Philosophy and Science – Or The Gospel According to Popper

by Nancy Swanson, Ph.D. Physics, Florida State University

Science and philosophy have historically been intertwined; science was developed as a discipline in an attempt to answer some basic philosophical questions. These are: How do we know that there is anything “out there” independent from our subjective experience? And What is real (or true)? The astute reader may have noticed that other disciplines have also set about answering these questions. They are called religion. The difference supposedly being that religion relies on belief while science and philosophy rely on logic.

Philosophers have debated for millennia about what we can and can’t know and about what is and isn’t real. Plato, Socrates and Pythagoras claimed that the only true reality is idea. Physical reality is a manifestation of our ideas [1]. Whatever we experience is, by definition, less than our ideas (idealism). More recently, philosophers have debated about what can or cannot be proved. How can we justify using past observations as a basis for generalizations about what we have not yet observed. It has been acknowledged by most philosophers that nothing can be proved true. One illustration of this can be seen in what is known as The Ravens Problem.

Induction versus deduction

The philosophy game is played by making a statement (hypothesis) then using the rules of logic to prove or disprove the statement. The ravens problem goes like this:

Statement #1: All ravens are black.

How do we go about proving our statement is true? We employ the scientific method of observation. We go out in the world and look for ravens. We need find only one non-black raven to prove the statement false, but how many black ravens need we find to prove it true? No matter how many black ravens we tally, we cannot guarantee without doubt that the next raven we see will be black. We can, at best, assign a probability that the next raven we see will be black, but we cannot prove the statement is true by observing black ravens.

You might say, “Come now, don’t be ridiculous. Of course there is some small probability that the next raven we see will not be black, but we have observed so many black ravens we can be confident that we have proved our statement.” Okay then. By strict rules of logic, the statement

Statement #2: All non-black things are not ravens

is equivalent to the previous statement, all ravens are black. Therefore, if finding black ravens “proves” statement #1, then finding green tennis balls “proves” statement #2. Because statement #2 is equivalent to statement #1, I can “prove” that “all ravens are black” by finding green tennis balls. Or white shoes. I don’t even have to go outside and find ravens to prove my hypothesis. Abandon hope, all ye who worship logic!

The Ravens Problem highlights the weaknesses of inductive logic. Inductive logic is when you have some data (observation of black ravens) from which you form a hypothesis that explains the data (all ravens are black). A large number of possible resolutions to the ravens problem have been proposed. The various arguments for and against each will make your head spin, but I don’t believe there has been consensus among philosophers. And anyway, other paradoxes concerning inductive reasoning exist, along with more gnarly discussions. The bottom line is that it seems that we cannot confirm or prove a statement (hypothesis) made by inductive reasoning.

The situation is not improved by switching to deductive logic. Deductive logic starts with a premise or premises that are assumed true. Given that the premises are true, we then determine what can logically follow and call these conclusions. The premises themselves are unproven and unprovable, they must be accepted at face value, or by faith, or for the purpose of exploration. For example, using deductive reasoning we simply state without proof that:

1. All ravens are black.

We have a friend who has a pet raven called Berthe. We can logically deduce that Berthe is black, without ever having observed Berthe. We have no way of knowing whether or not the premise is true, but assuming that it is, Berthe is most certainly black. Notice that the reverse is not true, i.e. just because Berthe is black does not mean she is a raven. All scientific theories are built up in one or the other of these two ways.

One of the jobs of the philosophers is to establish the rules for what does and does not constitute an acceptable scientific theory. While the philosophers are by no means in agreement, Karl Popper’s philosophy is most widely accepted. Among scientists, that is. Popper’s view of science is highly idealistic, which is probably why it appeals to scientists. In fact, I would venture to guess that most scientists don’t know that there are any other philosophies concerning science! They’ve probably never heard of Popper either.

Popper believed that induction has no place in science. According to Popper, scientific theories should be formed by deduction. He agreed that we cannot confirm or prove a theory true, but that a good scientific theory has the possibility of being falsified or disproved. The experimental data must trump the theory, no matter how beautiful the theory. Popper had an ax to grind with Marxism, which he said cannot be falsified because no matter what happens, a Marxist can somehow fit it into her theory. Popper referred to such theories as pseudo-science. Any theories that cannot be subjected to falsification tests do not qualify as scientific, in Popper’s view. For scientific theories, we set up experiment after experiment to try to prove them false. If we fail to falsify the theory, then it is not necessarily true, but it is robust and can be relied upon to give accurate predictions. The theory then becomes a working model.

