(Swans - August 9, 2010) Science is not a question of truth and falsehood, but is a collection of theories and experiments, operations that we, or our machines, do. Theories are recipes for experiments and predictions about their results. We add theories that provide experiments that come out as predicted to our storehouse of scientific theories. That, and that alone, is science. What are theories and experiments?
Looked at from the outside an experiment often looks just like a measurement. The determination of a particular distance might be an experiment or just a simple measurement. For example, if we were measuring the diameter of a type of atom we would be doing an experiment. When we determine the size of a type of thing we can count on other tokens (examples) of that type being that size later or elsewhere. The result of these later measurements is predictable in advance. Theories, once established, make predictions about the outcome of experiments. These experiments then become routine operations. An experiment is either a measurement used to test a prediction or a measurement of a type of thing used to establish a standard for making a prediction. G in Newton's law of gravitation is a constant, that is, such a standard. The ability to make predictions is the purpose of the theory and our justification for accepting it. All the rest is giddy imaginings. Predictable outcomes of operations allow technological advances. For all technological production is a string of repeatable operations whose outcome is known in advance.
In the most common case in the "hard" sciences the theory will predict a number that a measurement of an operation will produce. For example, Newton's law of gravitation predicts how much time weights balanced at the ends of rods and in proximity will take to swing together. It is usually written like this.
Now when we look at this formula it is hard to see just how we get from it to the prediction. We might define all the symbols, but instead let us try another way. The formula, when interpreted correctly, supplies us with a picture. Heavy balls accelerate towards each other. The acceleration of each ball is directly proportional to the size of the other ball. The balls' acceleration towards each other is inversely proportional to their masses and proportional to the square of their distance apart. Thus Galileo's famous experiment, which showed two different sized cannon balls falling to earth from the Tower of Pisa at the same rate, is explained. For both accelerate towards the same other ball, the earth.
There is also a second picture. If these masses are in motion to start out with, then depending on this motion they will move in orbit around each other or in a parabola, approaching and then shooting apart. We imagine certain conditions and using the formula, predict, for example, a planet's or comet's location at a certain time. Given the proper interpretation, the formula, with the right values, can describe any elliptical orbit or parabolic trajectory. But in this case we calculate the values for the masses by use of the formula, and can make predictions only after this calculation.
We should note in passing that Galileo's experiment is not quite what the theory requires. If the cannon balls were dropped at different times they would fall, or appear to fall, at minutely different rates. For, according to the theory, the earth must also move. And its motion is different in the two cases, for the cannon balls are of different sizes and the earth's motion is proportional to the cannon balls' sizes. Newton's formula predicts that each mass will accelerate in proportion to the size of the other mass. We can't see the earth's motion because we are standing upon the earth, but its motion would shorten the time the heavier ball would appear to take to fall. The movement of the earth because of the cannon balls is, of course, minuscule, but this doesn't change that the setup of the experiment is wrong. Were one of the cannon balls a lot larger, say the size of the sun, it would fall to earth, or appear to because we are standing on the earth, a lot faster than a bowling ball. For then the earth's movement towards it would be, to say the least, palpable. To do the experiment correctly we must measure the movement of both the earth and the balls, both masses. That means we must somehow stand apart. Hence the elaborate setup with balls balanced at the ends of rods and swinging together.
Clearly, just looking at the symbols in the formula, it is impossible to know how to do the experiment correctly. How would we know that Galileo's experiment is not quite right? Can the formula tell us? How? If we imagine the picture as a photograph it doesn't really help us. Somehow the balls pulling at each other must be in the picture. But how? Do we want to imagine rubber bands between them? This wouldn't work because the rubber bands would get slacker as the balls got closer together. No, the picture is not just a picture, but more like a mental rehearsal of the experiment. We recount, in imagination, what ought to happen, then use the formula to calculate what actually did.
The formula, couched in mathematical symbols, is actually a description of one or more measurements and a calculation that predicts the value of a final measurement to be made at the end of the operation. So the measurements made at the beginning of the experiment allow us to predict the measurement made at the end. The picture is a description of the apparatus necessary to produce the final measurement. It is often said that Newton, through this formula, unified our understanding of the motion of the planets and the motion of falling bodies. It was this unification of what at the time seemed completely different phenomena that so impressed everybody. And it was Newton's unification that encouraged physicists later to search for physics' holy grail, the unified field theory, a single theory that explains everything. How did Newton do it?
