Science                                                                     (Problems printing? Click here.)

What Is Science?

For the purposes of this text, "science" will be defined as the set of human activities that jointly contribute to expanding and refining human knowledge of the universe, and of the laws, processes, and events that have shaped that universe. In my view, science can usefully be seen as having an organized body of knowledge, a method, a process, a set of standards and an ethos. To my mind, the question of what is and isn't scientific can only be answered on the basis of a thorough understanding of all aspects of science. So let's look at the various aspects of science.

Body of Knowledge

Introductory science textbooks tend to produce a somewhat simplified view of scientific knowledge. Usually, the reader is presented with two groups of claims. There are the scientifically established claims, which are all treated as equally true, and there are the scientifically discredited claims, which are all treated as false. This is perfectly fine, just so long as you can trust the book. (Which you can, ......... usually.) There is no practical problem with regarding well-established scientific claims as true, and scientifically discredited claims as false. (In fact, I think this attitude amounts to just about the best way to define "true" and "false" for all practical purposes.) However, for the theoretical purpose of understanding science, this language is somewhat inadequate, mainly because some people tend to see the word "true" as denoting claims that could not possibly turn out to be false, and "false" as donating claims that could possibly not turn out to be true. (I, of course, think that this is the wrong way to look at truth and falsity.) So, instead of science dividing claims into two piles, one unequivocably labeled "true" and the other unequivocably labeled "false," there are in fact many piles, blending into each other, and running the gamut from extremely well established claims at one end to utterly discredited claims at the other end. Notice that I do not talk about this in terms of "degrees of truth." A more accurate description would be in terms of strength of evidence. The most well-established claims are the ones for which we have the strongest evidence, and the most discredited claims are the ones against which we have the strongest evidence. I think this is an excellent way of looking at scientific knowledge, and will serve for almost all theoretical purposes, but there's one more element that should be discussed. A claim, or theory, can be utterly discredited even if there is no specific evidence against it. This is because a claim can turn out to be useless. A person who accepts this kind of claim as true is in exactly the same position as someone who rejects it as false. Such claims can be true or false without making any difference in present or future observations whatsoever. For this reason, science is not interested in claims that cannot generate testable predictions. These claims cannot need to new knowledge, and they simply do not constitute knowledge in themselves.

Scientific Method

As far as anyone can tell, the scientific method was first developed by the philosopher Thales, who lived a god-awful time ago. Sir Isaac Newton is credited with developing a more precise formulation of the method, but it wasn't really called the "scientific method" until after the scientist, logician and philosopher Charles Sanders Peirce called it the "method of science." It is important to note that people made an awful lot of scientific progress before anyone started thinking about the scientific method. Heck, people did an awful lot of science before it was even called "science." But just because they didn't know what the scientific method was, doesn't mean that they were not using it. Our understanding of what the scientific method is comes from looking at what successful scientists did that produce the best results. The "scientific method," is just our present term for "whatever method it was that produced all that lovely reliable knowledge." Thus, in my view, the scientific method as we understand it formally is an abstraction from everyday scientific practice. It may be that nobody follows the scientific method exactly, but it is true that scientists can only be successful to the extent that they follow the scientific method. As I understand it, the scientific method can be seen as having four phases, which I will call observation, hypothesis, deduction, and testing. In the observation phase, people collect together all of the most reliable and possibly relevant data about some pattern of experiences. Say that a particular kind of cancer appears to become significantly more common in a certain population. People wishing to understand this phenomenon would, in the observation phase, gather the most reliable information the code about the past and present incidence of cancer in that population, as well as information about anything else that may have become more of less common in that population, such as an increase in illegal drug use, a chemical spill, a decline in smoking, or an increase in recent personal appearances by C-list celebrities. In the hypothesis phase, people would basically make guesses about what they thought the underlying explanation was that for all that observed pattern of experiences. In the cancer case, one person might hypothesize that the unexpected cancers were being caused by the decline in smoking, and another might think that the cancers were caused by increased exposure to C-list celebrities. In the deduction phase, people develop conditional predictions, based on logically deducing what would happen in different circumstances, if a given hypothesis was true. For instance, the first person mentioned above could deduce that the abolition of the tax on cigarettes in that area would lead to a decline in those cancers, if his hypothesis was true. And the second person could deduce that if the production of Hollywood Squares was moved into the area, the incidence of cancers would increase, if his hypothesis was true. Finally, in the testing phase, experiments are conducted and the results analyzed. This is rarely a simple matter, because of the number of variables that have to be taken into account. For instance, if the cancer rate fails to increase after Hollywood Squares start shooting in the area, this could be because the hypothesis is false, or it could be because the presence of Hollywood squares Squares has discouraged other celebrities from making personal experiences appearances. And, if the incidence of cancer went down after C-list celebrities were banned from the area, the experimenter would have to make sure that his experiment did not have some other effect that accounted for the difference. Thus, in theory, the scientific method consists of observation, hypothesis, deduction and experiment, but in practice it consists of that plus a whole lot of other work to make sure that the observations and experiments were done and interpreted correctly.