If any one scientist is too egotistical and too attached to his or her particular conjecture to reject it based on experimental data, that should pose no problem as presumably there are other scientists who are wedded to different conjectures, so that it should all work out fine in the end. Even if everyone refuses to believe the new data, a new theory will eventually emerge when the old crowd dies off. “For Popper, a good or great scientist is someone who combines two features… The first feature is an ability to come up with imaginative, creative, and risky ideas. The second is a hard-headed willingness to subject these imaginative ideas to rigorous critical testing. A good scientist has a creative, almost artistic, streak and a tough-minded, no-nonsense streak. Imagine a hard-headed cowboy out on the range, with a Stradivarius violin in his saddlebags. Perhaps … you can see … the reasons for Popper’s popularity among scientists.” [2]

The criteria for a good scientific theory therefore, are that it has held up to intense scrutiny and has not yet been falsified. This does not make the theory “true” or “real” but it qualifies as a good map of reality. It is unfortunate that many, scientists included, mistake the map for the territory. And it is shocking that many, scientists especially, make the claim that thus and so has been confirmed or proven.

Popper’s criterion of falsifiability in practice

A criticism of Popper’s philosophy of science is what I will call Popper’s Decision Problem. It goes like this: Suppose you want to build a rocket. There exists a model for thrust and lift and so forth that has been used successfully in the past to build rockets. But a new model has come along that has never been tested. According to Popper, there is no reason to believe one model is more true than the other, as confirmation is not possible. Neither model has been shown to be false. Which model would you use to build your rocket? Popper was aware of this problem and was unable to give a response that did not involve some sort of confirmation.

The criteria that scientists claim to use for a good scientific model are very close to Popper’s ideal [3]:

1. The model must agree with observed data.
2. The model must make predictions that allow it to be tested. It must be possible to disprove the model. The model may need to be modified to fit new data or it may need to be discarded entirely.
3. The model should be aesthetically pleasing. It must be simple, neat and contain the fewest possible assumptions.

Criterion #3 is not strictly necessary, but is used in the same manner as Ockham’s Razor, i.e. when two competing theories explain the same data, choose the most aesthetically pleasing, which is also the simplest. It is supposedly also used as a bellwether, i.e. if a theory gets too cumbersome with many ad hoc adjustments and modifications then it is ugly and needs to be replaced. Such was the case with the Ptolemaic geocentric model of the solar system with its epicycles and recalibrations.

I claim that Popper’s view of how science is done is idealistic for the simple reason that, while this may be how science should be done, it is not. Suppose our statement, All Ravens Are Black is given as true. It follows that if I go outside and see a raven, it will be black. If Popper were correct, then a scientist trying to falsify our statement would go outside and actively look for white ravens. Or green and purple ones. What in fact would happen is that our scientist would go outside and look for black ravens in support of the statement. Suppose he is out there looking for black ravens and a white one flies by. He likely wouldn’t see it because he is only looking at black birds and then determining if they are ravens. He is paying no attention whatsoever to white birds. If he does happen to see one, he will wonder if he made a mistake. The sun was in his eyes and it only seemed white from that angle, so it must’ve been black after all. Thus he will justify discarding that data point and continue his search for black ravens.

Suppose further that our intrepid scientist runs across yet another white raven. Now he is in a pickle because all the other scientists in the world also believe that ravens are black. If he publishes his results they will not believe him. He will either abandon the project to go do something else, or he will go ahead and try to publish. The reviewers will recommend that his paper not be published because everyone knows that ravens are black. They will question his methods or impugn his integrity. Was he wearing sunglasses or another optical device when the so-called observation was made? Did he record the time, date, temperature, atmospheric pressure and GPS coordinates so other scientists can repeat his observations? Scientists who have a stake in the theory (a lifetime of work and publications) will claim that he misunderstood or misinterpreted his observations. The raven probably had some snow on it or something. Kill the messenger.

Suppose still further that a second scientist, being more open-minded than her fellows, went out where our first scientist was and she also spotted a white raven. Now we have a newspaper headline: Scientists Find White Raven; Black Raven Theory In Trouble. Then the public never hears about it again, but six months later the Raven Theory is still intact. It now looks like this:

Theory: All Ravens Are Black

Codicil: On the first Sunday after the first Full Moon after the Spring Equinox, white ravens might possibly be observed.

