Talk:Scientific method/Archive 19

From Wikipedia, the free encyclopedia

Archive 15←Archive 17Archive 18Archive 19Archive 20Archive 21→Archive 23

Article section 2

Hello! So my edits from yesterday got reverted, which I guess I’m not entirely surprised by, since it involved deleting some text – but in any case, here is my explanation. I’ve tried to go in order for ease of reading (although that didn't entirely work out since my edits included some rearrangement), and I hope this isn’t too long to read.

Insertion: I decided after writing my justification to undo the reversion, mainly based on the misuse of the word "theory" (see below). Please let me know if this is not appropriate. If somebody reverses this, please leave the changes that involve correcting this particular mistake. Arc de Ciel (talk) 21:52, 6 May 2012 (UTC)

Onto my justifications - starting with the first sentence:

-“In the same way that Alhazen sought truth during his pioneering studies in optics 1000 years ago, arriving at the truth is the goal of a scientific inquiry.”

I removed this sentence. I object mainly to the second clause, “arriving at the truth is the goal of a scientific inquiry.” The goal of a scientific inquiry is to produce models of reality with predictive capability, and the text that I used reflects that. The actual words were adapted from the definition on the page Science.

It is possible that scientific knowledge represents truth (the kind of Truth with a capital T that I would think is evoked when such a charged word is used), but we cannot assume that except in a colloquial sense.

I also brought forward some text from below, about the scientific community and how science produces knowledge. It was hard for me to see how this paragraph fit under the heading “Certainty and myth,” which is where it was. (Of course, I am willing to hear explanations.)


Next (and most importantly): the text uses the word “theory” in a scientifically inappropriate way – specifically, where it should use “hypothesis” or “model.” I apologize for the bolding, but this is incredibly important, and if any part of my edits gets kept, this would be the part.

For example:

-“theories become accepted by a scientific community as evidence for the theory is presented, and as presumptions that are inconsistent with the evidence are falsified.”

A scientific model is not considered a theory until it has been accepted by at least a significant fraction of the scientific community. It is possible that the author of this had string theory in mind, but this is a special case due to (among other things) the fact that it is extremely difficult to test it by experiment.

-“Most scientific theories don't result in large changes in human understanding.”

There aren’t actually that many theories in existence – theory of gravity, theory of evolution, atomic theory, cell theory, etc, and I cannot think of any for which this statement is true. In this case, I replaced the term “scientific theories” with “experimental results,” for which the statement is undeniably true, and which I think is a more important point that people should understand in any case. If the intent of this statement was to say “most putative scientific theories do not end up becoming accepted,” I agree with that statement as well, but that is not what the text says as written. I could also agree with the use of “scientific models” instead of “theories” (though that is more contestable), or that the full strength of theories is often not completely appreciated for some time, but these are also different statements.

-“Theories vary in the extent to which they have been experimentally tested and for how long, and in their acceptance in the scientific community.”

This is trivially true – but again, (assuming we are not talking about theories that were previously held but are now rejected), they have all been tested rigorously and for a long time, and enjoy a very high level of acceptance in the scientific community. If they did not have these properties, they would not be called theories. To be honest, when I read it my first thought was that this is the kind of statement used by an evolution- or global-warming-denier (though of course I AGF, and I have no idea who it was who added the statement anyways).


Continuing on, in the section “Beliefs and biases”: The most important relation of beliefs and biases to the scientific method is that science tries to eliminate them, to be objective in interpreting experimental results, etc. As a result, I moved this to the top of the section. The rest of the section as written essentially comprises examples (namely confirmation bias) which help to make this point, so I made this more explicit.

I also removed some redundant text, like the following:

-“This image illustrates Ludwik Fleck's suggestion that people be cautious lest they observe what is not so; people often observe what they expect to observe. Until shown otherwise; their beliefs affect their observations (and, therefore, any subsequent actions which depend on those observations, in a self-fulfilling prophecy)”

This is a restatement of confirmation bias, which has already been explained. Nevertheless, I incorporated some of this wording (“people observe what they expect to observe”) into my new first paragraph for this section. (I would also point out that “Until shown otherwise” followed by a semicolon is a sentence fragment.)

This next part is not precisely a restatement, but I don’t think it is useful in helping to describe the scientific method:

-“Researchers have often noted that first observations are often somewhat imprecise, whereas the second and third were "adjusted to the facts". Eventually, factors such as openness to experience, self-esteem, time, and comfort can produce a readiness for new perception.”

The first sentence I think is a reasonable to include (and it is in the self-correcting nature of science to expose such things, after all), but I think it’s just a logical example of confirmation bias. I also object to the use of the weasel words “researchers” and “many,” even if they are in the source. The second sentence is extremely vague and seems to basically state that people are sometimes able to overcome confirmation bias.

(That was the last of the sentences that I removed. The rest is rearrangement and clarification.)

(As a side note, I also ended up removing the names of particular authors from my discussion. They’re redundant as long as the statement is sourced normally, and furthermore, the person who promoted a particular idea is not relevant to the merits of the idea itself. I suppose this may be a convention in certain areas of the humanities, but one that I have never understood.


Continuing:

I also felt that the short discussion on myths fit better under “Beliefs,” especially since I was expanding the rest of the section it was in. To correspond to this, I then renamed the section “Scientific certainty.” I then noticed that a) scientific certainty achieved through the scientific method is more central to this topic than the way the scientific method guards against biases and b) the beginning of the “Certainty” section follows well right after the introductory section, so I reversed the order of the two sections.

With regards to the “myths” discussion, I also relocated the unrelated “theories become accepted...” statement, as I already discussed above. I thought that the “a posteriori versus a priori” statement was also unnecessary, but I decided someone might think it is an important point so I moved it with the rest of the paragraph.

I also moved the statement about the appeal to novelty (which was not mentioned by name) – I think this is appropriate even without the section name change, since it is one of the biases that science attempts to guard against.


For the “Certainty” section, I then made a couple of further changes in addition to the ones already discussed:

-I added more information to the first paragraph about how theories are modified over time, thus increasing scientific certainty (as the predictive power of theories increases).

-I then added more detail to the third (now second) paragraph, and clarified the statement “the body of independent, unconnected, scientific observation can diminish,” which I felt could be interpreted as science having the ability to reduce in knowledge over time. Of course, this is not the case, and I assumed that it was referring to the unification of scientific knowledge by more and more powerful theories.

-I then wrote a new paragraph describing how theories unify. This was based around the core idea from the Goldhaber quote, although I didn’t think that the quote itself was useful so I just restated it differently (although I now notice that I didn’t put another citation in, which I think I should have). I also included a couple of sentences describing one of the best-known examples of explanatory power and unification.


Finally, I decided to change the title of the entire section, from “Truth and belief” to “Scientific inquiry.” I’m not attached to this particular change one way or the other, but I felt that it was a better description given the changes that I had made (and possibly even without them). I think that if someone wants to discuss “truth,” it deserves its own section, perhaps under “Philosophy and sociology of science.” Some of the material currently here could fit under that section (perhaps the comparison to myths, for example), but I don’t think the rest of it would be suitable.

