Realism versus Constructivism:
Which is a More Appropriate Theory for Addressing the Nature of Science in Science Education?


Brian Campbell
Southwestern Oklahoma State University



    Though the term "constructivism" is relatively recent with respect to epistemology (Phillips, 1995), components of it have been used for years by empiricists, instrumentalists, operationalists, idealists and others in arguments against humans being able to unambiguously know reality (Matthews, 1994; Mechling, 1990). Science history is replete with many colorful examples of debates concerning reality and the nature of science: The realist Aristotle and empiricist Plato, Copernicus’ realist and Osiander’s instrumentalist conviction of the heliocentric universe; Newton’s realistic and von Leibniz’s skeptic view of a gravitational force; Planck’s realistic and Mach’s empiricist arguments concerning the actual existence of atoms; the realist Einstein and the instrumentalist Bohr over the Copenhagen interpretation of quantum mechanics (Matthews, 1994). My argument will be that by addressing science education from a constructivist approach, one may more accurately capture the nature of science than if a realist approach is used.

    A great many papers and books discuss the various forms of constructivism such as: developmental constructivism, feminist constructivism, radical constructivism, social constructivism, trivial constructivism, etc. (Osborne, 1996; Phillips, 1995; von Glasersfeld, 1989; Speed, 1991; Held, 1990). While the various faces of constructivism have many differences, they also have much in common (Osborne, 1996; Phillips, 1995). Much the same can be said for realism. Realist epistemology is not new. There are many different variations of realism: critical realism, modest realism, naive realism, objective realism, strong realism, and weak realism, (Osborne, 1996; Matthews, 1994; Yarusso, 1992; Speed, 1991). These various versions also share differences and similarities (Matthews, 1994).

    From an examination of articles, it appears that many if not most authors take a favorable view of a constructivist pedagogical approach. Particularly in the discipline of science education, a constructivist-based instruction is frequently hailed as a great improvement over the didactic lecture, textbook reading (Osborne, 1996; Trotter, 1995; Gil-Perez & Carrascosa-Alis, 1994; Black & Ammon, 1992; Mechling, 1990; Elkind, 1989). Even though many agree that constructivism is an improvement for science education, several journal articles and recent books have been critical of a constructivist-based pedagogy for appropriately addressing the nature of science (Osborne, 1996; Matthews, 1994; Speed, 1991; Held, 1990).

Variations of Realism

    Speed (1991) defines realism in a succinct, yet informative way as "the position that reality exists, can be discovered by people in an objective way and thus determines what we know." This definition contains two explicit tenets and one implicit tenet held by many realists:

    Outside the mind of humans there exists an independent reality (Matthews, 1994; Yarusso, 1992; Speed, 1991). Reality is not altered by human perception, thought, or interaction (Osborne, 1996; Matthews, 1994; Klemke, et al., 1980). It is composed of absolute truths that are inflexible (Matthews, 1994). This reality can be discovered by any cognizant being with the capacity to understand it. Reality and all knowledge of it is revealed to, not invented by, the observer. As Yarusso (1992) states it, "The key (realist) position is that there is an external reality that is, ultimately, accessible to the human mind." Reality can be perceived objectively. "In a perfect world, we would all have the same interpretation of what is true, what is right, what is false, and what is wrong" (Yarusso, 1992). This information, these inexorable discoveries and facts about the universe, once unearthed, can be taught to, or learned by, others in a way that will allow their being transferred without alteration of meaning. Since "external reality is, ultimately, accessible to the human mind," it follows that those capable would perceive it in exactly the same way. With some effort, every human (or sentient being for that matter), could understand the same information about the universe in exactly the same way since that information can be discovered objectively and with the same interpretation.

    One realist’s view is humorously summed up in the following poem from Chemistry for Changing Times (7th Ed.) by Hill, J. W. & Kolb, D. K. (1992).