As we have seen, the formula considered as a picture depicts bodies that accelerate. Acceleration is described like this in my old physics book: often the velocity of a moving body changes either in magnitude [speed], in direction, or both as the motion proceeds. The body is then said to have an acceleration. So velocity has two components, speed and direction. If either or both changes there is an acceleration. Now when bodies fall their direction doesn't change but their speed does, and when planets move in a circular orbit their speed doesn't change but their direction does. So if acceleration is a change in speed or direction, both falling bodies and planets can be explained by the same formula. Thus a single formula unites two apparently disparate experiments because one of the essential quantities can be measured in either of two ways. [If a planet is in an elliptical orbit both its speed and direction change.] A unified field theory would also have to have quantities of a similar sort to accommodate a variety of clearly quite different operations.
But, we should ask, does Newtonian mechanics explain these motions? No, Newtonian mechanics gives us, at best, an ability to make certain predictions. It offers no explanations of why anything happens. Science makes predictions and that is all. We might predict the speed of a falling body after, say, ten seconds. Or we might predict the position of a planet on the fourth of July 2010. We will then perform measurements, experiments, to see if we are correct. Why the body falls or why the planet is there is of no scientific concern.
Experiments of the second kind, those that are used to establish quantities for predictions, do not result in flashy theories, and what they do show might not be called theories at all. For example, it is essential for engineers to know how strong various materials are. Experiments are used to establish the compression strengths of various kinds of cement, various kinds of steel, those of some highly artificial kinds of materials. These measurements, unlike, for example, measurements of the width of a particular hallway, are measurements of a type of thing. It is a measurement of an example of this or that type of cement. By establishing the strength of a type of cement we can predict just how strong a pillar or post made of that cement will be. With this knowledge we can make elaborate structures. The measurement of a type of thing looks no different from the measurement of a particular thing. What makes something a type of thing rather than simply a particular thing? Why, that we produced it (or if you want identified it) with a repeatable operation.
Should we call this list of compression strengths a theory? Well, it is as you will. It does seem possible to determine strength (and there are various kinds) through what might be called "more theoretical" means. But it is not to my purpose to trace how that might be done here. My point is a simple one: scientific theory, for those who understand it, supplies a recipe for carrying out an experiment and predicts what the result of that experiment will be. Experiments of one kind test theories by producing results that either agree or disagree with theoretical predictions, while those of a second kind establish these predictions for a class of operations through the measurement of a token of a particular type. Real theories must lead to experiments that either agree or disagree with the theories' predictions. A theory must supply us, as Einstein put it, with an experiment that can "cook" the theory. Such an experiment tests the prediction.
Prediction of the result of our own operations is the be all and end all of science. Taken together theories and experiments provide a catalog of operations we can reliably do, a collection of "steps" whose results we can reliably predict. We make our technological marvels by performing elaborate sequences of these steps, the results of which are predictable. Production lines are collections of machines that do these steps repeatedly. Theories that do not supply such experiments, and through them predictable steps, are not scientific theories.
Given these introductory comments let us take a look at a number of scientific controversies. How about the idea of "intelligent design"? When an interviewer asked Stephen C. Meyer, a Cambridge educated leading proponent of intelligent design, "What would be your main argument for the evidence of intelligent design in the cell?" he said:
The DNA molecule is literally encoding information into alphabetic or digital form. And that's a hugely significant discovery, because what we know from experience is that information always comes from an intelligence, whether we're talking about hieroglyphic inscription or a paragraph in a book or a headline in a newspaper. If we trace information back to its source, we always come to a mind, not a material process. So the discovery that DNA codes information in a digital form points decisively back to a prior intelligence. That's the main argument of the book.
Now if intelligent design is to have the form of a scientific theory it must predict the result of an experiment. What is the experiment of which intelligent design will predict the result? I suppose we might imagine ourselves discovering the workshop of the designer, but how would we know that it was such a workshop? Would it differ from a human laboratory? Meyer's answer is in the form of a logical argument. It does not supply a "theory" of intelligent design that describes an experiment one might do. I for one cannot imagine any such experiment, but if someone comes up with one the nature of the experiment will reveal what the theory of intelligent design means. As it stands now the theory of intelligent design fails because it is simply not a theory that is in the proper form to be a scientific theory. Meyer's logical argument is no more convincing than something like this. "Everything that moves must have a mover, but there must be a first mover, so God is the unmoved mover." In any case such logical arguments do not make scientific theories.
As an aside I note that Meyer argues that "information always comes from an intelligence." However, the same argument might be made that "information always comes from a human being." So Meyer's argument could equally support the claim that a human being is the intelligent designer of DNA. But again logical arguments do not make scientific theories unless they produce a prediction of an experimental result.
