Prompted by my discussion of Medawar and recalling that once in the past I called him a gadfly (although obviously I meant it in the good way), Bill Hooker drops another Medawar quotation on me and asks if I'll bite:
If the purpose of scientific methodology is to prescribe or expound a system of enquiry or even a code of practice for scientific behavior, then scientists seem to be able to get on very well without it. Most scientists receive no tuition in scientific method, but those who have been instructed perform no better as scientists than those who have not. Of what other branch of learning can it be said that it gives its proficients no advantage; that it need not be taught or, if taught, need not be learned?
Bill's take is "scientific methodology" here can be read "philosophy of science". So, what do I think?
First off, I'm not entirely sure Medawar was using "scientific methodology" to mean "philosophy of science," so I'll consider the quotation both ways. And of course, I'll have to give some sensible description of what philosophy of science (which is a fairly big and bustling subfield of philosophy these days) might be.
There are lots of things philosophers of science study, but one central set of concerns is what is distinctive about science -- how science differs from other human activities, what grounds its body of knowledge, what features are essential to scientific engagement with phenomena, etc. This means philosophers of science have spent a good bit of time trying to find the line between science and non-science, trying to figure out the logic with which scientific claims are grounded, working to understand the relation between theory and empirical data, and working out the common thread that unites many disparate scientific fields.
If you like, you can think of this set of philosophical projects as trying to give an account of what science is trying to do -- how science attempts to construct a picture of the world that is accountable to the world in a particular way, how that picture of the world develops and changes in response to further empirical information (among other factors), and what kind of explanations can be given for the success of scientific accounts (insofar as they have been successful). Frequently, the philosopher is concerned with "Science" rather than a particular field of science. As well, some philosophers are more concerned with an idealized picture of science as an optimally rational knowledge building activity -- something they will emphasize is quite different from science as actually practiced.*
Practicing scientists pretty much want to know how to attack questions in their particular field of science. If your goal is to understand the digestive system of some exotic bug, you may have no use at all for a subtle account of scientific theory change, let alone for a firm stand on the question of scientific anti-realism. You have much more use for information about how to catch the bug, how to get to its digestive system, what sorts of things you could observe measure or manipulate that could give you useful information about its digestive system, how to collect good data, how to tell when you've collected enough data to draw useful conclusions, appropriate methods for processing the data and drawing conclusions, and so forth.
A philosophy of science course doesn't hand the entomologist any of those practical tools for studying the scientific problems around the bug;s digestive system. But philosophy of science is aimed at answering different questions than the working scientist is trying to answer. The goal of philosophy of science is not to answer scientific questions, but to answer questions about science.**
So, if Medawar means to claim that a working scientist doesn't need to have learned philosophy of science in order to get the scientific job done, I think he's right. Neither does a scientist need to have studied Shakespeare or history to be a good scientist -- but these still might be worthwhile endeavors for the scientist as a person. Every now and then it's nice to be able to think about something besides you day job. (Recreational thinking can be fun!)
Now, there are some folks who will argue that studying philosophy of science could be detrimental to the practicing scientist. Reading Kuhn's Structure of Scientific Revolutions with its claim that shifts in scientific paradigm have an inescapable subjective component, or even Popper's view of the scientific method that's meant to get around the problem of induction, might blow the young scientist's mind and convince him that the goal of objective knowledge is unattainable. This would probably undermine his efforts to build objective knowledge in the lab.
My graduate advisor in chemistry had a little story he told that was supposed to illustrate the dangers for scientists of falling in with the philosophers and historians and sociologists of science: A centipede is doing a beautiful and complicated dance. An ant walks up to the centipede and says, "That dance is lovely! How do you coordinate all your feet so perfectly to do it?" The centipede pauses to think about this and eventually replies, "I don't know." Then the centipede watches his feet and tries to do the dance again -- and can't!
The centipede could do the dance without knowing precisely how each foot was supposed to move relative to the others. A scientist can do science while taking the methodology of her field for granted. But having to give a philosophical account of or a justification for that methodology deeper than "this is what we do and it works pretty well for the problems we want to solve" may render that methodology strange looking and hard to keep using.
