Science...ultimate source of knowledge?--Not

"Towards an Initial Understanding of Science" (1)

by

Ian Johnston

2000

Now, here's an odd question: What exactly is science? The question is comprehensible enough, but the answer is far less clear. Our life and our thinking are so dominated by science, we have so much faith in science, and we refer so much of our education and our problems to people called scientists or technicians, that anyone would assume we have a clear idea about what this powerful thing is that stands at the centre of our culture. But it doesn't take much asking around, even among practicing scientists and students of science, or much reading in the local library to discover that there's a fairly lively dispute going on, some confusion, and a great deal of ignorance about just what we mean by the term.

But the really odd thing is that we (or most of us) are even not aware of this dispute. We don't discuss the matter much, at least in general public forums or the classrooms (not even in the science classrooms). There seems to be a general assumption that we all know more or less what science is, and the few things that we might not quite understand are all taken care of by a consensus of experts or academic philosophers with time on their hands. There are so many scientists around, and they do so many good things, obviously they know what they are doing. So we can turn our attention to more disputatious things like politics, art, or the next basketball dynasty.

In some things it may be the case that ignorance is bliss, but in this issue that might not be the best solution. Many of most important social and political decisions involve scientific questions, we are constantly being asked to evaluate scientific studies of particular issues, and repeatedly we are required to think about imposing limitations on science (or at least the technological implications of science). In one way or another, an understanding of science appears central to the exercise of our political responsibilities, to say nothing about our desire to know. So, if we have any wish to understand our world and ourselves better, it might be important for all of us to explore a little bit more beneath the surface of our immediate understanding of the term so that we begin to develop, if not a precise definition of what science is, at least an awareness of some of the complexities and ambiguities contained within a term that is perhaps too often misunderstood as something determined once and for all, even if we are not totally sure just what that determination adds up to.

Science as Fact Gathering

I suspect that most people asked to provide some initial definition of science would describe it as a means of understanding certain problems; science, they would claim, is characterized by a set procedure called, appropriately enough, the scientific method. To practice science is to apply this method to the analysis and explanation of events. And to the extent that one applies this method rigorously, one is doing science, thinking scientifically, being a scientist.

And what does this method require? Well, here again, there would seem to be widespread agreement that science is the business of carrying out measurements, collecting data, conducting experiments. Then, once a certain amount of quantifiable information has been gathered, the process of science constructs a theory to account for the problem under investigation. On the basis of that theory, certain predictions are made, which are tested. To be scientific is thus pre-eminently to concern oneself with the objective search for quantifiable information about something, constructing theories on the basis of that information, and then testing the theories with further experiments, and thus refining an understanding of the world. An important part of this method, so defined, is that the pursuit of science produces public results based on replication of quantifiable measurements and thus provides objective knowledge, free of the biases and subjective uncertainties of other forms of knowledge.

Now, while it doesn't take much observation of what a scientist actually does in his or her daily work to confirm that this description accurately conveys a sense of a large part of science, it is also quickly evident that this account of science is seriously inadequate. For no one can set out to gather data, make measurements, conduct experiments, and so on, until she is clear about what she is looking for, about what counts as evidence and what does not, about what to include in the data and what to exclude. And that information can only be provided by some theory constructed in advance of the observations. In that sense, theory must come before measurement.

It may well be true that in much of science, and especially in the education of young scientists, the theory is more or less taken for granted, so that all attention is focused on the best methods of collecting and presenting data within the context of that theory. However, it is clear that all such objective data gathering must be guided by a theory determined in advance, whether the scientist or would-be scientist understands the full complexity of the theory or is even aware of it.

Science as Falsification

The logical inadequacy of seeing science as simple fact gathering has led some thinkers to focus on theory checking as the essence of science. According to this view, science is an activity characterized by the constant attempts to prove theories about the world incorrect, to falsify theories. The concept goes something like this. Science begins with the development of a theory which contains in itself a means of being confirmed (i.e., tested). The major work of science involves subjecting such theories to rigorous and repeated testing, by making predictions based upon the theory and then conducting experiments or field work to see if these predictions do, in fact, take place. If they do not, then the theory has been falsified. It must be modified, recast, or abandoned, and the process begins again.

