Indeed… most wouldn’t consider mathematics as science, I don’t think. But I think that makes the net too small… mathematical pursuit is no more pure than science (IMNSHO—but that is probably a debate in itself), and we might even consider that (to support the Popper def’n against my better judgement) mathematical proofs are experiments, and they are acceptable so long as we find no contradiction (which would correspond to bad data) with the rest of mathematical theory.
Well, anyway, that is another way of looking at it. Not one I would subscribe to.
But I believe this is the bulk of scientific work. Bad data means your experiment was screwed up, there was a temperature fluctuation, the compound dissociated, there were impurities, there was a faulty computer connection, a faulty observation, a calculation error… the bulk of what is done under the term of “science” has nothing to do with falsifying theories; I think the whole “falsification” is a lip service to an epistemological foundation for scientific pursuits, but not one that is really followed to a large degree (again, for what we mean when we say “science”). Working chemists in general are a segment of the scientific population that usually do not operate under falsifying conditions. I would never claim these men and women are not scientists.
Wanna make a bet? I guess that depends on what one considers “bad data” based on a flawed theory and “bad data” based on “asumed tolerances of measurement”, doesn’t it?
Not that we need measurement for science, of course (sarcasm)
I’m sincerely flattered by all the excellent responses to the OP thus far. I can’t possibly answer every post, but I’d like to address some of the more outstanding points: Jerevan:
Regarding taxonomy: well, I suspect that if you were to review Linneus’ original work on the subject, you would notice that he does quite a lot of measuring. It seems to me that comparative morphology almost requires meticulous measurements of body size, mass, bone-length ratios, and so forth. Certainly, since Linneus’ time, measurements of various kinds have played a central role in zoology. (By my definition, even if you use a yardstick, you are investigating Nature by means of instrumental measurement.)
However, I would not be so quick to give up on your objection, if I were you. You do have a point, perhaps, that in some cases taxonomy does not depend upon strictly upon measured characteristics – and if so, at least in this instance, my definition fails. Do you have any concrete, specific examples in which a taxonomic scientific study did not in any case rely on measuring? If so, I would be very interested to hear about them.
Tris:
This is precisely the point of my definition: science should be understood as a specific method of knowledge production, differentiated from all other forms of such production by virtue of the fact that it relies upon instrumental measurement.
Regarding astrology: I probably should have addressed this point in the OP, so as to save ya’ll all this jawin’. On the other hand, the astrology angle brings up a lot of other interesting points, so maybe it’s just as well.
Let’s start with Popper: it is in fact a myth that Popper’s criterion actually serves as useful method of differentiating astrology from the natural sciences. To begin with, despite Popper’s claims to the contrary, astrology does, in fact, produce falsifiable observation statements. These statements are called “predictions,” and historically, they have sometimes proven right, sometimes wrong – but they have most definitely been falsifiable. Only the shallowest, “newspaper horoscope” version of astrology is thus ruled unscientific by Popper’s criterion.
In contrast to Popper’s demarcation criterion, on the other hand, I submit that my definition of science provides a perfectly clear method of differentiating astrology from the natural sciences. Basically, I would argue that your account of astrology fails to take note of the fact that astrology consists of two acts: 1) casting the horoscope, and 2) interpreting the horoscope. The first act, which involves a careful measurement of the position of various planetary bodies at a specific point in time, is, in actuality, highly scientific. (Astronomers are in fact deeply indebted to the work performed by early astrologers, and especially to their meticulous measurements and observations, as precursors of their modern field of study.) The second act, however – interpreting the horoscope – relies upon a complex web of mystical associations: Venus/Aphrodite = green = love = sex = the Evening Star = emotions = genitals, etc. This interpretive web is most certainly not derived from an investigation of nature by means of instrumental measurement, but rather by an investigation of the soul of man by means of spiritual illumination. Thus, I argue that while the actual casting of a horoscope might be considered scientific under my definition, its interpretation – which is really the central focus for the body of astrological thought – is not.
I don’t know much about numerology. Phrenology fails because it never involved instrumental measurement – the measurements were almost invariably done by hand.
