Talk:Scientific method/Archive 10

This is an archive of past discussions about Scientific method. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 5

←

Archive 8

→

Note

Unfortunately, there are multiple threads of conversation on the same talk page. I am attempting to at least separate the threads by adding more of an outline structure to the headings. My first change is to push this linear structure into more of a tree. Ancheta Wis 08:51, 30 Jul 2004 (UTC)

Changes,post-archive 9

Fwappler

AAh, Ancheta, unfortunately a good article is a passing thing in the Wiki. Already the rot has set in. Banno

What the devil is a 'reproducible meaning'? Does the meaning of 'dog' get reproduced each time someone uses the word? In that case, what would be an example of a meaning that was not reproducible? Banno

What could it mean for an 'idea' not to be reproducible? Perhaps it means that the idea was forgotten? Banno

The edits to the material on special relativity add nothing, and the use of the term 'infamous' is POV. Gotta love the claim that certain terms were both 'irreproducible' and 'widely repeated'. Banno

And finally, it is surely possible for a scientist to change their mind about the subject of their research. Doing so is revisiting the 'characterization' stage.Banno

Can't see anything worth keeping in these edits. Reverted the lot. Have I just been Trolled? Banno 02:35, Jul 11, 2004 (UTC)

Banno, An interesting problem in determination, is it not? And there are levels of scale in time and perspective in such problems as well. I see no way to be sure one has not been. ww 14:21, 13 Jul 2004 (UTC)

Banno, (Ancheta Wis 10:21, 11 Jul 2004 (UTC))I propose the following:

Page protection

This article has taken years to get to the current state; over a thousand edits have been made and dozens of contributors have taken part, the vast majority well-meaning. As in the Scientific Method itself, which took centuries to develop, and a specific set of societal conditions, which protect the scientists who practice it, we are asking that this article be taken under the stewardship of the system administrators. If there are meaningful, substantive changes, please insert them in the talk page, and let the combined talents of the wiki community decide on the proposed changes.

(William M. Connolley 12:29, 11 Jul 2004 (UTC)) There is absolutely no justification for this. I oppose it strongly.

No, Ancheta. That would take all the fun out of it. The real merit of the Wiki is in the process, not the product. Banno 12:46, Jul 11, 2004 (UTC)

That really isn't going to happen while it's on Wikipedia - David Gerard 13:18, 11 Jul 2004 (UTC)

Thank you, all. Never Mind ;-0) Ancheta Wis 18:05, 11 Jul 2004 (UTC)

There is another article to be cleaned up

A professional historian has determined that scientific enterprise is in need of editorial help. This topic is based on the 'scientific method' article. If you want to contribute your knowledge about the 'scientific method' to scientific enterprise, it is currently open for edit, with our best wishes to you.

I'm glad that a 'professional historian' has taken an interest. But, in the nature of the WP, extra-WP position or status does not take priority. One must take one's chances, in a sense. This is a 'community effort' (self-selected really) and it includes both professional and, ahmm, non-professional members. More or less like life, and there are disadvantages to that approach. But it's the approach we've all accepted. ww 14:10, 13 Jul 2004 (UTC)

Banno

Ouch, Ancheta -- no worse rot than attempting to steer millions of (would be- ?) practitioners. Frank W ~@) R

What the devil is a 'reproducible meaning'?

Where in Wiki is this answered more completely than in the corresponding articles on Reproducibility and Meaning (and links therein)?

(As far as those articles are not yet explicit enough for you to answer your following more specific questions yourself, they can be be improved correspondingly.)

Most concisely perhaps: what remains invariant under any (one-to-one) relabelling of variables;

meaning which can be expressed equally and distinctly in any (sufficiently rich) language;

meaning which can be known and agreed upon by all (who are capable to distinguish "knowing" from "not knowing"), at least in principle.

Does the meaning of 'dog' get reproduced each time someone uses the word?

No -- hence the distinction between the word or label ('dog') and any of numerous notions (for instance reproduced as dog) to which this word may be attached.

However, from repetitive specific use (e.g. under certain societal conditions) it may at least be expected that a word is and remains consistently attached to one particular notion, and that it is at least insofar meaningful.

In that case, what would be an example of a meaning that was not reproducible?

As the case is, and as presently in the article, for example Newton's meaning of 'length'. Apparently he suggested that this word ought to refer to some notion which should be "well known to all", and surely other than dog, for instance. If so, which notion ?? (If any besides "just that".)

What could it mean for an 'idea' not to be reproducible?

To be a more or less vague mental reconstruction of (one's own) perceptions, as an Idea in the sense of Hume;

to be a thought which is not guaranteed to be comprehensible and agreeable to all (who are capable to distinguish "agreement" from "disagreement"), at least in principle.

Perhaps it means that the idea was forgotten?

Worse: that several interchangable but contradictory ideas were held together.

The edits to the material on special relativity add nothing

As noted in the Edit summary, they add emphasis to Einstein's advance, over Newton's failing, along with more precise terminology.

the use of the term 'infamous' is POV.

Of course -- the point being that, and why, Newton skipped what Einstein recognized as essential. Is this POV in dispute ?

Gotta love the claim that certain terms were both 'irreproducible' and 'widely repeated'.