Though single atoms are too small

For you to see or feel,

There isn’t any question that

These particles are real.

Variations of Constructivism

    The term "constructivism" has been used so extensively by such a large number of people and for a wide variety of purposes that there is almost no consensus as to its actual meaning (Phillips, 1995). There are, however, several elements of these theories that Phillips identified and assembled into a three dimensional "framework for comparing constructivisms". In this framework, each axis represents a major component of constructivism.

    The first axis discussed is labeled "individual psychology verses public discipline". On one extreme of this axis are the researchers concerned only with the individual and how (s)he constructs knowledge (ex. Phillips, Piaget, Vygotsky). On the other extreme are those unconcerned with the individual, but consumed by the idea of "the construction of human knowledge in general" (ex. Acoff, Barnes, Bloor, Collins, Potter). This is the social constructivist view. In the middle ground are philosophers such as von Glasersfeld, Popper, Kant; concerned both with the individual and the society.

    The second axis is "humans the creator versus nature the instructor". This addresses the issue of whether knowledge is constructed " the mind or creative intelligence of the knower..." or is "imposed from the outside" and the human mind is nothing more than an empty basket awaiting facts and knowledge to fill it. The views of Ernst von Glasersfeld represent the former while those of John Locke are a good example of the latter. The comparison of constructivist’s ideas now begins to become somewhat convoluted because not all (it seems very few) agree when these two axes are combined. Two researchers may agree on one topic yet be diametrically opposed on the other. This convolution is increased when the third dimension is added.

    The third axis is "physical activity versus mental activity". At one extreme it is believed that only by doing, by being physically active, can knowledge be constructed. The other extreme views knowledge construction as a mental process only. Most of the constructivists mentioned in this work fall somewhere in the middle ground believing that knowledge is constructed by a combination of physical and mental activities.

    A review of these axes demonstrates that many combinations of convictions can be held and still considered constructivist (Phillips, 1995). For purposes of this paper I will hold three of the more conventional constructivist components:


Comparison of the Two Theories for Addressing the Nature of Science

    Let us now compare and contrast the two epistemologies in question with each of the five general components of the nature of science. During this analysis, several natures of science components are considered at once. The history and philosophy of science, because of their close association (Matthews, 1994), and the two components addressing how the scientist actually performs science are discussed simultaneously (Kuhn, 1970). Finally, the goals of science are addressed separately.

    Throughout recorded history, humans have created or invented explanations for observed natural phenomena. In some cases this led to the invention of magical beliefs, myths, or religions (Campbell, 1988). In other cases this led to more scientific explanations. These explanations were tested by whatever means the culture either deemed appropriate or had the technological ability to confirm. Support for these theories was, therefore, at the mercy of cultural biases and technologies (Matthews, 1994). In some cases the culture did not value or practice confirmation in any modern sense. Quoting authority, saving the appearances, and/or argument were all that was required to keep many explanations viable for years (Campbell, 1988).

    When a more powerful* theory emerged, the old theory was amended to incorporate the new. Often if the competing theories were incompatible, the new completely replaced the old in a "revolution" of some kind (Matthews, 1994; Wallace, 1989; Richards, 1987; Kuhn, 1970). History is teeming with examples of theories that have been replaced or modified: Aristotle’s four fundamental elements and his crystal spheres, Hipparchus’ geocentric universe, Newton’s absolute space and time, Hoyle’s steady state theory of the universe (Hawking, 1988; Hoyle, 1962).

    The strong realist view would have us believe that once the truth is known, no further changes need ever be made. After the answer to the question has been found, what is the purpose of further investigations? (Matthews, 1994). The reader is asked to remember Speed’s definition of realism, "...that reality exists (and) can be discovered by people in an objective way...". This objective reality, once known, can be experienced and shared precisely as it is.