Science, thus, is not so much a means of discovering the truth as it is a means of removing error, and its conclusions, the statements which survive repeated testing, have a progressive character, that is, we become increasingly certain that they are not false and they become increasingly sophisticated and secure and, perhaps, closer approximations to the truth of nature, so long as they have not been falsified.

Now, this view has much to recommend it. In the first place, it emphasizes a crucial point: science concerns itself only with claims about the world which can, in fact, be checked independently and objectively (i.e., by experiment, measurement, and objective fact gathering). Any statement which cannot be so checked does not constitute a scientific statement and therefore is excluded from the enquiry. For example, the claim that the old Titan Prometheus made human beings out of mud is not a scientific claim. It may be true (please note this point), and the story may well encourage other stories, cults, poems, prayers, and so on. Many people may derive considerable emotional satisfaction from it. But this account generates no predictions which can be tested to confirm it independently by any form of experiment or measurement, and hence there is no way it could ever be falsified. Thus, that claim lies outside the realm of science.

By contrast, a claim that an object dropped from a certain height will fall towards the earth with a standard acceleration or that a set mixture of zinc and hydrochloric acid will produce a certain amount of hydrogen under certain conditions are claims that can be checked repeatedly by different people. If their results differ from the claim, then the statement has been falsified. Hence, those two claims above are scientific. The key idea here, one should note, is not whether the claims are true or false, but whether there is some means of checking them by observation publicly, repeatedly, and independently.

For this reason, a falsificationist would insist that the scientific status of many social science theories is suspect, if not spurious, because they cannot be checked against specific predictions. Freudian psychology or Marxist political theory, for example, claim to be based on scientific principles, yet it is impossible to test them against specific predictions. These theories do make predictions, but no matter what the outcome, the theories have an explanation. If class war does break out, that confirms Marx's theory (the proletariat has been turned into a revolutionary class); if class war does not break out that also confirms Marx's theory (the proletariat is not yet ready). In fact, with a great deal of social science theory, it is, in practice, very difficult to imagine an outcome that might not be covered by some detail or, alternatively put, it is difficult to organize a specific testable prediction which might serve to falsify the theory. Most scientists object to claims from psychics on related grounds. Psychics claim to have special powers which enable them to make predictions. Many of these predictions are scientific claims, in the sense they can be tested and falsified. But many psychics either refuse to take the tests under suitable conditions or, when the prediction does not occur, claim that there must have been some interference.

Incidentally, this notion that science deals only with statements which are falsifiable is one key to understanding the seemingly interminable debate between Creationism and the scientific community. This debate often presents itself as an argument about whether Darwin's account of how evolution works or the Genesis account of how the world was made is true or whether they both deserve to be considered equally. For reasons which will be apparent later in this essay, framing the issue in this way is extremely misleading, perhaps even impossible to resolve. In fact, the major focus of the argument in the context it which it is usually carried out is not which of these two claims is true, but rather the following question: Should we teach the Creationist account in the science curriculum of the schools? The vast majority of scientists answer this question with a resounding negative, because the Creationist account is based on statements which cannot be falsified, that is, they cannot be tested. The Genesis account generates no predictions of the sort generated by Darwin's theory (2). Therefore, it is not scientific and has no place in a science curriculum.

Curiously enough, many Creationists, while rejecting the point I have just mentioned about the importance of falsification, then use that concept to accuse evolution of not being scientific: since it cannot be observed (the time periods are too lengthy), it cannot be measured and checked. In other words, they claim that evolution is not scientific because it cannot be falsified. This, of course, is not the case. It is possible to make predictions on the basis of evolutionary theory and to test them against the fossil record. Such testing goes on all the time.