I think I may have skipped one or two minor things that I did, so please let me know if you’d like me to discuss those as well. I didn’t think that my edits would be controversial in any way, but I suppose that was not the case. :-) Arc de Ciel (talk) 21:32, 6 May 2012 (UTC)

Arc de Ciel, thank you being bold. The edit pattern we are exemplifying is called 'BRD'
B (Wikipedia:Be_bold)
R (revert -- while adhering to the 3RR rule)
D (discuss= Wikipedia:Editing_policy#Talking_and_editing)
Right now, the article shows two conflicting points of view simultaneously, a contradiction which needs to be cleaned up. My first problem is that the previous editorial consensus was that Scientific Certainty is a myth. Christiaan Huygens was the first to say this. But right now, the headings proclaim otherwise, by seeming to claim Certainty. We should discuss this, and arrive at consensus on the talk page, first. --Ancheta Wis   (talk | contribs) 00:31, 7 May 2012 (UTC)
I agree with Ancheta about certainty, which people usually take to mean "absolute certainty." This article is for the general public.
  Replacing the idea of truth with that of predictive power merely veils the idea of truth which is in the idea of predictive power for example. I.e. predictive statements coming true or turning out false and thereby tending to support or overturn hypotheses, i.e., tending to support that a hypothesis is true or show that it is false. We owe it to the general reader to say things as plainly as we can; the idea that scientists do not aim at truth, and do not think of themselves as aiming at truth, requires empirical support that I don't think will be forthcoming.
  The explanatory hypothesis can't just point to a candidate cause, it needs at least a mini-theory of the causal mechanism, the opportunity, etc. Under what common usage of the word "theory" is it, that the theory of continental drift was not even a theory until it became one by virtue of winning some acceptance in the mid-20th Century? When a "model" or theory dependably generates true predictions rather than false ones, we say that the theory is true, or at least near the truth; it consists of an account, a representation, of things happening which we consider to be more or less accurate, from which account we deduce predictions, not as from an oracle, but as from a picture of real things as they are understood. The Tetrast (talk) 01:25, 7 May 2012 (UTC).

I’m definitely willing to enter a discussion. :-) This is a reply I composed to Ancheta before I saw the next reply – I’ll have time to read and reply to Tetrast a little later.

I agree that science cannot be “certain” in the sense of “proven,” like for logical truths – I would say that is an inevitable result of a process using empirical data, and I do not intend to imply otherwise. I have no objection to changing the subheading “Scientific certainty” to, for example, “Certainty and science” or “Scientific uncertainty”. (When I hear the term “scientific certainty,” I think of “the level of certainty afforded by science” rather than “science can be certain about things.” I would suggest that the statement “no theory can ever be considered completely certain” should be sufficient to clarify any misunderstanding, but I think I understand your concern. Is this the conflict you were referring to?)

Ironically, this is actually one of the reasons that I didn’t like the use of the word “truth” (for example, in the previous opening sentence to this section) which I would say implies certainty. I don’t think many scientists would say that they are “seeking truth” – as I mentioned above, the goal is predictive power. (Scientific modelling) The concept of "truth" is simply not very useful to a scientist, partially because it can be interpreted in so many different ways.

I think one point that I felt was missing was that scientific knowledge can still reach a level where we can be very confident that our predictions are correct (so long as we accept the basic assumptions of science that reality exists, etc). More generally, scientific knowledge has a range of certainty, and is able to approach although not reach a probability of 1. So for example, we can be “certain for all intents and purposes” that if we drop an object, then ceteris paribus it will fall according to the Theory of Gravity, although I would not personally use the word “certain” for that without qualification. Arc de Ciel (talk) 02:26, 7 May 2012 (UTC)

There are several competing factors in play here.
  1. One factor is Aristotle's long shadow. Truth is a goal for most scientists, and this article rests on it. But that does not mean that Truth means only Aristotle's Correspondence theory of truth.
  2. A second factor is the search for Certainty. Imre Lakatos said of himself, '[his] search for certainty has restricted him to the boring problems and has blinded him to the interesting problems' (my paraphrase from Proofs and Refutations).
  3. A third factor is the use of models (in the twentieth century) as opposed to laws (which date back to the Roman Lex, and which lasted for thousands of years, up to the nineteenth century). When a statement is a model, then our attention shifts to the assumptions which that model rests upon, as well put by Goldhaber and also by Brody (both cited in the article). Thus, a model need no longer be categorically true, but only probable (as Huygens put it).
  4. A fourth factor is the problematic use of 'theory', as in ptolemaic theory, which gives reliable, repeatable predictions (a theory with predictive capability, just the sort of thing that Alhazen criticized as not true, but in fact error. Alhazen did not criticize Ptolemy for making errors, since that is our lot as human beings, but that Alhazen still sought the truth, by using logic, geometrical proof, autopsies of cadavers, and other experimentation.
  5. Ludwik Fleck pioneered the idea that scientific facts are socially constructed. The article on truth is leading up to it, step by step, in a pragmatic / constructivist way.
--Ancheta Wis   (talk | contribs) 03:59, 7 May 2012 (UTC)

Apologies again for the length of this response – I’m trying to make sure I get my ideas across coherently. Please let me know if you would like me to summarize.

First, on the subject of “truth” – I have always assumed that this means something which is absolute (that is, unchanging unless the state of nature itself changes), and this seems to be the important distinction here. (Also, if “truth” is not absolute, I don’t see how you could say that the weaker word “certain” is itself absolute.) The other thing is that even if we agree, the general public may not. For example, many religions claim some form of “truth” but this is not related to scientific knowledge. Again, my solution is simply to avoid using the word, which I don’t really think is necessary in a discussion about the scientific method in any case. (This is also the point where I would usually call rationalist taboo on the word if we were going to keep discussing it :-) )

The scientific method guarantees predictive power – it is the major criterion on which scientific theories are evaluated. To say that the predictive power then guarantees (or approximates) truth is an additional step which (if it is included) I think should be clear. Personally, I do not think the second step is necessarily justified (I think you need to assume it). Predictive statements coming out true or false are evidence for or against the hypothesis, not evidence for truth or falsity as you suggest. The more results come out in one direction, the stronger the confidence in the conclusion; however, any conclusion could still be affected by mistakes, uncontrolled variables, a deeper theory that has not yet been discovered, etc.

As an extreme example, if we are living in a simulation then no current scientific knowledge would be true (unless it applied to the higher universe by coincidence), but its predictive power would be intact. Similarly, if general relativity were close to the “truth,” then Newtonian mechanics and the Flat Earth model would look just as wrong if the metric we used was not prediction. If you are willing to define truth as “high but not absolute certainty,” then I would go along, but I am not sure that this is a well-accepted viewpoint or would be taken as such by most readers. I would suggest that this is why scientists tend to talk about “scientific knowledge” or “scientific fact” rather than “scientific truth.”

On continental drift: fair example. Many putative theories (especially those initially proposed by a single person or group) go through a period after being proposed where some refer to it as a theory and others as not, partially because of different people’s varying opinions about the strength of the evidence, and also a tendency to describe one’s work to be as important as possible. I will point out that at least some people still cited continental drift as a hypothesis in the early 1960s – for example, 1 and 2. This should not happen if the definition did not contain at least some component related to the strength of the evidence. Another source which claims that the hypothesis became a theory in the 1960s is here (search the phrase “continental drift”).

I would not object to a description of this being included. I also agree that part of the problem is that the terms themselves (theory, etc) are not well defined.

Now, a response to Ancheta specifically: could you describe what you think the roles of the competing factors are? (how they are in competition, how they are relevant to the article, how my changes may have affected them) - also, are there any specific changes you would propose to make?

I've made a few comments, but I’m not sure what you are trying to communicate overall.

1. I would like to think that the results of my experiments reflect some underlying truth to reality, but I really have no way to know that, and the scientific method cannot tell me. Personally, my main motivation comes from doing things that will save and improve the lives of other people.

