Alternatives to the mainstream scientific method that have generated real, practical output.

March 17, 2009 11:47 AM

Alternatives to the mainstream scientific method that have generated real, practical output.

Hi all. I appreciate this is a wide-ranging question, and simply asking the question introduces a lot of points en-route that might provoke debate. The overall question, though, is whether any real, tangible, practical output has come from modern alternatives -- or complements -- to the "scientific method".

A debate about the value of the scientific method is probably out of place here -- whether it represents a pinnacle of achievement; if it's a "good" or "bad" thing; etc -- these are conversations that don't fit MeFi's purposes. Ditto the nature of truth, experience, and so on. All valuable stuff, but I'm looking for something specific, for now. I'll be investigating epistemology in my own time.

Both proponents of, and opponents to, the scientific method recognise that it has limitations. But are there any alternatives that have provided as much? I know of very few places to look. Goethean science seems interesting, in its acknowledgement of both the subjective and objective. But how many Goethean scientists have cured a disease?

If we consider the scientific method to be a "gold standard" (and I appreciate that some do not) then we'd recognise that it grew from a distinctly non-scientific background. Proponents would consider it as a pinnacle of sorts, emerging as increasing clarity from confusion. I'm less interested in the tangential off-shoots from this process, more interested in parallel or independent systems of thought -- if that makes sense.

Measuring the "value" of the outputs of any alternative system is tricky, if we're not to use the yardsticks of the scientific method itself, so perhaps that's a sub-question here! But, back to the main thrust: what are the alternative systems, and what is the value of their output?

I appreciate that by asking the question, I'm taking a rationalist approach; but every journey starts somewhere.

posted by ajp to science & nature (27 comments total)

Science is predicated on the notion that causative principles of the world can be deduced from sensible results that (a) can be repeated and (b) point to correlation. Therefore, the conditions wherein science is invalid are conditions where (a) results cannot be repeated or (b) correlation is not equivalent to causation.

If the latter is true, I can't think of any way to know it that is available to us in general; at least not any way that I guess I can believe in. As far as the former, it does happen; the frontiers of our knowledge in general fall into this category. It's quite impossible to know anything about the beginning of the world with the same certainty that we know about the properties of water or air because it's impossible for us to repeat the beginning of the world; we can infer conclusions about it from things we can deduce, but the thing itself is untestable. I would submit that there are more 'unrepeatable' realms than we currently focus on; science has shifted our focus to that which it can test.

Rationalism isn't an alternate approach so much as a more prior approach; science takes part in rationality to some degree, but rational thought can reach beyond the bounds of science in some matters. Mathematics, for example, although arguing this would take some effort.

In the matters you're asking about, I would highly recommend this book: The Great Dialogue of Nature and Space, by Yves R. Simon. The author covers the approach to science from Aristotle through the modern world, and argues cogently that the foundation of science on Space as Descartes founded it and as it has remained ever since lacks some fundamental insights contained in Aristotle's foundation of science on Nature. It is one of the best books I've ever read.
posted by koeselitz at 12:04 PM on March 17 [2 favorites]

I think koeselitz hits the nail on the head here. I do have a question about your answer though - I thought it was a basic premise of science that correlation does not imply causation, and that you cannot make the jump to presuming a causation until you do further experiments examining the nature of the correlation? Am I misunderstanding something here?
posted by idiopath at 12:18 PM on March 17

real, tangible, practical output … from modern alternatives

As far as I know, the scientific method hasn't helped produce superior output in the arts. Science can investigate interesting questions that relate to the arts, but so far, art has relied on an arbitrary, innate process that resists replication. Is art practical? Another subject for debate, but I would say that in the broad sense of "making lives better," it is.

Inspiration has also played an indispensable role in science, in conjunction with the scientific method. If you've read Kuhn's Structure of Scientific Revolutions, you understand that scholarship in a field will plod along until it piles up some number of inconvenient inconsistencies that can't be explained within the prevailing model for that field. At which point someone needs to come up with a new model, and that is a product of inspiration.
posted by adamrice at 12:20 PM on March 17

A fundamental block in your question is that any non-scientific alternative, if it works, gets wrapped up into science. Science is not "the scientific method", that's a classroom fairytale, science is anything where it is possible to see or somehow deduce that it works. So by asking for something that works but which isn't science, you are asking "what is something that is science but which isn't science".