    Aristotle’s crystal spheres and Hipparchus’ geocentric universe were considered truth by the human race for far longer than the presently accepted theories. Who is to say that what is presently considered scientifically sound will remain so in the future? Has not history taught us humility about our ability to discover the ultimate truth? As was so concisely expressed by Carl Sagan " of ultimate truth we (meaning scientists, author’s inset) leave to politics and religion" (Sagan, 1980).

    The reader might realize that this is also an argument for another set of components under consideration: How the scientist discovers or invents patterns and explanations for those patterns. These can be lightheartedly paraphrased as, "What do scientists actually do?".

    One function of the modern scientist is to explain and make predictions concerning natural phenomena (Medawar, 1982; Kuhn, 1970). One recent example of this is geologist’s attempts to explain why earthquakes and volcanoes occur. For centuries it was known by some societies as a "fact" that these two events were unrelated. The former were known to be caused by huge animals moving underground while the later were seen as proof of the existence for a fiery underworld for wrongdoers (Thompson & Turk, 1991; Campbell, 1988).

    Over the centuries many different societies invented explanations for these two disparate events. The citizens of these societies went about their daily lives in relative comfort knowing these questions had been put to rest. But not everyone was satisfied with the accepted explanations.

    In 1912, Alfred Wegener published his theory of Continental Drift. Bitter debates ensued. After a short time, geophysicists defeated a major component of the theory. It was correctly demonstrated that Wegener’s explanation for his proposed continental movement was wrong. He suggested for the mechanism causing this drifting the gravitational influence of the Sun and Moon on the Earth. This gravitational force was calculated to be far too weak to move the continents (Thompson & Turk, 1991; Leet & Judson, 1971).

    During the following decades many other geologists including Hess, Vine, Matthews, and Lawrence investigated, altered, and revised Wegener’s initial theory into the presently generally accepted form of plate tectonics (Thompson & Turk, 1991). This theory elegantly draws together volcanism, earthquakes, mountain building and other once separate events into a single entity. But is this the ultimate truth? Have all of the questions been answered?

    How did the scientists in this example construct their newest theory? They improved their equipment and observations. They made serendipitous finds. They made and then tested predictions. They debated and discussed with each other. When possible they went to locations and investigated. When it was not possible to visit a site for first hand experience, existing equipment was modified or new equipment was invented that extended their senses (e.g. sonar). Many different scientists performed many different types of investigations. Not everyone followed the same steps. There is no one way to do science. There is no one scientific method (Millar, 1989).

    Modern scientific theories are held to rigorous standards. They must be internally consistent. They must agree with observable phenomena. They must make predictions that can be tested. Tests must be repeatable by others (Rhodes & Schaible, 1989; Trusted, 1979; Kuhn, 1970).

    Theories are considered by the strong realist scientists to be approximations of an absolute knowledge that is ultimately attainable precisely and objectively by the mind of man. The first half of this last statement should cause no controversy between realists and some constructivists. As the reader will remember from the description of the constructivist theories, some also maintain that their theories actually approximate a reality. They would argue, however, that this reality can never be objectively known since all of the perceptions of the outside world are filtered through experiences (Matthews, 1994) and assimilated into existing cognitive structures (Wadsworth, 1996). As the individual tries to fit (assimilate) these impressions from the senses into existing structures, it is often the impressions, not the structures that are altered (Wadsworth, 1996). Also, to many constructivists it is not whether the theory is actually true, for theories can never be proven (Kuhn, 1970). What is of interest is how well the theory in question works; how well it enables one to cope.

    For the present, the theory of plate tectonics works well in explaining why many geological features are as they seem. But it is not a completed theory. There is much it can not do. For example, it can predict where earthquakes are more likely to occur (ex. California). It can not predict with absolute accuracy when and where earthquakes caused by strike-slip faults will occur nor how destructive they will be. It can explain why many volcanoes exist around the Eastern Rim. It can not predict with perfect accuracy when or where a subduction zone induced volcano will erupt. We do not yet know reality absolutely.