3. When we have a scientific model, our attention generally first shifts to the predictions made by that model and how they might be tested. If the model’s predictions are validated by experiment, then we may start building on that knowledge or looking for connections to different models. Since science is iterative, any assumptions that go into a model are generally ones that have already been demonstrated by the underlying theories – the existence of further assumptions may reflect a poor model, unless evidence is specifically given to show that they are reasonable. Secondly, it is not just that a model “need no longer be categorically true” – it cannot be (because science cannot be absolutely certain). Or at least, it may be exactly true but we can never know this fact with absolute certainty.

4. I am not sure that Ptolemy’s theory would be called a “theory” by today’s standards. As far as I am aware, it was not open to possible disproof – it predicted only one thing (planetary orbits) and was constructed in such a way that deviations from prediction could be “fixed” by postulating any number of additional epicycles. The scientific method did not have a chance to act, and when it did (mainly through increasing instrumental precision) his theory was falsified.

5. I’m not sure what “socially constructed” is supposed to mean. If it refers to the “social process of science” and the way that human interactions affect the scientific enterprise, I think that’s uncontroversial, but if it means something about reality (which scientific facts are about) not being objective, then it can be demonstrated false by experiment. Arc de Ciel (talk) 09:44, 7 May 2012 (UTC)

Arc de Ciel, thank you for clarifying your motivation.
It may seem obvious, but it needs to be said: truth is subjective to each observer (this was established first in economics, and has spread to the other sciences in turn). Our individual observations, actions, beliefs, and motivations are generally partial and not absolute. But an organizing theory can be disproven, if not fully proven. So when Alhazen demolished Ptolemy's theory, his own beliefs took several centuries and continents to propagate, until finally most of us are convinced that Ptolemy's theory is false. Alhazen's subject in this case was astronomy (the first science). I agree that it is politically perilous to announce the truth about something. Now we use a synonym: reality. However, reality has some system beneath it; for Fleck, the Wasserman test for blood samples was his reality. Today, of course, this particular test for syphilis is outdated, but Fleck's concept of using a procedure, to validate and to clarify a model, remains. It has taken decades for Fleck's concepts to take root and become accepted, first in the medical laboratory, then to the other sections of the community of scientists. ('social process of science') Then the general public can accept the viewpoint/model as authoritative, until the model falls apart. Thus the cost of maintaining a laboratory model becomes part of the cost of some scientific enterprise.
It is a hallmark of acceptance in a community when the word 'theory' is elided. It's simply the Big Bang, or General Relativity, when the community has accepted it. When a theory is accepted, we acknowledge the researcher who started it by attaching his name to it in our articles, equations, and laws. When an experimentalist disproves some theory, we note the researcher or the institution in our articles, and we conversely attach the name of the scientist, even to a failed theory. In today's world, of course, a belief system intrudes all too often. But today, a consensus of interested parties will generally win out (in a socially constructed process). For the general/consuming public, this suffices until the model falls apart.
A scientific method is most useful when something is unknown. The process of clarifying what is known and what is still unknown, forces us to reexamine our assumptions, which we can then state up front. For example, simple imaginative acts can finesse the need for a costly experiment by replacing it with a thought experiment. Similarly, strong philosophical beliefs can guide research. For example, Gerard 't Hooft's position that a black hole does not yet have a simple quantum mechanical formulation, to me, is very clear and instructive.
So now we are back to the predictive power of an explanation (read 'model' here, if you like). If the model fails, we reexamine its assumptions, costs, etc. If the cost is too great, that science must go dormant until a better idea (or model) emerges. That is where an attachment to truth comes into its own, and the cost of some experiment can even be brought up for funding by the general public, if the idea is appealing enough. What has happened is that a consensus explanation (even if only between two parties, or between two observers -- this is Max Born's idea) is progress.
Finally, this brings us to the problem of rigged experiments, where the outcome is known, before announcement of the prediction. This is where the cost of competition between groups of experimentalists justifies itself. The desire for Certainty, of course, would be the motivation for a rigged system, where a commitment to truth would suffice for a lone researcher. What I am raising, of course, are the ethical issues, such as conflict of interest.
Might I suggest that Scientific Certainty be rewritten in the article.
--Ancheta Wis   (talk | contribs) 12:09, 7 May 2012 (UTC)

I'm not really sure where this discussion is going. :-)

I think it should be clear from my previous posts that I do not consider it obvious that truth is subjective! I also notice that the page on "Truth" agrees with me – it states in the lead (final sentence) that there is at the least some controversy. I will just state that this illustrates again the point that the word is unnecessarily ambiguous.

I would agree with use of the phrasing “understand how reality works” or “understand the mechanisms underlying reality” to describe science (it seems that you might consider that equivalent, even though in my experience scientists would not). However, I will point out that not all science has this as its goal – curing diseases would be a counterexample.

I should also clarify that I don’t necessarily have an objection to most of the other uses of “truth” in the article. There are a couple of other cases where I would also have avoided it, but I am mainly concerned with the section that describes how the scientific method works. (When searching through, I was also somewhat amused by the dichotomy in the Philosophy section that would seem to categorize me as a postmodernist. I am not.) I also don't have any objection to the anecdote that Alhazen considered himself to be pursuing truth, even though I removed it. In fact, on thinking about it I think it's quite an interesting point to make.

Just a few points about science that address some things that you said.

1. It is not possible to disprove a theory in the same way that it is not possible to prove a theory - there is no absolute certainty in either direction.

2. While scientific ideas often take a while to gain acceptance, that does not mean that social forces play any major role in determining the final outcome - they can only speed or delay it. The final outcome is the one that has been shown by experimentation to have the most predictive power; science will not return to Ptolemy's model and decide that it is a better explanation than what we know today. Science moves in one direction only.

3. With regards to elision of the word “theory”: sorry, but that’s also incorrect. In fact, the two statements are not the same thing; dropping the word "theory" refers to one of the specific phenomena postulated and/or explained by the theory, whereas maintaining it refers to the entirety of the explanation and its predictions. For example, the Big Bang was a specific event which is predicted by the Big Bang theory - the theory itself involves lots of mathematics and makes many different predictions, including about what the universe looks like today.

4. If a model fails (in a non-trivial way) and is subsequently rejected, then we generally ignore that model except as a historical note. It is sometimes the case that a model is later revisited, but invariably that is because new data has come to light (without which it would have been unreasonable to accept the model), or because the model itself has been significantly updated.

There are a number of other things in your response that I could explain or otherwise address, but I’m trying to keep it short(er) and (please don’t take this the wrong way) I’m not able to see the relevance for much of it to the changes that I made. For example, I could see a paragraph or two about conflict of interest being included in the article, but it wasn’t included in the first place so I’m not sure why you’ve brought it up.

Continuing on,

-“Might I suggest that Scientific Certainty be rewritten in the article.”

(I assume that you now refer to the text rather than just the title, which we were discussing earlier.)

What part(s) do you feel are inappropriate, and/or what do you feel is missing? I have already tried to explain my justifications above.

The most important points I was trying to incorporate in this section were:

-Science cannot reach absolute certainty (which was already there before my changes, although I expanded on it)
-Nevertheless, scientific knowledge can reach different degrees of certainty, including very high degrees of certainty. (Contingent, of course, on the assumptions of science – reality existing, etc, as described in Science).
-The scientific process is iterative, and moves consistently towards unification and greater explanatory power.

With reference to our discussion thus far:

-I have said that I am not particularly attached to the title that I chose, and suggested a couple of alternatives to avoid the concern that you raised (and which I agree with). Another suggestion might be “Properties of the scientific method.”
-Another suggestion might be to keep the title but to begin the first paragraph with the (current) second sentence instead “No theory can ever be considered completely certain...” – I think this doesn’t flow as well, but I think either way maintains the core point.
-In fact, I’m not particularly attached to the organization either. For example, the first paragraph could be moved upwards to be the third paragraph under “Scientific inquiry,” or merged into the “Introduction to scientific method” section since they both relate to the same topic.
-As I described above – I have suggested the use of “an understanding of reality” or something similar to that in place of “truth," which seems to me to be a reasonable compromise.