The openness is, as you note, also difficult. Eg
"But are there any alternatives that have provided as much?"

As much what?
Impressive feats of engineering? Reliableness of prediction? New understanding of the previously mysterious? Technological advance? Pages of mumbo jumbo? :-)

Some possible answers:

I would suggest that trial-and-error is still going strong in the modern world. For example, with many years of experience, some craftsmen can create things suggestive of scientifically-underpinned engineering, and/or push the envelope of what is technologically possible.

(In terms of your mentioned example - curing a disease, trial and error has almost certainly had some success there too.)

Along similar lines, genetic algorithms, where artificial selection allows random noise to find elegant solutions to problems.

Along similar lines again, I'd also suggest that within certain limited spheres, the intuition of experienced people can be a fairly reliable path to knowledge.
posted by -harlequin- at 12:21 PM on March 17 [1 favorite]

I'm not sure if this is what you mean, but some science is done not in the traditional hypothesis-driven way, but as what scientists (derisively) call a "fishing expedition" where huge amounts of data are trawled for some sort of pattern. Of course, once patterns emerge, they are expolored in a more hypothesis-driven way.
posted by twoporedomain at 12:29 PM on March 17

Alternatives to the mainstream scientific method that have generated real, practical output.

Science already includes all known methods of generating real, practical output. To put it another way, if a method generated real output, then that method is now science.

If you can come up with a new method of generating same, then your method too will become a part of science. But don't forget to patent it. ;-)
posted by TurnedIntoANewt at 12:31 PM on March 17 [2 favorites]

Along similar lines again, I'd also suggest that within certain limited spheres, the intuition of experienced people can be a fairly reliable path to knowledge.

Expanding on this. Intuition is all we have in developing knowledge where the complexity of the subject is difficult to deduce from mathematical/scientific principles. Examples?:
- Military strategy
- Economic policy
- Weather prediction
- Guidelines on sales, courting, leadership
- Game tactics (e.g. Chess)
etc.

This is a huge shared knowledge basis which is rarely possible to reproduce precisely.
posted by vacapinta at 12:31 PM on March 17

Also, when someone talks about alternatives to science, I suspect that I will soon see a claim that cannot be verified, empty speculation and tautology dressed up as philosophy, or some new religion they invented. More often than not you will also find that they have a profit motive somewhere in there too.

Even creative inspiration or random guesses a-la genetic algorithms are part of a scientific process if what you take from them is a testable conclusion of some sort.
posted by idiopath at 12:33 PM on March 17

People do talk about the "scientific method."

But the scientific method that can be spoken of is not the scientific method.

The "scientific method" is a toy to amuse philosophers and the like, and to keep them out from underfoot while scientists get on with science.
posted by jamjam at 12:35 PM on March 17 [4 favorites]

I think you first need an answer to "What are the alternatives?"

harlequin has it right: if some obvious piece of non-science mumbo-jumbo turns out to be genuine after all, it ends up in the history book as an example of "successful science."

How about a contrived example ...imagine that a non-scientist experiences a religious vision where Jehovah/Thor/etc. reveals a novel mechanism, or a drug, or natural principle, etc. If the revelation proves genuine and actually works, then we'll rewrite history and define it as a hypothesis which survived empirical testing. Probably the scientifically-embarrassing religious vision won't end up in the history books.

Example: dose yourself with LSD, have visions explaining the structure of the DNA molecule, win Nobel, but then keep the origin secret, since visionary revelation isn't part of "reasoning."
posted by billb at 12:44 PM on March 17

jamjam: The "scientific method" is a toy to amuse philosophers and the like, and to keep them out from underfoot while scientists get on with science.