If you really want to take my definition to the limit, then try considering the case of racial biology. Pochacco:
As I indicated in my OP, (although perhaps not emphatically enough), I reject Popper’s criterion. The reasons are legion. For a real doozy of a critique, check out the Duhem-Quine Thesis sometime.
**As you are probably aware, Popper was not too concerned with what he deemed to be “questions of meaning.” He was simply concerned with finding a means of separating “scientific statements” from all other sorts of statements. Thus (according to Popper), theories (or, to be even more correct, hypotheses) may have more or less “meaning,” but to be considered “scientific,” they must generate falsifiable observation statements.
On the contrary, unless I have misunderstood Popper, his criterion is not concerned with “epistemology.” It is a demarcation criterion; it tells us the difference between science and non-science, and thus cannot be “applied to fields outside of science.” Don’t mistake the technique of “falsification” with its function a marker. I suspect that Popper would reject out of hand your equation of music history with science proper – he was quite allergic to sociology, psychology, and so forth. He even wrote an entire book as a polemic against such piffle, entitled The Poverty of Historicism. starryspice:
First off, welcome to the SMBD! You wouldn’t happen to be related to Scary, would you?
Undoubtedly true. My scientific friends often react in fact to precisely this point, and I admit that, perhaps, my definition misses something essential about the scientific project. However, let me put it to you this way: how do you differentiate your guesswork concerning “the nature of the system” in one of your laboratory experiments, from your guesswork concerning, say, your next door neighbor’s love life? After all, both are theories that you may have about phenomena that have presented themselves to you as an experiencing subject.
I submit that the major difference between these two kinds of guesswork, which allows us to classify the first as science and the second as speculative gossip, is that the former builds primarily up from a basis in instrumental measurement. On the other hand, I think that if you try to classify theories as themselves as “scientific” or “non-scientific” solely on the basis of their characteristics, well… that way lies madness. ultrafilter:
As I pointed out above, in my response to Pochacco, the point of Popper’s criterion is rule out “non-sciences,” such as psychology and evolution. As you may know, for example, Popper came to the conclusion that evolution theory wasn’t scientific precisely because it didn’t produce falsifiable observation statements. (Since then, of course, Popper’s judgement on this issue has been criticized.) erl and Jerevan:
My definition of science allows us classify taxonomy as science, and shelf-stocking as not-science. The key is the word “investigation.” As I pointed out in the OP, the “investigative” character of the scientific project allows us differentiate it from other disciplines, or crafts, that also rely on measurement – such as engineering, and grocering.
And Jerevan, your point about who gets to define science is most timely. There is a distinctly political aspect to this debate. erl:
Try not to be too hard on Popper. Even though I disagree with him, I think he’s a very good philosopher, and very interesting. And in fact, most of the really interesting work that’s been done in the history, sociology, and philosophy of science has been done in relation to his theses, often in the form of a rebuttal to them.
Actually, Popper’s thesis is not, in fact, a scientific hypothesis. It belongs to what Popper would probably term “the field of metaphysics.” It is a statement, or claim, about the difference between “science” and “non-science;” hence, in order to make sense, it must stand outside the scientific frame of reference.
Stated simply: because the falsification criterion is not itself a scientific hypothesis, nor claims to be one, it does not need to be falsifiable. Even though it is not, strictly speaking, a scientific statement, it can nevertheless have meaning. erl:
You make some very astute points, and I agree that your observations concerning the day-to-day, nuts-and-bolts work of scientific research are right on target.
I will have to reflect a bit about my definition of “instrument.” Naturally, it includes all sorts of measuring devices: clocks, yardsticks, scales, etc. It might also include telescopes and microscopes. In addition, it might also include survey forms, such as are used in statistical research on popular opinions and so forth.
I mentioned in the paragraphs following the definition that experimentation is one form of falsification. Falsification is a more general (and more powerful) concept. It explains why certain activities that don’t rely strongly on experimentation (archaeology) can still be considered science. And, as Jerevan Somerville pointed out, it establishes the criterion for what make an experiment rigorous (the potential to falsify an existing theory).
I’ll point out that if the experimenters were certain that the theories were corrent there would be no point in doing the experiment. Experimenters tend to work in areas where many competing theories are in play and where falsification is useful.