Books for burning

At first glance, it appears that scientific theories cannot be falsified. We have already established that they cannot be confirmed. What a deal.

Another reason that scientists agree with Popper’s philosophy of science is that it gives them an excuse to reject anything they don’t like or don’t understand. Sort of like Popper and Marxism. Just say it isn’t science and move on. A case in point is Rupert Sheldrake.

Rupert Sheldrake is a renegade biologist. He started out as a perfectly respectable biologist until he got obsessed with all the things we don’t understand that we pretend to understand. Or things we pretend don’t exist or just plain don’t try to understand. Such as: how do pigeons home; how do pets know when their owners are coming home; how can we tell when we are being stared at and so on. He has written several books, one of which, A New Science of Life, proposes a “morphogenetic field” containing information that can be accessed via “morphic resonance.”

In Popper’s world scientists would grab hold of this theory, figure out what sort of predictions can be made and go test them (by falsification, of course). What actually happened was that the late Sir John Maddox wrote in an editorial in Nature (1981): “This infuriating tract… is the best candidate for burning there has been for many years.”

Who among us has not experienced the sense of being stared at? Yet scientists will tell you it does not exist (never mind the explanation). For many years scientists claimed that asteroids do not exist because rocks do not fall from the sky. Suppose that a rock fell from the sky into your yard. You pick it up and take it to the local university where the scientific priests can be found and they tell you that it could not have fallen from the sky. Maybe a bird dropped it. Do you believe them? Or do you believe your own direct experience? If you believe them, you begin to doubt that you saw what you saw. And if any scientist dares to explore rocks falling from the sky, she runs the risk of being publicly humiliated, denied funding, and encountering difficulty publishing. She’d better have another source of income. What part of this do you consider logical, reasonable, or rational?

There are other philosophers who expound different ideas about how science is or should be done. Because in these days philosophers are not scientists, they don’t actually know how science is done. And scientists don’t listen to philosophers anyway; with one notable exception.

Thomas Kuhn to the rescue with the ‘paradigm’

In 1962 Thomas Kuhn published a book on the philosophy of science called The Structure of Scientific Revolutions. Kuhn started out as a physicist; then wandered into the history of science and from there into the philosophy of science. Kuhn claimed, as I do, that our world view is a result of the current scientific paradigm. He actually coined the term “paradigm” in this sense. The paradigm is the big picture, which may include many theories and hypotheses. Classical Newtonian physics is an example of a paradigm that includes kinematics (baseballs and such), optics, electrodynamics, gravity and the like. Each of these has various theoretical models that yield predictions and so forth.

According to Kuhn most all of science is practiced as “normal” or “paradigm-based research,” i.e. research within the currently accepted scientific paradigm [4]. No-one actively tries to falsify or overthrow the paradigm, nor should they according to Kuhn. One needs to have a paradigm as a starting place within which to operate. It’s our belief system. “No part of the aim of normal science is to call forth new sorts of phenomena; indeed those that will not fit the box are often not seen at all. Nor do scientists normally aim to invent new theories, and they are often intolerant of those invented by others. Instead normal-scientific research … seems an attempt to force nature into the preformed and relatively inflexible box that the paradigm supplies.”

“Normal science” consists of three major experimental and observational endeavours, all of which constitute fact gathering in support of the current paradigm. They are:

1. Determining physical properties with greater precision (stellar positions, specific gravity of materials, compressibility of materials, spectral properties, electrical conductivity of materials, etc.).

2. Demonstrating agreement between theoretical predictions and experimental data to greater accuracy by improving experimental apparatus or finding new ways to demonstrate agreement.

3. Articulation of theory:

a. Determining physical constants with greater precision (speed of light, universal gravitational constant, Coulomb’s constant etc.)

b. Exploring empirical laws (Boyle’s law, for example)

c. Determining the limits of applicability of laws within the paradigm.

Normal science also consists of three corresponding theoretical pursuits:

1. Using the existing theory to calculate expected values of physical properties that can then be measured.

2. Adjusting the theory to more closely agree with experimental observations or modifying the theory for special cases (adding air resistance to kinematic calculations, or modifying the value of a constant in the calculations, for example)

3. Reformulating the theory to extend the model (adding another mass to calculate the gravitational interaction for three masses instead of two, for example).