Science assumes several things without ever proving them (for example: that true principles of the world can be tested, and that such tests can be isolated so that they aren't affected by outside forces) while philosophy at least attempts not to assume but to know. Science makes up for its lack of rational basis and cognitive rigor through scholarship and thoroughness, but that doesn't remove the fundamental uncertainty about it. What's more, as the book I referenced above points out, it could be argued that the conceptions which scientists have heretofore formed about the world are incomplete in serious ways. Science doesn't necessarily need to listen to philosophy to come to see itself and recognize its mistakes, but it must form some idea of its own history, which is woefully neglected these days.
posted by koeselitz at 12:44 PM on March 17

idiopath: I do have a question about your answer though - I thought it was a basic premise of science that correlation does not imply causation, and that you cannot make the jump to presuming a causation until you do further experiments examining the nature of the correlation? Am I misunderstanding something here?

Further experiments can only reveal deeper correlation; in fact, you're absolutely right when you mention the nature of the correlation, but modern science doesn't aim at the nature of anything. It aims to see what things do in certain conditions, and to correlate an effect with a cause. If an occurance correlates with another occurance in a rigorously reproducible way, they can be said to be in a causative relationship.

Think about it: if an egg falls out every time we slap a chicken, leading us to believe that slapping chickens creates eggs, and we go to do experiments to test this, there are only two possibilities: either all the chickens we slap will pop out eggs, or not all the chickens we slap will pop out eggs. We're looking for a deeper correlation between the slapping and the laying.

billb: harlequin has it right: if some obvious piece of non-science mumbo-jumbo turns out to be genuine after all, it ends up in the history book as an example of "successful science."

I disagree entirely. There's a false idea some people have that the whole of history has been a long uphill climb, an eons-long progression toward scientific knowledge; but history is actually more diverse and confusing than that. Many students of the world have had vastly different approaches to it; and it's provincialism to assume that all students of the world have either been exactly like modern scientists or just mystical weirdo dreamers.

The present example is a good illustration of this: in a case where we don't know if a correlation really represents causation, Aristotle, who is very rigorously rational, would respond by thinking about that case; he would try to consider what causation is, what its characteristics are, and whether it can be identified in the case at hand. This is not the approach of modern science; modern science says it's fine if you want to sit and think, but real results come with more testing.

It's a small example, but it shows that there are many ways to go, and that the contemporary paradigm isn't necessarily just an aggregation of everyone who's been right in the past.
posted by koeselitz at 12:57 PM on March 17 [1 favorite]

More of the Scientific Method:

In "The Art Of Scientific Investigation", WIB Beveridge gives the formula for solving intractable problems: beat your head against them for weeks. Then give up. Choose a new problem, or perhaps go on vacation. Suddenly the solution to the unsolved problem will appear full-blown and in great detail in your mind while you're taking a shower (or perhaps running to catch a streetcar.)

In "Advice to a young scientist", P. Medawar recommends tolerating Bipolar people on research staff, since their immense creative output in the manic phase justifies the Director's efforts required to keep them from turning to alcoholism while depressed!
posted by billb at 12:59 PM on March 17

I love your contributions, koeselitz, in case anything I've ever said on Metafilter has led you to think otherwise.

The mistakes of science are its growing points, and science has a protean quality that no conceivable philosopher-Odysseus will ever be able to overcome, in my opinion.
posted by jamjam at 1:02 PM on March 17 [1 favorite]

I wanted to mention what I think is the biggest and nearest-at-hand example of an art or study that science has failed at. This guy (like him or hate him, he's actually quite brilliant if somewhat misused, like Nietzsche, who was thought erroneously to be a Nazi for decades after his death) spent his life arguing, I think, that there was one, and probably only one, realm in which science had utterly failed the modern world: while science has saved lives and conquered much pain, political science still does not even deserve the name of science. Unfortunately, not only is the political realm one in which it is effectively impossible to conduct experiments, but even many of the principles involved (sociology, psychology) are at least hazy and shadowy to science. This is not to say that there is no way to think about it or even to be certain about some of its principles; but this is not a scientific certainty.
posted by koeselitz at 1:04 PM on March 17 [1 favorite]

jamjam: if I sound like a pompous boor, it's not out of dislike, but just because that's what I am. ;)
posted by koeselitz at 1:06 PM on March 17 [1 favorite]

To play devil's advocate, koeselitz, it seems to presume much to think we can know the metaphysical why and how of events. I see a certain admirable humility in saying "we know nothing about where an egg comes from, but I can show you these charts of the probability of one coming out under the following controlled circumstances".