The definition emphasizes theories over data collection because ultimately the goal of science is to produce theories, not data. This is not meant to trivialize the role of data collection in science. Gathering empirical evidence is one of the primary ways that theories can be falsified. But collecting piles and piles of data and never connecting them into an explanatory body of knowledge is pointless.
**
This may not be how you define science but Popper’s views are widely accepted in the scientific community.
**
Mathematical theories are most definitely falsifiable. As I pointed out in my original post experimentation is only one method of falsification. Logical inconsistency is another. A mathematical theorem can be falsified by demonstrating that it contains a logical error. A scientific theory can be falsified on logical grounds as well.
**
Your example is specious. Einstein himself suggested ways that his theory could be tested and the first experimental confirmation came only a few years later. He was doing very good science by Popper’s criteria.
The fact that Einstein’s physics fail to explain all aspects of reality does nothing to diminish their explanatory power over a wide range of phenomena. All scientific truths are provisional. Some day Einstein’s physics will be subsumed by a new theory with great explanatory power, just as Newton’s physics were subsumed by Einstein’s.
**
But the OP’s definition of science clearly excludes fields that are commonly considered science (zoology, for example, or archaeology), while my definition doesn’t. What grounds do you have for rejecting it as inadequate?
Oh no? Hmmm, I think that most scientists (myself included) count four basic sciences: biology, chemistry, physics and mathematics. Of these mathematic is fundamental.
… “science is 10% inspiration and 90% perspiration”. Yes, there is good deal of grunt work in it…
I beg to differ, and quite strongly. As a “working (analytical) chemist”, a lot of my effort is spent on problem-solving. Not in the sense of what you mention above – sorting through what you call “bad data” which resulted from flawed experimental design, unforeseen variables, parameters which could not be controlled as planned – but rather examining “good data”, that which falls within acceptable limits of error/tolerance generated from well-planned, well-controlled experiments, and then determining what this tells me is not so: “No, sir, I can’t say what you did make in your chemical reaction, but I can tell you with certainty that it is not the material you expected [theorized] this reaction would produce.” I don’t always have enough information to say what did happen; but I only need one piece of good data to prove what didn’t happen.
So, as for this:
… A good experiment constructed around flawed theory does not generate what I would call “bad data”. It simply produces data which were unexpected – not predicted from the theory. This leads to the conclusion that the theory might be flawed.
“Bad data” result from experiments which are flawed in design or execution, as I touched on above: parameters not controlled as planned, certain variables not taken into account, extrapolation from theory to concrete experiment flawed, instrumentation (when used ) not functioning correctly, and so on. “Bad data” may lead to a re-design of the experiment, but generally speaking conclusions ought not to be drawn from “bad data”.
There are ways of assessing, independent of a given experiment, whether one is looking at unexpected results or “bad data.” One of them, to which you allude, is determining whether measurements fall within acceptable tolerance/limits of error.
eris, I would go into more detail but I simply do not have the time right now. Tomorrow, perhaps.
Mr. Svinlesha, I will read your post at length, and at leisure, later this evening or tomorrow morning. For now I have to run!
Certainly. And what it means, stated simply, is that the scientific method, whose foundational principle is falsifiability, is epistemologically no more valid than any other method of inquiry into the principles and properties of nature.
[QUOTE]
*Originally posted by Mr. Svinlesha *
**
Don’t mistake the technique of “falsification” with its function a marker. I suspect that Popper would reject out of hand your equation of music history with science proper – he was quite allergic to sociology, psychology, and so forth. He even wrote an entire book as a polemic against such piffle, entitled The Poverty of Historicism.
**
I believe The Poverty of Historicism is concerned with the difficulty of constructing testable hypotheses of historical progress. My wife’s work revolves around determining the provenence of certain historical manuscripts. If one carefully limits one’s inquiries it’s possible to construct falsifiable theories about a manuscript’s origins which can then be tested in a variety of manners – analysis of letter forms, DNA analysis of the parchment, etc.
Merely because Popper railed against “Historicism” don’t jump to the conclusion that all history is “piffle”.
Then you would be disputing a claim I never made. Note that I said that falsification is the foundational principle of the scientific method, and not of science.