The Scientific Process

In practice, scientific progress is rarely, if ever, accomplished by a single experiment. Even the simplest hypothesis usually takes a large number of experiments to establish as supported or unsupported. Generally what happens is that a hypothesis is tested multiple times in multiple ways. Even if an experiment goes as predicted, it does not necessarily follow that the hypothesis is confirmed or disconfirmed. Rather, it might be true that the result could be explained in a number of ways, and further experiments must be designed and carried out in order to determine which explanation is correct. The scientific process is thus a continuous series of iterations of the scientific method applied to successions of progressively more refined hypotheses. Generally, the most effort is applied in the most promising direction, but that does not mean that no effort is spared for hypotheses that are less likely to be true. Rather, all reasonable hypotheses tend to be pursued, each according to how likely to be true it appears at the time. An important part of the scientific process is carried out by the established scientific journals. Scientists who think that they have something interesting to say will write papers and send them to journals. The editors of these journals will send these papers out to other scientists that they think are competent to judge the quality and integrity of the work presented. There then follows a messy process of criticism, comment and revision as the authors of the papers attempt to meet the standards set by the reviewers, but eventually some of these papers make it into the pages of the journals. Publication by a refereed scientific journal does not guarantee that the conclusions expressed in the paper off well-founded. Rather, what the journals are supposed to guarantee is that the authors have given a clear and complete description of their experimental procedures, have correctly interpreted their purported results, and have properly cited the relevant previously published work. This process, when it works right, makes sure that the journals only publish papers that have a reasonable chance of being right, and which give enough information for other scientists to critique the experimental method, or repeat the experiment.

The Scientific Community

Scientists have a phrase, "standing on the shoulders of giants," which they use to indicate that individuals typically only make great discoveries after others have laid a great deal of groundwork. Sir Isaac Newton, for instance, used this phrase to indicate his debt to previous scientists and astronomers. (Some of whom should more properly be called astrologers than astronomers.) The scientific method is often a group effort, with one person providing the observations, another the hypotheses, a third the deduction and perhaps even a fourth designing the experiments. This means that scientists tend to have a lot to do with other scientists, and deal with other scientists in ways that they do not deal with non-scientists. We can thus define a pattern of "scientific interaction" as the set of interactions and behaviors that scientists typically direct towards fellow scientists, and not to other people. Obviously, this will include reading each others' relevant papers, corresponding on points of mutual interest, organizing and attending conferences, and arguing with each other both in person and via the pages of scientific journals. In general, or at least ideally, these interactions will all be characterized by one overriding theme. That theme is an effort to make sure that the scientific method is applied as thoroughly as possible to whatever questions are currently of interest to science. In sociological terms, members of the scientific community generally engage in a political process, both between themselves, and in regards to others, to ensure that the scientific process is protected and promoted as much as reasonably possible.