    The three general constructivist factors being discussed match the scientists' behavior quite well:

And what of the last component of the nature of science, the goals of science?

    As for the realist notion that the goal of science is to discover exactly what nature is and how it works, Jacob Bronowski in his book The Assent of Man (1973) remarked, "One aim of the physical sciences has been to give an exact picture of the material world. One achievement of physics in the twentieth century has been to prove that that aim is unattainable" (Bronowski, 1973).

    This comment was made with reference to quantum theory and in particular the Heisenberg Uncertainty Principle. This principle demands that there is a limit to the amount of accurate information it is possible to obtain about subatomic particles. The more accurately one measures, for example, the position of a particle, the less accurately one knows its momentum (Giancoli, 1991). When humans attempt to measure these particles, the measurement process itself changes them. Our attempt to understand nature actually alters it. Whether there is a reality independent of the observer or not, there is a limit to the amount and accuracy of that which is knowable by that observer. All information about nature can not be known precisely (Elkana, 1970).

    The philosophical implications of this principle have long been known. To quote a prolific science writer:

"Philosophically, this is an upsetting doctrine. Ever since the time of Newton, scientists and many nonscientists had felt that the methods of science, in principle at least, could make measurements that were precise without limits. One needed to take only enough time and trouble, and one could determine the nth decimal place. To be told that this was not so, but that there was a permanent wall in the way of total knowledge, a wall built by the inherent nature of the universe itself, was distressing" (Asimov, 1966).

    With the goal of ultimately discovering nature objectively dashed, a realist might argue that this itself is understanding the truth about nature. An ugly and confusing argument that would be: The ultimate reality of the universe is that humans can never know reality precisely. How do we know that the universe does not function in such a way as to prohibit our minds from ever understanding it precisely? Since the human mind is part of the universe and presumably follows its laws, it must also be a part of prohibiting itself from the knowledge that it seeks!

    This is a clear advantage for a major component of many constructivist theories: Human knowledge must always be incomplete. Our actions taint the results hence the conclusions we draw from them. Our actions influence what we can and do know about reality. This says nothing about the existence of an independent reality, only our ability to precisely know and copy it.

"There is no absolute knowledge, and those who claim it, whether they are scientists or dogmatists, open the door to tragedy. All information is imperfect. We have to treat it with humility. That is the human condition ..." (Bronowski, 1973).

Or to quote one of the most revolutionary and prolific researchers in developmental constructivism, "Knowledge is not a copy of reality" (Piaget, 1964).

    To review three of the most powerful arguments constructivists use when debating realists on this topic (as discussed in Matthews, 1994):

1. The history of science is replete with abandoned and disproved theoretical entities that earlier were firmly ensconced in the best science of their time. (What is accepted as fact today may be discredited tomorrow.)

2. Theoretical provisions are always limited by the amount of evidence available, and consequently the same evidence will also support other existing or potential, theoretical entities. (Theories can never be proven. A good theory is a useful theory. There are often more than one theory that will fit the existing data.)

3. Scientific conceptions are determined by the theory in which they occur as well as by the reality they purportedly describe. (If our knowledge of reality is ultimately imperfect and incomplete, so must be our theories of reality.)

Implications for Science Education

    "The study of science as an intellectual and social endeavor - the application of human intelligence to figuring out how the world works - should have a prominent place in any curriculum that has science literacy as one of its aims" (Rutherford, et al., 1993).

    In the previous section I argued that some elements of constructivist epistemologies more closely approximate components of the nature of science. Having done this, how can this information be used? What are some implications for a science education curriculum?

    One implication is obvious; science courses should not be taught as a list of known facts to be memorized. How does the teacher know what facts will still be considered as such in the years to come? How can the teacher be sure of what each and every student will need to know in the coming decades? To quote from the popular and influential book, National Science Education Standards, "In learning science, students need to understand that science reflects its history and is an ongoing, changing enterprise" (National Research Council, 1996). Science must be seen as a dynamic not a static endeavor.