What about the theory that contends with it, Darwin's theory of evolution? The intelligent designers object primarily to the idea that life evolved through a series of random accidental mutations. Darwin, or his present defenders, claim that evolution occurs when a random mutation proves beneficial and those individuals with it gain an advantage. Now this too describes no experiment. What experimental result does this theory predict? Suppose we allowed fruit flies to reproduce endlessly in some controlled environment and a mutation occurred that benefited those who had it so that the population gradually all had this mutation. Could we then say that this change was accidental? What would tell us that this was so? Perhaps many mutations would occur and the population would retain only a few, those that proved beneficial. Would this convince us that the changes were accidental? If it did convince us then, scientifically, it should convince us only of what the experiment shows, namely that a population retains beneficial mutations. If the fact that there are many mutations and only some retained convinces us that the process is "accidental," then this is the meaning of accidental in this context. No more and no less. But if someone denied that this meant that the process was accidental, how could we tell whether the change was purposive or accidental? Finally, this experiment tells us nothing about what happened in the past. Populations might retain accidental beneficial mutations and also produce intentional beneficial ones. Certainly scientists in the laboratory have intentionally produced mutations.
In the interview with Dr. Meyer the interviewer quotes John Walton, a Biblical scholar. "Science is not capable of exploring a designer or his purposes. It could theoretically investigate design but has chosen not to by the parameters it has set for itself." Walton is quite right that science cannot examine the question of a designer because of the "parameters it has set for itself." Whether or not an experiment produces a result because of purpose or chance is not a scientific question. Science only asks what the result is. Mr. Walton is quite wrong to think that science could do otherwise. Explanations that do not lead to predictions are outside its scope.
Of course, we can create populations of plants and animals in the laboratory through forcing mutations. Now some anomalies are passed on to offspring and some aren't. Can we identify which is which? Yes, to some extent. And in this way we can make predictions about experiments. People might say that this is how Darwin's theory is scientific, even though such experiments do not involve the past or "how evolution happened." The theory of evolution, the idea that species evolved from other species through this process of chance mutation "back then," remains without an experiment. For a scientific theory must say that the experiment will come out this way and not that way.
Proponents of intelligent design do not object to evolution completely, but to evolution of all life forms from a single source through chance mutation. Meyer has this to say:
I think small-scale microevolution is certainly a real process. I'm skeptical about the second meaning of evolution -- the idea of universal common descent, that all organisms share a common ancestry. I think the fossil record rather shows that the major groups of organisms originated separately from one another. But that's not what the theory of intelligent design (ID for short) is mainly challenging.
I do not know what Meyer is pointing to in the fossil record. Presumably it is large structural differences or differences in DNA. Would such differences convince us that the two species had different origins? It might convince some and not others, but no predictive experiment emerges. So, whether it convinces us or not, it is not scientific. Meyer has a point here.
Did humans evolve from "pre-humans" through chance mutation? What experiment does such a theory describe? I see none. Then does any predictive experiment emerge from Darwin's work? Well, yes. Through the use of radiocarbon dating and perhaps other means we can give a "date" to fossil remains. Through this method we have been able to date, roughly, the origin of humans and other species. We have created a whole timeline, a natural history, through our methods of dating fossil remains. From this we can make claims that are experimentally testable. Darwinians can predict that any discovered human fossil remains will not be twenty million years old. They might even be able to claim that they will be no more than 100,000 years old. Through various characteristics picked up through earlier experiments they can probably date remains quite accurately and predict, with reasonable accuracy, what the radiocarbon dating will be. The appearance and disappearance of various species within this constructed past allows us to make predictions about "dating" experiments. In this sense Darwin's theory is scientific.
But we should realize that only in this sense is Darwin's theory scientific. There is no real way to say that one species evolved into another. Even if we found the so-called "missing link," some fossil that had some of the characteristics of one species and some of a later one, this would not prove that one evolved into the other. A shrug, as if to say, "how could it be otherwise?" is not a scientific experiment. Again, no experiment is indicated.
The validity of radiocarbon dating rests upon a theory and an experiment. Simplified it goes something like this. Living beings take in carbon and so have the same amount of Carbon 14 as the atmosphere. When they die they stop breathing and the carbon 14 decays. So the amount of carbon 14 in a fossil indicates its age. By measuring the ratio of carbon 14 to carbon 12 we can tell, roughly, when in the past the organism died and stopped exchanging carbon with the atmosphere. Does radiocarbon dating really tell us when in the past something happened?
All we can say is that it fits into the complex web of tests of the past that we have, and this complex web of tests is the scientific past. We picture the past as if it were a time and place "back then." It's hard to resist the idea that it still exists, "somewhere." Scientifically, the past, like everything else, is a complex of experiments that can only be done right now. For that is where we always are, right here, right now. In reality, our image of the past is a set of rules and a scientific theory, a recipe for a multitude of experiments that permit us to locate events within "past time." Prior documents can't refer to posterior ones. Certain inks only came into use at certain times, as did certain kinds of paper. Experiments determine whether these are in a given sample. Mothers and fathers must be older than their children. Documents tell us of birth dates. Things cannot be in two places at the same time. Reports tell us of the location of people and things. Cause and effect must move from earlier to later times. And much more. Radiocarbon dating must fit into the complex of experiments we can do to locate a person or event in the past. That it does so is our justification for incorporating it into this web of experiments.