Scientific Standards

In the world outside of science, the word "certainty" generally means that an emotional conviction that some claim is correct, or even a determined unwillingness to accept the claim could be false. Many of the beliefs that are held by nonscientists to be "certain" are in fact not even particularly likely to be true, and in many cases are in fact very unlikely to be true. Unlike most nonscientists, scientists recognize that no conclusion about the state of the universe or the laws of nature can ever be held with absolute certainty. Thus even the best established theories of science, which are far better supported than any nonscientific claim, are still held by scientists to be tentative conclusions about the universe that could possibly be wrong. Thus when a theory is accepted by the scientific community, that does not mean that anyone thinks that it is absolutely certain to be true. Rather it means that scientists think it is very likely be true, that it has no unrefuted competitors, and that scientists think that accepting it as true is likely to lead to productive work in the future. When a piece of scientific work is accepted by other scientists, it does not mean that there are absolutely no holes in its arguments, or that absolutely every step in its explanations has been nailed down tight. What it means it is that scientists think that its arguments are good enough, that its explanations are detailed enough, for it to rank with all the other work that has been generally accepted. It is always possible to raise trivial objections to any scientific paper, but if scientists were required to specifically spell out absolutely every step in a process, no scientific work would ever be completed. Scientific work is judged by a very high standard, but it is never required to be perfect. This means that there are two ways to abuse scientific standards. First, of course, scientific standards can be abused by accepting or promoting work that does not rise to the standards generally imposed. But second, scientific standards can be abused by dismissing work that meets those standards merely because it is not absolutely perfect.


Terminology

Scientists have their own way of talking about things, but I'm really not going to talk about that. Instead, I'm going to develop a vocabulary for talking about science, which may not be the same as the vocabulary that scientists use to talk about their work, but will hopefully enable me to be more precise in how I talk about what I think science is.

Ad Hockery. An "ad hoc" assumption is an assumption made purely to explain some diversion from predicted or hoped-for results, or (some contradiction of well-established theory.) If an experiment does not go the way the scientist wanted it to go, he is perfectly free to try to save his theory by coming up with a plausible explanation that allows his theory to still be true, even though the results of the experiment appear to disconfirm that theory. If this explanation appears plausible to other scientists, they may continue work on this theory. And sometimes these ad hoc assumptions even turn out to be true. However, when a theory's multiple failures are covered only by a series of ad hoc assumptions, it is a strong sign that the theory is in serious trouble.

Consistency. An experimental or observational result is "consistent" with a theory if it does not contradict the theory. Take the theory that the "fixed" stars are all embedded in a giant crystal sphere that has the Earth at its exact center. The observation that the stars appear to rotate around an axis running through the North Star is consistent with this theory, but it is also consistent with the theory that the stars are distributed theough an immense volume of outer space, and their apparant rotation is merely the result of the Earth turning on an axis running through the North Star. Thus the same observation can be consistent with a large number of different, mutually contradictory theories. (We can also imagine that, contrary to what we actually see, the fixed stars don't appear to rotate. This imaginary, but logically possible, observation would be consistent with the "crystal sphere" theory, but it would not be consistent with any theory that required the Earth to rotate.)

Confirmation. An observation or experiment "confirms" a theory if the result of the observation or experiment is logically deducible from that particular theory and from no other competing theories. An experimental result confirms the theory if both the following things are true. 1. This theory predicted that particular result. 2. All other competing theories predicted different results from that experiment, or made no predictions about that situation. A confirmation does not necessarily prove a theory, but it does contribute to such proof. Please note that confirmation is much stronger than consistency. All confirming results are consistent, but not all consistent results are confirmations. Suppose, in the example above, we had somehow eliminated the "crystal sphere" theory and were attempting to decide between two theories attempting to expain the Sun's apparant motion around the Earth. The first "Sun Moves" theory is that this motion is real, and that the Sun really does whiz around a fixed Earth. The second, "Earth Rotates" theory asserts that the Sun is basically fixed, and that the Earth rotates on its axis to produce the appearance of motion. The mere observation that the Sun appears to circle the Earth is no help here because it is consistent with both theories. But the observation that the fixed stars also appear to rotate about the Earth on a 24-hour cycle confirms the "Earth Rotates" theory because is is both consistent with that theory, and inconsistent with its competitor.