Then again, I'm told what Einstein did for physics had as much to do with proposing a (philosophical) reorganization of the theoretical territory as it did with new empirical data. So perhaps the odd scientist can put some philosophical training to good scientific use.
Could Medawar have possibly meant, in the passage quoted above, to say that scientists don't need to learn the scientific method? This wouldn't be an outrageous claim if the "scientific method" in questions was an idealized system of steps (like those in the narrow inductivist picture of scientific inquiry Hempel dismisses) that isn't what scientists actually do. There might even be something to the idea that people really learn the kind of scientific that matters by doing science rather than by reading about it or listening to someone lecture on it.
The wee concern I have with this is that, despite the fact that scientists in different fields have expertise with different specific tools and methods, there seems like there should be a common core that unites the patterns of inference they all deploy -- else what would put them all under the umbrella of science? Couldn't that common core be a "scientific methodology" worth learning, if only so the scientist studying bug guts understands how her inquiry is similar to that of other scientists?
Or is Medawar's idea that the needed patterns of inference don't need teaching for some other reason (maybe because they are part of our common sense)?
_____
*For the record, I find science-as-actually-practiced -- in particular scientific fields, rather than generalized as 'Science" -- more philosophically interesting than the idealized stuff. But, as one of my labmates in graduate school used to put it, "One person's 'whoop-de-doo' is another person's life's work."
**Really, to answer philosophical questions about science, since historians and sociologists and anthropologist also try to answer questions about science.
- Log in to post comments
Interesting post. Most philosophers of science, in my experience, would not ask that question. :-)
One place philosophy of science can actually harm science is when devotees of some particular school of philosophy of science get into positions of influence in granting agencies or institutions, and then try to impose their model on working scientists. This has happened in NIH with 'hypothesis driven research' and 'paradigm shifts', with STM in some branches of NSF, etc..
I'm not a big fan of Feyerabend, but he was right that science is pretty anarchistic. 'The scientific method' varies from individual to individual and field to field. I'm not saying that's always good, by the way. Some fields can become quite pathological, but fortunately, natural selection usually has its way and corrects things.
I personally think learning about many of the issues in the philosophy of science is more useful to non-scientists than it is to scientists.
So many people have no clue what science is, how it works, and what makes it different from non science (even generally). All they know is that scientists are white men in lab coats, that science explains the "how" and religion explains the "why", and that there are "other ways of understanding" other than science, and so on. It would do some good to see myths like this dispelled.
Knowing what's distinctive about science is irrelevant to the practice of science - as long as you work within the methodology, it doesn't matter to your results one whit what other methodologies do.
And for trying to find the line between science and non-science, don't you think there's a bit of a circularity problem there? If there isn't a clear meaning intended when people call themselves philosophers of science, then there's no meaningful distinction being made between them and other philosophers.
If the response to "what do you study?" is "I don't know, that's what we're trying to find out", you're wasting your time.
Caledonian, I take it that one can identify "science" by enumerating the fields included in the category (biology, chemistry, physics, astronomy, geology, ...) and then pose the question of what commonalities (if any) these examples of scientific fields have with no risk of circularity. Especially since we're leaving it to the scientists to tell us what the sciences are.
You're quite right that one could practice science without caring whether other folks are using the same methodologies -- the issue of what makes science distinctive is a question that occupies the philosophers.
Your centipede analogy is good for describing the practice of normal science. By the same token, however, this discussion reminds me of my favorite quote from James D. Watson's The Double Helix:
"One could not be a successful scientist without realizing that, in contrast to the popular conception supported by newspapers and mothers of scientists, a goodly number of scientists are not only narrow-minded and dull, but also just stupid."
Thinking about academic scientific research, as opposed to the kinds and types of science that go on in all sorts of other places: is it possible to think of it as somewhat like a large set of craft guilds? There isn't any one scientific method they all share that everybody gets instructed in. You find someone to take you on as apprentice and you learn the methods of the local guild.
This is meant to be an analogy and not a literal descriptive but I think there's something to it.
So the question remains, what do all the different guilds have in common that allows us to call them all "science"?
As a sociological observation, I think Zuska is spot-on. Perhaps a sociological answer is most apposite for her question --
what do all the different guilds have in common that allows us to call them all "science"?