[Parenthetically, one might also observe here that at attention to falsification helps to clarify the nature of a scientific proof. Science does not prove things in the sense most people mean, especially when they talk about proving evolution. Science as falsification is designed to prove things wrong, to discredit theories, to locate weaknesses or anomalies in explanations. A scientific theory is established or confirmed or, I suppose, proven, if it repeatedly survives all the attempts to disprove it.]

Falsification theory does accurately describe a great deal of what scientists do much of the time: they make predictions on the basis of existing theories, and then seek to confirm or falsify those predictions. The results strengthen or weaken the case for a particular theory. A famous example is the prediction of the existence of the planet Neptune from the theoretical details of Newtonian celestial mechanics. When, on the basis of this prediction, astronomers searched in a certain portion of the sky where, according to the theory, the planet should be located, they discovered a planet whose existence they had not known before. This was a powerful reinforcement of the value of Newtonian physics as a basis for understanding celestial movements. A well-known negative example is Weissmann's experiment of cutting off the tails of twenty-two successive generations of mice in order to test whether the shapes of the tails in the new generations would change. The theory of the inheritance of acquired characteristics suggested that there should be some modification in the length of the tails of future generations under such conditions. When that did not occur, the result served to falsify the concept that acquired characteristics could be inherited.

The concept of falsification helps to illuminate also why there are so many important areas of human life which science cannot deal with, those areas where we cannot make clear predictions and evaluate them on the basis of observation and measurement. Our moral and aesthetic and legal judgments and our religious life cannot be quantified or framed in a language which enables us to make statements which we can confirm or refute in the normal scientific way. For about two hundred years (from about 1660 on) we made a concerted attempt to put such questions on a mathematical footing, in order to bring these difficult decisions under the authoritative banner of science, but the attempt failed. Science cannot tell us what we ought to do, or whether a particular art work is good or not, or whether a certain person is innocent or guilty, or whether God is a benevolent and loving presence. That is not to say that science is irrelevant in such debates, for it can clarify the facts upon which our decision often rests. But there is no method in science for telling us what we must or ought to do in the face of such facts. For example, science can provide extensive details of embryonic development, but it cannot inform us whether or not abortion is morally justified.

The falsification idea of science, however, also runs into logical difficulties for two reasons. First, there is the problem of a false observation. If a scientist makes a prediction on the basis of the theory, seeks to test it, and observes that the prediction does not occur, the problem may be in the theory, but it equally may be in the observation which results from the test. Hence, it would be premature and wrong to reject the theory. A well-known example is the apparent size of Venus. According to the theory of Copernicus that the Earth orbits the sun rather than the other way around, Venus must appear different sizes at different times. But observation did not confirm this prediction. Copernicus had no answer to this damaging point. Only years later, when observational techniques improved through the use of the telescope could people see that Venus did, in fact, change its apparent size. Strict falsification would have required the rejection of Copernicus's theory once observation failed to confirm a prediction. Thus, since science often has no immediate way of knowing if the failure of a prediction results from an inadequacy in the theory or from an inadequacy in the test, falsification often provides insufficient logical grounds for rejecting a theory out of hand.

A second problem with the falsificationist understanding of science is that a scientific theory may well survive falsification in some tests and not in others. If it is to be rejected the first time it is successfully falsified, then virtually no theories would survive their early stages. Scientists, in fact, often continue to work with theories some of whose predictions have not worked out properly if they are useful in other areas, partly for the reason already mentioned (there may be a fault in the testing experiments) and also because the theory is useful in other ways. A well-known contemporary example is Darwin's theory of natural selection. On the basis of this theory, Darwin himself predicted many intermediate types indicating the transition stages in the evolution of all species. Although numerous such transition types have been found, they have not occurred in the step-by-step fashion and in the numbers that the theory seems to demand. And it is no longer possible to argue, as Darwin did, that we would find such evidence with more investigation. Nevertheless, science does not reject Darwin's theory because it is so spectacularly successful at explaining so many other things (and certain modifications in the theory can take care of some of the problems created by this prediction), and it generates all sorts of other predictions which scientists wish to investigate. Similarly, modern physics has revealed important inadequacies in Newtonian mechanics in certain forms of enquiry. However, Newtonian mechanics is still extraordinarily successful at explaining and predicting so many things that it has retained much of its authority in modern physics.