Aristotle and his peers frowned upon empirical testing because they literally saw the material world as an imperfect copy of an intellectual ideal, so a thought experiment was a more reliable insight into the nature of things than a material experiment. While I may enjoy occasionally pondering unprovable abstractions, I don't think they will do me much good.

Mathematics are another thing entirely in this case, I see them as being formal intellectual games, where the demonstrably internally consistent results may be useful as a framework for understanding future concrete phenomena. Once again we have this distinction of testability, two mathematicians working from the same sphere of mathematical expertise can verify or fail to verify one anothers work.
posted by idiopath at 1:38 PM on March 17

have visions explaining the structure of the DNA molecule, win Nobel, but then keep the origin secret, since visionary revelation isn't part of "reasoning."

Friedrich Kekule's famous/infamous story of how he elucidated the benzene ring after a dream vision of a snake with its tail in its mouth is a staple of every "Science for Kids" book, though.

Today, an awful lot of scientists and historians of science emphasize intuition as an important part of the scientific process. Frederick Grinnell's new book, Everyday Practice of Science: Where Intuition and Passion Meet Objectivity and Logic sums up a ton of thinking and research on this issue.
posted by Sidhedevil at 1:57 PM on March 17

Vacapinta lists

Expanding on this. Intuition is all we have in developing knowledge where the complexity of the subject is difficult to deduce from mathematical/scientific principles. Examples?:
- Military strategy
- Economic policy
- Weather prediction
- Guidelines on sales, courting, leadership
- Game tactics (e.g. Chess)
etc.


One of these things is not like the others. Except for weather prediction, all of the above, like koeselitz's additional example of political 'science' and a further example of economics, would fall more or less in the realm of social science.

Weather prediction isn't perfect but it's come a long, long, LONG way in the last 30 years with the advances in meteorological technology. Meteorology also has the advantage of being able to continually collect data that affirm or contradict their predictions, which leads to better and better forecasts.

The rest can use numbers to try to tart up their legitimacy, but there's still a lot of bullshit and guesswork. You can politely refer to this as "intuition", but let's be real here.
posted by Sublimity at 2:07 PM on March 17

I think twoporedomain's mentioned the most important frontrunner. A core principle of "the Scientific Method" is that it is hypothesis driven (btw, I'm in biological research, but never hear the term "scientific method" used in everyday conversation, rather "hypothesis-driven research"). We now have technology at our disposal that could, arguably*, one-day replace hypothesis-driven research (at least in biology). Microarrays, large-scale mutagenesis programs and other high-throughput data-generating systems essentially allow scientists to pinpoint genes and proteins involved in various biological systems without needing any special a priori knowledge, and hence, by extension, without needing a hypothesis.

Approaches to problems by "trial and error" (as previously mentioned by -harlequin-) are similarly non-hypothesis driven, and are still employed to greater and lesser extents in various disciplines. In biological research, trial and error still has it's place for solving technical problems. I can't really speak for other fields.

*Arguably, not convincingly.
posted by kisch mokusch at 2:25 PM on March 17

idiopath: Aristotle and his peers frowned upon empirical testing because they literally saw the material world as an imperfect copy of an intellectual ideal, so a thought experiment was a more reliable insight into the nature of things than a material experiment. While I may enjoy occasionally pondering unprovable abstractions, I don't think they will do me much good.

First, Aristotle has no peers. (Hah!) I don't think the medievalists even really plumbed his depths, though they had their own interesting ideas.

Second, Aristotle's approach was, I think, more like this: at the beginning of On The Soul (which is, I think, his most honest treatise about metaphysics) he asks a question. One can fairly say when a man becomes angry that that person has become angry because he was hurt by an unkind word, or because he can't stand bagpipe music, et cetera. At the same time, someone else could say that the man becomes angry because of 'the boiling of his blood' (a common physical explanation of the time) or because of some other chemical imbalance. Now, Aristotle asks: which is the truest explanation? And, he asks, which person is really a scientist? Is it the person who investigates what it was inside the man, what he says made him feel that way? Or is it the person who investigates what can be seen about the man's body and his change in state?