“Normal science” is not aimed at disproving the dominant theory

Two things are immediately noticeable about normal science. One is that much of it is excruciatingly boring. Even scientists find it boring. This is why they invented positions now filled by graduate students and post-docs. The other, more important thing is that no-one is trying to disprove anything. All of the effort is going into supporting the current paradigm. Popper’s Decision Problem is moot because there are no theories that do not support the current paradigm.

How then do theories and paradigms change? In the course of normal science small puzzles pop up; little pieces of data that do not fit within the paradigm. Often these are simply not seen. Kuhn compares this with a psychological experiment published in 1949 by J.S. Bruner and Leo Postman [5]. In this experiment subjects were asked to look at a sequence of short exposures of playing cards and identify them. Inserted into the deck of playing cards were anomalous cards, a red six of spades and a black four of hearts. Responses to the anomalous cards were divided into four categories: they were reported as normal (i.e. the red six of spades was simply reported as a six of spades—the authors called this a “perceptual denial” [6]); they were reported as anomalous but incorrectly (e.g. a purple four of hearts); the subject knew there was a problem but couldn’t figure it out (one subject said, “I don’t know what the hell it is now, not even for sure whether it’s a playing card.”); and finally some subjects reported what they saw (a red six of spades) though they had to be exposed to it many times before they saw it correctly. Kuhn argues that scientists, being human after all, do the same.

Once the anomalies are seen and identified, scientists work to improve the detection apparatus. If that doesn’t resolve the puzzle, they modify the existing theory to accommodate the data for the special case. Usually the great discoveries of science come from solving these little puzzles.

Crisis of anomalies instigates a paradigm shift

Occasionally an anomaly arises that resists all efforts to resolve it within the paradigm. Every so often these resistant anomalies build up into a critical mass where scientists start to lose faith in the paradigm. Kuhn calls this a crisis. Only during a crisis are new theories formed outside of the working paradigm. Once a new theory arises that can satisfactorily predict the anomalies, along with the data that were not in question, then a “paradigm shift” occurs that changes our world and our world view. This was the case in shifting from the geocentric view to the heliocentric view and in the revolution of quantum mechanics vs. Newtonian mechanics.

According to Kuhn it is a good thing that scientists are resistant to change because our knowledge is added to considerably during the course of normal science and it wouldn’t do to distract the scientists too much from their belief system. We can’t be having paradigm shifts every time our data doesn’t come up to snuff. And anyway everything works out in the end because science is a self-correcting system that will root out its problems even if entire generations must die off before new ideas can be accepted.

This would be fine if you believed that our knowledge is always increasing and we are always getting closer to Truth than we were before. People assume this is so; that we are standing at some sort of pinnacle of knowledge and all who have gone before us are poor, ignorant fools. They grab the moral high ground, bludgeoning us with their version of reason and logic and insinuate that our only other choice is superstition, weeping and gnashing of teeth. They accuse anyone who dares to question science of trying to throw us all back into the dark ages.

Paul Feyerabend Against method

Not so, says Paul Feyerabend. First of all, Feyerabend saw science as a creative endeavor. The great scientists were anarchists, they were unafraid of breaking rules laid down by philosophers and they used any and all means available to them to aid in scientific discovery. In Feyerabend’s ideal world anything goes and any attempt to force scientists into a prescribed method will only serve to dampen creativity and create stultifying science [7]. But, while science started out to free us from the thought police in the One True Religion (Christianity), it has itself turned into the oppressor. He argues that we now need to be freed from the grip of the scientific establishment and he advocates a separation of science and state, like any good democracy [8].

Feyerabend claims that there are no objective reasons for preferring science to other traditions and ways of knowing. He says that [9], “Most intellectuals have not the foggiest idea about the positive achievements of life outside Western civilization. What we [have] in this area are rumors about the excellence of science and the dismal quality of everything else. …Western science has now infected the whole world like a contagious disease… Western civilization was either imposed by force, not because of arguments showing its intrinsic truthfulness, or accepted because it produced better weapons. … [R]ationalists have devised [arguments] to overcome difficulties. For example, they distinguish between basic science and its applications: if any destroying was done, then this was the work of the appliers [the politicians], not of the good and innocent theoreticians. But the theoreticians are not that innocent. They [emphasis his] are recommending analysis over and above understanding, and this even in domains dealing with human beings; they [emphasis his] extol the ‘rationality’ and ‘objectivity’ of science without realizing that a procedure whose main aim is to get rid of all human elements is bound to lead to inhuman actions. Or they distinguish between the good which science can do ‘in principle’ and the bad things it actually does. That can hardly give us comfort. All religions [emphasis mine] are good ‘in principle’ – but unfortunately this abstract Good has only rarely prevented their practitioners from behaving like bastards.” Feyerabend dismisses arguments in defence of science such as ‘science knows best’ (feeble) and ‘science works’ because [10], “Science works sometimes, it often fails and many success stories are rumors, not facts.”