Science as Social Process

The various logical problems encountered in trying to define science by specifying a universal method of enquiry (e.g., by objective fact gathering or by falsification) have led some people to attempts to clarify what science is by focusing on the work scientists do. In other words, science is defined by the activities of the scientists; what they carry out is what science, in fact, is.

What does this involve? Well, scientists carry out research in groups (research teams) usually at large institutions (universities, corporate research and development centres, and so on). What does that mean? In practice, this amounts to something like the following. Once he becomes a member of a research team, a scientist accepts as a working agenda a particular theory for a natural phenomenon which is shared by all members of the team (usually he has been specially educated for his position on such a team). This working assumption defines not only the scientist's specific area of interest but also his fundamental theoretical understanding of that area, which is accepted, for the purposes of the research program, as established and beyond dispute (accepting the common theoretical assumption is a necessary qualification for membership on the research team). Science is the activity he then undertakes to expand, refine, and improve the ways in which the theory provides an understanding of the particular area under scrutiny.

Such activity does not involve repeatedly testing the core assumptions which constitute the heart of the theory. To abandon these, the scientist would have to abandon that research project and join another guided by a different fundamental theory. Hence, in any one field of enquiry there might be competing research teams working with different theoretical assumptions, each seeking, in competition with the others, to establish its core theory as the most fruitful (i.e., the best source of useful and confirmed predictions). These research teams might share a common concern with scientific statements (of the sort mentioned above), but having different theories they could have important disagreements about what constitutes an effective test of their theory.

The competition between these different research teams would continue for as long as they all continued to attract new participants. If one emerged as richer and more fruitful in useful predictions, it might over time displace the others by attracting most or all of the participants in that area of enquiry (especially if it gained a dominant position in the university science curriculums). In other situations, rival research groups might continue to exist for a long time.

Consider, for example, two rival theories of inheritance in the history of modern genetics: one claimed that the reproductive material was directly affected by the living experiences of the adult parent, and thus the offspring could display characteristics acquired by the parent during the parent's lifetime; another claimed that the reproductive material was independent of the parent's experiences, and thus the offspring's inherited characteristics would not be affected by anything acquired by the parent.

On the basis of these two very different assumptions, different groups of scientists exploring inheritance organized and carried out different research programs. Each member of a particular group enthusiastically endorsed without question the fundamental assumption of the group's theory and repudiated the fundamental assumptions of its rival. They produced different accounts of particular phenomena, and each continued to prosper so long as it attracted adherents. Historically, the former theory (of acquired characteristics) dominated Soviet biology, and the latter theory (no inheritance of acquired characteristics) dominated biology in the West. The preponderant success of Western genetics over time led increasing numbers of international biologists to enter research programs where the basic assumption was that inheritance was independent of the parent's experience. And this has now become the dominant theory in modern genetics.

This emphasis on the sociological dimension of science (that is, on the ways in which science is essentially characterized by group decisions made by professional scientists often in a climate of competing theories) introduces something crucially different from the idea that science is a universal methodology agreed upon by all or justified by universal rules which transcend human groups. The emphasis on science as the product of a human activity makes science something much more complex and ambiguous. For if science is what scientists say it is, if it is the product of human decision making in groups of people who themselves determine what the rules are, then, in some sense, science is a form of knowledge constructed for particular human purposes in a cultural and social context, rather than a single correct method independent of individual or group human motives.

If this is the case, it raises important issues. For human beings, after all, are motivated by many things both individually and collectively, and we are entitled to wonder just how much these enter into decisions about research projects, about the fields of enquiry, about the preferred theories for understanding something, and about what constitutes confirmation or falsification of a theory. The decided preference in the Soviet Union for a genetics based on the inheritance of acquired characteristics, for example, clearly had a direct relationship with the political beliefs of the communist state, with which such a view of inheritance could be so easily reconciled, so as to give scientific endorsement for the ideals of communism. Natural selection, by contrast, has an equally clear affinity with liberal capitalism.