A juxtaposition of quite distinct concepts, which in less sensitive settings are however often not well separated; much like '...mass and weight...'. What's not to love.

And finally, it is surely possible for a scientist to change their mind about the subject of their research.

Certainly in principle.

Doing so is revisiting the 'characterization' stage

That's starting afresh with the 'characterization' stage of some other subject of research, without any consequence for the 'characterization' stage of the previous one.

At issue is instead whether or not 'experiment (test)' applies to 'characterization' at all, as the present article claims.

Example (somewhat elaborate, mostly to be sure about what constitutes one experimental trial) -- suppose:

Categorization: per Wikipedia the notions dogs, cats, squirrels and sparrows are given; based upon which I (as anybody, at least in principle) am able to recognize and distinguish them when I see them. I've observed each of them before in my neighborhood, and I'd expect to observe them again. Also: I (as anyone, at least in principle) am capable of ordering my observations.

Hypothesis: I never see a dog and a cat together, at once, and I never see a squirrel and a sparrow together, at once. Further, whenever I see a dog, then I see at least one squirrel but no sparrow before I see a cat; whenever I see a squirrel, then I see at least one cat but no dog before I see a sparrow; whenever I see a cat, then I see at least one sparrow but no squirrel before I see a dog, and whenever I see a sparrow, then I see at least one dog but no cat before I see a squirrel.

Thus the point there may be a 'necessary' or 'characteristic' quality which has not been investigated enough (or 'characterized' enough) to even think about ( or 'hypothesize' about). What is the relationship of these natural, existential statements to the universal statements that would arise in the 'hypothesize' stage of the article.

For example, a natural restatement for the hypothesis stage might be

Every cat is anti-dog; every sparrow is anti-squirrel; every dog is pro-squirrel; no dog is pro-sparrow; every cat is pro-sparrow; every squirrel is neutral-to-sparrow;

Caution - characterize is not the same as categorize. Another researcher might seek qualities about the subject, rather than predicates about cats, dogs, sparrows, squirrels. Occam's Razor might even suggest narrowing the investigation into fewer items, cats, squirrels and dogs, (all mammals) say. That would simplify the investigation. Another plan of attack might be to study anti- vs pro- vs neutral-to-.

Ancheta Wis 04:00, 12 Jul 2004 (UTC)

(Note that this 'Hypothesis' is expressed in terms of the notions given per 'Categorization' above.)

Prediction: The hypothesis will correctly describe the order of my observations in the next trial (in which I see both at least one dog and one cat).

Experiment: (can start as soon as I step out in my neighborhood; I hope it won't take too much time/resources to collect one trial worth of observations to evaluate the resulting measurement for comparision with the prediction).

Suppose now I had been fortunate enough to complete the scheduled trial, and I had obtained a result in contradiction to the prediction; e.g. from having observed first a cat, then a squirrel, and then a dog.

Note that this result value is expressed in terms of the notions given per 'Categorization' above.

Then the consequence for the hypothesis are surely undisputed:

The hypothesis is thereby experimentally falsified (and it had been experimentally falsifiable to begin with); it should be disregarded.

Using the notions given per 'Categorization' above, I'd proceed to express and consider any other hypotheses whose predictions are not in contradiction to the obtained trail result (and of course only such hypotheses which are experimentally falsifiable similar to the initial 'Hypothesis' above).

But: Do you (User:Banno, or anyone else) claim that there should be any consequences on the stated 'Categorization' as well ??

If so: then which? Should all or part of the above 'Categorization' be disregarded as well? How, if at all, should the result value then be expressed, based on which these consequences are considered? How could be expressed which, if any, 'Hypothesis' had been falsified in the trial under consideration?

If not: then I suggest that the article should make this point more visible.

Regards, Frank W ~@) R 23:29, 11 Jul 2004 (UTC)

Frank, I must admit I didn’t understand much of this. But tell me, what would be the implications for method if you found, on further observation, that what you had though were cats were actually a new breed of Fox? Would you reconsider your characterisation? Banno 01:58, Jul 12, 2004 (UTC)

Hallo Banno, (Frank W ~@) R 06:50, 12 Jul 2004 (UTC)),

I must admit I didn't understand much of this.

I'm sorry -- I found your above questions understandable enough to respond to them in kind. It might be valuable, though perhaps even more of an admission, if you could point out more specificly some of what you didn't understand (or what you don't distinguish); better yet, if you/anyone addressed my concluding questions from above.

what would be the implications for method if you found, on further observation, that what you had though were cats were actually a new breed of Fox? Would you reconsider your characterisation?

The problem or topic you're raising is very relevant, but I disagree with the ordering presumed in the first question:

A new characterization, containing the notion Fox, would have to be considered first anyways, before a experimental result involving 'Fox' could even be obtained at all (such as "my observation which was neither_Cat_nor_Dog_nor... was in fact of a Fox").

Without having defined and choosen or switched to the research subject 'Foxes (in my neighborhood)' already to begin with, I (in principle, or in practice: whichever legions of taxonomists it might require) could not have made a determination such as "Surprise! -- this one is actually a Fox breed" at all.

Therefore my answer to the second question is:

No -- once 'stage 4' claims a result, there can't be any re-consideration of the corresponding 'stage 1'.