--the fact that the guilds put the word "science" in their name. The Screen Actor's Guild, for example, does not.
Dear Dr. Stemwedel: May I call your attention to a most interesting paper by a physicist, Prof. B.K. Jennings, recently published on the arXiv:0707.1719 [physics.hist-ph] 11 July 2007. The paper is entitled, "The Scientific Method".
I would be most interested in your opinion.
Allow me to quote from the Abstract:
The nature of the scientific method is controversial with claims that a single scientific method does not even exist. However the scientific method does exist. It is the building of logical and self consistent models to describe nature. ...
As a physicist, I find this argument very impressive.
However, could it be that this is a physicist's view of the "scientific method" and that other scientific disciplines may have differing perspectives?
I look forward to your views.
Sincerely,
Dr Aadu Pilt
Isn't the thing that sciences have in common that they are evidence-based attempts to understand the world around us? I think that's broadly the same as methodological naturalism.
Actors don't do it, politicians don't it, astrologers don't do it (sorry Behe!), philosophers don't do it. That's why they're not scientists. Typically, engineers don't do it either: they take existing understanding of the world and apply it to solve problems, so they're not scientists either (and that's not intended to insult engineering in any way at all, and of course it has its scientific side).
As a scientist, you can play fast and loose in private ("bad science" if you like), doing quick-and-dirty experiments and assuming that correlation might indicate causation, but when you go public with analysis and theories, then the evidence has to stack up. And that rigor is unique to science; economists don't do it, historians don't do it, they're not scientists.
Good experiments help us to "unstitch" the world and push unwanted complexity to one side so we can see how it's put together.
Fine. But then you're presuming that there IS a unifying set of principles behind the things that are called science. Shouldn't the first task be to determine whether the category is an arbitrary listing or not? If not, the philosophers of science have set themselves on a snipe hunt.
Then I hope you have a way to identify 'scientists' besides "those who practice science". And how do you justify that identification? More circularity.
Then the question becomes: what are the philosophers good for? And if they don't practice science, why should we care what they think?
I once participated in a heated discussion between one individual and several others, in which people were arguing over what the definitive properties of Christianity were. Finally, the one individual decided that our arguing was moot and the only way to resolve it was to ask the Christians what defined their faith.
There seems to be a similar problem here. If we lack a definition for science, and we let scientists determine what science is, we need another definition to specify who counts as a scientist and who doesn't. And how do we justify that definition, especially given that the category is generally considered to be defined by the practice of science.
It's quite possible that the sciences won't end up having a common core -- that there won't be any there there.
And, I pretty much said in the post that scientists don't need to care about what the philosophers of science think.
I just re-read one of my favorite books, The Rise of Scientific Philosophy, by Hans Reichenbach. As classic today as when it was published in 1951.
Then why should anyone else?
I always figured that philosophy of science bears a similar relation to science that philosophy of language does to language. You won't generally use language any better for having studied Frege, just as you won't generally be a better scientist for having studied Popper. However, if you understand more about how the process works on a conscious (as opposed to subconscious) level, then you might see where it can lead you astray. (I guess that happens less often in language, except for deciding the difference between when someone was lying and when they were just being misleading.)
Then why should anyone else?
Why do you? (After all your hard work in this thread, you can't tell me you don't.)
Janet -- thanks for taking the bait! (Bzzzzzzzzzzz, or whatever noise happy gadflies make.)
FWIW, I don't think Sir Peter meant that scientists don't need to learn the(?) scientific method -- to quote himself again,
which goes nicely with your suggestion that one learns the methodology that matters by doing science.
The reason I've always thought Sir Peter meant, by "scientific methodology", what I call "philosophy of science", is that he was fond of contrarian and rather dismissive positions -- and, again as you point out, scientists really do not need philosophy of science for their day-to-day jobs.
My own response to this is two-fold:
1. Ethics matters, more than ever as science's power to do stuff increases; scientists may not need to study ethics but they would do well to take time occasionally to listen to someone who has.
2. There is little reason to think that scientific practice, defined as what scientists do, is optimal in its efficiency or its direction; if there is anything that could show the way to improvement, surely it is the thinking-about-science-from-an-angle that philosophers do?
Then why should anyone else?