I look forward to your reply. Arc de Ciel (talk) 06:08, 8 May 2012 (UTC)

I find Arc de Ciel's comments on truth quite unconvincing, for example:

Predictive statements coming out true or false are evidence for or against the hypothesis, not evidence for truth or falsity as you suggest.

Are you sure that you mean that? If evidence for or against a hypothesis is not evidence for or against the hypothesis's truth, then what other property or relation of the hypothesis is the evidence for or against? What is this non-truth favorability that a hypothesis has in light of evidence? Is it power, wealth, glamour, status? Is it beauty? But a beautiful or architecturally wonderful theory is still rejected if found false. You accept the idea of reality but not the idea of truth. Yet truth and reality are paired; hypotheses, theories, predictions, declarative statements, etc., aim to be true about the real. The idea of truth is not an unnecessary "second step," it's something that various things have in common and clarifies their kinship, so we use it to speak plainly to the general reader rather than engaging in bureaucratic euphemism and elision. One abstracts a common idea of what predictions, hypotheses, theories aim to be: and that is the idea of truth, not some other, or more general kind, of favorability or merit minimalistically alluded to by use of a fleeting preposition like "for" as in "evidence for"; nor should the forest which is truth be lost sight of amid the trees that are species of truth, even such key ones as predictive power; in a way truth is predictive power about potential investigatory results; a true statement truly predicts a common element of further truths that one will find, for example about the past, if one investigates far enough; and the meaning of each of those further truths is in its imaginable practical implications; yet truth is also correspondence; if you compare an accurate map to the mapped territory, you can see how the map and the territory have the same implications, lead to the same predictions; and one explains an account's or explanation's leading only to successful predictions as evidencing the account's or explanation's correspondence to that portion of the real which is its subject matter, for example the past. We need a forest word and that word is "truth". The idea that some religions might object to use of the word "truth" in regard to science goes, mutatis mutandis, for "reality", "certainty", and an awful lot else here; it's a red herring. It's also remarkable to say that "certainty" is a weaker word than "truth"; to assert that p is certain is stronger than to assert that p is true; it implies a stronger commitment by the speaker; it evokes the idea of doubts that have been overcome, or a probability that has been shown to be 100%. The Tetrast (talk) 00:34, 9 May 2012 (UTC).
I want to add a few things. Some times "truth" is used in a technical sense, as in formal logic, but usually in science it is not a term of art; even those who say they don't find it useful use it a lot. It's a common-sense word, with some of the vagueness natural to common sense - the true, the nearly true, the true enough, etc. Science is often said to be a development of common sense; it is more critical, and it increases in precision, but it starts from applying common sense. Phrasing away science's aim at truth obscures science's roots in the application of common sense and would lead this article down the path that too many Wikipedia articles have taken, written not as by knowledgeable editors for the general reader, but as if by (sometimes very) knowledgeable students for their professor, to let their professor know that they've absorbed the material, letting technical language eclipse the shared basis in common-sense understandings that makes articles more accessible to the general reader. There needs to be a balance - not between truth and falsehood, but between everyday English and technical, walking-on-glass language about science's aims that suits this or that school of thought or technical language that surrenders such words as "truth" in order to sidestep those schools' battles; one could surrender many important but nontechnical words that way.
And there's a bigger forest of which science is only a part, where the idea of truth-seeking helps distinguish science (and maths too, actually). This is another case of not missing the forest for the trees. When one says that science seeks to learn truths - seeks true representations that are bases for further true representations - one means that it is not primarily a game for seeking power, wealth, glamour, or status, and its method of settling questions is not the method that gives authority to power, wealth, etc. for settling questions; nor is it a ruling art, for truly knowing from what strengths one decides real things, though it's applicable there and has a planning dimension, nor is it a productive art, seeking to know truly by what means one handles and accomplishes real things, though it can be applied for such, as in medicine, and itself has a strong know-how dimension, nor is it an affective art, seeking to truly grasp in what (natural or artistic) effects or perfections one feels real things (subtle, complex etc.), though it is applicable there and has its own aesthetic dimension. The redoubled emphasis on truth means, that science seeks to truly know on what bases, what solid grounds of truth, one knows real things. The Tetrast (talk) 02:51, 9 May 2012 (UTC).

It is evidence for or against the ability of the hypothesis to make predictive statements in the future. Science is self-contained and recursive – it has no need for concepts not related to what we directly observe.

You appear to be trying to tell me what truth is. I do not “reject the idea of it,” although I might disagree with the sense (or parts thereof) in which you are using it. As I pointed out above, it seems quite clear that not everybody agrees with you (not even all philosophers). For example: see the lead of Truth. If you have a definition which can be easily determined by observation, universally agreed upon (especially by scientists), and helps in the formation or refinement of hypotheses, science may have a use for it. The same goes for “certainty,” which does have a generally accepted definition in science (it is very similar to “probability”), but which does not seem to be the one that you are using. An alternative suggestion would be to try and communicate your ideas without using any metaphors. Science does not speak in metaphor, except sometimes when it is educating others.

(I will also point out that you are using some very long sentences. For example, almost half of your first reply is a single sentence, from “one abstracts a common idea” to “its subject matter, for example the past” - and you also don't seem to be using paragraphs. I just want to suggest that this might be easier if you edited a bit more before submitting. :-) )

In any case: as I asked Ancheta above, what specific changes do you propose to make? Also, what do you think of the edits I have suggested (summarized at the end of my last reply)? Arc de Ciel (talk) 04:16, 9 May 2012 (UTC)

Update: I noticed that my suggestion about the Alhazen anecdote is not included in my summary above, but it should also be there. Arc de Ciel (talk) 04:37, 9 May 2012 (UTC)

There has been a debate about truth in science between the realists and the antirealists. From the Glossary of Curd's and Cover's Philosophy of Science:

Scientific realism Scientific realism has several dimensions: metaphysical, epistemological, and methodological. While there is no single, monolithic version of scientific realism that all scientific realists accept, scientific realism is generally taken to be the doctrine that the world studied by science exists and has the properties it does independently of our beliefs, perceptions, and theorizing; that the aim of science is to describe and explain that world, including those many aspects of it that are not directly observable; that, other things being equal, scientific theories are to be interpreted literally; that to accept a theory is to believe that what it says about the world is true, and that by continually replacing current scientific theories with better ones, science makes objective progress and its theories get closer to the truth.
Antirealism a diverse group of doctrines whose common element is their rejection of realism. In the philosophy of science, antirealism includes instrumentalism, conventionalism, logical positivism, logical empiricism, and Bas van Fraassen's constructive empiricism. Some antirealists (such as instrumentalists) deny that scientific theories that postulate unobservables should be interpreted realistically. Others (such as van Fraassen) concede that such theories should be interpreted realistically, but they deny that we should ever accept as true any theoretical claims about unobservables. (See constructive empiricism, conventionalism, instrumentalism.)

Another issue: the Duhem-Quine thesis says that a test of a theory may fail, not because the theory is false, but because one of the auxiliary hypotheses used in the test is false, e.g., there are seven planets in our solar system (testers of Newton's Law of Gravitation in the 1840's were unaware of the existence of Neptune). Ivar Y (talk) 08:42, 9 May 2012 (UTC)

Further edits

I know it’s only been a few days, but I’ve decided to implement some changes to try and address your replies. I really want to get on to fixing the problems in the rest of the article – for example, the introductory section, and indeed the whole article, doesn’t once mention the null hypothesis (!). I also see that the section “Pragmatic model” is very long in this article, yet the same heading has no content on the stated “main article” on the topic, and so forth.