Disconfirmation. An observation or experiment "disconfirms" a theory if it produces a result that is different from what the theory predicts would happen in those circumstances. Disconfirmation does not necessarily falsify a theory, but it does cast doubt on the disconfirmed theory, and definitely indicates that more work needs to be done.

Discredited. I shall refer to a claim as "discredited," when I think that it has been scientifically established that there is no point to entertaining that claim as a hypothesis any longer. The easiest way, in theory, to discredit a claim, is to show that it is inconsistent with our best observations. However, proponents of such a claim can often make excuses for why our observations turn out different, and some of these excuses can be valid. Therefore, a claim is really only completely discredited when all of the possible excuses have been ruled out, or it comes with so many excuses that trying to get anything useful out of the claim is so complicated as to be not worth the trouble. (Logically self-contradictory claims are, of course, automatically discredited.)

Fact. I shall use the word "fact" or "scientific fact" to denote any claim that is so well established that there really is no present practical point to doubting it. A scientific fact is a claim that is rightfully considered reliable by scientists in that they find that they can treat it as true under all circumstances. One way to know whether some claim is a scientific fact is to ask whether or not anyone is currently raising any credible arguments against that claim. If there are no logically credible arguments against the claim, then I'm just going to call it a fact. If there is even one even remotely reasonable argument against some claim, then I will hesitate to call it a fact, even if I personally think it is true.

Falsification. A theory is falsified if it has been disconfirmed so thoroughly that there is no reasonable explanation for all these disconfirmations other than to say that the theory is false.

Falsifiability. A theory is falsifiable if it is logically possible for some experiment or observation to turn out in some way as to prove the theory false. It is important to note that this is a logical possibility, not necessarily a practical one. It may be practically impossible to prove a certain theory false, merely because the proponents of the theory are very creative in coming up with excuses for their pet theory. This does not mean that the theory is not falsifiable. A theory will fail to be falsifiable if it comes with built-in excuses. The theory that disease is caused by demons is not falsifiable because there is no hypothetical observation that could prove the demons do not exist. The theory that disease is caused by bad-smelling air is falsifiable, because we can conceive of observing disease without bad-smelling air, observing bad-smelling air without disease, and finding that there is no significant correlation between the appearance of bad-smelling air and the appearance of disease. Both of these theories are discredited. The bad-smelling air theory is discredited because it has been abundantly disconfirmed. The demon theory is discredited because it is unfalsifiable, and thus cannot contribute to scientific progress.

Law. The term "law" has two main usages in science. First, it is used in names for particular ideas, as in "The Laws of Thermodynamics." Used in this way, it does not mean anything. If something is called a "law," that does not mean it is any more respectable than something that is not called a law. In fact, things that are called laws can be disconfirmed or discredited. The "laws" of thermodynamics are not scientific laws at all, but have in fact been shown to be special cases of Statistical Mechanics. Secondly, it is used more precisely to refer to what are believed to be universal regularities of nature. Einstein's theories of general and special relativity and Darwin's theory of evolution through natural selection are scientific "laws" in that sense.

Meaning. A statement is meaningful if there is some practically possible observation that can be made to determine whether or not is true or false. If the term "demon-infested" is defined in such a way that there is no practical way to tell a demon-infested object from a non-demon-infested object, then the term "demon-infested" is meaningless. This also applies to terms whose "meaning" is defined obliquely. If someone claims that your stereo must be infested with demons because it is not working correctly, and neither you nor he can figure out why, the term "infested with demons" is still meaningless because there is no way to tell the difference between a stereo that is infested with demons and one that merely has a malfunction of unknown origin. (In such a case, the term "infested with demons" just means ""not working.")