That leaves only consideration of any and all sorts of 'stage 1' a priori.

But just a moment ... there's the devellish problem of resources:

Of "any and all sorts of 'stage 1 Characterisations'", which one (or managably few) ought to be considered and invested in next?

Such decisions may well be based on all available observations, i.e. records of all preceding 'stage 4' experimental trials.

In the context of my example trial sketched earlier (and actually carried out just a while ago: I did indeed see some sparrows, one cat, one squirrel and two dogs, separately, in this order): importantly, I also collected plenty of observations which didn't result in me finding any one of these four instances.

How about therefore to change the example research subject profoundly, to include besides Dog, Squirrel, Cat, Sparrow also notions such as Human, Car, Tree, Leaves of Grass, and of course: Fox;

just as long as all those are indeed reproducible and distinct notions, which could and should be asserted before plunging into the corresponding 'stage 4', of course.

My immediate conclusions:

'stage 4' can indeed have implications for 'stage 1'. But, in disagreement with the present article:

The consequence is the revision (rather than affirmation) of the choice of 'characterization' for subsequent use,

not a revision of any one 'characterization' itself; and

Such a consequence to 'stage 1' is not only due to result values obtained in 'state 4', but also (and perhaps mostly) an attempt to obtain additional result values (e.g. from already available 'stage 4' observational data; a.k.a. "re-analyzing data with new techniques or objectives").

Both support the point made earlier (with which my above questions were concerned) that any one particular 'stage 1 characterization' is not subject to experimental test, as provided by 'stage 4' result values;

since, if it were then going to be deemed "false", the supposed result values should have become irreproducible, i.e. lacking any 'characterization' of the terms in which they may have been expressed initially.

IOW: once 'stage 4' claims a result, there can't be any re-consideration of the corresponding 'stage 1'.

Regards, Frank W ~@) R 06:50, 12 Jul 2004 (UTC)

You revised the characterisation in the light of observation. QED. Thank you. Banno 08:28, Jul 12, 2004 (UTC)

Hallo Banno, (Frank W ~@) R 02:52, 13 Jul 2004 (UTC)),

You revised the characterisation

No -- I revised (or rather: I expanded) the choice which of all characterizations to consider most urgently next. (As we had agreed right away, it is surely possible for a scientist to change their mind about the subject of their research.)

I did not revise the initially selected characterisation itself; it remains unchanged regardless of when, or even if, it will be used again.

Further, I considered a change in choice of characterization not as a consequence of any particular result of an experimental test, but because observations were available that didn't have relevance for the initially chosen subject.

Do you understand these distinctions ?

Consider Douglas Hofstadter who, after having done research (Ph.D.) in solid-state physics decided to focus on cognitive science instead. (While I, today, focussed on recognizing and counting 'Cars', 'Squirrels', and 'Trees'.)

Would you claim that Hofstadter's decision was the consequence of test results of solid-state physics experiments ?

Would you claim that Hofstadter's choice constitutes a revision of the characterization which underlies the science subject of solid state physics ?

in the light of observation.

Yes, in the light of observation; not in consequence of the result value of an experimental test.

QED.

For comparison, I repeat what's to be demonstrated or at least settled:

At issue is instead whether or not 'experiment (test)' applies to 'characterization' at all, as the present article claims.

I note that such a claim arises specificly from the present article line

"4. Experiment (test of all of the above)",

(i.e. including reference to

"1. Characterization")

Frank, the recursive and iterative nature of the scientific method allows revisit of any stage. However, the flow of time. prevents some revisits. Only the integrity of the researcher can distinguish what is known from what is not known at any time, especially when revisiting a stage in the light of later knowledge. Thus fudging the data in Stage 4 to make it conform to stage 2 is grounds for dismissal. Stage 1 and its relationship to Stage 2 is buried in the psychology of the researcher. For example, when Newton was investigating gravitation, only Galileo and Kepler's work was known. Hooke may have intimated an inverse square law, and he may even have believed that he innovated it, but Newton alone had the horsepower; Hooke tried and couldn't prove anything. Newton's character suffuses the subject of gravitation. If you insist on bringing in Einstein, even his work has not superseded Newton because our entire civilization uses Newton's work. Only now, a century later, are we able to begin using work that Einstein was trying to disprove, but in which he was seminal.

Experiment serves as the feedback for the Observation part of the Characterization stage. But there is more to Characterization than the machinery of perception/observation of a natural process, as the article points out; there is the researcher's thought during this stage. Ancheta Wis 10:54, 13 Jul 2004 (UTC)

while the entire present section Experiment

apparently doesn't elaborate or even repeat this claim at all.

Do you/Does anyone disagree that this a conflict, in substance or at least in representation ?

If so, I suggest it be resolved -- with all due disregard to mercy or prior preceding contributions, of course, as shown earlier already.

Thank you.