It bears on your definition because you attached to it a methodology. What method did you have in mind if not the scientific method?
And I am not a hijacker, thank you. I’m a bit sensitive about that particular ad hominem. You asked for criticism. I gave it. Please don’t ask for what you don’t want.
And see, I simply feel that most simply pay them lip service. Admittedly I don’t associate with a significant portion of scientists, I mainly work with analytical chemists in pharmaceutical research, but I would be somewhat surprised to hear even half of them assert that their work had anything to do with falisification of scientific theories.
Maybe they do. Appearances can be deceiving, after all.
Well, as I said, there are ways to conceive of mathematics as falsifiable. I don’t disagree that it can be conceived of that way, I just feel that mathematics are often not considered an element of science because they are not empirical. Admittedly that is reading into Popper’s perspective and thinking about standard interpretations more than taking that straight from the definition offered.
Easy for you to say in hindsight. Besides, it was meant to test the waters of the word “falsifiable”. Astrology is falsifiable. Bending spoons is falsifiable.
Quark theory? Well, now that’s a sticker, isn’t it, since quarks cannot exist independently and can never be directly observed. Is it falsifiable?
I don’t see how the OP’s definition excludes those. Could you elaborate?
Jeravan
Yes, well, I’ve always felt that mathematics was held seperate for not being empirical.
I object to the word instrument in the OP, unless one includes the human sensory aparatus as elements of the class. Then, of course, the definition will fail to differentiate between population biology and racial profiling.
I have great respect for Popper as an epistemologist. I think that his definition for scientific endeavor, however, fails precisely because it attempts to eliminate the subjective elements of what is manifestly a human endeavor. As others have pointed out, astrology, phrenology, et. al. create falsifiable predictions about nature. The difference between those fields and science is that science abandons (eventually) theories which fail to make reliable predictions. The hallmark of science is just this willingness to abandon long-held ideas in the face of contrary evidence–even to the point of questioning the fundaments of a field. Any definition for science which fails to consider this human element fails to adequately define science.
Science is what scientists do. (We might as well make the tautology obvious, eh?)
BTW, erl, I think that many of the day-to-day activities of scientists function under the Popperian view only through a broadening of “falsification” to include repeated performance of well-established techniques with the understanding that if the techniques fail to produce expected results it tells us something about the underlying assumptions of our experiment (which might be as simple as “our instruments are clean”.)
It is certainly a stretch, but that isn’t my real issue with Popper’s definition. I would actually be more than willing to concede that some things done by scientists in a lab are not “science” but “manufacturing”. Just because I drop two objects from the same height and watch them fall does not mean I am doing science. I might just be trying to annoy the people downstairs.
See, and I just have never had the impression that that really was an underlying assumption… it takes much more than falsification to turn science on its head. It takes falsification and something better at prediction. Remember our “absolute == best there is” little spat? Not that I didn’t end up coming around in that context, but in this I cannot, offhand, think of what theory was really just turned on its head through falsification alone.
I would prefer the investigation of Nature by means of empirical measurement. “Instrumental” for me comes too close to implying “instrumentalism” which denies any physical reality.
Okay, now that you’re all dancing let me slip in this definition of science according to Michael Shermer in The Borderlands of Science:
He goes on to describe cognitive methods as hunches, guesses, ideas, hypothyeses, theories, etc. Behavioural methods are research, data collection, collaboration, etc.
The key word is (as already mentioned in previous replies) “testable”.
Wait, now, let’s try to keep something in mind: scientific rigor rests on the potential for falsification, not necessarily the achievement of it. That is, you have to allow for the possibility that your theory is wrong, and proceed accordinly in your experimental approach. It does not mean that you must turn out to be wrong… After all, isn’t this partly why astrology is not accepted as a capital-S Scientific field, because it appears to be predicated on a belief in its own inherent rightness and does not appear to allow for the possibility that it is wrong?
eris, I’d have to do some digging tomorrow morning, but I think perhaps caloric theory was “turned on its head” when it was demonstrated to be inconsistent with observation, even before another, better heat theory was proposed in its place.
I guess that depends on what one considers “bad data” based on a flawed theory and “bad data” based on “asumed tolerances of measurement”, doesn’t it?
(sarcasm)