Introduction

Our text is an historically structured introduction, not to science, but to our theories about what science is, what distinguishes good from bad science, what is good about science, whether we should believe the consequences of our best scientific theories, what are scientific explanations, ..., in short, it is a historically structured introduction to the PHILOSOPHY OF science.

Why bother to do things this way? There is no general agreement about the answers to any of those questions. Many of them are the subject of serious continuing controversy. Much of that controversy is important, not just for philosphers of science, but for society as a whole.

Examples:

1. What should the government fund as scientific research?
2. What should parents be obligated to do in taking adequate care of their children?
3. Should health insurance pay for "alternative medicine"?
4. Should the belief that God created humans from mud roughly 6000 years ago be part of the curriculum of public schools?

Since there is no general agreement, saying what the current views and disagreements are is quite complicated, and one good way to structure the present state of affairs is to see how we wound up with it. For much of the 20th Century there was a broad consensus, at least in the English-speaking world, about what form the answers to the questions should take, though there was disagreement in detail. That consensus fell apart during what are called the "science wars," and seeing what the consensus was, and how it fell apart (that is, looking at the philosophy of science historically) is one of the best of ways of making sense of the present mess.

Of course, it probably isn't a very good idea to draw conclusions about what science is, what is good (and bad) about it, and so forth, without actually looking at some science. (There was a period, during the consensus, when the standard example of how we learn from experience in science was that of discovering that "all ravens are black" by looking at a lot of ravens. If any science is that simple, surely most of it isn't—not even ornithology, since, after all, there are albino ravens. Sometimes a simple toy example, like that one, is useful for making a point, but we will also look in some detail at some parts of science to see how our theories about science apply to actual science. (Our main examples will be in The Golem and Darwin's work.)

What are our core examples of science? The core examples are, and have always been, those that transformed our knowledge about the world and methods of gathering knowledge about the world in the Scientific Revolution: the period of development beginning with Copernicus and ending with Newton. What "everyone knows" about the world was completely transformed by those events, and how we learn it was also: the new ideas were taught in new institutions, which replaced the Scholastic universities of medieval times.

Throughout the last half of the 20th Century, molecular biology was probably the most successful science, and so it provides a more modern example.

There are many different ideas of what a philosophical characterization of science should look like, including at least the following:

Scientific method

Science is what is done according to specific method or procedure that needs to be spelled out in some detail. Today, many would say that science is too complex and various to conform to a single method, but the idea of a scientific method has been (and is) very important and influential, not least because many early scientists (including, for example, Galileo, Boyle, and Newton) wrote about the scientific method.

Scientist's thinking

Science is what is done by individuals who think (and maybe act) in a certain way. It includes not prejudging outcomes, recording observations meticulously, and a whole lot of other, related, stuff. The view is fostered by the "great scientist" myth: biographies leave the impression that Newton, Curie, Einstein and the rest made their discoveries on their own. Of course, there is a sense in which they did, but that neglects our next topic.

Edison didn't invent the light bulb. What he did was to invent the home lighting system: large scale electric generators, cable to transmit electricity to houses, outlets, and plugs, lamps with replaceable bulb, ... . Science too is a system, one in which there are individual scientists, but

Institutional structure

Science is what is done within a certain set of social institutions: refereed journals, grants, and so on. It is the way that the system is set up that makes science distinctive.

Logical structure

A good poker player may not know anything about the mathematics of probability, but a player who doesn't manage to act in a way that conforms to probabilities will lose. Similarly, science has an internal logic, and, whatever they think they are doing, scientists who don't act in accordance with it will fail. If that is correct, then the actual motivations and procedures of scientists are not of great interest for the philosophy of science. What is important is the logic of science. To find it, one would study, not how the latest science is being done (the "context of discovery") but the thoroughly analyzed and understood reasons for believing well-established science (the "context of justification").

There are a number of problems that will plague any use of the term "science" for our purposes. Let me start with the distinction between descriptive and normative.

A descriptive dictionary, for example, just lists words as they are used. A normative dictionary attempts to set standards. For example, a descriptive dictionary will have an entry for "judgement" that reads much like the one for "judgment." A normative dictionary would leave "judgement" out entirely or, perhaps, note that it is a common misspelling of "judgment." In fact, any dictionary is some combination of the two.

Similarly, when we describe actual events in science with an eye to figuring out what science is, or what good science is, we have two opposite temptations: leaving out all the parts we all agree really shouldn't be there, and giving a purely normative account. On the other side, we are so true to the actual descriptions of what goes on, that central points get lost in detail. Historians, psychologists, and sociologists of science are rather more interested in the "accidental" details than are philosphers looking for general criteria and principles. That will sometimes mean that philosophers miss important phenomena altogether, while screening out irrelevancies; and that historians, etc., will assign undue importance to particular details by only looking at a few events at a time.

Most people, and most philosophers, would, for example, assume that faking data is bad science, but it has been argued (chiefly by Feyerabend) that that is not always the case. Apparently Ampère faked many of his "experimental" results, but he is generally thought to have been a great scientist, and it seems that he could not have accomplished what he did without faking the results. It is, thus, at the very least, dangerous to assume that we already know what good science is before having a systematic account of it. The usual norms may rule out phenomena important to understanding science.

You should always be looking at everything we say in this class with an eye to when the claims are descriptive, when normative, and when some combination.

Three strands of theories of what science is:

Empiricism: "The only source of real knowledge about the world is experience." (p. 8)

Surely, except perhaps in pure mathematics or lexicography, all knowledge is in some way or other connected to experience, but the bald claim is also surely exaggerated. At least I hope it is, since I hope that is also possible to learn by taking college courses.

Note that the definition of empiricism doesn't mention science. Empiricists typically take scientific knowledge to be of a piece with the rest of our knowledge and so view the philosophy of science pretty much as a study of how knowledge (in general) is acquired with examples from the sciences.

Mathematics: What distinguishes science from other methods of investigating the natural world is that it makes use of mathematical tools.

It is worth investigating why mathematics is very useful for investigating the natural world, and whether that use is compatible with empiricism, but the broad claim just made is false: Darwin's theory of evolution is a counterexample, since Darwin's discovery of it, and many of the important developments from it, make little or no use of mathematics. (There are lots of other examples in medicine, biology, and geology.)

Social Structure: Grants, journals, refereeing, getting credit for being the first to propose or discover something, the need to make results public and reproducible.

-- ShaughanLavine - 24 Aug 2005 - 22 Aug 2007