You're welcome, Frank W ~@) R 02:52, 13 Jul 2004 (UTC)

You present a singular characterisation. You do not appear to deny that one might change to a different characterisation as a result of an experiment; this discussion appears to be a waste of time. Banno 05:44, Jul 13, 2004 (UTC)

Peer Review

This article, perhaps more than others in Wikipedia, attracts the attention and contributions especially from users who are particulary capable of, and used to, expressing, understanding, and respecting the various points of view on the article subject. These users can be expected to be responsive to questions, arguments and suggestions made on the attached discussion page, to be considerate to continued editing of the article in conclusion of such discussions, and to reach agreement on how to continue in improving this article.
As in the Scientific Method itself, which has developed for centuries and continues to be refined and re-evaluated, and which prospers in a specific set of societal conditions protecting the scientists who practice it, we ask the system administrators to support conditions of scientific discourse for those engaged in developing this article.

Being one, I'm at a loss how to 'support conditions of scientific discourse for those engaged in developing this article'. You appear to have an idea that intervention 'by authority' might help. On WP, admins don't have that sort of authority to begin with -- the WP belongs to its contributors of whatever stripe. Second, the continuing controversies with regard to this article would not be, were WP admins to have some sort of authority as here presumed, amenable to settlement by resort to that authority. The fundamental, and extremely longstanding, problem here would not be, and is not, accessible to fixing that way. ww 14:18, 13 Jul 2004 (UTC)

Couldn't help it. Hope you get the 'joke'.

Regarding God This article, perhaps more than others in Wikipedia, attracts the attention and contributions especially from users who are particulary capable of, and used to, expressing, understanding, and respecting the various points of view on the article subject. These users can be expected to be responsive to questions, arguments and suggestions made on the attached discussion page, to be considerate to continued editing of the article in conclusion of such discussions, and to reach agreement on how to continue in improving this article.
As for the worship of God himself, which has developed for centuries and continues to be refined and re-evaluated, and which prospers in a specific set of societal conditions protecting those who worship him, we ask the system administrators to support conditions of theological discourse for those engaged in developing this article. (20040302) David Gerard (next) is right of course.

See above. This really isn't ever going to happen ever on Wikipedia. You still greatly misunderstand the functioning of Wikipedia. If something like you describe is what you want, take your chosen 'good' version and save it somewhere else - David Gerard 16:15, 12 Jul 2004 (UTC)

To bludgeon home the point: you did read that disclaimer that shows up whenever you edit, didn't you? "If you do not want your writing to be edited mercilessly and redistributed at will, then do not submit it." - David Gerard 16:18, 12 Jul 2004 (UTC)

Hallo David Gerard, (Frank W ~@) R 03:11, 13 Jul 2004 (UTC)),

This really isn't ever going to happen ever on Wikipedia.

Is the "This" in your statement to be reproduced as "system administrators supporting conditions of scientific discourse for those engaged in developing this article" ?

If so, well -- perhaps such a prediction will be put to a test. In any case, "questions, arguments and suggestions" (along with "responses") seem to happen at least in some parts of Wikipedia already ...

You still greatly misunderstand the functioning of Wikipedia.

Possibly so. But as a Wikipedia user, I aspire to understand the functioning of Wikipedia merely insofar as it permits me to contribute my understanding of subjects such as the Scientific Method.

That appears to be patently not so, certainly not if you are responsible for: "we ask the system administrators to support conditions of scientific discourse for those engaged in developing this article", - a political stance deeply rooted in opposing the current methodology of Wikipedia. Moreover, the paragraph implies an elitist arrogance that reflects and reinforces only the demotic stereotype applied to scientists. What a can of worms that would be if the administrators acceded. Of course, they never would. (20040302)

If something like you describe is what you want

... yes, I would want the contents of Wikipedia itself to be capable of addressing questions such as those asked by User:Banno above, and I'd want it to be free of conflicts as the apparent one I'm asking about in my reply to him ...

One of the fascinating aspects of Wikipedia is that it's procedure is informal, and yet inquisitorial, adversarial and other procedures are adopted (albeit haphazardly) and accomodated by the community, without any rule-setting from the wikipedia administrators. The (tyrannical) idea of setting limits of behaviour on editors (of one system over another) of wikipedia is an anathema to the full-time wikipedia staff. However, if you feel that the overall approach to the development of a scientific encyclopedia, limited to procedure rooted in the scientific method warrants investigation, I suggest you set up your own wiki and give it a go. After all, the software is freely available, and the Internet is large. (20040302)

take your chosen 'good' version and save it somewhere else

'Goodness' and 'safety' and 'form' of what I've been trying to express are not choices of myself alone, but necessarily ... of (perpetual) committee. (See above.) I.e., of editors who disavow mercy and ownership, but who claim and respect instead:

accountability.

you did read that disclaimer that shows up whenever you edit, didn't you?

As thoughtfully as I'd expect, at least in principle, of any of the thousands of contributors preceding me, the hundreds with whom I may trade edits, and the millions to follow.

Regards, Frank W ~@) R 03:11, 13 Jul 2004 (UTC)

p.s. Hope you get the 'reference':

... him ...

