a professor. The students should take part in research programs beginning from the second part of their study, and should be obliged to choose at least three research programs during their study. The definition of their degree would depend on the distinctive field of their research. They should also write a thesis which would be judged by the extent to which it makes at least one of the research programs understandable in the technical sense described above.

Such a university structure would have at least two advantages: it would apply the principles of self-regulation and it would enable learning intellectual freedom in the above sense.

1. M. Polanyi, The Tacit Dimension (Garden City: Doubleday, 1966).

2. J.R. Anderson, Cognitive Psychology and its Implications (New York: Freeman, 1985).

3. F. Wallner and Thesen zu Maturana, "Eine Aufforderung zum Widerspruch", Research Report, Dept. of Epistemology and Cognitive Science, University of Vienna, IWTF-ECS-89-2.

4. Paul Feyerabend, Against Method (Revised Edition; London: Routledge, Chapman and Hall, 1988), p. 258.