    Another implication is that the students should be engaged in real investigations (National Research Council, 1996). They should be challenged to create their own problems and solutions (Elkana, 1970). These solutions should then be challenged and tested. The students should be given the freedom to work creatively on projects of their own interest (Phillips, 1994). "Experiments" should not be simply show and tell nor exercises in following directions (Phillips, 1994). They should be realistic examples of how science is done.

    "Since a key principle of education is to begin with what the learner already knows, finding this out is a very important initial step in any educational endeavor (Novak & Gowin, 1984)." This quotation illustrates a third implication one can gather from a constructivist epistemology: No student (or scientist for that matter) begins an investigation with a clean slate. That is to say, all students likely have some prior knowledge or ideas about what is to be investigated (Rauff, 1994). Students will either assimilate new data into existing structures or accommodate themselves to these new data (Wadsworth, 1996). An instructor working to further each student’s development should be aware that every individual in the class is bringing past experiences that taint present thinking. The instructor needs to question the student, learn some of these preconceptions, and then help to steer the student in a direction that will lead to growth* .

    Science education needs to be more than a study of known or presently believed facts, laws, and theories. It needs to be more than a history course into science’s past or a study of the philosophy, sociology, or psychology of science. To quote from Benchmarks for Science Literacy:

    "Acquiring scientific knowledge about how the world works does not necessarily lead to an understanding of how science itself works, and neither does knowledge of the philosophy and sociology of science alone lead to a scientific understanding of the world" (Rutherford, et al., 1993).


    Asimov, I. (1966). Understanding physics: The electron, proton, and neutron. Barnes & Noble Books, New York.

    Black, A. & Ammon, P. (1992). A developmental-constructivist approach to teacher education, Journal of Teacher Education, 43(5), 323-335.

    Bronowski, J. (1973). The ascent of man. Little Brown, Boston.

    Campbell, J. (1988). The power of myth. Doubleday, New York.

    Casti, J. L. (1989). Paradigms lost: Tackling the unanswered mysteries of modern science. Avon Books, New York. 15-48.

    Chalmers, A. F. (1976). What is this thing called science?: An assessment of the nature and status of science and its methods. University of Queensland Press, St. Lucia, Queensland. 113-133.

    Elkana, Y. (1970). Science, philosophy of science and science teaching, Education Philosophy and Theory, (2), 12-35.

    Elkind, D. (1989). Developmentally appropriate practice: Philosophical and practical implications, Phi Delta Kappan, 71(2), 113-117.

    Giancoli, D. C. (1991).(3rd. Ed.). Physics: Principles with applications. Prentice Hall, Edglewood Cliffs, New Jersey.

    Gil-Perez, D. & Carrascosa-Alis, J. (1994). Bringing pupils’ learning closer to a scientific construction of knowledge: A permanent feature in innovations in science teaching, Science Education, 78(3), 301-315.

    Glasersfeld, E. von. (1989). An exposition of constructivism: Why some like it radical, JRME Monographs, #224.

    Glasersfeld, E. von. (1989). Cognition, construction of knowledge, and teaching, Syntheses, (80), 121-140.

    Hawking, S. W. (1988). A brief history of time: From the big bang to black holes. Bantum Books, New York.

    Held, B. S. (1990). What’s in a name? Some confusions and concerns about constructivism, Journal of Marital and Family Therapy, 16(4), 179-186.

    Hill, J. W. & Kolb, D. K. (1992). Chemistry for changing times (7th Ed.). Prentice Hall, Upper Saddle River, New Jersey.

    Hoyle, F. (1962). Astronomy. Crescent Books, London.

    Klemke, E. D., Hollinger, R. & Kline, A. D. (Ed.).(1980). Introductory readings in the philosophy of science. (revised edition). Prometheus Books, Buffalo.