Let us dwell on this point longer. If we see science as a group activity carried on by professionals requiring adequate funding and dreaming of fame and fortune, with a large part of the work done in government-funded universities or international corporate research and development departments, and the scientists who make up the group, especially at the senior levels, consisting almost exclusively of men, then we might want to raise some of the following questions: To what extent are the selected fields of enquiry, the preferred theories, the methodology, and the results affected by factors which we thought extraneous, for example, by the political ideology of the government with large research grants at its disposal, by the possibilities for a quick market profit for a pharmaceutical company wishing to sell pills which chemically intervene in ways not yet fully determined, by the gender bias of men's perceptions of the really important problems, and so on? To what extent are theories and the evidence supporting them manipulated, consciously or otherwise, to provide the sort of knowledge which is most in demand by the research group or which is most profitable to its employer? What sort of a hearing do such groups permit rival theories which threaten their prestige? These questions become increasingly relevant, the more we realize that the practice of science, what really goes on in the research groups, is conducted by particular people often receiving generous financial support from institutions with a large stake in the outcome and surrounded by a professional culture with its own rules, hierarchies, and prizes.

In recent decades this view of science has sparked a number of penetrating critiques of the alleged objectivity of science. To take one common example: Is much of our scientific theory characterized by images of dominance (e.g., the brain dominating the body, the sun dominating the solar system, the nucleus dominating the cell) because that is really the best way of understanding nature, or are these models accepted as initial working assumptions and evidence found to support them because the image of dominance fits a preferred model of the western world's sense of our domination over nature or because it fits the male way of thinking first and foremost in terms of power, competition, antagonism, and dominance (rather than, say, in terms of co-operation) or because it meshes so easily with a capitalist economic and political ideology?

Similarly much attention has been paid to the way in which the interests of the human researchers select some topics for study rather than others (a common observation in connection with criticism of the lack of research into breast cancer, a serious illness for women). Another problem which has received much attention from this view of science is the various ways in which the research groups in a particular area seek to maintain their power (which can be enormous) by marginalizing alternative research programs or alternative methods of understanding particular areas of nature, anything which might pose a threat to an existing power base and the social and economic rewards it brings (a common objection to the hostility of the medical profession to non-Western treatments, like acupuncture and naturopathy).

It is not my purpose here to evaluate such critiques. But it is important to notice how they call attention to the ways in which what we call science is, in part, a social process, the meaning and methods of which are shaped by the activity of human participants. To the extent that such an approach reminds us that science is such an activity, it performs a useful service, especially for those of us who think that science can be easily and permanently defined in terms of a single specified objective method and who, thus, may too easily overlook human agendas (3).

Relativism

What are we to make of this? If science is the activity of scientists, and if they set the rules in accordance with a professional consensus in answer to certain cultural pressures, does this mean that science has nothing to do with the truth, that it is something constructed to meet the needs of human beings as these arise and that it can change to fit later needs or other social pressures? Well, some thinkers about such matters reply in the affirmative. Yes, they say, that is precisely the case. Science has nothing to do with the truth. It is simply a constructed means of knowing, just like all other forms of knowledge (e.g., poetry, philosophy, religion), created by human beings to meet particular needs, including the specific ambitions of those participating in it or the larger agendas of those encouraging it.

This is the case for a number of reasons, among which we can include the problem of language. Science claims that it describes reality, but descriptions require a language, and all languages are culturally determined, relative, and subject to change over time. Languages, in other words, are made by human beings, not given to us with the nature of things. A scientific theory is just another language we have made up. We believe that the Earth orbits the sun not because that is true, but simply because for historical reasons we decided we liked the language and imagery of Copernicus's model as modified by Galileo, Kepler, and Newton, and gradually began to use it, until it took over from the old language and images derived principally from medieval Christian scholasticism. Why did we make the transition? Well, that had something to do with the accidental shifts of European culture in the sixteenth and seventeenth centuries.