Or, Aristotle asks significantly, is it both?

I've heard it rightly said that Aristotle wasn't a 'materialist' or an 'idealist' but rather someone who saw the limitations of both views. To be precise, he saw, I think, that pure 'materialism' would mean that nothing had a nature or a principle about it, and would thus rule out knowing entirely; but pure 'idealism' would be incoherent and thus lead to the same place.
posted by koeselitz at 3:12 PM on March 17 [1 favorite]

Not sure if it's relevant but I posted this question awhile ago. May be some useful comments here: http://ask.metafilter.com/42964/StatsFilter-Evidence-challengingcontradicting-conventional-controlledplacebo-trials
posted by pallen123 at 7:55 PM on March 17

Studies of the efficacy of drugs routinely take into account the existence of the the placebo effect.

However, actually making use of the placebo effect for benefit of the patient is frowned on in western medicine. Many 'alternative' treatments make much use of the placebo effect.
posted by yohko at 8:51 AM on March 18

Hi all.

Thanks for many interesting answers. However, I think we're slightly off-topic.

A clarification: this is about science. Wikipedia defines science as "systematic knowledge or practice". In a more restricted perspective it continues: "Science is the effort to discover and increase human understanding of how reality works". "Experimental science" is about "investigating causal relationships among variables" whereas "applied science... is the application of scientific research to specific human needs".

The reason I mention Goethean science is that it seeks to acknowledge, rather than attempt to remove, subjective experience. There is no true objectivity; subjectivity is never truly removed. We deal in shades of probability, never in absolute certainty. So acknowledging this seems sensible, and may even be valuable.

So are there any alternatives to the "scientific method", that seek to achieve the same goals, by different means, and have produce valuable output? Are there any parallel systems that have produced a demonstrable improvement in our understanding of reality or causal relationships, and have addressed specific human needs?

Some specific feedback:

@koeselitz: Thanks for the book recommendation, I may well have a look. Does it steer clear of the author's Catholic worldview?

@admarice: During the Renaissance, science and art were largely inseparable. Was Leonardo an artist, a scientist, both together but separately, a mixture of the two, or something else? Does the distinction even make sense? Science may give us many insights into the arts, just as other disciplines can give us insight into science. On a mundane level, many artistic materials rely on chemical and physical science. Many sculptors rely on modern engineering science. Also, just because art is "arbitrary and innate" at the moment, does that mean it always will be? Is it unknowable or just unknown? The same qustions relate to inspiration.

@-harlequin-: I kinda agree. Modern Western science seems to be an agglutinative process, cherry-picking the best-of-breed from other disciplines by applying the scientific method to their output, and strengthen itself in the process. I don't mean to suggest this is by design, it's more of an artefact. koeselitz seems to give the lie to this suggestion.

@-harlequin- again: Trial and error is a good suggestion, but I'd discount it because it's not much of a process ;-) Also it seems that we've learned to apply repeatability and falsifiability to trial-and-error, thus discarding it (largely). So progression from trial-and-error to the scientific method seems sequential; I'm probably after more of a parallel approach, if that makes sense. Hopefully my clarification above helps too. (Also -- genetic algorithms stem entirely from scientific work......... but I quite like the suggestion.... not sure -- do they somehow replace the scientific method, or are they just another tool?)

@twoporedomain: I like the "fishing expedition" notion. But do people genuinely go looking for patterns without a prior agenda? Is this something parallel to the scientfic method, or is it just a statistical tool for scientists? I'm reminded of longitudinal qualitative research papers; eek.

@TurnedIntoANewt: "if a method generated real output, then that method is now science." Maybe not yet! :-)

@jamjam: Dodgy ground, surely? If "science is what scientists do" then it's a self-justifying description, which is exactly what scientists rail against so much. It all starts to get a little Dadaist -- have you been reading Feyerabend? Surely a consistent definition is required?

Might write more later... gotta run for now.