    Kuhn, T. S. (1970). The structure of scientific revolutions (2nd. Ed.). University of Chicago Press, Chicago.

    Leet, D. L. & Judson, S. (1971). Physical geology (4th. Ed.). Prentice-Hall, Edglewood Cliffs, New Jersey.

    Matthews, M. R. (1994). Science teaching: The role of history and philosophy of science. Routledge Press, New York. 137-178.

    Mechling, J. (1990). Theory and the other; or, Is this session the text?, American Behavioral Scientist, 34(2), 153-164.

    Medawar, P. (1982). Pluto’s republic. Oxford University Press, Oxford.

    Millar, R. (1989). Skill and processes in science education: A critical analysis. (Wellington, J. Ed.) Routledge, London.

    National Research Council (Ed.).(1996). National science education standards. National Academy Press, Washington, DC. 107-108.

    Novak, J. D. & Gowin, D. B. (1984). Learning how to learn. Cambridge University Press, New York.

    Osborne, J. F. (1996). Beyond constructivism, Science Education, 80(1), 53-82.

    Phillips, D. C. (1995). The good, the bad, and the ugly: The many faces of constructivism, Educational Researcher, 24(7), 5-12.

    Phillips, D. G. (1994). Sciencing toward logical thinking (2nd Ed.). Kendall/Hunt, Dubuque.

    Piaget, J. (1964). Development and learning, Journal of Research in Science Teaching, 2(3), 176-186.

    Rauff, J. V. (1994). Constructivism, factoring, and beliefs, School Science and Math, 94(8), 421-426.

    Rhodes, G. & Schaible, R. (1989). Fact, law, and theory: Ways of thinking in science and literature, Journal of College Science Teaching, 18(4), 228-232, 288.

    Richards, S. (1987). Philosophy and sociology of science: An introduction (2nd ed.). Blackwell, Oxford.

    Rutherford, F. J. (Ed.).(1993). Benchmarks for science literacy. Oxford University Press, New York. 3-22.

    Sagan, C. (1980). Cosmos. Random House, New York.

    Speed, B. (1991). Reality exists O.K.? An argument against constructivism and social constructionism, Family therapy, 18(13), 395-409.

    Thompson, G. R. & Turk, J. (1991). Modern physical geology. Saunders College Publishing, Chicago.

    Trotter, A. (1995). Classroom constructivism, The Executive Educator, 17(10), 25-27.

    Trusted, J. (1979). The logic of scientific inference: An introduction. Macmillan Press, London.

    Wadsworth, B. J. (1996). Piaget’s theory of cognitive and affective development. (5th Ed.). Longman, White Plains, New York.

    Wallace, B. A. (1989). Choosing reality: A contemplative view of physics and the mind. New Science Library, Boston, Massachusetts.

    Wolpert, L. (1993). The unnatural nature of science. Harvard University Press, Cambridge, Massachusetts. 1-18.

    Yarusso, L. (1992). Constructivism versus objectivism, Performance & Instruction, 31(4), 7-9.

About the author...

Dr. Brian D. Campbell is a 1997 Ph.D. graduate in Science Education from the University of Iowa, Iowa City.  His science education research was in the area of causality and how students invent explanations for observed phenomena.  His science research area was X-ray diffraction of clays and shales from Pennsylvanian and Jurassic Period exposures in central and western Iowa and X-ray diffraction of clays and shales searching in central Iowa for bentonite deposits of a concentration sufficient for time dating of the samples. In addition to the Ph.D., Dr. Campbell earned a B.S. from the University of Wisconsin, Platteville in education and a M.S. from the University of Iowa in science education. Dr. Campbell is currently teaching science education and physical science courses and performing research at Southwestern Oklahoma State University at Weatherford, OK.

To get to the top of this page, click here.

To get back to the current issue of the EJSE, click here.

To get back to the EJSE's Archive page, click here.