Such a view commits us to a complete relativism, namely, to the view that science has no special claim to the truth, any more than does any other theory or vision of nature. Furthermore, there is no appeal to some objective truth of nature which would enable us to approve some theories and discard others. All systems of belief are relative, that is, they are true if people believe in them, and we cannot adjudicate between them. People adopt systems of belief (in religion or science or something else) or discard them for reasons which have nothing to do with their truth (since they are all equally true and equally false). There is no firm structure to reality against which we can check the various fictions we construct in order to deal with the world. If a society finds a certain system of knowledge useful, it will call that the truth and dismiss rival theories. Later it will find something more useful (for historical reasons, which are accidental) and will discard the old truth and substitute a new one. But this is not progress. It is simply the substitution of one artificial system for another.

Realism

Such a critique, which disestablishes science's claim to the Truth (with a capital T) or, indeed, to any privileged position in comparison with any other system of belief, while common in many social science departments, generally is firmly repudiated by scientists themselves. Of course, science has a social component, they admit, and it is important that we remain alert to the ways in which that social component can affect what goes on. Nevertheless, there is a reality out there, it has a permanent structure, and the language of science does provide an increasingly accurate map of that structure. If there are mistakes and corruption of the ideals of science, created by the victory of greed over the desire to know or if we have had different beliefs about that structure, the process of carrying on the scientific enterprise will correct the mistakes. We really do know more about nature than we used to, even if there is still an enormous amount we do not know. The best indicator of this is our ability to make more sophisticated predictions about nature. These confirm the growing accuracy of our theoretical models and indicate to us that the understanding of the world which science provides does give us over time improved access to or at least better approximations of the truth of nature.

This claim that science does indeed provide an accurate map of reality (or brings us increasingly closer to reality) is called realism. According to such a view, the equations of physics really do correspond with the eternal laws of nature, the principles upon which God, the Divine Mathematician, created the world. And our images of molecules, electrons, and electromagnetic forces are, indeed, accurate depictions of what is really out there in the permanent structure of nature itself. While the picture is by no means complete and some parts of it may be inaccurate or still approximate, scientific realists claim, nevertheless, the project launched in the Renaissance of mapping reality is still dynamic, and we are gaining more and more secure knowledge of its details. Hence, it is entirely appropriate to think of science as a progressive search for and attainment of the truth of things. Some scientists (though by no means all) will even claim that we are only a few steps away from a set of equations which will explain finally the complex behaviour of matter.

Science and the Truth of Things

There are a number of problems with the claim of the realist that science maps reality (which is conventionally called the Correspondence Theory of truth, based on the notion that scientific theories and laws correspond accurately to reality). A major problem, well known to philosophers, is this: Can we be sure that the act of observing or experimenting with something gives an accurate picture? Any observation is a perception in the human mind, not an objective fact outside the human mind. So the reliable relationships our scientists establish are based, not on external reality itself, but upon perceptions of that reality (i.e., upon our inner processes of perceiving and imagining, not upon what is really out there). Since I have no way of contacting the external world except through my perceptions (which are inner), I have no certainty about what reality is really like. Even if perceptions are shared and confirmed by others, they are all still human perceptions, not the essence of the things being described (which we can never know in themselves). And the models we construct (like solar systems or molecules) are pictures in our minds. We have no way of checking them point by point against reality.

And this problem raises the question about the status of the regular scientific knowledge I do acquire. Consider, for example, the importance of mathematics in science. Why does so much of nature always seem to act in accordance with mathematical laws (e.g., gravity, acceleration, chemical reactions, planetary orbits, electrical discharges, inherited characteristics, and so on). There are three possible answers. First, it might be pure chance. Given the frequency and power of mathematical models, this seems unlikely, but it is possible. Second, it might be true that the world is run on mathematical principles and that, thus, mathematical equations describing the natural world are true, and the realist position is, indeed, correct: God is a complex mathematician. Third, it may be that our thinking processes are governed, in part, by mathematical principles. Hence, we can understand many aspects of the mysterious complexity of the natural world clearly only in mathematical terms (or at least that is the only method of securing wide agreement about how the natural world works). So far as I can see, there is no way of deciding between the second and third options, since there is no way of perceiving the world outside myself. If I wear red sunglasses, everything will appear as shades of red. If I am unaware of those glasses, does this mean that the reality is, in itself, shades of red?