Of course, if you find time to make suggestions, I will still be happy to discuss and/or explain. (Also, I don’t intend to come across as acerbic or anything like that, if I have – I’m just trying to improve the article to make it reflect how science actually works. This article is strongly related to the future of science education, after all!)

Anyways, I changed the heading “Scientific certainty” to “Properties of scientific inquiry” – which is the best title I’ve thought of so far, even though I still think it could be better. I then added a new sentence (second sentence of “Scientific inquiry”): “This allows scientists to gain an understanding of reality, and later use that understanding to intervene in its causal mechanisms (such as to cure disease).” I didn’t reinstate the Alhazen anecdote though, since I'm thinking it might instead fit best on the “Philosophy of science” page.

A couple of notes to Ivar Y:

-I identify mostly with the first of the groups you pointed out, mainly because I make the assumptions of science (reality exists, etc). Of course, they are both philosophical positions rather than being descriptive of science itself, which I think could be considered as part of what I was trying to explain above.
-Thanks for the edit to my statement about certainty – I didn’t notice that I had used the rhetorical phrasing “can always.” :-)
-With regards to “crucial experiment” – it’s a reasonably important concept in science, but I agree about its relevance (unless perhaps it could be incorporated somewhere in the “Elements of scientific method” section). I added a bit to clarify the relation to Bayesian analysis, although I think that this could probably be better placed elsewhere in the article as well.

Arc de Ciel (talk) 07:37, 12 May 2012 (UTC)

Arc de Ciel and Ivar Y, thank you for your development of this article. --Ancheta Wis   (talk | contribs) 11:19, 12 May 2012 (UTC)
Rather than pruning the pragmatic model section overly much, another approach might be to add more sections and content; for example I was planning to add more material on Francisco Sanches' skepticism That nothing is known. A physician, he had a biological viewpoint but was able to criticize physics (magnetism), which was still under the grip of Aristotle. --Ancheta Wis   (talk | contribs) 14:28, 12 May 2012 (UTC)
I wasn't suggesting "pruning" as such, only observing that the "main article" (Models of scientific inquiry) has no content under the same heading - so perhaps half or two-thirds of it could be moved from here to the main article. The section does seem quite long, especially for a model that is not the most recent one in use today. My other thought was that the key components of "abduction, deduction, induction" are quite deep within the section and could benefit from being more prominent. Arc de Ciel (talk) 17:49, 12 May 2012 (UTC)

I’ve now updated the Introduction, so please let me know what you think. :-) I’m not sure that it’s in the best format (for example, I have a mild preference for avoiding numbered lists, but I wouldn’t object to someone numbering them again; etc), and I’m quite sure improvements can still be made, but I’m quite sure that it’s better than the previous version. It’s at least much closer to the process that is practiced today.

Again, there’s been very little information actually deleted – in fact, the only thing I completely removed is the statement about Born (in the source, he is not talking about the scientific method itself, but the fact that you require induction in order to postulate a scientific law. “No observation or experiment...can give more than a finite number of repetitions.”) The Fleck statement is incorporated into my new “Other components” section, where I thought the point it made was stronger, although I didn’t maintain the wording (the original sentence used the word “experience(s)” three times). I also moved the original description of the hypothetico-deductive model to the main page and integrated it there, as well as some of the content below it (such as the paragraph containing the Bayesian analysis-probability section that I wrote yesterday). Almost all of the information from this (the main exception being the sentences about probability), is incorporated into the new text.

I didn’t change the DNA example, since it still reasonably correlates with what I’ve written. (It does need it though – for example, the relevant experiment was not “seeing photo 51” but the crystallization of DNA and subsequent X-ray diffraction; similarly, there isn’t any mention of the generation of the structural model; etc). I also see a few small overlaps with things that are said later, so there will need to be some adaptations to account for that as well.

It would be great if anyone could please modify instead of reverting (especially if it’s only a sentence or two!) or suggest changes for discussion. :-) Of course, I can specify what I felt was important for any particular statement, although as before I’m not especially attached to any particular wording or organization as long as it maintains the main points. Arc de Ciel (talk) 08:51, 13 May 2012 (UTC)

There could be a concrete example for avoiding confirmation bias: Ian Shelton was initially skeptical that supernova 1987a was real, but possibly an artifact of instrumentation (null hypothesis) so he went outside and disproved his null hypothesis by observing sn 1987a with the naked eye. This example was provided by Leon Lederman. --Ancheta Wis   (talk | contribs) 02:53, 14 May 2012 (UTC)
Good point, and I think SN 1987A is a great example for the null hypothesis (observing something so spectacular is extremely rare, far less common than instrumentation artefacts, so you would want to make a separate observation before being confident in your results). Were you suggesting adding it into the introduction or further down the page? Either sounds reasonable to me.
I'm not sure it's the best example for confirmation bias though, since Shelton didn't actually fall for it. Perhaps one of the examples discussed here might be better - the first two, N-rays and Jacques Benveniste, are the most well-known. I would lean towards N-rays since in that case many labs reported replication of the nonexistent results. (The second is also somewhat emotionally charged today since it has to do with homeopathy.) Arc de Ciel (talk) 03:06, 15 May 2012 (UTC)
Update: I see that you added it into the footnote. That works also. :-) Arc de Ciel (talk) 03:42, 15 May 2012 (UTC)

Null Hypothesis

Your (Arc de Ciel's) "Edit summary" comment that "all hypotheses have a corresponding null hypothesis" is arguable. A more common view is that null hypotheses apply only to statistical testing. Some definitions showing this:

From the NIH site that you referenced:

Null hypothesis: in hypothesis testing, the hypothesis that an intervention has no effect, i.e., that there is no true difference in outcomes between a treatment group and a control group. Typically, if statistical tests indicate that the P value is at or above the specified a-level (e.g., 0.01 or 0.05), then any observed treatment effect is not statistically significant, and the null hypothesis cannot be rejected. If the P value is less than the specified a-level, then the treatment effect is statistically significant, and the null hypothesis is rejected. If a confidence interval (e.g., of 95% or 99%) includes zero treatment effect, then the null hypothesis cannot be rejected.

From the Wolfram site that you referenced:

A null hypothesis is a statistical hypothesis that is tested for possible rejection under the assumption that it is true (usually that observations are the result of chance). The concept was introduced by R. A. Fisher.

From the on-line Oxford English Dictionary:

null hypothesis n. Statistics a hypothesis that is the subject of a significance test, esp. the hypothesis that there is no actual difference between specified populations (any apparent difference being due to sampling or experimental error).