... him -- who?? (Respectfully: reproducibly, please.) Yours? Theirs? The one (many?) being asked for supporting the relevant conditions, and/or the one (many?) having apparently maintained somesuch conditions (FWIW, thank goodness) so far ? ... (Frank W ~@) R)

Hawking radiation (a re-characterization)

Stephen Hawking has changed his position on Hawking radiation (a re-characterization) which for a 10^11 kg black hole, predicts an evaporation time of 3 billion years. Thus one possible experiment is to observe Astronomical objects at a distance of 10.7 billion light-years for Hawking radiation. However, such an object would have sucked in matter in the associated galactic cluster(s). How intense would the Hawking radation would have to be before being noticed above and beyond the activity from the associated materials (such as dust clouds)? Ancheta Wis 17:27, 16 Jul 2004 (UTC)

logic

I’ve placed a more formal account of the theory dependence of observation and the indeterminacy of theory at falsifiability. I think it is better there than here, or in induction (philosophy), but am open to suggestions. Banno 07:58, Jul 24, 2004 (UTC)

4 Fundamental steps in the scientific method

See Model checking for a statement of how the algorithmic model behind the scientific method works. Ancheta Wis 07:29, 30 Jul 2004 (UTC)

See Bayesianism for one philosophical school behind the scientific method. The recent re-characterization of Hawking radiation from a black hole shows clearly that Stephen Hawking and his co-bettors, in a light-hearted way, follow the tenets of this school of thought.Ancheta Wis 07:29, 30 Jul 2004 (UTC)

See also statistical hypothesis testing as one example of a model for hypothesis testing. Ancheta Wis 07:29, 30 Jul 2004 (UTC)

See the Double-slit experiment in quantum mechanics as an example of the wave-particle characterization of light. Ancheta Wis 08:24, 30 Jul 2004 (UTC)

Data for Creativity section

See Otto Loewi's example of his crucial experiment - how it came to him in his sleep Ancheta Wis 00:29, 30 Jul 2004 (UTC)

Content to be merged from Interpretations of the scientific method

The above article was listed on WP:VFD July 20 to July 26 200, consensus was to merge with this one and redirect. Not being a scientist I decided to add it here so that one of the regular editors of this page could merge it. The article the text is to be merged from has been redirect to this article. For the deletion debate see Wikipedia:Votes for deletion/Interpretations of the scientific method. -- Graham ☺ | Talk 23:29, 26 Jul 2004 (UTC)

Text from Interpretations of the scientific method

There are many interpretations of the scientific method. In general terms it is the way scientists investigate the world and produce knowledge about it. Many use the term to refer to an idealized, systematic approach that is supposed to characterize all scientific investigation. It is distinguished from other routes to knowledge by its use of controlled experiments and its requirement that results be reproducible. Many scholars do not believe in the existence of such a method, however, or they do not believe that it accurately describes science. The actual methods of scientists, they argue, are less ideal and more haphazard.

The question of how science operates is not only academic. In the judicial system and in policy debates, for example, a study's deviation from accepted scientific practice is grounds to reject it as "junk science." Whether they are diagnosing a patient, investigating a murder or researching a social trend, non-scientists cite "the scientific method" as a paradigm. Methodical or not, science represents a standard of proficiency and reliability.

A summary of the scientific method

Even if scientific investigation does not always proceed or advance in the same way, there are commonalities, including steps that tend to occur in roughly the same order.

Observe: Observe or read about a phenomenon.
Hypothesize: Wonder about your observations, and invent a hypothesis, a 'guess', which could explain the phenomenon or set of facts that you have observed.
Predict: Use the logical consequences of your hypothesis to predict observations of new phenomena or results of new measurements.
Verify: Perform experiments to test these predictions, to find just which prediction occurred.
Evaluate: Search for other possible explanations of the result until you can show that your guess was indeed the explanation, with confidence.
Publish: Tell others of your results. A reputable journal will have your draft reviewed by other, independent scientists in the field, before publishing the results. This process is known as peer review.
Reproduce: Other scientists will review your published work and attempt to repeat the results. If the results are not reproducible, the original hypothesis is not verified and weaknesses in the reasoning or methods must be sought.

These steps are repeated continually, building a larger set of well-tested hypotheses to explain more and more phenomena. These steps are not necessarily always followed in the pattern shown above. For example, theoretical physicists often develop multiple new hypotheses before selecting which phenomena to observe. See the article on Philosophy of science for more on this.

Do scientists really follow the scientific method?

Past episodes of scientific discovery have contained elements of ingenuity, opportunism and genius. Occasionally scientists have shown outright heroism. It is difficult to tell to what extent these features are necessary to the success of science and to what extent they are accidental and superfluous. The Vienna Circle answered the question by postulating two distinct processes at work during scientific research. The first a context of discovery, in which there are essentially no rules, acts only as input into the second process of justification. The context of justification is characterised by the application of logic and rigour we associate with scientific method.

Debate has raged about the practical utility of such a distinction. Further more, difficulties arise when trying to make the distinction. Whatever the outcome, the debate has provided new ammunition for both methodologists, who prefer to retain the distinction so that their methods have some quarry, and for philosophers of a more historical bent who feel that the success of science is not tied to simple rules.

Most philosophers of science are agreed that there are no definitive guidelines for the production of new hypotheses. The history of science is filled with stories of scientists describing a "flash of inspiration", or a hunch, which then motivated them to look for evidence to support or refute their idea. The anecdote that an apple falling on Isaac Newton's head inspired his theory of gravity is a popular example of this (there is no evidence that the apple fell on his head; all Newton said was that his ideas were inspired "by the fall of an apple.") Kekule's account of the inspiration for his hypothesis of the structure of the benzene-ring (dreaming of snakes biting their own tails) is better attested.