Then, there is the famous uncertainty principle. It is impossible to know anything scientific about nature without observing. But how can I be sure that my act of observation does not change the natural phenomenon I am observing? How do I know what really happens when I am not observing? This is, of course, the famous problem of the refrigerator light. How do I know with certainty that it really goes off when I shut the door? If I cut a hole in the door or hide in the refrigerator, then I have changed the situation, and I am no longer observing the process I am making conclusions about (the refrigerator's normal operation has been changed so that I can observe). This example may be trivial, but there are more serious possibilities. For example, to see something very small under high magnification, the scientist has to put a powerful energy source (e.g., light) through the object (e.g., a slide) and often introduce various dyes. But how can the scientist know that the high energy now passing through the object or the dye which enables her to see it clearly has not changed it in some way? Or, to take another example, if I put a measuring device into a river to study the river flow, I have, in effect, altered the flow of the river. What is it really like when the measuring device is not there? I have no way of knowing. If the very act of observing something means that I have changed in some manner what I am purporting to study, then, once again, I have no access to reality in itself (i.e., independent of my acts of observation, which change the situation).

Well, if, because of our perceptions and status as observers, science cannot deliver the truth of things, does this mean then that science is, as the relativists claim, just one more fiction about the world, one more story human beings have invented about nature, with no more claim to the truth than any other story from myth or folk tale? As mentioned above, some thinkers frequently make such claims. But this is a dangerous idea and misrepresents by oversimplification the status of science.

To raise some doubts about this relativist view, let us return to the question of mathematics mentioned above. Mathematics is central to the claim people make in defense of the privileged position of science, for it often serves to counter the argument about descriptions having to be framed in language which is always culturally determined and hence relative to particular cultural groups. Mathematics, many scientists claim, transcends culture differences. Equations, unlike poetry or prayers, remain the same in different parts of the world and in different translations. And mathematical explanations hold in any part of the world; there is not one equation for acceleration in North America and another in Asia.

Thus, an understanding of the world which expresses itself mathematically does not carry the same culturally determined and limiting range of meanings relative to very particular social contexts. Furthermore, the fact that a mathematically based understanding so often generates useful predictions strongly suggests that, even if it does not describe reality exactly, it give us a means of apprehending it which is more accurate (and certainly more useful) than other systems. Whether the universe really is exactly like Newton's model of it, we cannot know. We do know that Newton's mathematical model gives us access to that system, enables us to make predictions about it, and suggests all sorts of other things to look for. Hence, the language of mathematics has a universal quality unknown in other forms of description. Hence, the relativists are incorrect. To the extent that science relies upon the language of mathematics as the basis of its theories and predictions and tests, it occupies a privileged position among systems of knowledge.

To make this claim about mathematics is not therefore to establish that science must be true in the sense that a Correspondence Theory might demand. This cannot be the case, for science cannot deliver a map of the world which it can prove matches details of reality in itself. Nor does it make much sense to talk about getting closer to the truth or approximating the truth of things. We have no access to the truth by which to measure such progress in any public system of knowledge. Hence, questions about the relationship of science to the absolute truth of nature are, as mentioned before, meaningless. On the other hand, science does deliver some means of picturing reality, an understanding of things about which we can agree and upon which we can act without reference to particular cultural restrictions. Science enables us to grasp aspects of nature (whether they are really like our models or not), to make predictions about those aspects of nature, to confirm, abandon, refine—in short, to enlarge our knowledge of the world and our power over it (intellectual and practical power).