From the Merriam-Webster dictionary site:

Definition of NULL HYPOTHESIS: a statistical hypothesis to be tested and accepted or rejected in favor of an alternative; specifically : the hypothesis that an observed difference (as between the means of two samples) is due to chance alone and not due to a systematic cause

If all hypotheses have null hypotheses, what is the null hypothesis corresponding to Newton's Law of Gravitation? to Maxwell's equations? Ivar Y (talk) 17:57, 28 May 2012 (UTC)

Arc de Ciel, Your viewpoint is understandable and clear. However, there is another sense in which statistical hypothesis testing is arguable: from the viewpoint of mathematics. The article is clear that in scientific method, it is essential that the researcher understand what is known, and what is not known at the present time. This is true in both science and mathematics. In statistical work, there is an inherent subjective uncertainty (also denoted error), which is not part of the ontology of mathematics. One famous example of the divide between statistics and mathematics was the philosophical difference between Jerzy Neyman and Alfred Tarski, representatives of statistical hypothesis testing and categorical logic modelling, respectively. They despised each other's viewpoints, but came together on the matter of Julia Robinson upon her nomination to the National Academy of Sciences.
All I am saying is that the article ought to delineate the points of view being espoused in the respective sections. Surely there is room for both a statistical point of view and a mathematical point of view in the article. --Ancheta Wis   (talk | contribs) 18:56, 28 May 2012 (UTC)
Please call me Arc. :-)
The definition I used, “a general or default position,” was the one at the top of the the Null hypothesis page, and it is also the sense in which I use it myself (this is also the page where I got the citations from). The null hypothesis for an equation like the Law of Gravitation would be simply that the equation in question is incorrect. One way to look at this is that the likelihood of any arbitrary equation to describe reality (before any evidence is gathered) is extremely low, and as a result this is the most probable outcome a priori. (Although, to be precise: there can be even more proposals for what the null hypothesis is. For example, if there is a previous explanation of the same results with sufficient evidence behind it, as was the case when relativity superseded Newtonian mechanics, you could say that the previous explanation is the null hypothesis instead.)
That is, in non-population experiments, there is still a default position, which we do tend to call the null hypothesis. However, the definitions you gave are still correct in the sense that all experiments are statistical in nature - they all have a confidence value attached to them, otherwise you could achieve complete certainty in science. (I think this is what Ancheta is referring to – please correct me if I’m wrong.) The problem is that I would call it somewhat misleading to say it is only relevant to “statistical hypotheses,” a term that tends to be interpreted much more narrowly.
As an example, a single experiment using a similar triangle calculation is more than sufficient to test the hypothesis that the height of a particular tall tree is over 20 meters (null hypothesis: that it is not). Unless the tree is very close to 20, the experiment will give you far more than 95% confidence without any statistical calculations. In research, I would give the example of certain SDS-PAGE tests in which the result is binary (yes-no), subject only to possible mistakes in the experimental protocol or the interpretation of the evidence. In general, you would still do replications and use multiple sources of evidence (equivalent to attempting to falsify certain auxiliary hypotheses that would cause problems if incorrect), but this does not necessarily involve direct application of statistics.
All that being said ‒ I see based on your sources, and a few others, that most authors use the narrower definition. I suppose the article should probably reflect that, so feel free to change it back. (I would prefer to avoid the terms “so-called” and “assumes” though – I can explain that further if you would like. Also, if you remove my citations again, please replace them with ones you think are better!) Arc de Ciel (talk) 00:58, 29 May 2012 (UTC)

Method or methodology

In many places throughout the article 'method' is used synonymously with 'methodology'. Although this is also the case in common usage, I think we can afford to be more precise in an article about method. I suggest we use the word 'methodology' only when we are speaking about the study of method. I'll wait for comments before doing a search-and-replace. --ChrisSteinbach (talk) 22:10, 16 October 2012 (UTC)

I've now replaced 'methodology' with 'method' in those places where the meaning is unambiguous. --ChrisSteinbach (talk) 00:48, 20 October 2012 (UTC)

Descartes "framework"?

This taken from the history section:

Then, in 1637, René Descartes established the framework for scientific method's guiding principles in his treatise, Discourse on Method.

What framework did Descartes establish? To my mind the only thing that can be salvaged from his method, from the perspective of modern scientific method, is a mistrust of the senses. There are quotes on the history of scientific method page that show his thinking was clearly at odds with the conception of scientific method outlined here. --ChrisSteinbach (talk) 01:41, 20 October 2012 (UTC)

I'm not sure, but some searching shows that that one source for it could be this. Perhaps instead of using the vague word "framework" we could describe exactly which contributions he was supposed to have made (after cross-checking it with other sources, of course). Arc de Ciel (talk) 02:57, 20 October 2012 (UTC)

Boldface "the"

Someone recently changed the boldface to include "the" in "the scientific method." I thought that boldface was reserved for the article title and its synonyms (e.g. you would bold it in "The Beatles" because it's in the article title), but I could be mistaken and I can't find anything specifically about this in the MOS - the pages I found were MOS:BOLDTITLE and MOS:BOLD. Does anyone know of a precedent? Arc de Ciel (talk) 03:03, 20 October 2012 (UTC)

Updating Francis Bacon reference

Changing the following quote,

Bacon, Francis Novum Organum (The New Organon), 1620. Bacon's work described many of the accepted principles, underscoring the importance of theory, empirical results, data gathering, experiment, and independent corroboration.

to read,

Bacon, Francis Novum Organum (The New Organon), 1620. Bacon's work described many of the accepted principles, underscoring the importance of empirical results, data gathering and experiment.

Theory, or rather theorising, plays no role in Bacon's method; hypotheses are supposed to emerge from the process he outlines. Quite likely this is the root of Newton's Hypotheses non fingo. I removed "independent corroboration" from the list of principles handed down from Bacon. Certainly he wanted scientists to join him in using his method, but he was mainly concerned with adding to the body of scientific knowledge and not corroborating what was already established. --ChrisSteinbach (talk) 01:41, 20 October 2012 (UTC)

Chris, with your permission, I propose to augment the text in the reference as follows (from 1911 Encyclopedia Britannica, "Bacon, Francis"): [In Novum Organum, we] "proceed to apply what is perhaps the most valuable part of the Baconian method, the process of exclusion or rejection. This elimination of the non-essential, ..., is the most important of Bacon's contributions to the logic of induction, and that in which, as he repeatedly says, his method differs from all previous philosophies."
--Ancheta Wis   (talk | contribs) 02:54, 20 October 2012 (UTC)
Looks good to me: go ahead! --ChrisSteinbach (talk) 07:13, 20 October 2012 (UTC)

Propose replacing the GR example of a 'specific hypothesis'

Since, in scientific method, the hypothesis stage is best used for knowledge whose truth is unknown, I propose to replace the GR example with a hypothesis which is not yet corroborated. My citation for this proposal is to a review article, concerning a family of well-known hypotheses. One of them was proposed by Francis Crick and Christof Koch, among others, about the role of the claustrum in the function of the mammalian brain. (Crick came to this study after his prior success in the DNA story.)

There are other possibilities, for example a conjecture about the role of the arcuate fasciculus (AF) in the interaction between Broca's area and Wernicke's area for the human brain. My citations for this possible example are a textbook from the 1980s which cites Norman Geschwind's articles, and the newer reactions to this conjecture. (I see that Lateralization of brain function has a dubious tag on its AF statement which can be cleaned up in the process.)

I am hoping to capture the sense of danger (i.e., risk) in mooting a hypothesis, which a well-known example like GR does not engender in the article.