I don't see that this is so hard to figure out. For one thing, because other people are faced, even if not always in as rigorous form, with the same questions philosophers of science often ask about the nature and process of science. The reason scientists don't need to worry too much about it is that qua scientist they just need to do the science. But people trying to get a handle on (e.g.) how to teach children to think scientifically, or how best to approach scientific matters in government policy, or how best to filter out false expertise from genuine expertise in scientific matters, will be asking the same general kinds of questions philosophers of science look at more rigorously.
On the common core question, it's noteworthy that 'philosophy of science' originally was understood wholly to mean 'philosophy of the sciences' (it essentially has this meaning in Herschel, Whewell, and Mill, the great pioneers of the field). That is, originally work in philosophy of science was a wholly pluralistic approach, and there was no assumption that chemists (for instance) were doing the same thing as biologists. So the inquiry doesn't presuppose a common core; it just recognizes our tendency to collect these fields together and, starting there, tries to understand them. If there are commonalities, it is to find them; if not, then it takes a more pluralistic approach.
We can accept for the sake of argument the idea that philosophy of science might be useful in increasing the efficiency and effectiveness of the scientific process. We can accept for the sake of argument the idea that philosophy of science has no applicable uses for scientists. But we cannot accept both simultaneously. The two possibilities are exclusive.
>>And, I pretty much said in the post that scientists don't need to care about what
>>the philosophers of science think.
>Then why should anyone else?
Scientists don't need to care what historians, economists, politicians, psychologists, and so on think. Does this mean no one else should care?
Fine. But then you're presuming that there IS a unifying set of principles behind the things that are called science.
She isn't. She's saying that there are certain disciplines that people call scientific disciplines. But they might be mistaken in giving them a common name. Philosophers of science try to elucidate if there is such a unifying set of principles.
People could only be mistaken in calling a set of things by a particular name if that name had a pre-existing definition attached to it. If not, we can construct whatever set we like with whatever elements we like.
But the set may not have any simpler definition than itself. There might be no principles that would allow us to derive the association of elements without deriving the act of its construction.
Make up your minds! Is the philosophy of science potentially useful or informative, or not? Are you examining how people implement a set of defining principles, or trying to produce a set of defining principles by studying how people behave?
Is there anything of substance here, anything besides an endless progression of smoke and mirrors?
If those fields of study had no implications for people taking part in the endeavors being studied, then no, I don't think anyone should care about them. Not the people endeavoring, nor anyone else. The process of study wouldn't lead to practical applications or even a better understanding of what was being studied - it would be completely worthless.
Make up your minds!
Answer my question!
People could only be mistaken in calling a set of things by a particular name if that name had a pre-existing definition attached to it. If not, we can construct whatever set we like with whatever elements we like.
This is straightforwardly false; it involves a confusion about what is involved in naming. Definition in the strict and proper sense is in fact very difficult; achieving definitions that are both accurate and useful for a given set of purposes is often extraordinarily difficult, and is more often than not the end, rather than the beginning, of an investigation. This does not mean in the meantime that the names in question do not refer, only that the referring is imprecise (and may at times involve a confusion of more than one thing) and governed less by definition than by associated, and sometimes officially standardized, usages. For instance, suppose you are interested in studying jade. You do not need a definition of jade in order to begin studying jade; nor, from the fact that you cannot give a definition of jade at the beginning of your investigation, does it follow that you can just, for no good reason, start including milk among the things called 'jade'. And in studying jade, you start learning the things that are useful for definition: for instance, the fact that what seemed to be a single object, jade, turns out to be at least two, since nephrite (an amphibole) and jadeite (a pyroxene) are certainly not the same. As you continue looking into this, you might start uncovering puzzles about jade worth solving -- e.g., nephrite turns out not to be easily classifiable, since it can be either actinolite or tremolite (my understanding is that they are very similar and under fairly common conditions tend to change into one another over time). Thus, nephrite, which has a standard usage and therefore a meaning, turns out to be resistant to rigorous definition. This might lead in any number of directions in one's inquiry, e.g., to an attempt to come up with a better classification, or a rethinking of the inquiry itself. And so it is here, I would think.