Furthermore, science achieve all these things in a way that enables us to conduct many of our arguments about the world in a context free from the immediate limitations of any single culture and to base our decisions together on a shared understanding of the world. To say this is not to deny that sometimes science is abused and manipulated, consciously and unconsciously, for various reasons or that scientists can often be arrogant, ambitious, and misleading. Nor is to deny that the wholesale application of science to the understanding of and solution to human problems has created special difficulties. It is, however, a reminder that science is a uniquely privileged way of understanding the world and that the best way of dealing with the various problems occasioned by the pursuit of science is not a rejection or a demotion of science but rather better science. And the first step toward that worthy goal might be a better understanding of what the activity is and is not.

So Where Does that Leave Us?

This essay began with a question: What is science? It has not fully answered that difficult issue, but on the basis of the above review, let me suggest some characteristics of the activity we call science. These I offer here as items to discuss, rather than as confirmed certainties free of any ambiguity.

First, science involves formulating theories about the world and framing those theories in statements which yield predictions which can repeatedly, publicly, and independently be tested by observation. Any theory which cannot be so tested, which cannot, that is, potentially be falsified in this way is not part of science. Science teaches us to be very skeptical about any system of knowledge which cannot meet this criterion.

Second, science is carried out by scientists working in groups, often in competition with each other, in a cultural context. This cultural context plays a large role in all aspects of their work. We need, therefore always to be very skeptical about science itself, assessing as best we can how a particular theory or test may be shaped by the self-interest or the bias of the scientists or their employers, a human motive lurking beneath the disinterested language of the theory or the results.

Third, attempts to evaluate theories by some appeal to the truth of nature are inherently pointless. Such evaluations should carried out on the basis of how rich, coherent, and productive a theory is (how many useful and reliable predictions it generates, how comprehensively it explains what we wish to know about, how much it stimulates potentially interesting or useful new research). On this basis, we are entitled to evaluate science, to rank theories, and to discard old ones. It is not that case that all theories are equally good.

Fourth, at its best, science should be open ended, that is, always ready to test new theories which meet the criteria mentioned above. The public nature of scientific testing requires as much access as is practically possible for new challenges. This, in fact, is the best remedy for some of the most important problems arising out of the practice of science. It should be clear that this point does not mean that we should consider all theories equally valuable.

Finally, we should devote far more time than we do in our public education system, especially at the post-secondary level to making sure that all students, including especially students in science, have some introduction to these issues, rather than continuing on as we do now, for most Arts students simply ignoring the issues and, for most Science students, plunging them into the approved methods with no reflection on the wider implications of and justifications for the activity.

Notes:

(1) Details of this essay are drawn from a number of sources, particularly the excellent book by A. F. Chalmers, What is this thing called Science?, Second Edition (Indianapolis: Hackett, 1994). Anyone wishing to pursue some basic ideas about the nature of science at greater length and in more complex detail should read this well-known and justly popular introduction to the topic. The present essay deals with some very complex issues in a necessarily cursory fashion.

(2) This claim is not quite true. It is possible to generate at least one very specific prediction on the basis of the Genesis account, namely, that if all the species were created at once, then we should find their remains at every level of the fossil record. The growing (and eventually overwhelming) evidence against this prediction was a major reason why the doctrine of the fixity of species was abandoned by virtually all biologists.

(3) An amusing and instructive example of the ways in which social and political processes can interfere with scientific and technical work occurred in July 1975, in the link up between the American Apollo 18 and the Soviet Union's Soyuz space vessels. Normally such link-ups were made with a coupling rather like an electric plug, with a "male" component being inserted into a "female" component (the terminology is standard in electrical appliances). Both nations had used such couplings before in their own space programs. In this instance, the engineers had to design an extremely expensive and complex interlocking system, because (in the words of F. Gwynplaine MacIntyre, "on that historic 'first date' between the two rival space agencies, neither participant was willing to take the 'female' role, which would require its spaceship to be penetrated by the other nation's 'male' hardware" (Letter to the Editor, Atlantic Monthly, February 2000).

A similar story concerns rival interpretations of behaviour in a primate group. The male researchers described the group in terms of "competition," "dominant males," and so forth. The female researchers, observing the same primate group, described it in terms of "cooperative behaviour," "sharing," "assistance," and so forth.