Do any editors object? --Ancheta Wis   (talk | contribs) 19:59, 6 November 2012 (UTC)

Since this is a broad overview article, I would say that an example being well-known is a point in its favor, and I don't think that's exclusive with the idea of risk. :-) But practically every subject still has open hypotheses - I would first want to consider how illustrative an example is, how easy it is to explain, how easy it is to explain the significance of the question, number of available references, etc.
For example, why not abiogenesis instead? The question is very easy to describe, a summary of the current hypotheses and research approaches should be quite easy, there is a lot of literature about it, it is still an open question, etc. This is the first thing that comes to mind, but of course I'm sure there are others. Also, I think that a well-chosen closed question that used to be open could also achieve your goal, provided that the time frame during which the question was open is adequately described, and there would likely be many more sources available. (I don't think the current GR version is good in that regard - although we don't have to have only a single example, so perhaps both sections could be included.) Arc de Ciel (talk) 07:03, 8 November 2012 (UTC)
Arc, I like your suggestion. Upon reading abiogenesis I had some immediate questions for further possible further inquiry. This could be inspirational or even aspirational for lay readers.
1. Might water be a prerequisite for life to arise? This would save billions or even trillions in the future research budgets.
2. If water ( or other similar earth-based constraint ) were not necessary for life to arise, we would then arrive at Fermi's question "where is everyone else?"
3. Are we alone?
4. If we are alone then why are we hell-bent on destroying ourselves?
Again, it's a fertile question. --Ancheta Wis   (talk | contribs) 10:26, 8 November 2012 (UTC)
I would actually say that 2/3/4 are aspects of the same question - the Fermi paradox, which you referenced - for example, the self-destruction of intelligent species is one of the proposed answers. :-) For the first question, there is some discussion at Alternative biochemistry#Non-water solvents which could be used. Analogous questions discussed on that page include the possibilities of non-carbon-based life (e.g. silicon-based) or other forms of DNA (e.g. the arsenic-containing DNA that was in the news last year but which hasn't been corroborated). I think expanding a discussion on abiogenesis to include those could be very informative for readers. I probably wouldn't be able to get involved in much writing (I'm currently trying to extricate myself from another topic area right now as well!) but I'm happy to discuss. Arc de Ciel (talk) 12:01, 9 November 2012 (UTC)
Arc, upon reading Christian De Duve's Vital Dust, I realized that the central dogma of molecular biology actually clarified some of the questions about the primordial soup, and that the central dogma is a good example. The holes in the central dogma have always bothered me, so the holes serve additionally as an reminder that the central dogma is still an open question. --Ancheta Wis   (talk | contribs) 08:54, 11 November 2012 (UTC)
Hi! I haven't read the book you refer to, but the central dogma (DNA -> RNA -> protein) was almost certainly not present in the earliest life (for example, see RNA world hypothesis). Also, the original hypothesis actually isn't an open question in biology any more (all known life depends on the process), except insofar as there is some doubt about how prevalent the important noncanonical processes are.
Anyways, I included both versions, since I think it's important to have a quantitative example in this initial section (I had misinterpreted you as referring to a later part of the article). Also, since it was Crick who first came up with the idea, I changed the attribution and described Nirenberg's role in the citation, then shortened the quote to the more concise phrasing. Arc de Ciel (talk) 21:22, 11 November 2012 (UTC)
Arc, it's important to separate hypothesis from prediction. In the GR example, Einstein's calculation of the precession of the perihelion of Mercury was not a prediction, since the precession had been known since 1859. Rather his calculation was a confidence builder for him, a corroboration that his theory of GR was right (Newton's corroborating calculation in classical mechanics was the acceleration of the apple -- known since the time of Galileo -- and the corresponding acceleration of the moon under the influence of gravitation "and I found them to answer, pretty nearly"). If we are talking GR, Einstein's hypothesis was his equivalence principle: the equivalence of inertial mass and gravitational mass. I propose replacing the perihelion statement in the overview with the equivalence principle, instead. For the prediction stage of scientific method, a corresponding prediction for GR would have been the prediction of the measurements at Príncipe,which was a Royal Astronomical Society expedition planned in 1917 in anticipation of the May 29, 1919 eclipse (this is already in the article). --Ancheta Wis   (talk | contribs) 01:37, 12 November 2012 (UTC)
As a tangential comment, hypothesis and prediction are generally not that distinct, since the two of them (at least ideally) are logically equivalent (also, you're right that I should have been more careful about my use of the word "prediction" - there is a colloquial use which can refer to anything derived from a hypothesis). I agree that it is better to use the equivalence principle, so I've added it. Arc de Ciel (talk) 04:35, 12 November 2012 (UTC)
If the objective is to make this article accessible to as many people as possible, then a historical scientific discovery limits readership to those familiar with that discovery and the scientific context. What about an every day example that illustrates the process rather than a historically significant discovery?
For example, on entering a room and switching on the light, we observe there is no light. The hypotheses might be: the bulb is broken, the fuse has tripped, there is a powercut. Experiments might be: try other lights to see if there is power, replace the bulb, check the fuse, etc.
Fixing a broken light has many elements of scientific method and is sufficiently concrete for most readers to understand. pgr94 (talk) 10:04, 12 November 2012 (UTC)

Positivism and the scientific method

In order to define the area of use of the scientific method as described within this article in science, I tried adding some not necessarily perfect text about its relation to positivism, social sciences, qualitative research and the positivism dispute. I'm used to such edits being flatly reverted by mathematicians and physicists with no regard for other academic disciplines than their own, so I'll add this little notice in case someone is interested in taking a look at what I added in the article history. Narssarssuaq (talk) 12:18, 17 November 2012 (UTC)

You are again adding more original research. Look at WP:UNDUE and WP:OR, WP:SYN. "Antipositivism (also known as interpretivism or interpretive sociology) is the view in social science that the social realm may not be subject to the same methods of investigation as the natural world", i.e not science. IRWolfie- (talk) 13:02, 17 November 2012 (UTC)
You need to reference your claim that social sciences (soft science) not based on positivism are counted outside the realm of science. If true, it is of vital importance to include this information in the article, as the very terms social sciences and soft science indicate otherwise. Narssarssuaq (talk) 13:46, 17 November 2012 (UTC)
The editor adding unsourced claims is the one who has to provide the sources. Also, I don't think the content of your edits is suitable for the lead even if it were sourced. Arc de Ciel (talk) 21:24, 17 November 2012 (UTC)
Narssarssuaq, you're failing to WP:AGF, but that's not why editors are flatly reverting contributions like this, the WP:LEAD should summarize what's in the body of the article... You've tried to frame the discussion of (21st century) scientific method against the unexamined hubris of 19th century positivist sociology... and following a brief declamation of the article's non-applicability to antipositivist "branches of science", i.e. the social sciences, because of a methodological dispute in 20th century Germany, you imply it somehow exposes a lack of concern for the ubiquitous methodology of hard science: not only in philosophy, ethics and theology, but all the traditionally non-scientific academic studies. See WP:NPOV/WP:OR and shoot for not necessarily a problem.—Machine Elf 1735 04:07, 18 November 2012 (UTC)

Discussion to restore pseudoscience and antiscience as part of definition in Alternative medicine article, using sources Annals of New York Academy of Sciences, etc.

A discussion involving retoring content from sources describing alternative medicine as being based on pseudoscience, antiscience, tradition, and bad science, including the first 14 sources of this version, such as Journal of the Association of Medical Colleges, Annals of New York Academy of Sciences, Academic Medicine, Canadian Medical Association Journal, Medical Journal of Australia, Nature Medicine, etc., to the Alternative medicine article is now going on here. ParkSehJik (talk) 02:57, 22 November 2012 (UTC)

No mention of Burden of proof/burden of evidence

The disambiguation page Burden of proof has an entry for Scientific burden of evidence which redirects to Scientific method but the word burden does not appear in this article, nor the main Science article. I DID find the word burden in Talk:Scientific method/Archive 1 but only as part of a discussion of some other aspect of the article. I've read and heard many times in relation to science and the scientific method that "the burden of evidence is on the positive" (meaning that one needs to show evidence that something exists, rather than just claiming something and asking others to show evidence that it does NOT exist) - shouldn't this phrase be used in the article?