Thanks again for all the thoughts and info, great stuff. But are we reaching the conclusion that the scientific method is the only option out there?
posted by ajp at 11:28 AM on March 18

ajp: ... I think we're slightly off-topic. A clarification: this is about science... The reason I mention Goethean science is that it seeks to acknowledge, rather than attempt to remove, subjective experience. There is no true objectivity; subjectivity is never truly removed. We deal in shades of probability, never in absolute certainty. So acknowledging this seems sensible, and may even be valuable.

Ah... now I see what you were driving at with Goethean science.

I have to say, however, that I don't see how what you're describing could be called 'science' in any meaningful way. Science, by definition, seeks to approach understanding, despite what some positivists might tell you. If there is really em>no true objectivity, then we don't deal even in shades of probability, or even in shades; probabilities are predicated on certainty about some aspect of a thing, and without at least some objectivity, it's not so much that we can't do science; it's that science has no meaning whatsoever. If there is no seeing whatsoever, the question of whether there is anything to be seen is something of a moot point.

You make what I think is a valid criticism of so-called 'western' science, what might more properly be called 'Cartesian' science. The book I mentioned deals with this very well;
Descartes instituted the new sciences by building them on the foundation of space, and therefore perspective - the position of the observer, the categories of subjective and objective - became paramount. It became necessary to assume that the observer's perspective could be absolute, or that the observer's influence could be removed, from any experiment. There are ways in which it can be argued that this outlook on the world is being shown to be wrong by the scientific researches of today.

But the question is this: is there any way to know beyond this apparently limited framework? Aristotle seems to believe that this is possible, and he outlines the way in what is probably the most important section of his great On The Soul, the eighth chapter of the third book:

And now, bringing together what has been said about the soul under one main point, let us say again that the soul is in a certain way all beings, for beings are either perceptible or intelligible, while knowledge in a certain way is the things it knows, and perception is the things it perceives...

This is how science is possible for Aristotle, who saw that perception has the fatal flaw of limiting perspective and therefore can never be a complete or absolute picture of the truth about the world: because there is a sense in which these faculties become the thing they encounter, there is a sense in which our minds become the things they know. In that sense, all the barriers that are inherent in perception are removed as knowledge becomes identical to the object known.

There are ways to argue against this possibility for knowing. However, it is a strong argument. And I think that this is the best answer I can give you: Aristotelian science, philosophic science, is the only thoroughgoing alternative to the modern sciences that I can think of.
posted by koeselitz at 2:21 PM on March 18

ajp: Thanks for the book recommendation, I may well have a look. Does it steer clear of the author's Catholic worldview?

The author is a follower of St. Thomas of Aquinas and a medieval scholar, although he's pretty well-rounded in other areas. I don't remember him mentioning religion in that text.
posted by koeselitz at 2:24 PM on March 18

@koeselitz:

"Without at least some objectivity, it's not so much that we can't do science; it's that science has no meaning whatsoever. If there is no seeing whatsoever, the question of whether there is anything to be seen is something of a moot point."

That sounds heretical; I like it :-)

Are there degrees of objectivity? Can you have "some" objectivity?

If so, then scientific investigation has meaning within its definition of objectivity. But if there are degrees of objectivity or alternative (and "valid") definitions of objectivity, it seems arbitrary to pick any one objectivity.

If however there is "one true objectivity" -- if degrees of objectivity, or alternative definitions, are invalid -- but we recognise that objectivity is impossible to define, or might even be unknowable, then "objective science" has no more meaning than non-science.

So it seems that, regardless of objectivity, non-science has as much validity as science.

(Just to set something straight... I don't much like this conclusion, but I want to see where it goes).

Favouring one -- "science" or "non-science" -- rather than the other seems arbitrary. I guess you could pick one based on its results, and scientists might point to science's impressive track-record. But anyone can point to a flawless track-record; it doesn't mean that the system used to produce that track record is flawless. The ends don't justify the means. Also, perhaps they've just got lucky; or perhaps they're covering up the flaws in the track-record.

Is that a fair conclusion? I suspect this is well-worn ground.

Could anyone recommend thinkers who have written about this? I'd like to investigate the demarcation problem; Feyerabend; confirmation holism; notions of objectivity in science and the philosophy of science.
posted by ajp at 7:53 AM on March 20