Scientists tend to look for theories that are "elegant" or "beautiful"; in contrast to the usual English use of these terms, scientists have a more specific meaning in mind. "Elegance" (or "beauty") refers to the ability of a theory to neatly explain all known facts as simply as possible, or in a manner consistent with Occam's Razor.

In 1962 Thomas Kuhn published his essay The Structure of Scientific Revolutions, a seminal work on the practice and process of science. Kuhn suggested that sociological mechanisms significantly affect the rejection of older scientific theories and the acceptance of new ones. According to Kuhn, when a scientist encounters an anomaly that is not explained by the scientific community's currently accepted general paradigm or theory, that community can ignore it (the increasing problems with Ptolemaic epicycles in accounting for the motion of the planets was a long standing case), but is often compelled to accommodate it by either modifying the existing theory or replacing it with a new one. A paradigm shift occurs when a new paradigm gains wider acceptance than a pre-existing one. It is at this point that sociological factors may partly influence that abandonment. Kuhn postulates that "normal science" continues on after the adoption of a new paradigm, punctuated with occasional scientific revolutions as later anomalies arise and paradigm shifts occur.

The typical example used in Kuhnian explanations is the development of astronomical theory that began, more or less, with the Aristotelian model of the universe: "The earth is the center of a pristine, perfect universe," and all motions in such a universe must be circular. The Aristotelian model was afflicted with various anomalies, such as the apparent retrograde motion of the planets, which were accommodated by modifications of the model. Nicolaus Copernicus's model differed by placing the sun at the center of planetary motion. Both Kepler and Galileo found evidence that supported the heliocentric model. Aristotle's laws were replaced by Isaac Newton, and eventually by Albert Einstein's General Relativity. This example demonstrates that much time may pass before a substitute paradigm is widely accepted. The Aristotelian model dominated Western thought for more than 2000 years before Newton's viewpoint took its place.

Late 20th century study on the scientific method has focused on quasi-empirical methods, such as peer review, the spread of notations, which are the key common concern of philosophy of science, and the philosophy of mathematics.

History is replete with examples of accurate theories ignored by peers, and inaccurate ones propagated unduly, due to social factors. The establishment of "official" scientific doctrine in the former Soviet Union is a case in point.

Scientists differ on how 'real' their models of reality are - the traditional concern of philosophy of science itself. Some writers involved with deconstructionism adopt a position of extreme skepticism, and argue that no empirical methods can validate any given theory, and therefore all of science must be seen as quasi-empirical. They argue that science is just a social construction; it is only a way that human cultures come to agree on facts, notations, and predictions.

History of the scientific method

The earliest foundations of the scientific method are often credited to Roger Bacon in England and Galileo Galilei in Italy. Later contributions by Francis Bacon, Rene Descartes and others have added to the understanding of scientific method. Some historians of science believe that their developing the collection of practises which comprise the scientific method may have been inspired by preceding tradition, either developed in the Islamic world, or at least conserved by it.

See also

External links

An Introduction to Science: Scientific Thinking and the Scientific Method by Steven D. Schafersman.
[1] Introduction to the Scientific Method
The Myth of the Scientific Method by Dr. Terry Halwes
Rational Reconstruction and Historical Reconstruction, Horus Publications

References

Feyerabend, Paul 1975, Against Method London: Verso. (ISBN 0860916464)
Feyerabend, Paul and Lakatos, Imre 2000. For and Against Method University of Chicago Press. (ISBN 0226467759)
Feyerabend, Paul Atoms and Consciousness, in Common Knowledge Vol. 1, No. 1 1992: 28-32
Karl Popper, The Logic of Scientific Discovery and Conjectures and Refutations.
Thomas Kuhn, The Structure of Scientific Revolutions.

End copied text -- Graham ☺ | Talk 23:29, 26 Jul 2004 (UTC)

Folks, the forked article, above, was forked a year ago, during one of the interminable debates over the scientific method. The fork feels like Groundhog day where the protagonist is forced to relive it until he gets it right. I tried to number each of the 90+ sentences of the forked article (0010 to 09x0), but database hasn't returned my Preview edit yet. My reason for attempting to number the sentences was to note which are in the present article, and which live in the archives. To my knowledge, the Vienna Circle paragraph is the only one that is not in the archives. The other sentences have been re-hashed a least a dozen times by a score of contributors. Ancheta Wis 01:44, 30 Jul 2004 (UTC)

aw, Great observation! WP's Scientific method = Groundhog Day (the movie). Made my day. ww 13:38, 30 Jul 2004 (UTC)

Continuing a thread from the forked article - Some historians of science believe that their developing the collection of practises which comprise the scientific method may have been inspired by preceding tradition, either developed in the Islamic world, or at least conserved by it. - my search has turned up 4 renowned medieval scholars who were physicians / mathematicians clearly, but not yet physicists: Ibn Rushd (Averroes), Ibn Sina (Avicenna), Al-Kindi, Al-Khwarizmi. They were still working on the First Cause, rather than using the 4-step method. Still on the lookout for experimenters, like the one who built the Baghdad Battery Ancheta Wis 07:38, 31 Jul 2004 (UTC)

Found another: Al-Batani improved on the measurements of the earth's precession made 1000 years earlier, by Hipparchus, who himself built on the measurements of researchers in Babylonia and Chaldea, the countrymen of Al-Batani.Ancheta Wis 01:16, 6 Aug 2004 (UTC)

What a mess!