Proof. A theory is proved, in my view, if it is so well confirmed, and its competitors are so well disconfirmed that there is no further real point in testing the theory. Individual scientists make different individual judgments on whether or not particular theories are proved. And your judgment of whether or not a particular theory is proved can be different from mine. Since no theory is ever proved absolutely, there really isn't much point in trying to work out an objective definition of the word "proof."

Replicability. (Or reproducibility.) Results in science are supposed to be replicable, which is to say that other scientists who perform the same experiment in the same way will get the same result. If a scientist claims to have gotten a certain result with a certain experiment, but competent other scientists prove to be unable to get that result from that experiment, then it isn't replicable. When a result turns out to be unreplicable (or irreproducible), the most obvious explanation is that the original scientist made some kind of mistake, and so unreplicable results are generally ignored by science.

Reputation. Certain scientists, labs and areas of study have good reputations, and attract a lot of attention and resources from the scientific community. Certain other scientists, labs and areas of study have bad reputations, and attract little or no attention or resources from the scientific community. Ideally, scientists and labs get good reputations by producing useful theories, replicable results and other reliable information. Areas of study get good reputations by being areas where productive work can be done. Scientists and labs get bad reputations by producing irreproducible results, or otherwise doing shoddy scientific work. An air of study gets a bad reputation by constantly failing to be an area where replicable results can be gained, or otherwise useful work can be done. A field of study in which researchers consistently promise that they have or soon will have compelling results, and yet which produces no working devices, and no clear reproducible experiments, will have a very bad reputation. A scientist who has earned a bad reputation will not be taken seriously by other scientists. A lab that has earned a bad reputation will not be taken seriously by scientists. And a field of study that has gotten itself a bad reputation, will also not be taken seriously.

Result. A result is simply whatever is observed when an observation is made or an experiment is performed. Ideally, results rule science because results are direct parts of our experience, and sciences primary job is to explain our experience. A theory that contradicts well-established results should be abandoned. However, a mere claim that someone has achieved a particular result is never by itself enough to overturn a theory. When a theory-contradicting result comes from a lab without a proven track record, the fact that result contradicts an established theory actually gives very good reason to think that somebody at the lab has made a mistake. Thus results that contradict well-established theories tend to be very tightly scrutinized, and are only taken seriously if they are presented together with evidence that the relevant experiments were performed with a very high standard of care.

Theory. Most philosophers of science offer very precise definitions of the word "theory." In my view, this does not fully reflect the way scientists use this word. Scientists use the word "theory" in three main ways. First, they use it as part of names for particular ideas, as in the "Theory of Relativity." Used in this way, it does not mean anything. Secondly, scientists use the word in the same way that nonscientists use it, to mean any kind of hypothesis from a wild speculation to some idea that is actually been pretty well established. Finally, they also occasionally use the word in a very precise way to denote a theory that meets the rigorous standards demanded of the kind of theories that are worth investing the resources necessary to properly test them. A scientist might claim that some idea isn't really a theory as a way of saying that he thinks it's too vague to be tested, or that there's not enough experimental evidence to justify taking it seriously, or that it is already contradicted by existing, well-established results. Thus a scientist might refer to some idea as a "non-theory" as a way of emphasizing his opinion that it is not the kind of idea that ought to be taken seriously by the scientific community.

http://en.wikipedia.org/wiki/Scientific_method

http://en.wikipedia.org/wiki/Timeline_of_thermodynamics%2C_statistical_mechanics%2C_and_random_processes

http://en.wikipedia.org/wiki/Phlogiston

http://en.wikipedia.org/wiki/Caloric

Thompson

Rumford

Joule

Von Mayer

Thomson

http://en.wikipedia.org/wiki/Theory_of_heat

Homework Perform one of the following tasks. 1. Pick 5 to 10 of the above concepts, and explain them in your own words. or 2. Follow one of the above links, and write a short report explaining how the scientific method and/or other practices or values of science contributed to the creation of one particular scientific discovery.

In class we will discuss the development of heat theory, and then the film: Expelled:_No_Intelligence_Allowed




Copyright © 2006 by Martin C. Young

---

This Site is Proudly Hosted By.