And it has already been pointed out by more than one commentator that you can't use the term 'useful' as a non-relative term. The question is to whom it may be useful, and under what conditions; Janet's point only committed her to saying that it won't be useful to scientists for answering scientific questions. Not surprisingly: it doesn't study the things we have scientific questions about, but the scientific questioning itself.
Nonsense! Unless we've already invested the name with meaning, we can give anything we like any name we please. We could call cats 'dogs', 'cucumbers', or 'storks' if we felt like it, as long as we were consistent in our usage. If we wished, we could call them 'anteaters', keeping in mind that the formerly descriptive name would have its descriptiveness stripped away.
You need at least a preliminary functional definition to do so, or you have nothing to study. What, do you imagine that people can instantly translate foreign languages they've never seen before by examining the meaning? The words have no meaning other than those we agree to imbue them with.
I find myself agreeing with what Bill says. In fact, I am pretty sure that if the research group I was a member of during my thesis days had known more of "method", we would have sooner pinpointed what problems others used ad hocs was for the understanding of the field. Some of them looked fine at the surface, but they weren't necessary to explain the main phenomena in the system we looked at.
I can't agree with this description. Development engineers and others that "push the envelope" are establishing new knowledge, as well as novel ways of problem solving are. The difference is that engineers don't often do it systematically and extensively.
Think "craft" and "immediately interesting and narrow questions". Or think of it this way, where do (most) new technology come from?
Of course, there is a lot of truth in the characterization of "typically". Many of the areas tasks are routine. But the area as such is empirically based.
Brandon already gave a good example of something where the preliminary meaning is governed by vague usage instead of precise definition. I give you another (a classic in philosophy): Try and define what a game is, and see how far you get. There is no precise, all-encompassing definition for a game, yet we can easily classify given activities as games or not-games. The notion of a game is governed by how we use the word 'game'; not the other way round.
Do not confuse a lack of precision or a lack of completeness with a lack of meaning. There are no grounds for requiring language (which has rather organic origins after all) to be precise. Instead, we should look at language itself, like how we look at natural phenomena, and observe if it is precise.
There are two essentially complementary abilities. I use complementary in the sense that they are almost conjugate variables of each other (in so far as that concept makes sense in this context), that of knowing when a hypothesis is wrong, and being able to convince someone a hypothesis is correct.
The latter skill is necessary to get a paper published, to get a research program funded, to achieve "success" in a scientific career. The former is much more important in the day-to-day work of a scientist.
With "enough" data, it becomes obvious which hypotheses are wrong and which ones are correct. The amount of data may not be actually realizable, but in principle with "enough", there would be no doubt.
The skill of a scientist is measured by how much data he/she needs to make those determinations correctly.
At one extreme you have people who can convince others (or themselves) with zero data (or fraudulent data), and on the other hand are those who can make reject incorrect hypotheses with little or no data. For example Einstein formulated his theory of the photoelectric effect, and Millikan spent 10+ years trying to prove it was wrong, collecting ever more data that fit it better and better. Einstein was able to reject the "light is a wave" and the "luminiferous aether" hypothesis whereas Millikan was not, even when Milliken's own data was the best "proof" of the photoelectric effect.
Problems in science occur when some of the paradigms are wrong. Thomas Kuhn's work is extremely important in this, and quibbling about what a paradigm is and isn't doesn't take away from it.
Unfortunately for Millikan, he was too good at convincing himself that "light is a wave".
In essence, that is the problem with superstition. People have convinced themselves and others that certain things are "true", in the complete absence of data, or even with abundant data that directly contradicts their belief (for example young Earth creationism).
I think the "convincing others of the hypothesis" is fundamentally a human social need/ability, where the "rejecting a wrong hypothesis" is the fundamental characteristic of a "scientist".
I think these two characteristics are quite fundamental in the structure and function of a person's brain, and are not things that can be "learned" in the common sense of learning things. You have to be able to modify the paradigms of your own thinking processes, and not everyone can do that, or can choose the paradigms that lead to this.
People could only be mistaken in calling a set of things by a particular name if that name had a pre-existing definition attached to it. If not, we can construct whatever set we like with whatever elements we like.
So I can say that philosophy is a science and the only thing that makes me wrong is that people don't agree with me? If so, how many people have to agree with me for philosophy to be a science?