Both the article and this discussion page. No wonder there are nine archives!

And what a contrast to what I learned 30 years ago when I studied science in college. Then I was taught by professors and in numerous textbooks that there is a specific scientific method, one that is objective, able to evaluate whether or not what anyone does is science or not. The summary of that method was fairly well described in the now removed section copied below in "Content to be merged from Interpretations of the scientific method".

As the article now stands, it looks like the proverbial pan of speghetti thrown against the wall to see what sticks.

Maybe this article should be renamed "Philosophy concerning scientific method" and a simple "scientific method" article to appear under science, and not philosophy.

(William M. Connolley 09:12, 6 Aug 2004 (UTC)) Hey, that was my idea. I suppose it disappeared into one of the archives.

Then that would describe the method actually used by scientists and not that theorized by philosophers. Even when I was in college, scientists would look over at the philosophers and snicker at the strange interpretations philosophers would make of scientific theories, the most famous I heard of being how philosophers misrepresented Heisenberg's Uncertainty Principle. Or has the definition of scientific method changed from an objective standard to "science is what scientists say is science", thereby making themselves a priesthood of an arcane study to be accepted on faith by the laity?

I am an anonymous new user, not yet signed in, but as Banno commented below, rot has set in. This article is so bad that I felt constrained to make a comment.

My present work is neither that of a scientist nor philosopher, so I doubt my qualification to do the rewrite.

Fiddlesticks. Anyone can research! Big articles go through many cycles before finally becoming stable. Read up a bit first, so you know what you're talking about (this might take the most time ;-) ). After that , you can check the article to see what can be salvaged, then move the contents to Talk:Scientific Method/older version, put an {{inuse}} up on the page, and have at it! Worst thing that can happen is that you get reverted. ;-)

If that's too much for you at first (I can imagine), you could try to clean up some of the more obviously bad parts first. You might get some ideas on how to fix further from there. One rainy sunday afternoon you might find yourself with some time on your hands, and there you go. Be bold!

Hmm, note that our understanding of science has improved in the past 30 years. I always point people to the rant I wrote to Lord Kenneth at the end of the page at Talk:Scientific skepticism.

The pan of spaghetti thrown against the wall is basically what every open collaboration looks like at midlife. It means that a lot of good things have been added, but lot of bad things haven't boiled out yet. I guess that's the next step for this article.

Remember to have fun! Kim Bruning 17:27, 29 Jul 2004 (UTC)

As for who should do the rewrite, as I understand it, while there is no central authority who can rule from above, apparently there is an ad-hoc committee that decides which articles are to be retained, and which to be deleted. Could not that same committee choose a willing person to do a complete rewrite, submit it to the committee for peer review, then republish it? Or are disagreements so strong in the committee that that is not possible?

I think you just nominated yourself. You can just make the changes you want anytime you like, just click on Edit this page.

I bet you think that sounds crazy, right? Well, the concept is 'stolen' from the design processes that were used to create the internet, and that seems to be working okay I suppose. There's a bunch of hidden plumbing (both organisatorial and technological) that makes sure that you can't make too disasterous a mistake. If you do anything wrong, folks can roll back any mistakes you make all the way back to the point where it first went wrong.

Edit a (couple of) paragraphs, or a section at a time, and then wait a bit to see if it gets accepted.

To get the hang of using the system, it might be a good idea to use Wikipedia:Sandbox to learn to use the editing tools, and maybe edit a couple of smaller articles before cutting your teeth on this big monster :-)

Go ahead, try it, it's addictive! Kim Bruning 17:43, 29 Jul 2004 (UTC)

You are correct that this article is a mess. While the scientific method I was taught in school (elementary and secondary school) bore little resemblance to the state of the art in understanding it (largely Baconinan induction which had been logically exploded long before (eg Hume in mid 1700s)), there is indeed -- in my view not shared by some contributors here -- a scientific method (more or less) which could be described here. It isn't, as you observe. But Kim is right. Pitch in and improve it. This article needs new blood.

However, his observation that every open collaboration looks like this at midlife is not very accurate. This article is hardly at midlife. It's been hacked at and worked at and hashed over, and over and over and over. Surely it's reached an elder status, if only in effort expended, by now?