Suppose someone said (as people do) "God plopped down the whole past all at once." Well, fine, but scientifically irrelevant. It is nothing but a fantasy picture and affects none of the operations that allow us to locate things in the past.
Darwin's theories allow us to predict, from knowledge of just what species left certain remains, the age of those remains as calculated through radiocarbon dating or any other method within the web of experiments that supply us with knowledge of the past. To this extent, because it allows us to make such predictions, Darwin's is a scientific theory.
But we must note that the notion of "survival of the fittest," a notion that has so perniciously encouraged the belief in the goodness of the "free market system," is not scientific, for it allows no predictions. The problem here is simple -- to identify fitness one must already know those who survive. Nothing marks one individual as more fit than another except its survival. Chickens are more fit than extinct eagles. Their fitness consists in their being good to eat and easy to control, so humans protect and breed them, guaranteeing that they survive. DDT wiped out the eagle. But there is no predictive element, no trait identified ahead of time that will guarantee that a particular individual will survive as opposed to another. Big beaks in birds might make individuals more formidable in fights with rivals or simply add extra weight that hinders successful migration. "Fitness" can only be identified after the fact. An individual was fitter because it has survived, so no prediction is possible. "Fitness" and "survival" are interesting because they seem to be concepts that relate to each other as cause and effect whereas they are in fact synonyms. "Survival of the fittest" is thus a tautology, true in all worlds, and hence having no predictive power.
Again, a theory is a scientific theory only if it allows us to predict the result of an experiment. Thus intelligent design fails and Darwin's theories succeed only partially in being scientific theories.
Let's look at one more controversy. What about global warming? Is this a scientific theory? Yes, it is. The theory says global average temperatures are a function of CO2 concentrations in the atmosphere. The experiment is as follows: place thermometers all around the world. Record daily temperatures and take the average. Compare these temperatures year over year. Produce a regression graph of these temperatures year by year. Also, daily, record the CO2 level in the atmosphere, and make a regression graph of these measurements. The slope of the regression graph of average temperatures will be a linear function of the level of CO2 in the atmosphere provided all else remains constant. This last limitation is completely legitimate, and Newton's law of gravitation has the same limitation. "All else" includes levels of solar radiation, the presence of other greenhouse gases, arctic ice cover, and the level of particulate matter in the atmosphere.
The problem? The experiment takes years to perform, and the regression graph, with only a few data points, does not convince everybody. In order to add more data points, scientists have tried to use proxies to measure temperatures in the past. Also, unlike with experiments associated with Newton's theory, we cannot keep all else constant, so variations in solar radiation, other greenhouse gases, and atmospheric particulate matter have affected the measurements. But, since we have theories and experiments that predict just how these phenomena affect global temperatures we can incorporate them in our calculations.
The theory of global warming also requires yet another experiment. This part of the theory says that the atmospheric levels of CO2 are a function of human production and release of CO2. Again, all science can show is a correspondence between human emissions of CO2 and atmospheric concentrations of CO2. And also again, other factors will not remain constant, but their measurement and effect are possible.
But the problem here is not merely in determining a scientific outcome. For here it is important that the result of an experiment change our way of life, and the connection between this result and this change is not scientific. Warming, in the sense described above, is scientifically demonstrable. That people ought to do something about it is not. Nothing, scientifically, requires human beings to people the earth. We might just as easily engage in an end-of-the-species party. Nothing scientific opposes it.
If you find Michael Doliner's work valuable, please consider
Feel free to insert a link to this work on your Web site or to disseminate its URL on your favorite lists, quoting the first paragraph or providing a summary. However, DO NOT steal, scavenge, or repost this work on the Web or any electronic media. Inlining, mirroring, and framing are expressly prohibited. Pulp re-publishing is welcome -- please contact the publisher. This material is copyrighted, © Michael Doliner 2010. All rights reserved.
Have your say
Do you wish to share your opinion? We invite your comments. E-mail the Editor. Please include your full name, address and phone number (the city, state/country where you reside is paramount information). When/if we publish your opinion we will only include your name, city, state, and country.
About the Author
Michael Doliner studied with Hannah Arendt at the University of Chicago (1964-1970) and has taught at Valparaiso University and Ithaca College. He lives with his family in Ithaca, N.Y. (back)