Scientists today are largely ignorant of what the philosophers think or say. Feyerabend notices this and he calls us “uncivilized” because of it. I think that Popper’s version of science is unrealistic but it’s how scientists see themselves, although they don’t seem to understand the full implications (nothing can be known for sure). Kuhn’s version of science is how science is actually done, whether it should be (as Kuhn argues) or not. I agree with Kuhn that theories should not be thrown out at the first sign of trouble because they are still useful as models (and they are all only models) as we shall see. Feyerabend’s version of how science should be done seems the most fun and promising, not to mention open-minded. I leave this chapter with Godel’s Theorem (Kurt Godel, 1906-1978).

Gödel for science?

Gödel’s Incompleteness Theorem is a mathematical theorem that he did not intend to be applied outside of mathematics [11]. However many have argued that his theorem is applicable to any system of logic, which includes physics. Gödel’s Theorem states that we cannot prove the veracity of a system from within the system. Another way of saying that is: We cannot prove that the system itself is true using the axioms of the system. Now, in plain English: We made up the rules but we cannot use the rules to prove the truth of the rules.

For example, algebra is a system with made-up axioms, namely the commutative and associative properties of addition and multiplication of numbers (numbers, themselves are made up!). This system of manipulating numbers is internally consistent, so long as we follow the rules. However, we cannot prove the truth (or usefulness in the physical world) of the system of algebra using the axioms of algebra. We cannot even prove that algebra is the only useful way of manipulating numbers in the world (uniqueness) using the axioms of algebra. If, using the axioms, we were to encounter an inconsistency we should begin to suspect the veracity of the system, or at least the axioms. Thus we can only show that a system is false from within the parameters of the system.

Once again in plain English: as long as we are in physical reality, there is nowhere to stand outside of physical reality to determine the truth of our ideas about physical reality. We cannot know that anything is true. We can’t even know that there is a physical reality “out there” independent from our thoughts and ideas about it. So, let us not get carried away with our ideas and take our models too seriously. They are road maps, nothing more. The terrain may be entirely different when viewed from the ground. Direct experience is the only thing we can know for sure, and even that can be rationalized away.

References:

1. Godfrey-Smith P. Theory and Reality, University of Chicago Press, Chicago, IL, 2003.
2. Ref. 1, p. 62.
3. See for example, Karl F. Kuhn KF and Koupelis T. In Quest of the Universe, 3rd edition, Jones and Bartlett Publishers, Sudbury, MA, 2001, p.37.
4. Kuhn T. The Structure of Scientific Revolutions, 2nd edition, University of Chicago Press, Chicago, 1970, chapter 3.
5. Ref. 4, chapter 6.
6. Bruner JS and Postman L. On the perception of incongruity: a paradigm. Journal of Personality 1949, XVIII, 206-23.
7. Feyerabend P. Against Method, 3rd edition, Verso, London, 1993 (first published in 1975 by New Left Books).
8. Ref. 7, appendix 2.
9. Feyerabend P. Farewell to Reason, Verso, London, 1993, chapter 12 section 4.
10. Ref. 9, chapter 12 section 3.
11. Hofstadter DR. Gödel, Escher, Bach: an Eternal Golden Braid, Random House, New York, 1980.

~~~~~

*Nancy Swanson, Ph.D. Physics, Florida State University, worked for eleven years as a staff scientist for the United States Navy, then Research Associate and part-time professor at Western Washington University for five years, and ran a private consulting business over the past ten years. Her research interests were in laser light transmission and scattering through optically dense media, optical properties of nanoparticles, photonic crystals, optical phase conjugation, and quantum optics.

Dr. Swanson holds five US patents and is author of over 30 scientific publications. She has also written three books, Penetrating the Tungsten Barrier: A Survival Guide for Female Scientists and Engineers (2005), Avenging Agnodice: The Struggles and Successes of Female Scientists, Antiquity to Present (2005) and The Religion of Science (2012).

Recently she has been researching and writing about the harmful effects of genetically engineered crops and pesticides.