That it is not better and more adequate than it is is due (crudely) to conflict between two views of science. One is (again crudely) post modern and holds that science is, as all else, a social activity. To the extent that this is not tautological (ie, what we humans do is a social activity since we are social animals), it is, in my view, wrong. That scientists do many things, and say they do many things (the lists being only somewhat overlapping), and confuse us all isn't evidence that there is not actually an underlying method. Only that some scientists are not over reflective. It is that method (largely Popperian/Peirceian) which is objected to by the 'science is a social activity' advocates. The situation is at loggerheads and has been for eons (in WP time anyway). That's why this article is flagged as NPOV. Your input may move it off the dime. ww 17:53, 29 Jul 2004 (UTC)

The notion that there is a single, stateable “method actually used by scientists” is a myth. As Feyerabend noted, scientists are obsessed with their own myths. Observe how IP address fails to level any real criticism of the content of the article, instead using an ad hominem argument against philosophers? Very scientific. Observe that Ww still believes that a messiah will re-write this article, and declare the one true method? Such faith is a central part of the scientific enterprise. Yes, the article is a mess, but it is a mess because it is forced to accommodate scientific mythology. Banno 18:20, Aug 5, 2004 (UTC)

(William M. Connolley 09:12, 6 Aug 2004 (UTC)) The problem with this article is exactly what anon noted: too much philosophical wurble.

Whether one likes it or not, philosophy has had a lot to do with the creation and development of what we call scientific method. -- Chris 10:57, 6 Aug 2004 (UTC)

By which, William, you mean that the story presented here doesn't match the one you expected. Read it again, carefully, and you might learn something new about science. Banno 22:00, Aug 6, 2004 (UTC)

Every few months someone comes along wishing to remove the philosophical material from the article. The fact is that the study of scientific method is a branch of philosophy; that philosophers have done most of the work on the topic. That science professors and teachers do not point to this material, preferring to repeat their own old myths, is sad. Banno 22:14, Aug 6, 2004 (UTC)

You're kidding, right? ;-) I sure did learn it! :-)

Some of those science professors and teachers at my uni were kind of adamant about it. (disclaimer: I've done some ethology, which really needs a bit of scientific philosophy background to eliminate certain kinds of bias) Kim Bruning 07:42, 7 Aug 2004 (UTC)

Vienna Circle : 'context of discovery' vs 'context of justification'

Banno, there was elided material in the forked article that did not survive in the pruning process. The Vienna Circle and Wittgenstein's influence clearly state elements of steps 1 and 2, but your idea of 'characterization' being bigger than 'observation' alone, is a distinction which is not in the Vienna Circle's 2 contexts. If we view them as mental processes, 'characterization' seems to include a ruminative, contemplative side, in which a prepared mind discovers a clue, and runs with it. I know for a fact that this step 1 cannot be scheduled or forced. But once an 'aha' moment occurs, then consequences can follow. Andrew Wiles said that it is like entering a completely dark room, being forced to fumble around, bumping into objects, and then discovering the light switch where everything becomes clear. But it took work to get to the light switch, and for the uninitiated (not the unenlightened :-) ), some explanation of what a light switch is. Ancheta Wis 01:36, 6 Aug 2004 (UTC)

Ancheta, Characterization was a compromise – I still prefer definition. There would be considerable overlap between characterisation and hypothesis development if this material was included, which I think would be less clear. Banno 22:09, Aug 6, 2004 (UTC)

Demarcation section merged and extracted

I have merged the demarcation section with that on the Scientific Method article and extracted it to its own article: Demarcation Problem. I think that there is still some material to merge from The Criterion of Demarcation, but I'll wait to see if there are any complaints so far. -- Chris 08:27, 6 Aug 2004 (UTC)

Since Chris has moved some material to the Demarcation article, which mentions the Vienna circle, I now see that they are given credit somewhere in Wikipedia. Does everyone now agree that they do not need to be mentioned in this article, as they are covered elsewhere? If that is the case, then the forked article has been completely covered, and does not need to be further re-examined. It has been already covered in the past. Agreed? Anyone? Ancheta Wis 08:28, 7 Aug 2004 (UTC)

Agreed. Banno 22:30, Aug 7, 2004 (UTC)

All three demarcation sections are now merged from Pseudoscience, Falsifiability and the Scientific Method. -- Chris

NPOV msg restored

I just noticed it had been deleted. Still should be there.... If nothing else the comments under What a mess immediately above are sufficient. A review of the assorted archives will provide further context for the msg. Those changes to date have changed little in this respect. ww 14:28, 9 Aug 2004 (UTC)

The history shows that the NPOV message has been gone since Revision as of 10:18, 3 Jul 2004. That NPOV message, I believe is a little outdated. I believe that Banno addressed the "post-modern" concern that you brought up some months ago, and that is in the archive. Ancheta Wis 16:14, 9 Aug 2004 (UTC)

What about listing the issues? Posting the banner implies that you will do work on the article, as well. If we remove the banner from the article, and post it to the talk page, then that is a little better, if you do not wish to work on the article yourself.Ancheta Wis 16:14, 9 Aug 2004 (UTC)

The thrust of the argument in ‘what a mess’ appears to be that the content of the article is different to what someone without a name learned 30 years ago. Yes, Ww, time to re-state your case for the NPOV banner. Banno 21:54, Aug 9, 2004 (UTC)

Very quiet in here, isn’t it… Banno 21:38, Aug 14, 2004 (UTC)

Anonymous deletion of annotated list.

I plan to revert the deletion. Does it make sense to convert this to a reference to the list of topics on scientific method, and use this change to stimulate discussion? Ancheta Wis 12:13, 4 Sep 2004 (UTC). I alerted RC patrol, I can ask them to revert the deletions as well?