Also I've seen a similar problem in at least one other article, some word or phrase forwards to another related article in which there's absolutely no mention of the original word or phrase. I'd think the forwarded-to article should actually say something about the word/phrase. Is there any meta-policy related to this? Benbradley (talk) 04:25, 22 November 2012 (UTC)

The article is clear that certainty is not definite in science or in scientific method. Why is it necessary to assign burden in that case? Who wants to know? Why do they want to know? How much do they want to invest or expend in order that they might know? What is the benefit of knowing? Exactly who would be benefitting if the result were known? --Ancheta Wis   (talk | contribs) 09:50, 22 November 2012 (UTC)
Ben, if you've read and heard many times in relation to science and the scientific method that "the burden of evidence is on the positive", then you should be able to find sources for discussion. Please take care to find reliable sources which fully meet the verifiability policy. Thanks, dave souza, talk 11:06, 22 November 2012 (UTC)

Discussion of "so called scientific method" at WP:MEDRS talk

Following discussion re MEDRS, ontologic status of psychiatric categories, and controversy re the scientific methodologies for attaching the term "disease", "disorder", and :lifetime" to the categories (if they really exist, e.g., Penis envy), in the psychiatry, forensic psychiatry, Bipolar disorder, and related article talk pages, FiachraByrne correctly wrote (bolfaced added by me for emphasis of most relevant part, and whose comment I may have distorted by excerpting just a part of it in order to raise the following issue) -

"Psychiatry is one of the oldest medical specialisms. It's designation as medical practice is a disciplinary/professional attribute that has little to do with the actual content of psychiatric knowledge or the nature of psychiatric practice. To establish this it is unnecessary to evaluate whether in any or all instances psychiatry adheres to the so-called 'scientific method'."

However, the designation of psychatry always being medicine, and not just some parts of it, with the associated implications of established efficacy in healing real diseases, at Wikipedia, is a WP:MEDRS issue, not just a matter of determining the common usage on the street. The part of FiachraByrne's comment quoted above raises issues being glossed over by other editors at those multiple talk pages, where it is declared to be "common knowledge" that psychiatry is for the most part evidence and science based, that its designated categories (eg., penis envy and bipolar disorder) are real, that the DSM designation of their being "disorders" estabishes with MEDRS that they are, and that they are lifetime, and questioning this violated WP:COMMONSENSE, and is WP:BATTLE because it is unquestionable, even with MEDRS and RS saying otherwise, all because Diagnostic and Statistical Manual of Mental Disorders (DSM), the self-proclaimed "bible" for practitioners, is always unque3stionably MEDRS. Furthermore, RS and MEDRS content is being totally deleted from any WP:MOS (lede) "controversy" paragraph as being UNDUE, by simply citing the declarations in DSM, even when contradicted by other MEDRS sources.

The same WP:MEDRS standards should be applied to psychiatry as to alternative medicine articles. Traditional Chinese Medicine (TCM) is also one of the oldest "medical" practices. There is rigorous enforcement by WP:MEDRS hawks (of which I am one) that assertions re TCM being healing "medicine", as defined in that article and by MEDRS standards. The only allowable edits are that TCM practitioners "claim" to heal. TCM uses supernatural etiological objects ("qi" flow blockage causing qi, not the heart, to propel the blood inadequately), and outright false statements about anatomies, developed without the "cutting" of the "tom" in "anatomy" (Greek "tom" means "cut", as in "a-tom" – meaning not further able to be cut, as atoms were thought to be), has also historically been designated "medicine". MEDRS has different standards than accepted common usage, and for good reasons well argued in setting up the policy.

  • Should the psychiatry, forensic psychiatry, Bipolar disorder, and related articles be held to a lower WP:MEDRS standard than alternative medicine and its related articles, as to its designation as a healing "medicine", with implications to claims of efficacy and intent of all areas of its practice (e.g., forensic psychiatry, or psychiatry practiced under the color of being "medicine" at Guantanamo), when there are substantial MEDRS sourced content that at, least part of psychiatry, is not based on science at all, and other parts are not intended to heal anything?
  • Should Diagnostic and Statistical Manual of Mental Disorders V be continued to be unquestioned as MEDRS, and citec as "common knowledge" which, if questioned with MEDRS or RS, is claimed to be WP:BATTLE and violate WP:COMMON SENSE, as was DSM IV, especially in light of comments such as that of Allen Frances, chair of the DSM-IV Task Force - "DSM 5 will accept diagnoses that achieve reliabilities as unbelievably low as 0.2-0.4 (barely beating the level of chance agreement two monkeys could achieve throwing darts at a diagnostic board".

In light of the above discussion and related discussions at associated talk pages, discussion re uniform application of MEDRS standards to all WP articles is here.

ParkSehJik (talk) 19:51, 29 November 2012 (UTC)

History cont.

Ancheta Wis, deep thanks for yr research; I'm puzzled. You suggest Popkin is mistaken in thinking Sanches' Spanish phrase means 'scientific method' because the term was first used in the C19th. This is circular because if he wasn't mistaken then the first use wasn't in the C19th. Can you say how he is mistaken and the basis for saying the first use was in the C19th? On the big picture, I still think "innate skepticism" misrepresents both skepticism as an argued philosophy and the fact that 1580ish was a very particular moment in intellectual history because of skepticism's revival, which matters if skepticism was formative of modern science. Pertin1x (talk) 21:43, 22 June 2012 (UTC)

Pertin1x, let's transfer the discussion to Talk:History of scientific method. What we are learning fits there better. Briefly, Popkin cites the title of a lost book. The commentary on this lost book, such as
  • Guy Patin (1701) Naudaeana et Patiniana, pp. 72-3 cites Sanchez Metodo universal de las ciencias, a lost, projected, or unpublished title, according to Limbrick. This is the best evidence so far for Popkin's statement.
  • Émilien Jean Marie Senchet (1904) Essai sur la méthode de Francisco Sanchez Paris: V. Giard & E. Briere
I do not disagree that Sanchez Metodo universal de las ciencias means that he originated this term and that he even called it a 'universal method'. But the text of the Sanches book is currently lost to us.
Popkin breaks new ground for the English-speaking world, and I propose that we introduce the topic on the history page with your Popkin citation. I have been basing everything on the Limbrick/Thomson edition of Sanches, and the Google books link, above, agrees with what is in the Sanches article already, not with Limbrick. She states that Sanches was in Rome 1571-1573, but the Google books link says 1569. She states that there was a botanical garden at La Sapienza, but it was not started til the 1700s, too late for Sanches. Patin introduces errors about Sanches as well; he got his death date mixed up, an error corrected by Limbrick's research.
From our perspective, I think we both agree skepticism is 'in' scientific method. The current best evidence is Descartes' Discourse on Method, but Descartes does not cite Sanches, although Limbrick suggests that Descartes saw Quod Nihil Scitur in Germany. I think we both agree that Sanches' modus sciendi is his method, as suggested by the titles above. It's just not for sure, from my point of view, because it's a lost work. What Sanches does say in Quod Nihil Scitur is 'get down to the details, the specifics, the thing itself' before you can say you know something. --Ancheta Wis   (talk | contribs) 10:01, 23 June 2012 (UTC)

Pertin1x, I have just posted a proposal on Talk:History of scientific method. All editors, please feel free to correct any misstatements I may have inserted on that talk page. --Ancheta Wis   (talk | contribs) 02:18, 25 June 2012 (UTC) ≠≠≥

Ok, meet you there.Pertin1x (talk) 12:08, 25 June 2012 (UTC)

edits re skepticism as per approved alteration to History of Scientific MethodPertin1x (talk) 08:26, 30 November 2012 (UTC)

Nature speaks for itself??

Related Articles

Wikiwand AI